Environmental Carcinogens and Mutagens

Environmental Carcinogens and Mutagens Edward L Loechler,

Boston University, Boston, Massachusetts, USA

Cancer is the second leading cause of death in the United States. A third of all Americans will get cancer, with about 1 220 000 new cases, and about 550 000 deaths in the year 2000 according to the American Cancer Society. General information of this type can be found at the American Cancer Society’s Web site (www.cancer.org). When different organs and cell types are considered, there are over 100 different kinds of cancer. If skin cancer is excluded, prostate and breast cancer are the most prevalent for American men and women, respectively, although lung cancer causes the most deaths per year for each. Carcinogens are agents that cause cancer. Cancer is an abnormal, enhanced growth of cells, frequently into a mass or tumour. Most carcinogens are mutagens, which are substances that cause mutations in the genetic material, DNA. A mutation is any permanent, heritable change in DNA (i.e. in the sequence of the DNA). The structures of several important environmental carcinogens and mutagens are given in Figure 1. The fact that mutagens are carcinogens is understandable given that cancer cells grow in an improperly regulated fashion if they have mutations in particular genes whose function it is to provide proper growth control for the cell. Carcinogens can be classified into agents that are chemical, viral or physical (e.g. radiation, such as UV light, X-rays and g-rays). Approximately 60–90% of all cancers are now generally believed to be due to environmental factors, to which humans are exposed in food, water or air. It is important to mention that ‘environmental factors’ include both natural and human-made agents. The term ‘carcinogen’ refers to any substance that can contribute to the process of tumour formation. Carcinogens do this in at least three ways – as mutagens (or genotoxins), as co-carcinogens and as tumour promoters – and the term is most often associated with substances that are genotoxic (meaning ‘genetic toxins’), which initiate the process of carcinogenesis by causing a mutation in DNA (i.e. as mutagens). A variety of endogenous processes that

Article Contents . Introduction . DNA

Environmental agents (chemicals, radiation and viruses) can be carcinogens (cancercausing substances) and mutagens (substances that alter DNA). The steps from mutagen/ carcinogen to tumour involve mutagen/carcinogen activation by metabolism, reaction with DNA to give a DNA adduct and DNA replication of the adduct, resulting in a mutation; a mutation in a cancer gene may result in enhanced growth potential, and ultimately in tumour growth.

Introduction

Secondary article

. Historical Breakthroughs in the Understanding of Environmental Carcinogenesis . Classifying Carcinogens . Identifying Carcinogens . Steps in Carcinogenesis

occur spontaneously inside cells can also contribute to mutagenesis and carcinogenesis, including spontaneous DNA damage, such as hydrolysis of or oxidative damage to the DNA bases, as well as mistakes (errors) being made during the duplication of DNA by DNA polymerase. A ‘co-carcinogen’ is a substance that by itself does not cause a tumour to form but that, when present at the same time as a genotoxin, enhances the potency of that genotoxin. A ‘tumour promoter’ is a substance that by itself does not cause a tumour but that enhances tumour formation when it is given (usually repeatedly) after exposure to a genotoxin. Unless stated otherwise, the term ‘carcinogen’ will hereafter be used to refer to a mutagen (genotoxin). Chemical mutagens/carcinogens, which are probably most important to human cancer causation, will be the focus. There are a number of excellent references that comprehensively discuss most aspects of environmental carcinogenesis and mutagenesis (e.g. Eaton and Groopman, 1994; Kitchen, 1998; Singer and Grunberger, 1983; Searle, 1984; Balmain et al., 2000).

DNA Understanding of mutagenesis and carcinogenesis is intimately tied to understanding various aspects of the genetic substance DNA. DNA is a polymer, whose backbone (a DNA ‘strand’) is composed of alternating sugar (deoxyribose) and phosphate groups. Linked to each sugar is one of the so-called ‘DNA bases’: adenine, cytosine, guanine and thymine (abbreviated A, C, G and T, respectively), The exact linear sequence of these four DNA bases provides the blueprint for all of the functions in all of the cells in an organism. DNA in cells typically has two sugar-phosphate backbones wrapped around each other in the form of a helix (the ‘double helix’). The two strands are held together by interactions between complementary bases in each strand: A opposite T, and G

ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net

1

Environmental Carcinogens and Mutagens

PhIP (2-amino-1-methyl-6phenylimidazo-[4,5-b]pyridine)

