Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Carcinogenesis vol.26 no.8 pp.1323--1334, 2005 doi:10.1093/carcin/bgi079 Advance Access publication March 31, 2005 COMMENTARY Tumorigenesis: the adaptation of mammalian cells to sustained stress environment by epigenetic alterations and succeeding matched mutations Tatiana V.Karpinets1, and Brent D.Foy2 Introduction 1 The genome of cancer cells usually displays a large number of mutations, but causative mechanisms underlying the origin of these mutations are still debated (1). A recent model of tumorigenesis (2) suggests that cells may generate mutations by themselves as a result of the mutator response (MR), an error-prone cell-cycle progression with participation of errorprone polymerases and aberrant mitosis. According to the model, the MR may be activated in the cellular environment with persistent production of proliferative and stress-related survival signals. This environment, referred to as the sustained stress environment (SSE), gives rise to arrested cells because of DNA damage or replicative senescence. The induction of intracellular proliferative signal in the arrested cells in combination with the blockade of apoptosis by stress-related survival signals can abrogate the cell-cycle arrest and activate error-prone cell-cycle progression thereby generating mutations in the genome of emerged cells. The objective of this paper is to show the role of epigenetic alterations (EA) in inducing the conditions for activation of the MR in SSE and in providing a tagging system that may direct mutations in the MR to specific genes in the genome. The combination of these epigenetic effects ensures the high adaptive potential of mutational modifications in the genome of tumor cells and can explain many features in cancer origin and progression. Epigenetic control of gene activity is widespread in the genome of eukaryotic cells and leads to persistent gene silencing or gene expression. This control is implemented by changes in the methylation status of DNA (the methylation of cytosine residues at 50 carbon) and by chromatin modifications (e.g. histone acetylation, phosphorylation, methylation and ubiquitination). The epigenetic modifications efficiently control the metastable state of gene expression in the genome by inducing stable silencing of some genes and promoting activation of others (3,4). The epigenetic control of gene expression is heritable through cell division, but reversible, because it does not involve DNA sequences. DNA methylation is the best studied EA. It is commonly implicated in the programming of the genome during embryonic development providing its stability by silencing genes of inactive X chromosome, ensuring parent-of-origin specific expression of imprinted genes, regulating tissue specific expression, and silencing of interspersed repetitive retrotransposons and heterochromatic satellite repeat sequences (5). Recent studies reveal the involvement of DNA methylation and histone modifications in the reprogramming of the genome of mammalian cells during many age-related human diseases including cancer. These EA have been mainly elucidated by comparing cellular genomes in healthy and pathological states and therefore, terms hypermethylation and hypomethylation refer to the degree of methylation when compared with the healthy state. Well known changes in methylation related to cancer are summarized in several recent reviews (3,4,6--8). Department of Plant Sciences, University of Tennessee, 2431 Center Drive Knoxville, TN 37996-4500, USA and 2Department of Physics, Wright State University, 3640 Colonel Glenn Hwy, Dayton, OH 45435, USA To whom correspondence should be addressed. Tel: þ1 865 974 0710; Fax: þ1 865 974 5365; Email: [email protected] Recent studies indicate that during tumorigenic transformations, cells may generate mutations by themselves as a result of error-prone cell division with participation of error-prone polymerases and aberrant mitosis. These mechanisms may be activated in cells by continuing proliferative and survival signaling in a sustained stress environment (SSE). The paper hypothesizes that longterm exposure to this signaling epigenetically reprograms the genome of some cells and, in addition, leads to their senescence. The epigenetic reprogramming results in: (i) hypermethylation of tumor-suppressor genes involved in the onset of cell-cycle arrest, apoptosis and DNA repair; (ii) hypomethylation of proto-oncogenes associated with persistent proliferative activity; and (iii) the global demethylation of the genome and activation of DNA repeats. These epigenetic changes in the proliferating cells associate with their replicative senescence and allow the reprogrammed senescent cells to overcome the cell-cycle arrest and to activate error-prone replications. It is hypothesized that the generation of mutations in the error-prone replications of the epigenetically reprogrammed cells is not random. The mutations match epigenetic alterations in the cellular genome, namely gain of function mutations in the case of hypomethylation and loss of functions in the case of hypermethylation. In addition, continuing proliferation of the cells imposed by signaling in SSE speeds up the natural selection of the mutant cells favoring the survival of the cells with mutations that are beneficial in the environment. In this way, a stress-induced replication of the cells epigenetically reprograms their genome for quick adaptation to stressful environments providing an increased rate of mutations, epigenetic tags to beneficial mutations and quick selection process. In combination, these processes drive the origin of the transformed mammalian cells, cancer development and progression. Support from genomic, biochemical and medical studies for the proposed hypothesis, and its implementations are discussed. Abbreviations: EA, epigenetic alternations; EP, epigenetically reprogrammed; MR, mutator response; SSE, sustained stress environment. Carcinogenesis vol.26 no.8 # Oxford University Press 2005; all rights reserved. 1323 T.V.Karpinets and B.D.Foy They include (i) tumor or age related hypermethylation of specific CpG islands and coding regions resulting in silencing of previously active genes, (ii) changes in methylation of differentially methylated regions leading to loss of imprinting and (iii) hypomethylation of previously silent genes and repetitive sequences resulting in chromosomal instability, mitotic recombination, chromosomal loss and aneuploidy. The exact cellular mechanism that provides epigenetic transformations is not established, but many observations indicate a relationship between the level of gene activity in the genome, the methylation state of related DNA and chromatin modifications. Several lines of evidence suggest that gene expression is inversely correlated with DNA methylation (9,10). Mammalian cells tend to methylate promoter sequences of those genes in the genome that cease their participation in cellular activity. It is proposed that chromatin remodeling as a result of histone deacetylation and methylation is the primary event in initiating gene silencing, whereas DNA methylation, a subsequent event, establishes a permanent state of gene inactivation (11). It was also postulated that methyl-binding proteins, a component of the promoter methylation mechanism, have to compete with transcription factor abundance for the establishment of gene repression (12). The existence of this competition may explain why steady cessation of a cellular function and a concomitant lack of transcription factors steadily decreases gene activity related to this function. Conversely, a steady presence of the transcription factors shifts the balance to demethylation of the promoter. Therefore, while hypermethylation of gene promoters relates to transcriptional silencing of unwanted genes, hypomethylation indicates an increase in the level of gene activity (4). We present a hypothesis that considers EA in cells imposed by stress-related proliferative and survival signaling in a SSE as starting events in cancer development. The hypothesis describes potential epigenetic mechanisms involved in priming cells for the generation of beneficial mutations in the genome and quick adaptation to SSE. We provide support of the hypothesis from the published literature and discuss its potential implications. Hypothesis Tumorigenesis is considered a process of the adaptation of mammalian cells to SSE by means of EA of the genome, subsequent mutations matched to EA and natural selection of originated mutant cells via apoptosis (Figure 1). The process of adaptation to SSE involves the emergence of senescent epigenetically reprogrammed (ER) cells with specific tumorrelated EA in the genome. The EA prime the cells for errorprone replications and provide epigenetic tags in the genome for beneficial mutations to arise. Emergence of the ER cells. We hypothesize that tumorrelated EA occur before mutations and reprogram the genome for their origin. The initiation event for these EA is long-term exposure of cell tissue to SSE that induces continuing replication of some cells in the tissue. Persistent proliferative and stress-related survival signals in the environment induce in these proliferating cells a stable activation of biological processes related to promotion of replication and stress response. On the other hand, processes related to inhibition of proliferation, activation of cell-cycle arrest and apoptosis, and DNA repair pathways will be constantly suppressed in these cells. 