Oxidative stress, DNA methylation and carcinogenesis

Available online at www.sciencedirect.com

Cancer Letters 266 (2008) 6–11 www.elsevier.com/locate/canlet

Mini-review

Oxidative stress, DNA methylation and carcinogenesis Rodrigo Franco a,b,1, Onard Schoneveld b,1, Alexandros G. Georgakilas c, Mihalis I. Panayiotidis d,* a

Laboratory of Cell Biology and Signal Transduction, Biomedical Research Unit, FES-Iztacala, UNAM, Mexico, USA Laboratory of Signal Transduction, National Institute of Environmental Health Sciences, National Institutes of Health, Research Triangle Park, NC 27709, USA c Department of Biology Thomas Harriot College of Arts and Sciences, East Carolina University, Greenville, NC 27858, USA d School of Public Health, University of Nevada-Reno, MS-274, Reno, NV 89557, USA b

Received 12 February 2008; received in revised form 12 February 2008; accepted 14 February 2008

Abstract Transformation of a normal cell to a malignant one requires phenotypic changes often associated with each of the initiation, promotion and progression phases of the carcinogenic process. Genes in each of these phases acquire alterations in their transcriptional activity that are associated either with hypermethylation-induced transcriptional repression (in the case of tumor suppressor genes) or hypomethylation-induced activation (in the case of oncogenes). Growing evidence supports a role of ROS-induced generation of oxidative stress in these epigenetic processes and as such we can hypothesize of potential mode(s) of action by which oxidative stress modulates epigenetic regulation of gene expression. This is of outmost importance given that various components of the epigenetic pathway and primarily aberrant DNA methylation patterns are used as potential biomarkers for cancer diagnosis and prognosis. Ó 2008 Elsevier Ireland Ltd. All rights reserved. Keywords: Oxidative stress; Free radicals; Reactive oxygen species; Oxidative DNA damage; DNA methylation; Epigenetics; Gene expression; Cancer

1. Introduction Oxidative stress has long been known to be involved in the pathophysiology of many human diseases including, but not restricted, to cancer. The term ‘‘oxidative stress”, refers to a cell’s state characterized by excessive production of reactive * Corresponding author. Tel.: +1 775 682 7082; fax: +1 775 784 1340. E-mail address: [email protected] (M.I. Panayiotidis). 1 Both authors have contributed equally to the preparation of the manuscript.

oxygen species (ROS) and/or a reduction in antioxidant defenses responsible for their metabolism. This generates an imbalance between ROS production and removal in favor of the former. Cancer is a multistage process and often involves ‘‘alterations” or ‘‘changes” in the transcriptional activity of genes associated with many critical cellular processes for tumor development including proliferation, senescence, inflammation, metastasis, etc. Furthermore, there are known to be both genotoxic and non genotoxic mechanisms contributing to malignant transformation. Generally, genotoxic mechanisms involve changes in genomic DNA

0304-3835/$ - see front matter Ó 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2008.02.026

R. Franco et al. / Cancer Letters 266 (2008) 6–11

sequences that ultimately lead to mutations whereas non-genotoxic include mechanisms (other than directly affecting DNA) capable of modulating gene expression. ROS have been implicated at all stages of the carcinogenic process by involving both types of mechanisms [1]. On the other hand, the term ‘‘epigenetics” refers to altered levels of a gene’s transcriptional activity without directly affecting its primary DNA nucleotide sequence. It involves alterations in DNA methylation patterns that together with specific histone modifications (methylations, acetylations, deacetylations, etc.) contribute to a transcriptional inactive chromatin state [2]. In this respect, epigenetic regulation of gene expression can be viewed as a non genotoxic mechanism for promoting tumor formation. Throughout this review article, we will present the evidence for the involvement of oxidative stress in carcinogenesis, consider the evidence for its role in inducing DNA methylation changes and assess the importance of these changes in the multistage process of human carcinogenesis. 2. Sources of ROS There are both endogenous and exogenous sources of ROS generation. Endogenous sources include those of: (1) mitochondrial oxidative phosphorylation, (2) P450 metabolism, (3) peroxisomes and (4) activation of inflammatory cells. It has been postulated that during oxidative phosphorylation 1–2% of molecular oxygen is converted to ROS primarily through a series of sequential one-, twoand three-electron reductions giving rise to superoxide, hydrogen peroxide and hydroxyl radical formation consecutively [3]. In addition, activation of P450 metabolism has been proposed as a very significant source of ROS formation by mechanisms that involve: (1) redox cycling, (2) peroxidase-catalyzed drug oxidations and (3) ‘‘futile cycling” of cytochrome P450. In particular, induction of P450 2E1 and 2B has been proposed to contribute significantly to ROS formation primarily through metabolism of ethanol and phenobarbital, respectively. Furthermore, in a number of studies using chemicals such as peroxisome proliferators, an increased production of hydrogen peroxide was observed that contributed to increased levels of oxidative stress and their association with cancer induction. Finally, inflammatory cells like neutrophils and macrophages are perhaps the most significant sources of

