Induction of active demethylation and 5hmC formation by 5-azacytidine is TET2 dependent and suggests new treatment strategies against hepatocellular carcinoma
- 2.3k Downloads
Global deregulation of DNA methylation is one of the crucial causes of hepato cellular carcinoma (HCC). It has been reported that the anti-cancer drug 5-azacytidine (5-AZA) mediates the activation of tumor suppressor genes through passive demethylation by inhibiting DNMT1. Recent evidence suggests that active demethylation which is mediated by ten-eleven translocation (TET) proteins may also be an important step to control global methylation. However, there exists a controversial discussion in which TET proteins are involved in the demethylation process in HCC. Therefore, we firstly wanted to identify which of the TETs are involved in demethylation and later to study whether or not 5-AZA could trigger the TET-dependent active demethylation process in HCC. HCC cell lines (Huh-7, HLE, HLF), primary human hepatocytes (hHeps), and tissues from both healthy (55 patients) and HCC patients (55 patients) were included in this study; mRNA levels of isocitrate dehydrogenase (IDH1, 2) and TETs (TET1–3) were studied via qPCR and confirmed by Western blot. The expression of 5hmC/5mC was determined by immunohistochemistry in human HCC tissues and the corresponding adjacent healthy liver. HCC cell lines were stimulated with 5-AZA (0–20 μM) and viability (Resazurin conversion), toxicity (LDH release), proliferation (PCNA), and 5hmC/5mC distribution were assessed. In addition, knockdown experiments on TET proteins in HCC cell lines using short interference RNAs (siRNAs), in the presence and absence of 5-AZA, were performed.
Our data applying qPCR, immunofluorescence, and Western blotting clearly show that TET2 and TET3 but not TET1 were significantly decreased in HCC tissue and different HCC cell lines compared to non-tumor liver tissues and hHeps. In addition, we show here for the first time applying knockdown experiments that 5-AZA is able to trigger an active TET2-dependent demethylation process with concomitant significant changes in 5hmC/5mC in HCC cell lines and hHeps.
Our data clearly show that the expression and activity of TET2 and TET3 proteins but not TET1 are impaired in hepatocellular carcinoma leading to the reduction of 5hmC in HCCs. Furthermore, this study identified a novel function of 5-azacytidine in promoting a TET-mediated generation of 5hmC suggesting that the availability of 5-AZA in cancer cells will have various effects on different epigenetic targets. These findings may open new therapeutic strategies for epigenetic drugs to treat HCC.
KeywordsTET 5-Hydroxymethylcytosine DNA methylation 5-Azacytidine Cancer biomarker
hepato cellular carcinoma
human primary hepatocyte
proliferating cell nuclear antigen
short interference RNA
In the past years, epigenetic studies have focused on DNA methylation in mammals. DNA methylation plays an important role in genomic imprinting, especially in the inactivation pattern of X-chromosome [1, 2], the silencing of retrotransposon, and activation of tissue-specific genes. Recently, it has been demonstrated that these epigenetic changes have an influence in various pathologies and gene therapies, for instance, in cancer , viral infections , and the activity of mobile elements .
Methylation at the C-5 position of cytosine (C) has long been considered to be the only biologically functional epigenetic modification of the animal genomic DNA. In mammalians, 5-methylcytosines (5mC) established by DNA methyltransferases (DNMTs), mainly at CpG dinucleotides, is associated with gene repression [6, 7, 8].
Loss of methylation was observed during the development processes. Therefore, it was suggested that DNA methylation is reversible either through passive demethylation via DNA replication or by active demethylation via DNA repair followed by base excision repair of modified nucleotides or DNA glycosidase [9, 10, 11]. Moreover, it has been shown in many studies that active demethylation does occur in many mammalian cells . The most well-known example is the rapid loss of 5mC immunoreactivity of the paternal pronucleus in the zygote after fertilization . The active DNA demethylation pathway is enabled by the ten-eleven translocation proteins (TETs), named after the ten-eleven translocation (q22, q23) (10, 11) occurring in rare cases of acute myeloid and lymphocytic leukemia, which all belong to the family of Fe(II)/2-oxyglutarate-dependent dioxygenases [9, 14]. 5mC is oxidized by TET proteins and thereby converted to 5hmC, 5fC, and 5CaC [9, 15]. TET proteins mediate the DNA demethylation process and increase the diversity of the epigenetic state of genomic DNA. It was suggested that 5hmC represents an intermediate product in the active DNA demethylation process possibly following a demethylation mechanism through the thymidine salvage pathway to repair DNA alkylation damages .