1324 Fig. 1. Scheme of processes involved in cellular transformation imposed by SSE. Most well-known oncogenes are crucial regulators in these processes. Stable changes in the expression of the genes are secured in the continuously replicating cells by proper EA in the genome, hypermethylation in case of upregulation and hypomethylation in case of downregulation. The EA in the oncogenes are also accompanied by global demethylation of the genome and stress-induced activation of DNA repast. In combination, these EA reprogram the genome for quick adaptation to stressful environments and give rise to ER cells bearing tumor-related EA in their genome. Mutator phenotype of the ER cells. Continuing stressinduced replication of the ER cells leads to their senescence. In normal cells replicative senescence is accomplished by cellcycle arrest, i.e. cells remain viable, but cannot subsequently divide. ER cells cannot activate the cell-cycle arrest, apoptosis or DNA repair, because these processes are silenced epigenetically in the cells. Therefore, they eventually continue the proliferation in an error-prone mode, regardless of replicative senescence or DNA damage in the genome. Tumor-related mutations in the ER cells. We hypothesize that mutations generated in the error-prone replications are not random. They match EA in the genome of the ER cells. Hypermethylation of CpG dinucleotide predisposes the ER cells to point mutation, frameshift mutation, missense mutation and deletions. These dinucleotides are often dominated in gene promoters. Therefore, hypermethylation of a promoter will lead to a loss of gene function. Hypomethylation predisposes cells to chromosomal mutations, rearrangements and aneuploidy. These mutations usually confer a gain of gene Tumorigenesis: adaptation to SSE by EA and succeeding mutations function. In addition, global demethylation imposed by SSE leads to a variety of mutational effects and to a diversification of the genome through the activity of DNA repeats. In this way, EA in the genome of the ER cells (hypermethylation and hypomethylation imposed by SSE) predispose to generation of certain stress-associated adaptive mutations in the error-prone replications. As a result, the ER cells give rise to mutant cells bearing the tumor-related mutations. The natural selection of the mutant cells is accelerated by their continuing proliferation in the stressful environment. It favors the survival of mutant cells harboring mutations that are most suitable for this environment, i.e. confer immortality and high proliferative activity. Multiple rounds of error-prone replication fix these mutations in the genome of progenies and lead to the emergence of transformed cells. In this way, EA in the genome imposed by proliferative and survival signaling in SSE drive the tumor development. The following experimental findings support the model and will be discussed in more detail in the following sections. (i) EA in the genome play a leading role in the origin of cancer and its progression. (ii) EA in cells exposed to SSEs are similar to those indicated at premalignant and early stages of cancer development. (iii) Activity of cell-cycle inhibitors is essential to accomplish stress-induced replicative senescence and their inactivation in the senescent cells can lead to tumorigenic transformations. (iv) Known molecular mechanisms may underlie the causative effects of EA on succeeding mutations in ER cells. Experimental studies that support mechanisms underlying the activation of the MR and the natural selection via apoptosis are presented elsewhere (2). EA in the genome play a leading role in the origin of cancer and its progression According to the proposed hypothesis, the EA induced by continuing replication in the cellular genome (epigenetic reprogramming) are indispensable preliminary events for cancer-related genetic mutation. Two stress-induced epigenetic events are a requisite for cancer development in the hypothesis. They are (i) epigenetic silencing/activation of oncogenes involved in proliferation, cell-cycle control and DNA repair, and (ii) global hypomethylation of the genome. The epigenetic silencing/activation of oncogenes is a necessary introduction to adaptive genetic mutation in the cell, because it primes the genome for the error-prone cell division and provides epigenetic tags to genes that are under selection pressure in the cellular microenvironment. Global hypomethylation is important for subsequent adaptation of the cells to a new environment and for a quick acquisition of new cellular functions, because demethylation of the genome leads to activation of transposable elements and facilitates chromosomal mutations in the cells including a chromosome loss or a gain. In support of a leading role of these mechanisms in tumorigenesis, many studies draw attention to hypermethylation and hypomethylation taking place at precancerous and early stages of cancer. Some of these studies are summarized in Tables I and II. Given a role of EA in the control of a stable state of gene expression, the early onset of hypermethylation and hypomethylation of the genes found in the studies indicates that persistent changes in the related cellular functions occur before tumorigenic transformations. In concordance with the hypothesis, studies show early epigenetic silencing of genes involved in the activation of cell-cycle arrest and apoptosis including cyclin-dependent kinase inhibitors p16, p14, p15, p73, APC gene and DAP-kinase (Table I). The other processes Table I. The role of hypermethylation in the origin of cancer and its progression Epigenetic alterations Tissue/organisms under consideration Ref. p16 An animal model of lung carcinogenesis and human squamous cell carcinomas Nonmalignant bronchial epithelial cells from current and former smokers Early cervical tumor cell populations Barrett’s epithelia (premalignant state of esophageal adenocarcinoma) Human primary tumors and normal tissue of breast, colon and bladder Leukemic cell lines, acute myeloid leukemia and acute lymphoblastic leukemia samples Multiple myeloma cell lines and patients with malignant plasma cell disorders Benign and malignant breast samples obtained from breast cancer patients and samples from unaffected women Summary of the data on ovarian cancer Human gastric cancer cell lines and primary tumors Lymphoid/hematopoietic tumors, potential premalignant specimens, control specimens, EBV-transformed B lymphoblastoid cultures and lymphoma/leukemia or multiple myeloma cell lines Nonneoplastic gastric epithelia Summary of the data on lung cancer Breast tumor, paired preoperative serum DNA, normal healthy women and patients with inflammatory breast disease or nonneoplastic breast tissue specimens Premalignant lesions in the colon Normal tissue adjacent to tumors (lung, prostate and pancreatic carcinomas) and normal tissue from noncancer patients Summary of the data on the role of BRCA1 hypermethylation in breast and ovarian cancer Early-onset sporadic colorectal carcinomas Ovarian primary tumors A summary of the data on a wide spectrum of human tumors Comparison of stages of mouse skin tumor progression including premalignant stages Colorectal cancer progression 13 14 15 16 17 18 19 20 p16 and p14 p15, p16 p16 and p73; SOCS-1; E-cadherin; DAP-kinase APC gene; RARbeta2 gene; RASSF1A FNACF TGF-beta receptor SHP1; SOCS1; SYK p16; E-cadherin; hMLH1 MGMT gene; TIMP-3; DAP-kinase; p16 and p15 APC gene; RASSF1A; DAP-kinase CRBP1, CDH13 (H-cadherin); MINT31 HIN-1 BRCA1 and BRCA2 hMLH1 and hMLH2 hMLH1 MGMT gene promoter MGMT and MLH1; E-cadherin SFRP 21 22 23 24 25 26 27, 28 29 30, 31 32 33 34 35, 36 37 1325 T.V.Karpinets and B.D.Foy Table II. The role of hypomethylation in the origin of cancer and its progression Epigenetic alterations Tissue/organisms under consideration Ref. Sat2 (satellite 2), global hypomethylation Breast adenocarcinomas of known tumor grade and stage, non-neoplastic breast tissues (mostly fibroadenomas), and a variety of normal somatic tissue controls The folate/methyl deficient model of multistage hepatocarcinogenesis (rat liver) Normal gastric mucosa, superficial gastritis and chronic atrophic gastritis specimens Non-small cell lung cancer samples, primary cancer samples, corresponding normal lung tissues and bronchial brush specimens from former smokers without lung cancer Gastric carcinoma specimens, corresponding non-neoplastic mucosae and gastric carcinoma cell lines Summary of the experimental data on alteration of DNA methylation in gastrointestinal carcinogenesis Human livers of different ages, cirrhosis and hepatocellular carcinoma Chronic liver disease without hepatocellular carcinoma, non-tumor liver tissues from patients with carcinoma, carcinoma tissues and control liver tissues Hepatocellular carcinoma fetal and normal human liver tissues Comparison of stages of mouse skin tumor progression including premalignant stages Comparison of colorectal cancer specimens with normal mucosa of non-cancer patients 38 Global hypomethylation Global hypomethylation Melanoma antigen-encoding genes (MAGE-A1, MAGE-A3 and MAGE-B2) Cyclin D2 promoter c-myc, c-Ha-ras, c-fos c-fos c-myc, c-Ki-ras c-myc, EGF receptor Snail The tumor associated changes in promoter CpG island methylation of 17 genes subjected to early hypermethylation in different tissues include the regulation of cell-to-cell adhesion (E-cadherin and H-cadherin, MINT31, SYK, TIMP-3, Snail), suppression of cytokine signaling (SOCS1, TGF-beta, SHP1, HIN-1), DNA repair (BRCA1, BRCA2, MGMT, hMLH1, hMLH2) and regulation of cell differentiation (SFRP, APC, RARbeta2, TGF-beta). Early hypomethylation (Table II) is found for transcription factors involved in activation of proliferation (c-myc, c-fos) and transduction of proliferative and survival signals (EGFR, Ras). Interestingly, when one of the two alleles is mutated in the germ line of a patient with a familial form of cancer, hypermethylation is commonly seen as the second inactivation event (48). Specifically, this epigenetic effect was reported for p16 gene and E-cadherin (49,50). Additional evidence on a leading role for DNA methylation at the preneoplastic stages of cancer development was reviewed by Jones and Laird (51) including the fact that reverse demethylation of methylated genes suppresses the formation of some tumors. In summary, although pathways and genes affected by early EA may be different in different tumor types, in almost all instances they relate to one of the following processes: (i) continuing