7

endogenous ROS formation. More specifically, activated macrophages through ‘‘respiratory burst” give rise to various ROS and primarily superoxide and hydrogen peroxide. For instance, activation of specialized macrophages in liver known as Kupffer cells has been implicated in tumor promotion through ROS release [1]. On the other hand, exogenous sources of ROS generation include those of various xenobiotics (with a variable degree of potency), chlorinated compounds, various transition metals (participating in Fenton-type chemical reactions) and radiation, all of which have been documented to cause ROSinduced damage to cellular macromolecules (DNA, RNA, lipids and proteins) both in vitro and in vivo [1]. 3. Involvement of oxidative stress in carcinogenesis In general, the multistage process of carcinogenesis involves the distinct phases of initiation, promotion and progression. The cellular and molecular events underlying each of these phases include DNA damage, increased proliferation, deficient cell death and further genetic instability, respectively [1,4]. Oxidative DNA damage can trigger tumor initiation. More specifically, studies using ionizing radiation have shown multiple hydroxyl radical-induced genotoxic by- products in the bases as well as the deoxyribose backbone of DNA. Among these byproducts, the most characterized one is 8-hydroxy2-deoxyguanosine (8-OHdG) [5,6]. This DNA lesion has been shown to be mutagenic by causing G ? T transversions in both bacterial and mammalian cells [7]. However, there are a number of studies indicating that oxidative DNA damage could not account entirely by itself for tumor development as they have shown that elevated levels of 8-OHdG do not reflect increased cancer rates [8,9]. So, what other mechanisms could account for the involvement of ROS in carcinogenesis? As it was mentioned earlier on, ROS can affect a number of cellular processes critical in tumor development such as cell proliferation, senescence, inflammation, metastasis, etc. In terms of cell proliferation, ROS have been shown to modulate cell cycle regulation through modulation of various cell cycle proteins including p53 [10] and ATM (ataxia telangiectasia mutated) [11]. ROS have also been shown to modulate the cell death (senescence) process [12] by acting either as an anti-senescence stimulus [13] or through

8

R. Franco et al. / Cancer Letters 266 (2008) 6–11

the specific induction of AIF (apoptosis-inducing factor) which suppresses apoptosis and consequently maintains the transformed phenotype of a cancer cell [14]. On the other hand, inflammation has long been known to act as a trigger in cancer development. For example, chronic hepatitis B viral infection and elevated liver levels of 8-OHdG can lead to development of hepatocellular carcinoma [15]. The mode of action by which chronic inflammation can trigger tumor initiation seems to involve NF-jB (nuclear factor jB) activation which, among its pleiotrophic effects, can cause inhibition of apoptosis [16]. Finally, metastasis is an integral part of tumor progression during which ROS have been documented to play a major role [17–19]. In fact, various studies have shown that metastatic tumor cells produce higher levels of ROS than primary malignant cells which together with increasing levels of ROS metabolizing enzymes and antioxidant compounds greatly reduce metastasis [20–22]. 4. The role of epigenetics in carcinogenesis In general, increased DNA methylation in the promoter region of genes causes gene silencing and in this respect can contribute to the multistage process of carcinogenesis. In addition, in mammalian cells, both DNA methylation and chromatin structure are interconnected in specific ways so that genes will either be transcribed or repressed. Briefly, the process is initiated by DNA methyltransferases (DNMTs) bringing together the DNA methylation machinery to the chromatin through recruitment of histone deacetylases (HDACs) and other chromatin-binding proteins to promoter sites. In this regards, the state of the chromatin’s acetylation status is important in regulating transcriptional activity in a way that histone acetylation maintains chromatin in a transcriptionally active state whereas histone deacetylation maintains transcriptional silencing. More specifically, during gene silencing DNMTs direct the binding of HDACs and other proteins [methyl-CpG binding proteins (MeCPs); and methyl-CpG binding domain (MBP)] to hypermethylated regions of the chromatin where they form complexes with other molecules and thus blocking access of the transcriptional machinery to the promoter. Specific histone modifications have been implicated in gene silencing including the following: (1) histone H3 methylation at lysines 9 and 27 and/or deacetylation at lysine 9, (2) histone H4 methylation at lysine 20 and/or deacetylation