The loss of 5hmC, together with a downregulation or mutation of TET proteins, have been reported in numerous types of cancer, e.g., melanoma, breast, prostate, myeloid, brain, and liver cancer [16, 12]. For example, mutations of TET2 protein have been recorded in patients with different myeloid malignancies [17, 18, 19]. A number of studies have provided an important link between TET2 dysfunction, an oncogenic metabolite 2-hydroxyglutarate (2-HG), myeloid differentiation, and malignancies [20, 21].
Hepatocellular carcinoma (HCC), which is frequently caused by a hepatitis virus B and C infection or alcohol abuse, is the most common type of primary liver cancer [22, 23]. The mechanisms that lead to HCC development are still unclear. Beside many possibilities that may lead to HCC development, it has been shown that accumulation of genetic abnormalities such as chromosomal alterations and gene amplifications, mutations along epigenetic alterations can trigger HCC development . Recently, a particular focus has been directed towards abnormal DNA methylation at the fifth position of cytosine (5mC) . Therefore, the detection of DNA methylation may provide a potential application for the clinical diagnosis of HCC. Liu and colleagues have reported that a reduction of 5hmC is associated with progression of liver cancer through downregulation of TET1 protein . In contrast, Gao et al. have studied nine samples and reported not only the downregulation of TET1 but also of TET2 mRNA levels with a concomitant decrease of 5hmC. The researchers, however, found no change in TET3 expression in hepatocellular carcinoma compared to normal liver samples . Moreover, in another study by Yang et al., the decrease of all three TET genes was shown in three pairs of frozen human hepatocellular carcinoma tissue compared to matched normal liver tissue . Despite accumulating evidence for the correlation between loss and decrease of 5hmC and progression of hepatocellular carcinoma, it remains totally unclear, which of the TET proteins seems to be responsible for the loss of active demethylation pattern in HCC.
In contrast to conventional or molecularly targeted therapies for inhibiting dysregulated genes or signaling pathways in HCC, epigenetic drugs may provide an alternative approach by reversing the methylation status. 5-Azacytidine is known as a DNA methyltranferase inhibitor (DNMTi), which is clinically approved for the treatment of myelodysplasia syndrome and acute myelogenous leukemia (AML) [28, 29]. However, the role of 5-azacytidine in active demethylation pathway is not clear.
Therefore, because of the apparent debate, which TET proteins are involved in 5hmC/5mC regulation in HCC, our primary aim of this study was to identify which TET protein play a crucial role in the regulation of 5hmC and 5mC in HCC. Furthermore, we wanted to know whether or not 5-AZA triggers a TET-dependent active demethylation process in HCC controlling 5hmC/5mC regulation.
Cell culture medium, DMEM medium, William’s medium E, and cell culture supplements were purchased from Sigma-Aldrich (Steinheim, Germany). Cell culture plastics, phosphate buffered saline (PBS), and fetal calf serum (FCS) were purchased from PAA Laboratories GmbH (Pasching, Austria). DNaseI (RNase—free) and first strand cDNA Synthesis Kit were purchased from Fermantas (Ontario, Canada). 5-Azacytidine (SLBH7350V) was obtained from Sigma-Aldrich (Steinheim, Germany). All other chemical compounds were purchased from Carl Roth (Karlsruhe, Germany). 5hmC (39769) rabbit pAB and 5mC (39649) mouse mAB were purchased from Active Motif (Carlsbad, CA, USA). Proliferating cell nuclear antigen (PCNA) (ab92552) rabbit mAB was obtained from Abcam (Cambridge, UK). Corresponding secondary antibodies goat anti-rabbit Alexa 555 and goat anti-mouse 488 were acquired from Invitrogen (Carlsbad, CA, USA). Anti-TET2, anti-TET3, and anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH) antibodies were used from Sigma-Aldrich (Munich, Germany). The HRP-linked anti-rabbit IgG secondary antibody was purchased from Cell Signaling (Beverly, MA, USA).