activation of pathways related to replication; (ii) silencing of pathways and genes with growth-inhibitory activities including the activation of cell-cycle arrest, apoptosis and cellular differentiation; (iii) silencing of DNA repair genes. These cellular processes indicate persistent activation of proliferative and survival signals in the cells even before premalignant stages of cancer and therefore support the plausibility of epigenetic reprogramming of the cells by these signaling as a preliminary event for tumorigenic transformations. Global hypomethylation of the genome was also reported as an early epigenetic event in several cancers (Table II). Though, if hypermethylation and hypomethylation of oncogenes are more often considered in the studies as important events in the origin of tumors, increased demethylation or hypomethylation strongly relates to progression of cancers and increased malignancy. A positive correlation between the degree of global hypomethylation, the grade of malignancy in tumors and chromosomal alterations are found in many cancers (52--55). Hypomethylation is very often observed in DNA repeats (4). The repeats that display tumor-associated hypomethylation include endogenous retrotransposons, 1326 39 40 41 42 43 44 45 46 35 47 LINE-1 repeats, Alu repeats, endogeneous retroviruses and moderately repeated non-viral DNA sequences. It is believed that hypomethylation is necessary for metastasis and invasion in cancer development. Several examples of this role of hypomethylation in breast cancer are given in the review of Szyf et al. (56). These experimental observations indicate a leading role of EA not only in the origin of cancer, but also in its progression. EA in cells exposed to SSEs are similar to those indicated at premalignant and early stages of cancer development According to the model the EA indicated in cells at premalignant stages result from their epigenetic reprogramming in SSE. The reprogramming may be induced by continuing production of cytokines in this environment because they activate proliferative and stress-related survival signals in the cells. Proliferative signaling activates genes and pathways involved in cellular replication. Survival signaling suppresses the activation of cell-cycle arrest and apoptosis. Inactivation of DNA repair pathways may follow persistent suppression of genes involved in the cell-cycle control (like APC gene and p53) that mediate DNA repair pathways (57,58). In long-term, these changes in gene expression lead to tumor-related EA in the genome. In support of the role of SSEs in epigenetic reprogramming, EA in oncogenes (similar to reviewed above), global hypomethylation and activation of DNA repeats are reported in non-malignant mammalian cells, if they are exposed to wellknown carcinogens (gamma and ultraviolet radiation, tobacco smoke, toxicants with potent carcinogenic effects, etc.) or to oncogenic cellular stresses (chronic inflammation, wounding, oxidative stress, genotoxic stress, etc.) (Table III). These stresses usually impose continuing cellular assault and death and are therefore accompanied by persistent production of cytokines. Although studies on EA after exposure to carcinogens are not as extensive as studies in tumor cells, they also indicate (i) hypermethylation of crucial regulators of cell-cycle arrest and apoptosis (p16, p53, APC); (ii) hypomethylation of oncogenes involved in proliferation and stress-related signaling (c-myc and raf ); (iii) global hypomethylation of the genome and activation of DNA repeats. These changes are similar Tumorigenesis: adaptation to SSE by EA and succeeding mutations Table III. Epigenetic alterations induced by SSE in mammalian cells Epigenetic alterations Nature of stress Tissue/organisms Ref. Hypermethylation of tumor suppressor genes (p16, APC) and decrease in the mean methylation index Hypermethylation of tumor suppressor gene p16 Hypermethylation and transcriptional silencing of p16 and p53 Tobacco smoke exposure Lung tissue of long-time smokers 59 Exposure with cadmium (potent carcinogen) Cell culture (BALB/c-3T3 cells) 60 Exposure with dioxin (an ubiquitous environmental contaminant and a highly potent carcinogen) Phenobarbital’s exposure Exposure with carcinogens (arsenite, dichloroacetic acid and trichloroacetic acid) Exposure with cadmium Exposure with nickel Gamma radiation Ultraviolet radiation (UVB 280--340 nm) Genotoxic stress induced by a variety of DNA damaging agents Endogenous oxidative stress Primary human keratinocytes 61 Rodent liver tissue Variety of tissues, including the liver 62 63, 64 Rat liver cells (TRL1215) Chinese hamster embryo cells Four non-malignant cultured cell lines Rat liver Mouse cells 65 66, 67 68 69 70 71 Anoxia (may accompany wound healing) Glucose 6-phosphate dehydrogenase deficient mice Rat fibroblasts 72 Chronic inflammation (chronic ulcerative colitis) Non-neoplastic mucosa 73 Proliferation under stress (cell culturing) Primary mouse embryonic fibroblasts 74 SV40 virus infection Human mesothelial cells 75 Hepatitis B virus and hepatitis C virus infection in liver with chronic inflammation Liver tissue specimens of human hepatocellular carcinoma with or without hepatitis virus infection Epstein--Barr virus-associated and negative gastric carcinoma 76 Hypomethylation including raf-gene Upregulation and hypomethylation of c-myc and c-jun Hypomethylation of the genome Decondensation of the heterochromatin Dose-dependent hypomethylation Dose-dependent hypomethylation Induction of Alu elements (short interspersed elements) Induction of long interspersed nuclear element-1 (LINE-1) Induction of VL30 elements (a multigene family within the class of retroviruses and retrotransposons) Hypermethylation of mismatch repair genes and microsatellite instability Epigenetic silencing of imprinted genes that are usually induced by tumorigenesis Progressive aberrant methylation of protooncogenes (RASSF1A, HPP1, DcR1, TMS1, CRBP1, HIC-1 and RRAD) methylated in malignant mesothelioma Hypermethylation of p16 promoter Specific high-density methylation of cyclin-dependent kinase inhibitors p14 and p16 Epstein--Barr virus-associated gastric carcinoma to changes in the cellular processes at premalignant and early stages of cancers indicated in the previous section. The induction of DNA repeats in SSE, in particular, has strong support in the various studies. In 1984, McClintock (78) first proposed the hypothesis that retrotransposons may be considered as stress-induced generators of genomic diversity. By now, there is substantial evidence of activation of DNA repeats by different kinds of stresses in various organisms including plants, microbes and humans. It is believed that DNA methylation and chromatin structure globally repress the transcription of the repeats, but sustained stress activates them inducing an opening of the chromatin structure and demethylation of the repeats (79). Recent genomic studies have revealed an important role of DNA repeats in the revival of cryptic genes, in diversification of the genome and in adaptation to stressful environments (80). Considering the fact that DNA transposons in human genome hold DNA records from the genomes of evolutionarily preceding species (81), we may speculate that demethylation of DNA repeats in SSE prepare tumor cells for diversification of the genome and for reactivation of cryptic genes involved in functions required for cellular survival in this environment. Thus, we conclude that EA induced by exposure to carcinogens and by oncogenic cellular stresses are similar to those observed at premalignant and early stages of cancer. Similar to cancer cells, these EA indicate continuing proliferation of the cells accompanied by silencing of cell cycle arrest and apoptosis, and activation of DNA repeats. These observations 77 are consistent with a central feature of the hypothesis that SSEs have a leading promotional role in the epigenetic reprogramming of the cells and their subsequent tumorigenic transformations. Activity of cell-cycle inhibitors is essential to accomplish stress-induced replicative senescence and their inactivation in the senescent cells can lead to tumorigenic transformations Further support for the hypothesis is provided by studies of replicative senescence. This process depends on the number of cell divisions and is triggered by short telomeres (the ends of linear chromosomes) that lose ~50--200 bp per round of replication. SSE is accompanied by cell death and renewal, and therefore, inevitably speeds up the cellular senescence. Many studies support the idea in the hypothesis that a stressful environment favors replicative senescence (82). In the recent review of Ben-Porath and Weinberg (83), cellular senescence is considered as a general stress response program of the cell with important biological functions including the onset of arrest on proliferation. Stress-related survival signal induced by Ras/Raf/MEK/ERK pathway caused cellular senescence in several cell types, and the activity of cell-cycle inhibitors p53 and p16 was indispensable to accomplish senescence by cellcycle arrest. Inactivation of these genes prevents cell-cycle arrest and results in tumorigenic transformation (7,84--86). According to the studies in human fibroblasts and mammary epithelial cells, cell-cycle arrest response in senescent cells is 1327 T.V.Karpinets and B.D.Foy maintained primarily by p53. P16 provides a dominant second barrier to the unlimited growth of human cells (86,87). Analysis of the methylation status of cancer-related genes in nonmalignant senescent human cells from different tissues was reported by Yuasa (88). This study demonstrates that though the methylation of the promoters of cancer-related genes may also take place in normal senescence, age-related methylation never includes hypermethylation of p53 and p16 genes, which are crucial for the establishment of cell-cycle arrest in senescent cells. A different output may be observed in senescent mammalian cells subjected to continuing proliferative