B[a]P (benzo[a]pyrene)

DMN (N-nitrosodimethylamine)

CH 3

H 3C

N H 3C

NH 2 N O

N

N

O

N

O NH 2

O

O

H 3C N

O O AFB1 (aflatoxin B1)

NH 2

OCH 3 Benzidine (3,3′-diaminobiphenyl)

N

O

N NNK (4-(N-nitrosomethylamino)1-(3-pyridyl)-1-butanone)

Figure 1 The structures of several chemical carcinogens discussed in the text. Benzo[a]pyrene (B[a]P) is a polycyclic aromatic hydrocarbon (PAH); PAHs are found in soots and are produced by incomplete combustion (e.g. in exhaust gases from internal combustion engines and power plants, cigarette smoke and charred foods). Aflatoxin B1 (AFB1) is a mycotoxin produced by moulds that grow on improperly stored foodstuffs (e.g. peanuts). Two aromatic amines are shown: PhIP, found in charred foods; and benzidine, an industrial chemical. Two alkylating (methylating) agents are shown: dimethylnitrosamine (DMN), found in certain foods; and NNK, found in cigarette smoke.

opposite C, principally involving hydrogen bonds between these complementary bases. Hydrogen bonding between the base pairs can be broken, which allows the DNA strands to separate, as must occur when DNA is duplicated (by a process called DNA replication) during cell division. Chromosomes are composed of the two strands of DNA coiled around each other, with numerous bound proteins that serve structural and regulatory functions. Chromosomes are typically very long; for example, humans have 46 chromosomes (or 23 pairs) that contain a total of 3 109 pairs of DNA bases.

Historical Breakthroughs in the Understanding of Environmental Carcinogenesis The notion that human cancer might be caused by environmental factors, in particular chemicals, began in 1761 when John Hill noted the development of nasal cancer as a consequence of excessive use of tobacco snuff; he was followed in 1775 by Percival Pott, who noted the unusually high incidence of scrotal cancer among chimney sweeps exposed to chimney soots. The experimental investigation of the role of environmental exposure and cancer did not begin until more than a century later, when in 1915 Yamagiwa and Ichikawa first showed that chemicals (a mixture of coal tars) caused skin cancer when painted on to the ears of rabbits. The first pure synthetic compound that demonstrated carcinogenicity 2

was dibenz[a,h]anthracene in 1930, and the first compound isolated from a complex mixture (coal tar) shown to be carcinogenic was benzo[a]pyrene (B[a]P, Figure 1) in 1933; much of this work was spearheaded by E.L. Kennaway. These compounds are examples of the large class of chemicals called polycyclic aromatic hydrocarbons (PAHs). Wilhelm Hueper wrote the first comprehensive book about exposures to chemical carcinogens and cancer in 1942, which began as an outgrowth of his pioneering work in the 1930s showing that certain aniline dyes (e.g. benzidine, Figure 1) caused bladder cancer in dogs. Hueper’s work was initiated because of concern about the observation, first made by Rehn in 1895, that workers in the aniline dye industry got bladder cancer at an alarmingly high rate. The pursuit of the observation of an acute hepatotoxic disease in turkeys (turkey ‘X’ disease), resulted in the isolation of the mycotoxins (e.g. aflatoxin B1) by Gerald Wogan and colleagues in 1963, which eventually led to the realization that these agents, produced by moulds that grow on improperly stored foods (such as peanuts), also cause liver cancer (Eaton and Groopman, 1994). The association between exposure to a chemical (or mixtures of chemicals) and human cancer has been established in numerous epidemiological studies, most significantly in the role of smoking in lung cancer. The first report of this association was by Muller in 1939, but five reports in 1950 were more definitive, with the most influential being by Doll and Hill, and by Wynder and Graham (Hoffmann and Hoffmann, 1997). The first report of an association between exposure to asbestos, the most important occupational carcinogen, appeared in 1942,