at lysine 16. Among these, the best-characterized histone modification(s) to date is methylation of histone H3 at lysine 9 and how it ‘‘translates” into DNA methylation in order to repress transcriptional activation [2,23–25]. But how is DNA methylation involved in regulation of transcription? In mammalian cells, the enzymes critical for DNA methylation reactions are DNA methyltransferases 1 (DNMT1), 3a (DNMT3a), and 3b (DNMT3b). These enzymes catalyze the transfer of the methyl donor group from S-adenosylmethionine (SAM) to the 50 -carbon of cytosines within CpG dinucleotide islands (regions of 0.5–4.0 kb in length) in genomic DNA. More specifically, these enzymes are involved in DNA methylation by means of either maintenance (DNMT 1) and/or de novo methylation (DNMT 3a and DNMT 3b). Thus, they could potentially contribute not only to increased promoter methylation status (through de novo methylation) but also ensure inheritance of gene silencing (through maintenance methylation) both of which could account for the acquisition of a malignant transformation phenotype [26,27]. In general, 5-methylcytosine (by being a heritable modification) has the potential to alter gene expression by means of: (1) causing an increase in mutation rate [28] given that the presence of 5-methylcytosine in the DNA promotes deamination of cytosine to uracil with 5-methylcytosine deaminating 2–4 times faster than cytosines [29] or by (2) silencing genes necessary for controlling gene proliferation [30,31]. Finally, many genes in tumor cells have been shown to contain alterations in their DNA methylation patterns. Two such alterations have been proposed to be the common epigenetic mechanisms underlying development of human cancer: (1) global hypomethylation occurring early in the progression phase of neoplasia and (2) regional hypermethylation of normally unmethylated CpG islands. Both events can result in the transcriptional silencing of tumor suppressor gene expression and the transcriptional activation of oncogenes respectively. In addition, increased expression of DNMT1 has been found in early and late stages of neoplasia suggesting its involvement in the generation of abnormal methylation patterns in cancer cells [32]. Finally, genome-wide hypomethylation has been shown to increase mutation rates and thus leading to genome instability [33]. It is therefore evident that alterations in DNA methylation patterns underlie aberrant gene expression that is associated with malignant transformation.

R. Franco et al. / Cancer Letters 266 (2008) 6–11

5. Involvement of oxidative stress in DNA methylation Oxidative stress can contribute to tumor development not only through genetic but also through epigenetic mechanisms. As it was mentioned earlier on, generation of the hydroxyl radical can cause a wide range of DNA lesions including base modifications, deletions, strand breakage, chromosomal rearrangements, etc. Such DNA lesions have been shown to interfere with the ability of DNA to function as a substrate for the DNMTs, resulting in global hypomethylation [34]. More specifically, X-rays [35], ultraviolet [36] and c-rays [37] have been shown to reduce the methyl-accepting ability of DNA. Also, the presence of 8-OHdG in CpG dinucleotide sequences has been shown to strongly inhibit methylation of adjacent cytosine residues [38,39], and interfere with the ability of restriction nucleases to cleave DNA [40]. More specifically, in a series of elegant experiments, it was shown that when 8-OHdG is substituted for either guanine on the HpaII methylase recognition site (CCGG) such substitution can inhibit DNA methylation of adjacent cytosines as well as binding to the methyltransferase. The extent of this inhibition is dependent on the position of 8OHdG [38,39]. In addition, 8-OHdG may not be recognized by proofreading enzymes and thus may persist as a mutation resulting in G ? T transversions [41]. Another potentially mutagenic lesion in ROS-induced DNA damage is O6-methylguanine. A number of studies have shown that its presence can inhibit binding of DNA methyltransferases and therefore can lead to hypomethylation by means of inhibiting methylation of adjacent cytosine molecules [42,43]. Alternatively, O6-methylguanine can be spontaneously mispaired with thymine and thus also contribute to DNA hypomethylation [44]. Finally, single-stranded DNA can signal de novo methylation and, thus, it may be possible that formation of single strand breaks by oxidative stress can contribute to the modifications of DNA methylation patterns observed in oxidant-transformed cell lines [45]. These studies suggest that oxidative DNA damage can affect patterns of DNA methylation leading to aberrant gene expression and possibly contributing to the development of malignancy. As it was mentioned earlier on, DNA methylation mediates gene expression via binding of MBPs through DNA–protein and protein–protein interactions followed by recruitment of histone modifying enzymes. Consequently, these interactions result in