and survival signaling, because this environment leads to epigenetic silencing of genes involved in the onset of cell-cycle arrest and apoptosis (Tables I--III). This silencing disables the onset of cell-cycle arrest in the senescent cells and leads to malignancy. Given the experimental observations, we can hypothesize that continuing replication of some cells (ER cells in the model) in SSE leads to their senescence in a way that cannot be accomplished by cell-cycle arrest, because, in addition to the replicative senescence, these cells have epigenetically silenced genes involved in cell-cycle control. This silencing in combination with persistent proliferative signals predisposes the ER cells to accomplish their divisions in an error-prone manner, resulting in the generation of mutations and subsequent malignant transformations of the cells. Known molecular mechanisms may underlie the causative effects of EA on succeeding mutations in ER cells Although mechanisms underlying the causative effects of EA on the succeeding mutations are poorly understood, some of them have been demonstrated in both hypermethylation and hypomethylation. A correlation between the methylation status of oncogenes and their mutations in tumor cells is also well documented (53,89,90). The most prominent evidence of the link between EA and specific mutations is provided by studies of CpG dinucleotides. Mutagenic mechanisms involving 5-methylcytosine in CpG dinucleotides are particularly common. Methylated CpG dinucleotides are mutational hot spots in human genes including well-known cancer-relevant genes, such as p53 and retinoblastoma gene (90,91). The best characterized mechanism of G:C!A:T transitions at CpG sites is deamination of 5-methylcytosine at CpG sites, resulting in substitution of 5-methylcytosine by thymine. This occurs spontaneously or is factor-mediated, e.g. by cyclobutane pyrimidine dimers after exposure to UVB and sunlight (92), by some environmental pollutants (93) or through the action of nitric oxide produced under conditions of chronic inflammation (94). The factormediated C to T transition mutations are particularly frequent in some codons of p53 gene at sequences containing 5-methylcytosines. It was also suggested that error-prone DNA polymerases play an important role in the mutagenic bypass of the lesions and generation of the mutations (92). Methylated CpG dinucleotides are mutation hot spots in the retinoblastoma gene (91). Mutational abnormalities indicated in Rb gene include G!A substitution in the promoter region (95). Germline mutations in this gene include frameshift mutations, missense mutation and in-frame deletions (96). Deletions (either homozygous or heterozygous) are a major mechanism of p16 and p15 genes inactivation (97). Loss of function mutations in these genes are well documented in tumors. Frequency of mutations in p16 is second only to mutation of p53. As it was reviewed above (Table I), 1328 hypermethylation of cell-cycle inhibitor p16 takes place at premalignant and early stages of many cancers and also may be imposed by exposure with carcinogens (Table III). The following mechanisms may underlie the induction of matched mutations in genes with methylated CpG dinucleotides. It is generally agreed that C ! T mutations in CpG dinucleotides may be caused by spontaneous hydrolytic deamination. Cytosine deamination produces uracil, which can be removed by glycosylase, whereas deamination of 5-methylcytosine generates a GT mispair that cannot be processed by glycosylase and escapes proofreading, because it is not obvious which strand carries the correct genetic information (98). This is a common phenomenon, and in healthy cells, this type of replication error is repaired by the DNA mismatch repair system that is directed specifically to the newly synthesized strand and in this way increases the fidelity of DNA replication by up to three orders of magnitude (99). According to the model and experimental evidence considered above (Tables I--III), epigenetic silencing of cell-cycle inhibitors imposed by SSE activates an error-prone cellular division without participation of the repair machinery. As a result, mispairs at methylated CpG dinucleotides cannot be repaired and therefore produce point mutations or deletions and frameshift mutations in the case of replication slippage of error-prone DNA polymerase (100). In this way, hypermethylation status of CpG dinucleotides, which often characterizes TS genes in the earliest stage of cancer development, may predispose them to certain mutations, namely point mutation, frameshift mutations, missense mutation and deletions. In addition to increased endogenous mutability of 5-methylcytosine, its interaction with carcinogens may also contribute to certain types of mutations (101). For example, many chemical carcinogens including aromatic hydrocarbons in tobacco smoke preferentially react with guanines flanked by 5-methylcytosine inducing the G ! T transversion mutations. The majority of mutations observed in skin cancer are C ! T transitions or CC ! TT mutations at dipyrimidine sequences resulting from UV-light-induced cyclobutane pyrimidine dimer formation. Methylation of cytosine increases this formation by up to 15-fold owing to higher energy absorption of 5-methylcytosine (102). These results suggest that hypermethylation of tumor suppressor genes, besides contributing to endogenous mutability, can also make these genes vulnerable to environmental mutagens and carcinogens. The causative effect of hypomethylation on succeeding mutations is also well documented in tumors. If hypermethylation of CpG islands predisposes affected genes to point mutations, deletions and frameshift mutations, leading to loss of gene function, hypomethylation predisposes to chromosomal instability and to reshuffling of the mammalian genome by the activation of DNA repeats. Hypomethylated repetitive sequences are mainly located in the vicinity of the centromeres of chromosomes, comprised of heterochromatin (4). It predisposes cells to increased chromosomal rearrangements in this region, as takes place in many types of cancer. This type of mutation in tumors often provides a gain of gene function. Chromosomal instability has been linked to DNA hypomethylation in the heterochromatin regions of chromosomes in several papers (52,53,103). For example, correlation of DNA methylation status in heterochromatin regions with tumor histology and with chromosome alterations was found in a study of human breast carcinomas. Hypomethylation of these regions was associated with the accumulation of a large Tumorigenesis: adaptation to SSE by EA and succeeding mutations number of chromosome alterations and promoted the development of breast carcinomas with aggressive histological features (53). Likewise, genomewide DNA hypomethylation is associated with alterations on chromosome 8 in prostate carcinoma (52). A high percentage of Wilms tumors, breast adenocarcinomas and ovarian epithelial carcinomas have cancer-associated hypomethylation of the main satellite DNAs in the centromere-adjacent heterochromatin of chromosomes 1 and 16 (satellites Chr1 and Chr16), and it is associated with the recurrent rearrangements of the satellites in the vicinity of the centromere. These rearrangements usually lead to 1q gains or 16q losses (104). The following mechanisms may underlie the effect of hypomethylation on chromosome instability. It was shown that in lymphoblastoid cell lines hypomethylation in the heterochromatin regions of chromosomes induced DNA uncoiling and rearrangements at these regions (103). In a study of human hepatocellular carcinoma samples, hypomethylation altered the interaction between the CpG-rich satellite DNA and chromatin proteins, resulting in heterochromatin decondensation, breakage and aberrant 1q formation (55). We may hypothesize that heterochromatin decondensation and DNA uncoiling in the centromeric regions of chromosomes lead to abnormalities in normal chromosome segregation during error-prone cell-cycle progression in the MR, resulting in karyotypic abnormalities during mitosis. In this way, the alteration in chromatin induced by hypomethylation may underlie chromosomal mutations in the ER cells. Mobility of DNA repeats including retrotransposons and endogenous retroviruses imposed by their hypomethylation may also induce a variety of mutational effects. All retroelements are able to form an RNA transcript that may be reversetranscribed and inserted into a new location in the genome. Different classes of retroelements may have their mechanisms for RNA formation, reverse transcription and integration (105), but in all instances they contribute to insertional mutagenesis that may modify promoter or coding gene sequences. Retrotransposition of long terminal repeats, like LINE1, may remodel the genome through the transduction of 30 flanking sequences. Such events have the potential to shuffle, e.g. exons and promoters, creating new genes or altering the function of old genes (106). LINE1 is also able to generate processed pseudogenes and chimeric retrotranscripts (107). Alu elements generate mutations by the insertion into mature mRNAs via a splicing-mediated process termed exonization (108). Repetitive human Alu elements may cause deleterious mutations both by retrotranspositional insertion and by unequal, homologous recombination (109). The variety of ways in which activated DNA repeats appear to affect genomic diversity in tumor cells lends additional support to the hypothesis, i.e. hypomethylation of DNA repeats in SSE prepare the genome for mutational modifications that are important for subsequent adaptation of the disseminated tumor cells to new environments. Changes in methylation status of genes occur throughout the genome and involve many different genes and different codons within the gene so that the epigenetic tagging system cannot guarantee the generation of only beneficial mutations in the error-prone replication. Rather, this system facilitates the emergence of