Tissue samples and primary human hepatocyte isolation and cell culture condition
Tissue specimens were obtained from patients undergoing resection of HCC according to the approval of local ethics committee. A tissue microarray (TMA) containing HCC samples and their corresponding noncancerous liver tissue was constructed. Primary human hepatocytes were isolated from human liver tissue according to the institutional guidelines of the Tubingen University from liver resections of tumor patients with primary or secondary liver tumor (ethics approval number: 368/2012BO2). The isolation and purification of primary human hepatocytes were performed as previously described . Culture condition of HCC cell lines (Huh7, HLE and HLF) and human primary hepatocytes (hHeps) was published previously [31, 32, 30]. HLE and HLF cells were purchased from ATCC, and Huh7 was purchased from JCRB (Japanese Collection of Research Bioresources Cell Bank). The HCC cell lines as well as hHep were plated onto 6-, 24-, or 96-well plates, treated for 24 h and subsequently incubated with different concentrations of 5-AZA (20, 10, 5, 1) for 48 h. Huh7, HLE, and HLF cells were cultured in DMEM (Lonza Group Ltd., Cologne, Germany) supplemented with 2 mM glutamine (PAA Laboratories GmbH, Pasching, Austria) and 10 % FCS (Invitrogen, Darmstadt, Germany) at 37 °C and 5 % CO2 in a humidified incubator. Experiments were conducted between passage numbers 2 and 12 without full confluency throughout experiments. Absence of mycoplasma contamination was confirmed by Venor®GeMtest (Minerva Biolabs GmbH, Berlin, Germany).
Immunohistochemical stains were performed using 3.5 μm sections of TMA. Slides were deparaffinized, dehydrydrated, and after antigen retrieval were placed in 2 N HCL for 60 min, rinsed with distilled water, and placed in 100 Mm Tris–HCl (pH 8.5) for 10 min. Tissue slides were incubated in Dako Dual Endogenous Enzyme Block (CA, USA) for 10 min. The sections were blocked in PBS solution containing 10 % FCS and 0.1 % Tween-20 for 1 h. The section was incubated with either anti-5hmC rabbit at 1:1000 (active motif, 39769) or anti-5mC mouse at 1:500 (active motif, 39649) overnight at 4 °C and then incubated with Plus HRP One-Step Polymer kit from Zytochem (Berlin, Germany) for 30 min. Tissue sections were washed and subsequently DAB Substrate Kit High Contrast from Zytochem (Berlin, Germany) was used according to the manufacturer’s instruction. After rinsing, the cells were counterstained with hematoxylin. Scoring for immunohistochemical staining was performed by SAS and BS. All specimens were evaluated according to the 0–4 grading criteria (based on the percentage of 5hmC positive cells) and 0–3 grading criteria (based on the staining intensity) . The intensity score was multiplied with percentage of positive cells of three different high power fields for each section. The total score was determined from 0 to 12.