although it took the monumental work of Irving Selikoff to build the case for this association more definitively (Selikoff and Hammond, 1978). The ability of chemicals to cause human cancer was also clearly, dramatically and regrettably demonstrated in the case of occupational exposures to a number of chemicals, such as bischloromethylether (lung cancer), and vinyl chloride (hepatic angiosarcoma). The notion that 60–90% of all human cancers are likely to be due to environmental factors was first apparent from the work of Higginson in 1969, who showed that cancer incidence varied between countries, probably because of differences in environmental factors. This conclusion was reinforced by Haenszel and his collaborators, who showed that immigrants and their descendants tended to adopt the cancer risks of the country into which they immigrated. On the biological side, the hypothesis that cancer might develop from mutations in the genetic substance of cells was first proposed by Theodor Boveri in 1914. The link between damage to the genetic substance (later recognized to be DNA) and mutations was provided in the pioneering work by H.J. Muller, beginning in the 1920s, showing that X-irradiation caused mutations in Drosophila. While suspected for many years, the strongest support showing that carcinogens are mutagens was provided in the work of Bruce Ames in the 1970s, who developed the so-called ‘Ames test’, which utilizes a bacterium that has been engineered to be extraordinarily sensitive to the induction of mutations. The correlation between mutagens and carcinogens in the Ames test has generally been found to be quite high ( 80%). The role of mutations in human cancer was further substantiated by the finding that certain genes in tumour cells are mutated. The first studies involved groups led by Robert Weinberg, Mariano Barbacid and Michael Weigler, who showed that a particular codon in a gene called Ha-ras (in cells from a human bladder cancer) was changed from 5’-GGC to 5’GTC, resulting in a change of glycine to valine in the 12th amino acid in the protein encoded by Ha-ras (Cooper, 1995; Varmus and Weinberg, 1993). Perhaps the clearest example of a role for specific aetiological agents in the cause of human cancer comes from the findings that the mutations in the p53 tumour suppressor gene show patterns that seem to give a clue about the aetiological agent responsible (Hainaut et al., 1998; Greenblatt et al., 1994); this has been referred to as a mutagenic ‘fingerprint’. For example, mutations in cells from certain skin cancers resemble the mutations known to be induced by UV light.

Classifying Carcinogens The International Agency for Research on Cancer (IARC) maintains a registry of human carcinogens and suspected human carcinogens. This has been developed on the basis

of the examination of epidemiological studies on humans and animal model studies and the use of a variety of shortterm tests, such as the Ames test. More than 860 compounds had been evaluated as at March 2001, and these are discussed in a series called the IARC Monographs, which currently includes Monographs 1–77, along with multiple supplements. A complete listing of the chemicals evaluated and the carcinogenic hazards they pose can be found at the IARC Web site (www.iarc.fr) by pursuing the IARC Monographs Database. Some information from this database is given in Table 1. In addition, the National Toxicology Program (NTP) conducts animal assays for long-term cancer studies of suspected hazardous chemicals. The chemicals studied have been chosen on the basis of human exposure, level of production and chemical structure. Information on the compounds studied can be found in the Abstracts of NTP Long-Term Cancer Studies or at the NTP Website (ntpserver.niehs.nih.gov). The number of compounds that are or may be carcinogens is extensive. However, these chemicals can be subdivided into different classes (Table 1), including polycyclic aromatic hydrocarbons (PAHs), heterocyclic amines or aromatic amines (HAs), alkylating agents, mycotoxins, metals and fibres (notably asbestos), as well as other chemicals and mixtures, such as tobacco smoke, certain anticancer drugs (cyclophosphamide and cisplatin), warfare agents (sulfur mustard) and oxidizing agents (hydrogen peroxide). In each of these classes, there are typically one or two chemicals that have emerged as representatives and are the most studied, and some of these are shown in Figure 1.

Identifying Carcinogens Epidemiological studies It has frequently been difficult to prove definitively that a particular chemical causes human cancer (Kitchen, 1998). However, this hypothesis is supported by numerous (and unfortunate) examples in which subpopulations have been exposed to chemicals (or mixtures of chemicals) that led to a higher cancer incidence; examples include lung cancer in workers exposed to asbestos, bladder cancer in workers in the aniline dye industry, and several other examples cited above. In some instances, very strong cases have been made based on so-called retrospective epidemiological studies; two groups of individuals are matched as closely as possible, except with respect to exposure to a particular substance. If the exposed group has a higher incidence of a particular kind of cancer, then this is evidence that the exposure is likely causative. This approach was crucial in showing that cigarette smoking causes lung cancer, and


3


Table 1 Substances selected from the International Agency for Research on Cancer Monograph Database about their overall evaluation of carcinogenicity to humans Categorya