beneficial mutations in accord with the EA in the genome. Subsequent natural selection of the mutants by apoptosis adjusts the mutational spectra to the environment. Therefore, a continuing proliferation of the cells, a requisite of SSE, is very important not only for the epigenetic reprogramming of the cells but also for a quick shaping of the mutational spectra in the genome to the needs of the cellular environment. In support of this mechanism, several studies present evidence that the cytotoxicity of some carcinogens (like UVB light or tobacco smoke), rather than their mutagenicity, may drive tumorigenesis (110,111). Discussion Support of the proposed hypothesis from stationary state mutagenesis in bacteria and cell culture models of cancer The mechanism of cellular adaptation proposed in the hypothesis challenges the foundations of neo-Darwinism, namely that the environment cannot affect the rate of mutations and cannot direct mutations to certain genes. By now, a long line of evidence indicates that in unicellular organisms, such as microbes, the stressful environment can affect the rate of mutations and direct mutations to beneficial genes in this environment. Studies of stationary state mutagenesis in bacterial cells provide the most prominent and well-documented examples of the effect (112--117). Microorganisms increase the rate of mutations during the periods of prolonged stress. Stress limits the growth of microorganisms and induces them to enter a stationary state and produce adaptive mutations that confer a growth advantage in the growth-limiting environment. This strategy hastens the evolution of microbial pathogens including antigenic variation (changes in surface antigens to avoid detection by the host immune system) and antibiotic resistance (118). The mechanisms that allow cells to increase their mutation rates under conditions of sustained stress and produce the mutations in beneficial genes are especially important in evolutionary terms. Therefore, it is plausible that eukaryotic organisms retained these mechanisms in their cells and made them even more sophisticated. According to our hypothesis such an adaptive strategy is employed in the somatic cells of mammals in response to sustained stress and plays an important role in the development of malignances. In this case, a continuing proliferative and survival signaling in the tissue microenvironment indicates a poor adjustment of the cellular genome for the environment and hastens cellular adaptation. These signals are responsible for the epigenetic reprogramming of some cells leading to an increased rate of tumorrelated mutations in the genome. Numerous studies support the fundamental role of the cellular environment in the transformations. This role is especially evident when we consider in vitro culture of multicellular organisms. Rodent cells can easily undergo ‘spontaneous’ neoplastic transformation in culture and are widely used as in vitro models of cancer (119,120). A well-known example is the rodent fibroblast. In a primary culture of mouse embryo cells, a small number of cells, survived after apoptotic crises by forming immortalized cell lines, can overgrow the culture. These cells invariably display abnormal karyotypes and may undergo ‘spontaneous’ transformations (120). This output is consistent with the proposed hypothesis. Culturing of the cells mimics SSE in a tissue, because of cellular assault and continuing proliferation of some cells. This signaling epigenetically reprograms the cells and primes them for malignant modifications in the genome. This transformation represents an adaptive response, which is similar to a bacterial one, to constraints on growth. Although human cells rarely undergo spontaneous 1329 T.V.Karpinets and B.D.Foy transformation in culture, they also may be transformed by a stepwise process. For example, the transformation may be accomplished by immortalization of the cells by DNA viruses, followed by conversion of the immortalized cells to tumorigenic ones by Ras oncogenes or by using chemicals or chemical agents (121). DNA virus in this transformation provides sustained activation of c-myc that reproduces a persistent proliferative signal in the SSE. Ras-activation reproduces a stress-related survival signal that is necessary to silence genes involved in the cell-cycle arrest and to activate a global demethylation of the genome. Delayed effects of environmental factors on cancer development may be explained in terms of the proposed hypothesis An important inference of the proposed mechanism is that the nature of the mutations in cancer cells is not quite stochastic. The mutations are not just caused by DNA damaging agents in the cellular genome; the cell itself creates conditions to the generation of proper mutations by epigenetic modification of the genome in SSE. From this standpoint, mutations observed in tumors are not the reason for the disease, but a far-reaching consequence of cellular exposure to SSE. The mutations accomplish epigenetic changes in genes imposed by this environment. This mechanism may explain the delayed effects of toxicant exposure and the bystander effect of radiation on tumor developments, which are inconsistent with the accepted mechanism of direct DNA damage. It was shown, for example, that perinatal exposure of diethylstilbesterol (a drug with estrogenic activity) changed the methylation pattern of the promoters of several estrogen-responsive genes associated with the development of reproductive organs and resulted in a high incidence of epithelial cancers of the uterus in mice at 18--24 months of age (122). Long-term effects were also revealed after the short-term exposure to arsenic. In this study, progeny of the exposed Chinese hamster cells showed an increased level of chromosomal rearrangements that were not seen in the parents after the acute exposure. Moreover, the mutations were related to genome DNA hypomethylation induced by arsenic after treatment (123). In a study of ionizing radiation, non-irradiated cells acquired mutagenesis through direct contact with cells whose nuclei had previously been irradiated with alpha-particles (124). It was also shown that the delayed effects of the ionizing radiation (delayed reproductive death or lethal mutation, chromosomal instability and mutagenesis) may be transmitted over many generations after irradiation through the progeny of surviving cells (125). Molecular mechanisms underlying these experimental findings are not known, but it is believed that it may be a consequence of bystander interactions involving intercellular signaling and production of cytokines (126), i.e. factors in the hypothesis that induce sustained proliferative and stress-related survival signals and epigenetically reprogram cells. In terms of the presented hypothesis, the exposure with carcinogens lead to accumulation of the reprogrammed cells with tumor-related EA. But their transformation may be triggered later by cellular senescence or another SSE. The same epigenetic mechanism may underlie the causative effect of parental exposure with carcinogens (pesticides, smoking and so on) on the development of childhood cancers (127). Etiology of human cancers in terms of the proposed hypothesis The proposed model of tumorigenesis is based on the analysis of principal common characteristics of different cancers. 1330 According to the model, these common characteristics arise as a result of sustained proliferative and survival signaling in a SSE. This signaling may be imposed by continuing production of cytokines in a tissue subjected to continuing cellular assault, for example as a result of the exposure with carcinogens. A similar stress-associated signaling may be induced by some other environmental and physiological factors including such well known inducers of cancer as inflammation, hormones and viral infection. The effect of these factors on cancer development also may be considered in terms of the proposed model. The implication of hormones in the origin and development of many cancers is well documented. It is also known that many hormones activate both proliferative and survival signaling in cells inducing their proliferation on the one hand and inhibiting their apoptosis on the other (128). This signaling recapitulates the SSE characterized by continuing cellular assault. Although different pathways and genes may be involved in the signal transduction, both environments will impose similar cellular responses that include continuing proliferation of the exposed cells and silencing of cell-cycle arrest and apoptosis. In the long term, it will be accomplished by epigenetic reprogramming of some cells and their subsequent malignant transformation. Recent studies support the plausibility of this mechanism in the origin of breast and prostate cancers. The implication of estrogens in the development of breast cancer is well established. It is also known that this hormone activates the estrogen receptor and stimulates cellular proliferation by activation of AP-1 transcription factors (129) and cellular survival by activation of MAP kinase signaling (130). The other example is the development of prostate cancer. This cancer initially requires androgen for the growth, and responds to the hormone deprivation therapy. This responsiveness indicates that cellular apoptotic machinery is functional at this stage of cancer progression. It was found that activation of Ras-mediated stress-related survival signaling is sufficient, and may be necessary for progression of the disease; it may also induce the development of its androgen independent form (131). These examples indicate the involvement of proliferative and survival signals, which are crucial components of SSE, in breast and prostate cancers development. The effect of transforming viruses on cellular processes also mimics many aspects of SSE. The expression of oncoproteins encoded by some viruses changes cellular processes in the host in the same way as continuing proliferative and survival signaling. Pathways