Cells were plated onto cover slips and treated according to the experimental setup. Cells were fixed with 4 % paraformaldehyde solution for 15 min at RT and then washed with PBS. For permeabilization, cells were incubated with 0.5 % Triton-X-100 PBS solution for 15 min at RT. To denature the DNA, cells were incubated with 4 M HCl for 15 min at RT, rinsed with distilled water, and placed in 100 mM Tris–HCl (pH 8.5) for 10 min. After washing with PBS, unspecific binding sites were blocked with blocking buffer (10 % FCS, 0.1 % Tween-20 in PBS) for 1 h at RT. Then, cells were incubated with primary antibodies such as anti-5hmC rabbit polyclonal IgG (Active Motif, CA, USA), anti-5mC mouse monoclonal IgG, (Active Motif, CA, USA) or anti-PCNA Rabbit mAB (Abcam, Cambridge, UK) at 1:1000, 1:400, and 1:200, respectively, in PBS solution containing 1 % FCS and 0.1 % Tween-20 overnight at 4 °C. After washing with PBS, cells were incubated with secondary antibody solution (Alexa-Fluor antibodies, Invitrogen, NY, USA) diluted 1:400 in PBS solution containing 1 % FCS, 0.1 % Tween-20 for 1 h at RT. The nuclei were counterstained by incubation with Hoechst 33342 solution (2 μg/ml in PBS) for 10 min at RT. After a final washing step with PBS, the stained cells were mounted with mounting medium (Fluoromount G, Southern Biotech, NJ, USA). Images of the staining were taken with an EVOS fluorescence microscope (AMG, Life technologies, MA, USA) under standardized condition, processed, and analyzed with Image J 1.45s software (NIH, USA) .
RNA extraction, cDNA synthesis, and quantification of gene expression
Total cellular RNA was isolated with Trifast (Peqlab, Erlangen, Germany); 2–3 μg of the total RNA was digested with DNase I in order to remove remaining genomic DNA, according to the manufacturer’s instructions (Fermentas International Inc., Ontario, Canada). Complementary DNA (cDNA) was synthesized by Revert Aid First Strand cDNA Synthesis Kit (Fermentas International Inc., Ontario, Canada).
Sequences of primers used in qPCR
Gene bank ID
Product length (bp)
Resazurin conversion, LDH leakage, and FACS measurement
To investigate the impact of 5-AZA on cell proliferation, the cells were incubated with different doses of 5-AZA at different time points. Cell viability (mitochondrial activity) was determined by resazurin conversion. Briefly, 1/10 volume of the resazurin stock solution (0.025 % in DPBS) was added to the cells. After 30 min of incubation at 37 °C, fluorescence was measured (Ex/Em = 544/590 nm) and corrected to background control (solvent mixture without cells). Viability is given as percentage of control (untreated cells).
LDH released into cell culture media as index of cell death was measured using an LDH assay kit from Analyticon (Lichtenfels, Germany) according to the manufacturer’s protocol. LDH released into the media was expressed as the percentage of total cellular LDH per well measured after cells had been lysed with 1 % Triton X-100.
Apoptosis was determined by fluorescence-activated cell sorting (FACS) quantification of hypodiploid cells. For FACS analysis of sub2N peaks, 1 × 106 HCC cells per well were seeded in six-well plates and treated with different concentrations of 5-AZA. After 48 h, cells were fixed in ethanol and stained in hypotonic solution with RNase and Propidium Iodid (PI) for 30 min. Hypodiploid (sub2N) cells were considered apoptotic .
HCC cell lines and hHep cells were lysed in ice-cold RIPA lysis buffer (50 mM Tris, 250 mM NaCl, 2 % Nonidet-P40, 2.5 mM EDTA, 0.1 % SDS, 0.5 % DOC, complete protease inhibitor, 1.0 % phosphatase inhibitor, pH 7.2). Protein concentration was determined by micro-Lowry. Fifty micrograms of total protein was separated by SDS-PAGE and transferred to nitrocellulose membranes (Roth, Karlsruhe, Germany). The membranes were blocked by 5 % BSA solution for 2 h and incubated overnight with rabbit polyclonal primary antibodies (TET2, TET3, and GAPDH) (TET2, TET3 antibodies prepared 1:500 and GAPDH 1:10,000 in 5 % w/v BSA 1× TBS, 0.1 % Tween-20) (Sigma Aldrich, Munich, Germany) at 4 °C. The following day membranes were incubated with the corresponding HRP-labeled secondary antibodies for 2 h at RT. Chemiluminescent signals were detected with the ChemoCam (INTAS, Gottingen, Germany).