Substance Polyaromatic hydrocarbons (PAH) TCDD (dioxin) (2,3,7,8-tetrachlorodibenzo-p-dioxin) Benzo[a]pyrene Benzo[a]anthracene Polychlorinated biphenyls (PCB) Benzo[b]fluoranthene

1 2A 2A 2B

Heterocyclic amines (HA) 4-Aminobiphenyl Benzidine IQ (2-amino-3-methylimidazo[4,5-f]quinoline) MeIQ (2-amino-3,4-dimethylimidazo[4,5-f]quinoline) PhIP (2-Amino-1-methy-6-phenylimidazo[4,5-b]pyridine) 4-Nitropyrene

1 1 2A 2B 2B 2B

Alkylating agents Sulfur mustard DMS, DES, ENU EDB (ethylene dibromide) NNK (4-(N-nitrosomethylamino)-1-(3-pyridyl)-1-butanone) Many nitrosamines DMN (N-nitrosodimethylamine)

1 2A 2A 2B 2B 2A

Mycotoxins Aflatoxins naturally occurring Aflatoxin M1 Sterigmatocystin

1 2B 2B

Metals (as compounds) Nickel Chromium Cadmium Arsenic Beryllium Cobalt

1 1 1 1 1 2B

Fibres Silica (asbestos) Wood dust Ceramic fibres Glasswool Acrylic fibres

1 1 2B 2B 3 continued

confirmed that asbestos exposure caused a variety of cancers. 4

Of equal value are so-called prospective epidemiological studies, which involve collecting information (principally



Table 1 – continued Substance

Categorya

Others Vinyl chloride Tobacco smoke Tobacco products Cyclophosphamide Sulfur mustard Nitrogen mustard Formaldehyde Cisplatin Methylene chloride Hydrazine Aciclovir Hydrogen peroxide Ethylene

1 1 1 1 1 1 2A 2A 2B 2B 3 3 3

a

Categorization by group is based on the following criteria of evaluation for agents as stated in the Preamble to the IARC Monographs. Group 1: The agent is carcinogenic to humans. This category is used when there is sufficient evidence of carcinogenicity in humans. Group 2: This category includes agents for which, at one extreme, the degree of evidence of carcinogenicity in humans is almost sufficient, as well as those for which, at the other extreme, there are no human data but for which there is evidence of carcinogenicity in experimental animals. Agents are assigned to either group 2A (probably carcinogenic to humans) or group 2B (possibly carcinogenic to humans). Group 3: The agent is not classifiable as to its carcinogenicity to humans. This category is used most commonly for agents, mixtures and exposure circumstances for which the evidence of carcinogenicity is inadequate in humans and inadequate or limited in experimental animals. Group 4: The agent (mixture) is probably not carcinogenic to humans.

about lifestyle) from a large number of individuals (involving questionnaires), following their health for long periods of time, and then analysing associations between potential risk factors and adverse health outcomes. The three most important studies in this regard are the ongoing ‘Nurses’ Health Study’ (121 700 women), ‘Health Professionals Follow-up Study’ (52 000 men) and ‘Nurses’ Health Study II’ (116 000 women). Examples of relationships that have been studied include a positive association between alcohol consumption and breast cancer but no relation with fat intake; a positive association between animal fat and red meat consumption and risk of colon cancer; strong inverse associations between vitamin E consumption and risk of coronary heart disease in both men and women; and a positive association between partially hydrogenated vegetable fats and coronary heart disease incidence. Other endpoints being examined in these studies with regard to diet include diabetes, cataracts, glaucoma, gallstones and other malignancies.

Animal studies and short-term tests Potential cancer risks have also been investigated by other methods, including studies to determine whether laboratory animals, such as rats or mice, develop cancer when exposed to particular agents. Although the correlation is not perfect, animal carcinogens are generally likely to be human carcinogens. Because animal studies are both time-consuming and expensive, so-called short-term tests have also been developed (e.g. the aforementioned Ames test, which can be conducted very quickly, in some cases overnight). There is a high correlation between mutagens in the Ames test and carcinogens, which in fact was one of the crucial pieces of evidence that carcinogens are mutagens.