9

chromatin condensation and transcriptional inactivation. Incorporation of 8-OHdG and 5-hydroxymethylcytosine (the oxidation by-product of 5methylcytosine) in the MBP recognition sequence resulted in the significant inhibition of the binding affinity of MBP [46]. These results provide further evidence about the mechanism underlying ROSinduced DNA damage in interfering with DNA methylation patterns and consequently with transcriptional activation. Only recently, there have been a number of studies investigating changes in gene expression levels of major antioxidant enzymes in an attempt to relate diminished levels with the pathophysiology underlying human carcinogenesis. In this respect, it has been shown that decreased expression of the enzyme manganese superoxide dismutase (MnSOD; which metabolizes superoxide anion) was associated with decreased proliferation in human pancreatic carcinoma [47] and breast cancer [48] cells and that its inhibition was underlined by DNA hypermethylation-induced silencing. In addition, in the case of the metallothionein gene expression (induced by a variety of oxidative stress stimuli), it was also found to be silenced through promoter-specific DNA hypermethylation events in both mouse lymphosarcoma [49] and rat hepatoma [50] cells. Finally, both the NAD(P)H: quinone oxidoreductase 1 (NQO1) and glutathione S-transferase P1 (GSTP1) genes (phase II xenobiotic metabolizing enzymes) have been shown to be inactivated via promoter hypermethylation in hepatocellular carcinoma [51,52], breast, renal [53] and prostate [54] cancers, respectively. These studies clearly demonstrate the association between repressed expression of key antioxidant enzymes (in metabolizing ROS generation) and tumor development by means of promoter hypermethylation-induced gene silencing. 6. Conclusions and perspectives It is evident that ROS-induced oxidative stress is involved in the multistage process of carcinogenesis by both genetic and epigenetic mechanisms. In particular, there seems to be a growing interest (by many investigators) in the involvement of oxidative stress in the epigenetic regulation of gene expression and specifically in controlling DNA methylation. Given the recent findings in detecting circulating tumor DNA in patients with cancer [55], it becomes of paramount importance to reveal the molecular mechanisms underlying aberrant DNA methylation