Protein depletion by short interference RNAs
Oligonucleotide sequences for siRNA primers
5′ or 3′ Mod
Statistical significance of differences between the individual treatments was evaluated by Student’s t test, one-way ANOVA, and whisker (Prism 5.01, GraphPad Software, San Diego, USA). Data are means ± SEM of three to five independent experiments. All statistical comparisons were performed two-sided in the sense of an exploratory data analysis using 0.05 (*), 0.01 (**), and 0.001 (***) level of significance.
5hmC is decreased in human HCC tissues and cell lines
Reduction of 5hmC correlates with decreased TET2 and TET3 gene expression in HCC tissues and cell lines
Effect of 5-azacytidine on HCC cells proliferation and viability
5-Azacytidine triggers active demethylation in HCC cells
Changing the balance of DNA methylation and demethylation related to epigenetic modification has become a main interest of cancer research in recent years . The loss of 5hmC as a demethylation marker has been suggested as prognostic marker in various types of cancer and even to play a crucial role in the pathogenesis of cancers [42, 43, 44]. It was reported that 5hmC, in addition to its role in DNA demethylation via TET proteins, has a role in the regulation of genes involved in development, pluripotency, and the regulation of RNA splicing processes [45, 40]. It was shown that 5hmC is a major epigenetic modification mark in an adult human liver, playing an important role in hepatic gene expression changes in hepatocytes [46, 47]. In our study, we demonstrated that the generation of 5hmC is significantly reduced in hepatocellular carcinoma tissues and cell lines compared with healthy liver tissue and human hepatocytes. In line with other papers, our data clearly support the link between 5hmC ablation and tumor development [48, 49]. Furthermore, our study indicates that 5mC levels are induced in HCC tissues as well as cell in HCC cell lines. Altogether, our study confirms that detection of 5hmC could be a crucial diagnostic marker for HCCs.
TET proteins (TET1, TET2, and TET3) mediate the oxidation of 5mC to 5hmC. TET proteins were identified as dioxygenases that utilize two substantial factors: Fe(II) and 2-oxyglutarate (2-OG), in order to oxidize the methyl group (CH3) of 5mC, and to form 5hmC, thereby mediating active DNA demethylation [9, 14, 28, 50]. The expression of these three TETs varies not only in different organs but also in the development of organs [4, 14]. In various tumor types, a different TET family member has been reported to function as tumor suppressor [5, 6, 7, 8]. Thereby, identification of the most effective TET protein leading to cancer in the corresponding tissue is a crucial topic to be investigated in different tumors. Recent evidence suggests that there is a remarkable correlation between decreased 5hmC levels and TET expression, resulting in progression of tumor and metastasis, suggesting that TET proteins might serve as a tumor suppressor in certain cancer types [37, 51]. Recently, it was reported that the gene encoding TET2 but not TET1 and TET3 was frequently mutated, and it was identified as the relevant tumor suppressor gene, which is mutated in leukemia . However, in breast cancer, low expression of TET1 correlated with advanced cancer stage . Furthermore, in some cancer types like colorectal cancer (CRC), the reductions of all three TET proteins have been reported [11, 53]. In addition, it was shown that IDH1 or IDH2 mutations could also result in the reduction of 5hmC in cancers [16, 33, 54]. In HCCs, however, the published results are contradicting each other regarding the role of TETs. Liu et al. reported that a reduction of 5hmC is associated with the downregulation of the TET1 protein in HCC  while Gao et al. claimed a downregulation of TET1 and TET2 but no change in TET3 gene expression with a concomitant decrease of 5hmC in HCC . Moreover, Yang et al. reported the decrease of all three TET genes in HCC; however, this group investigated only three pairs of frozen human hepatocellular carcinoma tissue compared to matched normal liver tissue . In our study, we clearly found that only TET2 and TET3 genes are significantly reduced in HCC tissue and in hepatocellular carcinoma cell lines. These changes were paralleled by a concomitant decrease of IDH genes and 5hmC and a significant increase of 5mC. The change in 5hmC and 5mC generation is confirmed by all papers [18, 26, 27], including our own results. But what may be the reasons for these conflicting results? Recently, it was shown that TET1 plays a crucial role in the early development process in self-renewal pattern in embryonic stem cells [4, 14] suggesting a more crucial role for TET1 in embryogenesis. Another hypothesis explaining the differences may be the infiltration of the so-called cancer stem cells with an embryonic pattern, which might be responsible for differences seen in TET1 expression . Another explanation could be simply the use of specific analytical tools. It has been reported that the expression of typical housekeeping genes including glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and beta-actin (ACTB) can be affected through HCC development, and these housekeeping genes are expressed significantly different in HCC compared to normal liver tissue . The use of instable genes for normalization may lead to an over- or underestimation of the fold changes or to misinterpretation of the results .