involved in activation proliferation, cellcycle arrest and DNA repair are common targets of the viral infection. A detailed summary of these effects for three wellstudied oncogenic viruses (human T-cell leukemia virus type 1, hepatitis B virus and human papilloma virus) are given in the review of Gatza et al. (132). Although viral oncoproteins do not necessarily affect the same pathways, the overall mechanisms are highly analogous among the viruses. They activate/ inactivate critical cellular processes to promote proliferation, suppress cell-cycle control and DNA repair. Moreover, they directly facilitate the activation of error-prone replication stimulating bypass of G1/S and G2/M checkpoints in the cellcycle progression. Approximately one-quarter of all malignancies are thought to arise in part to chronic inflammation (133). Similar to SSE imposed by carcinogens, this cellular environment leads to production of reactive oxygen species as an important element of cellular defense system. Therefore, inflammation, Tumorigenesis: adaptation to SSE by EA and succeeding mutations regardless of etiology, always causes cellular assault, DNA damage, cell death and renewal. It creates a tissue microenvironment rich in cytokines and growth factors that induce proliferative and stress-related survival signaling. Recent studies on the role of inflammation in cancer development are considered by Brower (134). It is believed that nuclear factor— kappa B (a key transcription factor in infection and inflammation) can provide a link between chronic inflammation and cancer. It was shown that in the premalignant inflammatory environment, this transcription factor induces epithelial cell proliferation by upregulating macrophage-produced cytokines and also provides a survival signal suppressing cellular apoptosis. In summary, the effect of well-known inducers of tumorigenesis associated with the etiology of many different cancers may be described in terms of the proposed hypothesis. Implications of the hypothesis The proposed hypothesis may be helpful in the explanation of some common characteristics of tumor cells associated with their high adaptive potential. According to the hypothesis, the increased rate of mutations and their beneficial nature are crucial components of this potential acquired after epigenetic reprogramming of the cells in SSE. They allow the cells to easily overcome subsequent stresses imposed by tumor progression or medication. Metastasis and drug resistance are examples of implementations of this adaptive potential. Epigenetically reprogrammed and then transformed cells faced with a shortage of oxygen or nutrients, induced by overcrowding, can easily express those genes in the genome that increase their mobility and allow them to migrate and proliferate in new environments. In this way, the tumor cells acquire the sequential specialized adaptations that confer metastatic capabilities including the expression of metalloproteinases, angiogenesis, lymphangiogenesis and adhesion molecules (135). Ability to metastasize overcomes the environmental stress imposed by overcrowding. Similar survival strategies, which are based on increased mobility, are implemented by some unicellular organisms, like yeasts and slime mould colonies deprived of nutrients (136). A quick development of drug resistance, which is typical for cancer cells, is the other illustration of their high adaptability. A wellknown example is a variety of mutational effects employed in the advanced prostate cancer in response to androgen ablation therapy (137). The proposed insight into tumorigenesis may improve and extend our approaches to diagnostics, prevention and treatment of cancer. Screening of EA in genes involved in cell-cycle control, DNA repair and repetitive sequences in heterochromatin may provide a generalized picture of the exposure of cells to sustained stress and predisposition to cancer development. It opens the prospect of early diagnosis of cancer and its prevention, because EA are reversible and correction of cellular environment at this stage by proper treatment or by changing a lifestyle may be the most efficient way to prevent cancer development. Studies of the effect of environmental factors on EA and cancer development give indirect medical support for this conclusion (138). It was shown that dietary deficiencies, chemicals and genetics may alter methylation processes and cause tumor formation. Even environmental exposures in early life (e.g. estrogenic, xenobiotic, nutritional factors, stress) may alter DNA methylation and in this way modulate the risk of cancer and other chronic diseases in adulthood (139). Although many experimental observations reviewed in the paper support the proposed model of tumorigenesis, direct evidence of the adaptation of mammalian cells to SSE by EA and succeeding matched mutations postulated in the model as the main driving force of tumorigenesis is actually absent. Additional studies are necessary to validate and refine the model. Acknowledgements We thank Dr Greenwood (Warwick University, UK), Drs Riggs, Roberts, Rodin, Chen and other colleagues in the City of Hope for the comments on the paper. This study was supported by a grant from the Air Force Office of Scientific Research. Conflict of Interest Statement: None declared. References 1. Sieber,O., Heinimann,K. and Tomlinson,I. (2005) Genomic stability and tumorigenesis. Semin. Cancer Biol., 15, 61--66. 2. Karpinets,T.V. and Foy,B.D. (2004) Model of the developing tumorigenic phenotype in mammalian cells and the role of sustained stress. J. Theor. Biol., 227, 253--264. 3. Feinberg,A.P., Cui,H. and Ohlsson,R. (2002) DNA methylation and genomic imprinting: insights from cancer into epigenetic mechanisms. Semin. Cancer Biol., 12, 389--398. 4. Dunn,B.K. (2003) Hypomethylation: one side of a larger picture. Ann. N. Y. Acad. Sci., 983, 28--42. 5. Plass,C. and Soloway,P.D. (2002) DNA methylation, imprinting and cancer. Eur. J. Hum. Genet., 10, 6--16. 6. Fruhwald,M.C. and Plass,C. (2002) Global and gene-specific methylation patterns in cancer: aspects of tumor biology and clinical potential. Mol. Genet. Metab., 75, 1--16. 7. Neumeister,P., Albanese,C., Balent,B., Greally,J. and Pestell,R.G. (2002) Senescence and epigenetic dysregulation in cancer. Int. J. Biochem. Cell. Biol., 34, 1475--1490. 8. Tycko,B. (2003) Genetic and epigenetic mosaicism in cancer precursor tissues. Ann. N. Y. Acad. Sci., 983, 43--54. 9. Bird,A. (2002) DNA methylation patterns and epigenetic memory. Genes Dev., 16, 6--21. 10. Clark,S.J. and Melki,J. (2002) DNA methylation and gene silencing in cancer: which is the guilty party? Oncogene, 21, 5380--5387. 11. Nephew,K.P. and Huang,T.H. (2003) Epigenetic gene silencing in cancer initiation and progression. Cancer Lett., 190, 125--133. 12. Curradi,M., Izzo,A., Badaracco,G. and Landsberger,N. (2002) Molecular mechanisms of gene silencing mediated by DNA methylation. Mol. Cell. Biol., 22, 3157--3173. 13. Belinsky,S.A., Nikula,K.J., Palmisano,W.A., Michels,R., Saccomanno,G., Gabrielson,E., Baylin,S.B. and Herman,J.G. (1998) Aberrant methylation of p16(INK4a) is an early event in lung cancer and a potential biomarker for early diagnosis. Proc. Natl Acad. Sci. USA, 95, 11891--11896. 14. Belinsky,S.A., Palmisano,W.A., Gilliland,F.D. et al. (2002) Aberrant promoter methylation in bronchial epithelium and sputum from current and former smokers. Cancer Res., 62, 2370--2377. 15. Nuovo,G.J., Plaia,T.W., Belinsky,S.A., Baylin,S.B. and Herman,J.G. (1999) In situ detection of the hypermethylation-induced inactivation of the p16 gene as an early event in oncogenesis. Proc. Natl Acad. Sci. USA, 96, 12754--12759. 16. Hardie,L.J., Darnton,S.J., Wallis,Y.L., Chauhan,A., Hainaut,P., Wild,C.P. and Casson,A.G. (2005) p16 expression in Barrett’s esophagus and esophageal adenocarcinoma: association with genetic and epigenetic alterations. Cancer Lett., 217, 221--230. 17. Dominguez,G., Silva,J., Garcia,J.M. et al. (2003) Prevalence of aberrant methylation of p14ARF over p16INK4a in some human primary tumors. Mutat. Res., 530, 9--17. 18. Chim,C.S., Wong,A.S. and Kwong,Y.L. (2003) Epigenetic inactivation of INK4/CDK/RB cell cycle pathway in acute leukemias. Ann. Hematol., 82, 738--742. 19. Galm,O., Wilop,S., Reichelt,J., Jost,E., Gehbauer,G., Herman,J.G. and Osieka,R. (2004) DNA methylation changes in multiple myeloma. Leukemia, 18, 1687--1692. 1331 T.V.Karpinets and B.D.Foy 20. Lewis,C.M., Cler,L.R., Bu,D.W., Zochbauer-Muller,S., Milchgrub,S., Naftalis,E.Z., Leitch,A.M., Minna,J.D. and Euhus,D.M. (2005) Promoter hypermethylation in benign breast epithelium in relation to predicted breast cancer risk. Clin. Cancer Res., 11, 166--172. 21. Olopade,O.I. and Wei,M. (2003) FANCF methylation contributes to chemoselectivity in ovarian cancer. Cancer Cell, 3, 417--420. 22. Kang,S.H., Bang,Y.J., Im,Y.H., Yang,H.K., Lee,D.A., Lee,H.Y., Lee,H.S., Kim,N.K. and Kim,S.J. (1999) Transcriptional repression of the transforming growth factor-beta type I receptor gene by DNA methylation results in the development of TGF-beta resistance in human gastric cancer. Oncogene, 18, 7280--7286. 23. Reddy,J., Shivapurkar,N., Takahashi,T. et al. (2005) Differential methylation of genes that regulate cytokine signaling in lymphoid and hematopoietic tumors. Oncogene, 24, 732--736. 24. Waki,T., Tamura,G., Tsuchiya,T., Sato,K., Nishizuka,S. and Motoyama,T. (2002) Promoter methylation status of E-cadherin, hMLH1 and p16 genes in nonneoplastic gastric epithelia. Am. J. Pathol., 161, 399--403. 25. Ferretti,G., Brichon,P.Y. and Moro-Sibilot,D. (2003) Early detection of lung cancer: role of biomarkers. Eur. Respir. J. Suppl., 39, 36--44. 26. Dulaimi,E., Hillinck,J., Ibanez de Caceres,I., Al-Saleem,T. and Cairns,P. (2004) Tumor suppressor gene promoter hypermethylation in serum of breast cancer patients. Clin. Cancer Res., 10, 6189--6193. 27. Luo,L., Chen,W.D. and Pretlow,T.P. (2005) CpG island methylation in aberrant crypt foci and cancers from the same patients. Int. J. Cancer, 115, 747--751. 