10

R. Franco et al. / Cancer Letters 266 (2008) 6–11

patterns since potentially they could be used as biomarkers in cancer prognosis and diagnosis. Acknowledgements This work was supported, in part, by the Intramural Research Program of the National Institutes of Health (NIH)/National Institute of Environmental Health Sciences (NIEHS) (Onard Schoneveld and Rodrigo Franco). Dr. Georgakilas has been supported by funds provided by the Biology Department of East Carolina University and a College Research Award (East Carolina University). Finally, Dr. Panayiotidis has been supported by funds provided by the School of Public Health of the University of Nevada at Reno. References [1] J.E. Klauning, L.M. Kamendulis, The role of oxidative stress in carcinogenesis, Annu. Rev. Pharmacol. Toxicol. 44 (2004) 239–267. [2] F. Fuks, DNA methylation and histone modifications: teaming up to silence genes, Curr. Opin. Genet. Dev. 15 (2005) 490–495. [3] A. Pappa, R. Franco, O. Schoneveld, A. Galanis, R. Sandaltzopoulos, M.I. Panayiotidis, Sulfur-containing compounds in protecting against oxidant-mediated lung diseases, Curr. Med. Chem. 14 (2007) 2590–2596. [4] B. Halliwell, Oxidative stress and cancer: have we moved forward?, Biochem J. 401 (2007) 1–11. [5] M.D. Evans, M. Dizdaroglu, M.S. Cooke, Oxidative DNA damage and disease: induction, repair, and significance, Mutat. Res. 567 (2004) 1–61. [6] L.J. Marnett, Oxyradicals and DNA damage, Carcinogenesis 21 (2000) 361–370. [7] M. Moriya, Single-stranded shuttle phagemid for mutagenic studies in mammalian cells: 8-oxoguanine in DNA induces targeted G:C ? T:A transversions in simian kidney cells, Proc. Natl. Acad. Sci. USA 90 (1993) 1122–1126. [8] B. Halliwell, Effect of diet on cancer development: is oxidative DNA damage a biomarker?, Free Radic Biol. Med. 32 (2002) 968–974. [9] T. Arai, V.P. Kelly, O. Minowa, T. Noda, S. Nishimura, The study using wild-type and Ogg1 knockout mice exposed to potassium bromate shows no tumor induction despite an extensive accumulation of 8-hydroxyguanine in kidney DNA, Toxicology 221 (2006) 179–186. [10] K. Bensaad, K.H. Vousden, Savior and slayer: the two faces of p53, Nat. Med. 11 (2005) 1278–1279. [11] K. Ito, A. Hirao, F. Arai, K. Takubo, S. Matsuoka, K. Miyamoto, M. Ohmura, K. Naka, K. Hosokawa, Y. Ikeda, T. Suda, Reactive oxygen species act through p38 MAPK to limit the lifespan of hematopoietic stem cells, Nat. Med. 12 (2006) 446–451. [12] J. Chandra, A. Samali, S. Orrenius, Triggering and modulation of apoptosis by oxidative stress, Free Radic. Biol. Med. 29 (2000) 323–333.

[13] E.C. Vaquero, M. Edderkaoui, S.J. Pandol, I. Gukovsky, A.S. Gukovskaya, Reactive oxygen species produced by NAD(P)H oxidase inhibit apoptosis in pancreatic cancer cells, J. Biol. Chem. 279 (2004) 34643–34654. [14] A. Urbano, U. Lakshmanan, P.H. Choo, J.C. Kwan, P.Y. Ng, K. Guo, S. Dhakshinamoorthy, A. Porter, AIF suppresses chemical stress-induced apoptosis and maintains the transformed state of tumor cells, EMBO J. 24 (2005) 2815–2826. [15] T.M. Hagen, S. Huang, J. Curnutte, P. Fowler, V. Martinez, C.M. Wehr, B.N. Ames, F.V. Chisari, Extensive oxidative DNA damage in hepatocytes of transgenic mice with chronic active hepatitis destined to develop hepatocellular carcinoma, Proc. Natl. Acad. Sci. USA 91 (1994) 12808–12812. [16] M. Karin, Nuclear factor-kB in cancer development and progression, Nature 441 (2006) 431–436. [17] P. Storz, Reactive oxygen species in tumor progression, Front. Biosci. 10 (2005) 1881–1896. [18] K. Mori, M. Shibanuma, K. Nose, Invasive potential induced under long-term oxidative stress in mammary epithelial cells, Cancer Res. 64 (2004) 7464–7472. [19] F. Okada, M. Kobayashi, H. Tanaka, T. Kobayashi, H. Tazawa, Y. Luchi, K. Onuma, M. Hosokawa, M.C. Dinauer, N.H. Hunt, The role of nicotinamide adenine dinucleotide phosphate oxidase-derived reactive oxygen species in the acquisition of metastatic ability of tumor cells, Am. J. Pathol. 169 (2006) 294–302. [20] M. Nishikawa, K. Hyoudou, Y. Kobayashi, Y. Umeyama, Y. Takakura, M. Hashiba, Inhibition of metastatic tumor growth by targeted delivery of antioxidant enzymes, J. Control. Release 109 (2005) 101–107. [21] I. Heirman, D. Ginneberge, R. Brigellius-Flohe, N. Hendrickx, P. Agostinis, P. Brouckaert, P. Rottiers, J. Grooten, Blocking tumor cell eicasonoid synthesis by GPx4 impedes tumor growth and malignancy, Free Radic. Biol. Med. 40 (2006) 285–294. [22] F. Okada, H. Shionoya, M. Kobayashi, T. Kobayashi, H. Tazawa, K. Onuma, Y. Luchi, N. Matsubara, T. Ijichi, B. Dugas, M. Hosokawa, Prevention of inflammation-mediated acquisition of metastatic properties of benign mouse fibrosarcoma cells by administration of an orally available superoxide dismutase, Br. J. Cancer 94 (2006) 854–862. [23] H. Santos-Rosa, C. Caldas, Chromatin modifier enzymes, the histone code and cancer, Eur. J. Cancer 41 (2005) 2381– 2402. [24] M. Esteller, DNA methylation and cancer therapy: new developments and expectations, Curr. Opin. Oncol. 17 (2004) 55–60. [25] S.B. Baylin, DNA methylation and gene silencing in cancer, Nat. Clin. Pract. Oncol. 2 (2005) S4–S11. [26] A. Bird, The essentials of DNA methylation, Cell 70 (1992) 5–8. [27] M.J. Hitchler, F.E. Domann, An epigenetic perspective on the free radical theory of development, Free Radic. Biol. Med. 43 (2007) 1023–1036. [28] W.M. Rideout, G.A. Coetzee, A.F. Olumi, P.A. Jones, 5Methylcytosine as an endogenous mutagen in the human LDL receptor and p53 genes, Science 249 (1990) 1288–1290. [29] G.P. Pfeifer, M. Tang, M.F. Denissenko, Mutation hotspots and DNA methylation, Curr. Top. Microbiol. Immunol. 249 (2000) 1–19.