It is well established that 5-AZA causes the reduction of the DNA methylation status by inhibiting DNMT through passive demethylation pathway . Recently, the active demethylation of DNA through oxidation of 5mC to 5hmC mediated by TET proteins was discovered . To our knowledge, our study is the first to demonstrate that 5-AZA reduces the methylation status of DNA not only by triggering the passive demethylation pathway of DNA  but also the active demethylation pathway through converting 5mC to 5hmC by inducing TET2 and TET3 proteins. However, our results showed that 5-AZA was unable to induce the 5hmC in TET2 knocked-down HCC cell lines suggesting a crucial role for TET2 in 5hmC induction and in the pathogenesis of HCC. This was further underlined by our findings that 5-AZA stopped HCC tumor cell growth through decline of 5mC and a strong increase of 5hmC. However, a detailed mechanism of how 5-AZA affects the expression and activity of TET proteins remains to be further elucidated.
The data presented here indicate that a drop of TET2 and TET3 expression and activity as well as IDHs is impaired in hepatocellular carcinoma with a concomitant reduction of 5hmC. Therefore, epigenetic drugs that mediate the conversion of 5mC to 5hmC through modulating TETs and/or IDHs activity may find broad use in the treatment of HCCs.
This study identified a novel function of 5-azacytidine in promoting a TET-mediated generation of 5hmC suggesting that the availability of 5-AZA in cancer cells will have various effects on different epigenetic targets.
We would like to thank Ms. K. Neumann for editing the manuscript. This study was partially funded by the Ministerium für Ländlichen Raum und Verbraucherschutz/Baden-Württemberg (BW 05110214) and the SFB/TTR77 (SD) funded by the Deutsche Forschungsgemeinschaft (DFG).
- 13.Surani M, Hajkova P, editors. Epigenetic reprogramming of mouse germ cells toward totipotency. Cold Spring Harbor Symposia on Quantitative Biology; 2010: Cold Spring Harbor Laboratory Press. doi: 10.1101/sqb.2010.75.010.
- 28.Griffiths EA, Gore SD, editors. DNA methyltransferase and histone deacetylase inhibitors in the treatment of myelodysplastic syndromes. Seminars in hematology; 2008;45(1):23–30.Google Scholar
- 32.Lin J, Schyschka L, Mühl-Benninghaus R, Neumann J, Hao L, Nussler N, et al. Comparative analysis of phase I and II enzyme activities in 5 hepatic cell lines identifies Huh-7 and HCC-T cells with the highest potential to study drug metabolism. Arch Toxicol. 2012;86(1):87–95.CrossRefPubMedGoogle Scholar
- 34.Polzer H, Haasters F, Prall WC, Saller MM, Volkmer E, Drosse I, et al. Quantification of fluorescence intensity of labeled human mesenchymal stem cells and cell counting of unlabeled cells in phase-contrast imaging: an open-source-based algorithm. Tissue Eng Part C Methods. 2010;16(6):1277–85. doi: 10.1089/ten.TEC.2009.0745.CrossRefPubMedGoogle Scholar
- 42.Kroeze LI, van der Reijden BA, Jansen JH. 5-hydroxymethylcytosine: an epigenetic mark frequently deregulated in cancer. Biochimica et Biophysica Acta (BBA)-Reviews on Cancer. 2015.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.