28. Wang,P., Wang,X. and Pei,D. (2004) Mint-3 regulates the retrieval of the internalized membrane-type matrix metalloproteinase, MT5-MMP, to the plasma membrane by binding to its carboxyl end motif EWV. J. Biol. Chem., 279, 20461--20470. 29. Krop,I., Player,A., Tablante,A. et al. (2004) Frequent HIN-1 promoter methylation and lack of expression in multiple human tumor types. Mol. Cancer Res., 2, 489--494. 30. Catteau,A. and Morris,J.R. (2002) BRCA1 methylation: a significant role in tumour development? Semin. Cancer Biol., 12, 359--371. 31. Yoshida,K. and Miki,Y. (2004) Role of BRCA1 and BRCA2 as regulators of DNA repair, transcription and cell cycle in response to DNA damage. Cancer Sci., 95, 866--871. 32. Kim,H.C., Kim,C.N., Yu,C.S., Roh,S.A. and Kim,J.C. (2003) Methylation of the hMLH1 and hMSH2 promoter in early-onset sporadic colorectal carcinomas with microsatellite instability. Int. J. Colorectal Dis., 18, 196--202. 33. Geisler,J.P., Goodheart,M.J., Sood,A.K., Holmes,R.J., HattermanZogg,M.A. and Buller,R.E. (2003) Mismatch repair gene expression defects contribute to microsatellite instability in ovarian carcinoma. Cancer, 98, 2199--2206. 34. Esteller,M. and Herman,J.G. (2004) Generating mutations but providing chemosensitivity: the role of O6-methylguanine DNA methyltransferase in human cancer. Oncogene, 23, 1--8. 35. Fraga,M.F., Herranz,M., Espada,J. et al. (2004) A mouse skin multistage carcinogenesis model reflects the aberrant DNA methylation patterns of human tumors. Cancer Res., 64, 5527--5534. 36. De Craene,B., van Roy,F. and Berx,G. (2005) Unraveling signalling cascades for the Snail family of transcription factors. Cell Signal., 17, 535--547. 37. Suzuki,H., Watkins,D.N., Jair,K.W. et al. (2004) Epigenetic inactivation of SFRP genes allows constitutive WNT signaling in colorectal cancer. Nat. Genet., 36, 417--422. 38. Jackson,K., Yu,M.C., Arakawa,K., Fiala,E., Youn,B., Fiegl,H., MullerHolzner,E., Widschwendter,M. and Ehrlich,M. (2004) DNA hypomethylation is prevalent even in low-grade breast cancers. Cancer Biol. Ther., 3 Epub ahead of print. 39. Pogribny,I.P., James,S.J., Jernigan,S. and Pogribna,M. (2004) Genomic hypomethylation is specific for preneoplastic liver in folate/methyl deficient rats and does not occur in non-target tissues. Mutat. Res., 548, 53--59. 40. Cravo,M., Pinto,R., Fidalgo,P., Chaves,P., Gloria,L., Nobre-Leitao,C. and Costa Mira,F. (1996) Global DNA hypomethylation occurs in the early stages of intestinal type gastric carcinoma. Gut, 39, 434--438. 41. Jang,S.J., Soria,J.C., Wang,L., Hassan,K.A., Morice,R.C., Walsh,G.L., Hong,W.K. and Mao,L. (2001) Activation of melanoma antigen tumor antigens occurs early in lung carcinogenesis. Cancer Res., 61, 7959--7963. 42. Oshimo,Y., Nakayama,H., Ito,R., Kitadai,Y., Yoshida,K., Chayama,K. and Yasui,W. (2003) Promoter methylation of cyclin D2 gene in gastric carcinoma. Int. J. Oncol., 23, 1663--1670. 1332 43. Fang,J.Y. and Xiao,S.D. (2001) Alteration of DNA methylation in gastrointestinal carcinogenesis. J. Gastroenterol. Hepatol., 16, 960--968. 44. Choi,E.K., Uyeno,S., Nishida,N. et al. (1996) Alterations of c-fos gene methylation in the processes of aging and tumorigenesis in human liver. Mutat. Res., 354, 123--128. 45. Aiba,N., Nambu,S., Inoue,K. and Sasaki,H. (1989) Hypomethylation of the c-myc oncogene in liver cirrhosis and chronic hepatitis. Gastroenterol. Jpn., 24, 270--276. 46. Kaneko,Y., Shibuya,M., Nakayama,T., Hayashida,N., Toda,G., Endo,Y., Oka,H. and Oda,T. (1985) Hypomethylation of c-myc and epidermal growth factor receptor genes in human hepatocellular carcinoma and fetal liver. Jpn. J. Cancer Res., 76, 1136--1140. 47. Xu,X.L., Yu,J., Zhang,H.Y., Sun,M.H., Gu,J., Du,X., Shi,D.R., Wang,P., Yang,Z.H. and Zhu,J.D. (2004) Methylation profile of the promoter CpG islands of 31 genes that may contribute to colorectal carcinogenesis. World J. Gastroenterol., 10, 3441--3454. 48. Jones,P.A. and Baylin,S.B. (2002) The fundamental role of epigenetic events in cancer. Nat. Rev. Genet., 3, 415--428. 49. Myohanen,S.K., Baylin,S.B. and Herman,J.G. (1998) Hypermethylation can selectively silence individual p16ink4A alleles in neoplasia. Cancer Res., 58, 591--593. 50. Grady,W.M., Wills,J., Guilford,P.J. et al. (2000) Methylation of the CDH1 promoter as the second genetic hit in hereditary diffuse gastric cancer. Nat. Genet., 26, 16--17. 51. Jones,P.A. and Laird,P.W. (1999) Cancer epigenetics comes of age. Nat. Genet., 21, 163--167. 52. Schulz,W.A., Elo,J.P., Florl,A.R., Pennanen,S., Santourlidis,S., Engers,R., Buchardt,M., Seifert,H.H. and Visakorpi,T. (2002) Genomewide DNA hypomethylation is associated with alterations on chromosome 8 in prostate carcinoma. Genes Chromosomes Cancer, 35, 58--65. 53. Tsuda,H., Takarabe,T., Kanai,Y., Fukutomi,T. and Hirohashi,S. (2002) Correlation of DNA hypomethylation at pericentromeric heterochromatin regions of chromosomes 16 and 1 with histological features and chromosomal abnormalities of human breast carcinomas. Am. J. Pathol., 161, 859--866. 54. Gaudet,F., Hodgson,J.G., Eden,A., Jackson-Grusby,L., Dausman,J., Gray,J.W., Leonhardt,H. and Jaenisch,R. (2003) Induction of tumors in mice by genomic hypomethylation. Science, 300, 489--492. 55. Wong,N., Lam,W.C., Lai,P.B., Pang,E., Lau,W.Y. and Johnson,P.J. (2001) Hypomethylation of chromosome 1 heterochromatin DNA correlates with q-arm copy gain in human hepatocellular carcinoma. Am. J. Pathol., 159, 465--471. 56. Szyf,M., Pakneshan,P. and Rabbani,S.A. (2004) DNA methylation and breast cancer. Biochem. Pharmacol., 68, 1187--1197. 57. Narayan,G., Arias-Pulido,H., Nandula,S.V. et al. (2004) Promoter hypermethylation of FANCF: disruption of Fanconi Anemia-BRCA pathway in cervical cancer. Cancer Res., 64, 2994--2997. 58. Sengupta,S. and Harris,C.C. (2005) p53: traffic cop at the crossroads of DNA repair and recombination. Nat. Rev. Mol. Cell Biol., 6, 44--55. 59. Toyooka,S., Maruyama,R., Toyooka,K.O. et al. (2003) Smoke exposure, histologic type and geography-related differences in the methylation profiles of non-small cell lung cancer. Int. J. Cancer, 103, 153--160. 60. Lei,Y.X., Joseph,P., Wu,Z.L. and Ong,T.M. (2003) Effect of aberrant DNA methylation on the expression of cancer-related genes in Cd-transformed cells. Zhonghua Lao Dong Wei Sheng Zhi Ye Bing Za Zhi, 21, 114--116. 61. Ray,S.S. and Swanson,H.I. (2004) Dioxin-induced immortalization of normal human keratinocytes and silencing of p53 and p16INK4a. J. Biol. Chem., 279, 27187--27193. 62. Counts,J.L. and Goodman,J.I. (1995) Alterations in DNA methylation may play a variety of roles in carcinogenesis. Cell, 83, 13--15. 63. Chen,H., Liu,J., Zhao,C.Q., Diwan,B.A., Merrick,B.A. and Waalkes,M.P. (2001) Association of c-myc overexpression and hyperproliferation with arsenite-induced malignant transformation. Toxicol. Appl. Pharmacol., 175, 260--268. 64. Tao,L., Yang,S., Xie,M., Kramer,P.M. and Pereira,M.A. (2000) Hypomethylation and overexpression of c-jun and c-myc protooncogenes and increased DNA methyltransferase activity in dichloroacetic and trichloroacetic acid-promoted mouse liver tumors. Cancer Lett., 158, 185--193. 65. Takiguchi,M., Achanzar,W.E., Qu,W., Li,G. and Waalkes,M.P. (2003) Effects of cadmium on DNA-(cytosine-5) methyltransferase activity and DNA methylation status during cadmium-induced cellular transformation. Exp. Cell Res., 286, 355--365. Tumorigenesis: adaptation to SSE by EA and succeeding mutations 66. Sutherland,J.E. and Costa,M. (2003) Epigenetics and the environment. Ann. N. Y. Acad. Sci., 983, 151--160. 67. Conway,K. and Costa,M. (1989) Nonrandom chromosomal alterations in nickel-transformed Chinese hamster embryo cells. Cancer Res., 49, 6032--6038. 68. Kalinich,J.F., Catravas,G.N. and Snyder,S.L. (1989) The effect of gamma radiation on DNA methylation. Radiat. Res., 117, 185--197. 69. Becker,F.F., Holton,P., Ruchirawat,M. and Lapeyre,J.N. (1985) Perturbation of maintenance and de novo DNA methylation in vitro by UVB (280--340 nm)-induced pyrimidine photodimers. Proc. Natl Acad. Sci. USA, 82, 6055--6059. 70. Hagan,C.R., Sheffield,R.F. and Rudin,C.M. (2003) Corrigendum: human Alu element retrotransposition induced by genotoxic stress. Nat. Genet., 35, 377. 71. Rockwood,L.D., Felix,K. and Janz,S. (2004) Elevated presence of retrotransposons at sites of DNA double strand break repair in mouse models of metabolic oxidative stress and MYC-induced lymphoma. Mutat. Res., 548, 117--125. 72. Anderson,G.R., Stoler,D.L. and Scarcello,L.A. (1989) Retrotransposonlike VL30 elements are efficiently induced in anoxic rat fibroblasts. J. Mol. Biol., 205, 765--769. 73. Brentnall,T.A., Crispin,D.A., Bronner,M.P., Cherian,S.P., Hueffed,M., Rabinovitch,P.S., Rubin,C.E., Haggitt,R.C. and Boland,C.R. (1996) Microsatellite instability in nonneoplastic mucosa from patients with chronic ulcerative colitis. Cancer Res., 56, 1237--1240. 74. Pantoja,C., de Los Rios,L., Matheu,A., Antequera,F. and Serrano,M. (2005) Inactivation of imprinted genes induced by cellular stress and tumorigenesis. Cancer Res., 65, 26--33. 75. Suzuki,M., Toyooka,S., Shivapurkar,N. et al. (2005) Aberrant methylation profile of human malignant mesotheliomas and its relationship to SV40 infection. Oncogene, 24, 1302--1308. 76. Narimatsu,T., Tamori,A., Koh,N., Kubo,S., Hirohashi,K., Yano,Y., Arakawa,T., Otani,S. and Nishiguchi,S. (2004) p16 promoter hypermethylation in human hepatocellular carcinoma with or without hepatitis virus infection. Intervirology, 47, 26--31. 77. Sakuma,K., Chong,J.M., Sudo,M., Ushiku,T., Inoue,Y., Shibahara,J., Uozaki,H., Nagai,H. and Fukayama,M. (2004) High-density methylation of p14ARF and p16INK4A in Epstein-Barr virus-associated gastric carcinoma. Int. J. Cancer, 112, 273--278. 78. McClintock,B. (1984) The significance of the responses of the genome to challenge. Science, 226, 792--801. 79. Kim,C., Rubin,C.M. and Schmid,C.W. (2001) Genome-wide chromatin remodeling modulates the Alu heat shock response. Gene, 276, 127--133. 