R. Franco et al. / Cancer Letters 266 (2008) 6–11 [30] P.A. Jones, DNA methylation errors and cancer, Cancer Res. 56 (1996) 2463–2467. [31] S.B. Baylin, J.G. Herman, J.R. Graff, P.M. Vertino, J.P. Issa, Alterations in DNA methylation: a fundamental aspect of neoplasia, Adv. Cancer Res. 72 (1998) 141–196. [32] W.S. El-Deiry, B.D. Nelkin, P. Celano, R.W. Yen, J.P. Falco, S.R. Hamilton, S.B. Baylin, High expression of the DNA methyltransferase gene characterizes human neoplastic cells and progression stages of colon cancer, Proc. Natl. Acad. Sci. USA 88 (1991) 3470–3474. [33] J.E. Klaunig, L.M. Kamendulis, Y. Xu, Epigenetic mechanisms of chemical carcinogenesis, Hum. Exp. Toxicol. 19 (2000) 543–555. [34] J.T. Wachsman, DNA methylation and the association between genetic and epigenetic changes: relation to carcinogenesis, Mutat. Res. 375 (1997) 1–8. [35] V.L. Wilson, P.A. Jones, Inhibition of DNA methylation by chemical carcinogens in vitro, Cell 32 (1983) 239–246. [36] F.F. Becker, P. Holton, M. Ruchirawat, J.-N. Lapeyre, Perturbation of maintenance and de novo DNA methylation in vitro by UVB (280–340 nm)-induced pyrimidine photodimers, Proc. Natl. Acad. Sci. USA 82 (1985) 6055–6059. [37] J.F. Kalinich, G.N. Catravas, S.L. Snyder, The effect of c radiation on DNA methylation, Radiat. Res. 117 (1989) 185–197. [38] S.A. Weitzman, P.W. Turk, D.H. Milkowski, K. Kozlowski, Free radical adducts induce alterations in DNA cytosine methylation, Proc. Natl. Acad. Sci. USA 91 (1994) 1261– 1264. [39] P.W. Turk, A. Laayoun, S.S. Smith, S.A. Weitzman, DNA adduct 8-hydroxyl-20 -deoxyguanosine (8-hydroxyguanine) affects function of human DNA methyltransferase, Carcinogenesis 16 (1995) 1253–1255. [40] P.W. Turk, S.A. Weitzman, Free radical DNA adduct 8-OHdeoxyguanosine affects activity of HPA II and MSP I restriction endonucleases, Free Radic. Res. 23 (1995) 255– 258. [41] Y. Kuchino, F. Mori, H. Kasai, H. Inoue, S. Iwai, K. Miura, E. Ohtsuka, S. Nishimura, Misreading of DNA templates containing 8-hydroxydeoxyguanosine at the modified base and at adjacent residues, Nature 327 (1987) 77–79. [42] P.A. Hepburn, G.P. Margison, M.J. Tisdale, Enzymatic methylation of cytosine in DNA is prevented by adjacent O6methylguanine residues, J. Biol. Chem. 266 (1991) 7985– 7987. [43] N.-W. Tan, B.F.L. Li, Interaction of oligonucleotides containing 6-O-methylguanine with human DNA (cytosine5-)-methyltransferase, Biochemistry 29 (1990) 9234–9240.