80. van de Lagemaat,L.N., Landry,J.R., Mager,D.L. and Medstrand,P. (2003) Transposable elements in mammals promote regulatory variation and diversification of genes with specialized functions. Trends Genet., 19, 530--536. 81. Lander,E.S., Linton,L.M., Birren,B. et al. (2001) Initial sequencing and analysis of the human genome. Nature, 409, 860--921. 82. Lloyd,A.C. (2002) Limits to lifespan. Nat. Cell Biol., 4, 25--27. 83. Ben-Porath,I. and Weinberg,R.A. (2004) When cells get stressed: an integrative view of cellular senescence. J. Clin. Invest., 113, 8--13. 84. Bond,J., Jones,C., Haughton,M., DeMicco,C., Kipling,D. and WynfordThomas,D. (2004) Direct evidence from siRNA-directed ‘knock down’ that p16(INK4a) is required for human fibroblast senescence and for limiting ras-induced epithelial cell proliferation. Exp. Cell Res., 292, 151--156. 85. d’Adda di Fagagna,F., Reaper,P.M., Clay-Farrace,L., Fiegler,H., Carr,P., Von Zglinicki,T., Saretzki,G., Carter,N.P. and Jackson S.P. (2003) A DNA damage checkpoint response in telomere-initiated senescence. Nature, 426, 194--198. 86. Jacobs,J.J. and de Lange. T. (2004) Significant role for p16INK4a in p53-independent telomere-directed senescence. Curr. Biol., 14, 2302--2308. 87. Beausejour,C.M., Krtolica,A., Galimi,F., Narita,M., Lowe,S.W., Yaswen,P. and Campisi,J. (2003) Reversal of human cellular senescence: roles of the p53 and p16 pathways. EMBO J., 22, 4212--4222. 88. Yuasa,Y. (2002) DNA methylation in cancer and ageing. Mech. Ageing Dev., 123, 1649--1654. 89. Gonzalgo,M.L. and Jones,P.A. (1997) Mutagenic and epigenetic effects of DNA methylation. Mutat. Res., 386, 107--118. 90. Pfeifer,G.P., Tang,M.S. and Denissenko,M.F. (2000) Mutation hotspots and DNA methylation. Curr. Top. Microbiol. Immunol., 249, 1--19. 91. Mancini,D., Singh,S., Ainsworth,P. and Rodenhiser,D. (1997) Constitutively methylated CpG dinucleotides as mutation hot spots in the retinoblastoma gene (RB1). Am. J. Hum. Genet., 61, 80--87. 92. Pfeifer,G.P., You,Y.H. and Besaratinia,A. (2005) Mutations induced by ultraviolet light. Mutat Res., 571, 19--31. 93. Zhang,N., Lin,C., Huang,X., Kolbanovskiy,A., Hingerty,B.E., Amin,S., Broyde,S., Geacintov,N.E. and Patel,D.J. (2005) Methylation of cytosine at C5 in a CpG sequence context causes a conformational switch of a benzo[a]pyrene diol epoxide-N2-guanine adduct in DNA from a minor groove alignment to intercalation with base displacement. J. Mol. Biol., 346, 951--965. 94. Goodman,J.E., Hofseth,L.J., Hussain,S.P. and Harris,C.C. (2004) Nitric oxide and p53 in cancer-prone chronic inflammation and oxyradical overload disease. Environ. Mol. Mutagen., 44, 3--9. 95. Zajaczek,S., Jakubowska,A., Gorski,B., Kurzawski,G., Krzystolik,Z. and Lubinski,J. (1999) Frequency and nature of germline Rb-1 gene mutations in a series of patients with sporadic unilateral retinoblastoma. Eur. J. Cancer, 35, 1824--1827. 96. Harbour,J.W. (1998) Overview of RB gene mutations in patients with retinoblastoma. Implications for clinical genetic screening. Ophthalmology, 105, 1442--1447. 97. Yonghao,T., Qian,H., Chuanyuan,L. and Yandell,D.W. (1999) Deletions and point mutations of p16, p15 gene in primary tumors and tumor cell lines. Chin. Med. Sci. J., 14, 200--205. 98. Wiebauer,K. and Jiricny,J. (1989) In vitro correction of G.T mispairs to G.C pairs in nuclear extracts from human cells. Nature, 339, 234--236. 99. Stojic,L., Brun,R. and Jiricny,J. (2004) Mismatch repair and DNA damage signalling. DNA Repair (Amst), 3, 1091--1101. 100. Gifford,G. and Brown,R. (2004) Microsatellite instability: theory and methods. Methods Mol. Med., 97, 237--250. 101. Robertson,K.D. and Jones,P.A. (2000) DNA methylation: past, present and future directions. Carcinogenesis, 21, 461--467. 102. Tommasi,S., Denissenko,M.F. and Pfeifer,G.P. (1997) Sunlight induces pyrimidine dimers preferentially at 5-methylcytosine bases. Cancer Res., 57, 4727--4730. 103. Almeida,A., Kokalj-Vokac,N., Lefrancois,D., Viegas-Pequignot,E., Jeanpierre,M., Dutrillaux,B. and Malfoy,B. (1993). Hypomethylation of classical satellite DNA and chromosome instability in lymphoblastoid cell lines. Hum. Genet., 91, 538--546. 104. Ehrlich,M., Hopkins,N.E., Jiang,G. et al. (2003) Satellite DNA hypomethylation in karyotyped Wilms tumors. Cancer Genet. Cytogenet., 141, 97--105. 105. Deininger,P.L. and Batzer,M.A. (2002) Mammalian retroelements. Genome Res., 12, 1455--1465. 106. Goodier,J.L., Ostertag,E.M. and Kazazian,H.H.Jr (2000) Transduction of 30 -flanking sequences is common in L1 retrotransposition. Hum. Mol. Genet., 9, 653--657. 107. Buzdin,A., Ustyugova,S., Gogvadze,E., Vinogradova,T., Lebedev,Y. and Sverdlov,E. (2002) A new family of chimeric retrotranscripts formed by a full copy of U6 small nuclear RNA fused to the 30 terminus of l1. Genomics, 80, 402--406. 108. Lev-Maor,G., Sorek,R., Shomron,N. and Ast,G. (2003) The birth of an alternatively spliced exon: 30 splice-site selection in Alu exons. Science, 300, 1288--1291. 109. Schmid,C.W. (1996) Alu: structure, origin, evolution, significance and function of one-tenth of human DNA. Prog. Nucleic Acid Res. Mol. Biol., 53, 283--319. 110. Rodin,S.N. and Rodin,A.S. (2005) Origins and selection of p53 mutations in lung carcinogenesis. Semin. Cancer Biol., 15, 103--112. 111. Blagosklonny,M.V. (2002) Oncogenic resistance to growth-limiting conditions. Nat. Rev. Cancer, 2, 221--225. 112. Bridges,B.A. (1994) Starvation-associated mutation in Escherichia coli: a spontaneous lesion hypothesis for ‘directed’ mutation. Mutat. Res., 307, 149--156. 113. Foster,P.L. (2000) Adaptive mutation: implications for evolution. Bioessays, 22, 1067--1074. 114. Hendrickson,H., Slechta,E.S., Bergthorsson,U., Andersson,D.I. and Roth,J.R. (2002) Amplification-mutagenesis: evidence that ‘directed’ adaptive mutation and general hypermutability result from growth with a selected gene amplification. Proc. Natl Acad. Sci. USA, 99, 2164--2169. 115. Kivisaar,M. (2003) Stationary phase mutagenesis: mechanisms that accelerate adaptation of microbial populations under environmental stress. Environ. Microbiol., 5, 814--827. 116. Rosenberg,S.M. and Hastings,P.J. (2004) Adaptive point mutation and adaptive amplification pathways in the Escherichia coli lac system: stress responses producing genetic change. J. Bacteriol., 186, 4838--4843. 117. Wright,B.E. (2004) Stress-directed adaptive mutations and evolution. Mol. Microbiol., 52, 643--650. 1333 T.V.Karpinets and B.D.Foy 118. McKenzie,G.J. and Rosenberg,S.M. (2001) Adaptive mutations, mutator DNA polymerases and genetic change strategies of pathogens. Curr. Opin. Microbiol., 4, 586--594. 119. Rubin,H. (1993) Cellular epigenetics: effects of passage history on competence of cells for ‘spontaneous’ transformation. Proc. Natl Acad. Sci. USA, 90, 10715--10719. 120. Martin,S.G. (1996) Normal cells and cancer cells. In Bishop,J.M. and Weinberg,R.A. (eds) Scientific American Molecular Oncology. Scientific American, New York, pp. 13--40. 121. Rhim,J.S. (2000) Development of human cell lines from multiple organs. Ann. N. Y. Acad. Sci., 919, 16--25. 122. Newbold,R.R., Bullock,B.C. and McLachlan,J.A. (1990) Uterine adenocarcinoma in mice following developmental treatment with estrogens: a model for hormonal carcinogenesis. Cancer Res., 50, 7677--7681. 123. Sciandrello,G., Caradonna,F., Mauro,M. and Barbata,G. (2004) Arsenicinduced DNA hypomethylation affects chromosomal instability in mammalian cells. Carcinogenesis, 25, 413--417. 124. Zhou,H., Randers-Pehrson,G., Geard,C.R., Brenner,D.J., Hall,E.J. and Hei,T.K. (2003) Interaction between radiation-induced adaptive response and bystander mutagenesis in mammalian cells. Radiat. Res., 160, 512--516. 125. Suzuki,K., Ojima,M., Kodama,S. and Watanabe,M. (2003) Radiationinduced DNA damage and delayed induced genomic instability. Oncogene, 22, 6988--6993. 126. Lorimore,S.A., Coates,P.J. and Wright,E.G. (2003) Radiation-induced genomic instability and bystander effects: inter-related nontargeted effects of exposure to ionizing radiation. Oncogene, 22, 7058--7069. 127. Linet,M.S., Wacholder,S. and Zahm,S.H. (2003) Interpreting epidemiologic research: lessons from studies of childhood cancer. Pediatrics, 112, 218--232. 128. Moggs,J.G., Tinwell,H., Spurway,T. et al. (2004) Phenotypic anchoring of gene expression changes during estrogen-induced uterine growth. Environ. Health Perspect., 112, 1589--1606. 1334 129. Kushner,P.J., Agard,D.A., Greene,G.L., Scanlan,T.S., Shiau,A.K., Uht,R.M. and Webb,P. (2000) Estrogen receptor pathways to AP-1. J. Steroid Biochem. Mol. Biol., 74, 311--317. 130. Song,R.X.-D., McPherson,R.A., Adam,L., Bao,Y., Shupnik,M., Kumar,R. and Santen,R.J. (2002) Linkage of rapid estrogen action to MAPK activation by ERa-Shc association and Shc pathway activation. Endocrinology, 16, 116--127. 131. Weber,M.J. and Gioeli,D. (2004) Ras signaling in prostate cancer progression. J. Cell Biochem., 91, 13--25. 132. Gatza,M.L., Chandhasin,C., Ducu,R.I. and Marriott,S.J. (2005) Impact of transforming viruses on cellular mutagenesis, genome stability and cellular transformation. Environ. Mol. Mutagen., 45, 304--325. 133. Coussens,L.M. and Werb,Z. (2002) Inflammation and cancer. Nature, 420, 860--867. 134. Brower,V. (2005) Feeding the flame: new research adds to role of inflammation in cancer development. J. Natl Cancer Inst., 97, 251--253. 135. Nguyen,T.H. (2004) Mechanisms of metastasis. Clin. Dermatol., 22, 209--216. 136. Liotta,L.A. and Kohn,E. (2004) Anoikis: cancer and the homeless cell. Nature, 430, 973--974. 137. So,A., Gleave,M., Hurtado-Col,A. and Nelson,C. (2005) Mechanisms of the development of androgen independence in prostate cancer. World J. Urol., 23, 1--9. 138. Poirier,L.A. (2002) The effects of diet, genetics and chemicals on toxicity and aberrant DNA methylation: an introduction. J. Nutr., 132, 2336--2339. 139. Li,S., Hursting,S.D., Davis,B.J., McLachlan,J.A. and Barrett,J.C. (2003) Environmental exposure, DNA methylation and gene regulation: lessons from diethylstilbesterol-induced cancers. Ann. N. Y. Acad. Sci., 983, 161--169. Received January 3, 2005; revised March 21, 2005; accepted March 22, 2005