11

[44] W. Xiao, L. Samson, In vivo evidence for endogenous DNA alkylation damage as a source of spontaneous mutation in eukaryotic cells, Proc. Natl. Acad. Sci. USA 90 (1993) 2117– 2121. [45] J.K. Christman, G. Sheikhnejad, C.J. Marasco, J.R. Sufrin, 5-Methyl-20 -deoxycytidine in single-stranded DNA can act in cis to signal de novo DNA methylation, Proc. Natl. Acad. Sci. USA 92 (1995) 7347–7351. [46] V. Valinluck, H-H. Tsai, D.K. Rogstad, A. Burdzy, A. Bird, L.C. Sowers, Oxidative damage to methyl-CpG sequences inhibits the binding of the methyl-CpG binding domain (MBD) of methyl-CpG binding protein 2 (MeCP2), Nucleic Acids Res. 32 (2004) 4100–4108. [47] E.M. Hurt, S.B. Thomas, B. Peng, W.L. Farrar, Molecular consequences of SOD2 expression in epigenetically silenced pancreatic carcinoma cell lines, Br. J. Cancer 97 (2007) 1116– 1123. [48] M.J. Hitchler, K. Wikainapakul, L. Yu, K. Powers, W. Attatippaholkun, F.E. Domann, Epigenetic regulation of manganese superoxide dismutase expression in human breast cancer cells, Epigenetics 1 (2006) 163–171. [49] S. Majumder, K. Ghosal, Z. Li, Y. Bo, S.T. Jacob, Silencing of metallothionein-I gene in mouse lymphosarcoma cells by methylation, Oncogene 18 (1999) 6287–6295. [50] K. Ghosal, S. Majumder, Z. Li, X. Dong, S.T. Jacob, Suppression of metalothionein gene expression in a rat hepatoma because of promoter-specific DNA methylation, J. Biol. Chem. 275 (2000) 539–547. [51] M. Tada, O. Yokosuka, K. Fukai, T. Chiba, F. Imazeki, T. Tokuhisa, H. Saisho, Hypermethylation of NAD(P)H: quinine oxidoreductase 1 (NQO1) gene in human hepatocellular carcinoma, J. Hepatol. 42 (2005) 511–519. [52] S. Zhong, M.W. Tang, W. Yeo, C. Liu, Y.M.D. Lo, P.J. Johnson, Silencing of GSTP1 gene by CpG island DNA hypermethylation in HBV-associated hepatocellular carcinomas, Clin. Cancer Res. 8 (2002) 1087–1092. [53] M. Esteller, P.G. Corn, J.M. Urena, E. Gabrielso, S.B. Baylin, J.G. Herman, Inactivation of glutathione S-transferase P1 gene by promoter hypermethylation in human neoplasia, Cancer Res. 58 (1998) 4515–4518. [54] D.S. Millar, K.K. Ow, C.L. Paul, P.J. Russell, P.L. Molloy, S.J. Clark, Detailed methylation analysis of the glutathione S-transferase P1 (GSTP1) gene in prostate cancer, Oncogene 18 (1999) 1313–1324. [55] R.E. Board, L. Knight, A. Greystoke, F.H. Blackhall, A. Hughes, C. Dive, M. Ranson, DNA methylation in circulating tumor DNA as a biomarker for cancer, Biomark. Insights. 2 (2007) 307–319.