Examples of the value of animal studies and short-term tests Animal studies and short-term tests do not provide proof that a chemical will cause human cancer. For example,


5


O Benzo[a]pyrene

Benzo[a]pyrene 7,8-epoxide

Benzo[a]pyrene 7,8-diol

Benzo[a]pyrene 7,8-diol-9,10-epoxide H

O

O O Less carcinogenic metabolites

N

O HO

OH

N

N N

(+)-trans-antiB[a]P-N2-dG

N

HO

O HO

H

O

Mutation

Cancer

HO OH

OH

OH

DNA repair

Benzo[a]pyrene tetraol

HO

O HO OH Figure 2 The carcinogenesis paradigm, using benzo[a]pyrene as an example (see text). The horizontal arrows lead towards greater toxicity (in particular toward cancer), while the vertical arrows lead toward lesser toxicity.

relatively high doses of chemicals are usually given to a small numbers of animals in an attempt to mimic the relatively lower doses to which large numbers of humans may be exposed. Legitimate concerns have been raised about whether high doses in animals might cause a nonrepresentative response and lead to an incorrect indication of hazard. Such issues lead to the question: If a chemical tests positively, is it more prudent to treat it as likely or unlikely to be a human carcinogen? In this regard, it is worth remembering that, even if a mutagen could be shown to be noncarcinogenic in humans for some reason, mutagens by definition cause mutations, which are deleterious and are associated with other human diseases, most notably birth defects. Regarding the usefulness of animal studies and short-term tests, it may also be worthwhile to consider several historical examples. It is of course logically impossible to document cases where human cancer was prevented because of a ban on or restriction of a chemical based on a positive test for carcinogenesis in an animal study. However, the following examples may be illuminating in this regard. In 1941, both 2-acetylaminofluorene (2-AAF) and diethylstilboestrol (DES) were shown to be carcinogenic in animals. 2-AAF, developed as a pesticide, was banned on the basis of this single animal experiment. Although it is impossible to know how many lives were saved or what benefits were lost from this ban, the consensus is that the correct choice was made, since AAF has become the best-studied carcinogen in the heterocyclic amine class, and has consistently been shown to be a potent mutagen and animal carcinogen. DES was not banned. In fact, numerous women took DES during pregnancy to prevent miscarriage (it is now 6

generally regarded to be ineffective in this regard), and daughters born to these women have a high incidence of gynaecological abnormalities and cancer. In this case the consensus today is that the wrong choice was made. The usefulness of short-term tests can be illustrated by the following example. In an unlikely series of events, the Ames test was employed to discover that carbon black used in photocopier toner contained a high concentration of potent mutagens, which were identified as nitropyrenes. A simple change in the manufacturing process virtually eliminated these compounds from toner.

Steps in Carcinogenesis The impact of chemical carcinogens on biological systems is exceedingly complex, but can be subdivided into a number of steps to simplify its presentation. While the details vary depending on the carcinogen, the paradigm shown in Figure 2, which one particularly well-studied chemical carcinogen (benzo[a]pyrene), illustrates many of the principles (reviewed in Singer and Grunberger, 1983; Conney, 1982; Balmain et al., 2000). Steps in the horizontal direction lead towards carcinogenicity and include metabolic activation, reaction with DNA (adduction), adduct mutagenesis and tumourigenesis. Steps in the vertical direction lead to diminished carcinogenicity and include metabolic detoxification, carcinogen deactivation and DNA repair. Diminished carcinogenicity is also associated with delaying the cell cycle and with apoptosis. Each of these steps is discussed below, as well as several other relevant topics.



Mutagen/carcinogen metabolism Most chemical carcinogens are not inherently cancercausing without first being covalently modified (or ‘metabolized’) inside cells by enzymes (Balmain et al., 2000). Most metabolism leads to the formation of less toxic or less carcinogenic substances, which are eventually excreted from the body. However, a small fraction of metabolism leads to the formation of more carcinogenic derivatives. Most potent carcinogenic substances are hydrophobic and tend to accumulate in cell membranes and fatty tissues in the body, where they have the potential to cause dysfunction and toxicity. Thus, living organisms have evolved to rid themselves of these xenobiotic (foreign), toxic substances. The strategy generally involves the enzymatic addition of hydroxyl groups to mutagens/ carcinogens, which increases their polarity and thus their solubility in water, and facilitates eventual excretion. Other enzymes can also be involved in detoxification, such as glutathione S-transferase, which conjugates xenobiotics to the intracellular thiol glutathione. Metabolism occurs in a variety of tissues of the body, but is frequently most extensive in the liver. Several types of detoxifying enzymes can be involved, but quantitatively the most important are the ubiquitous cytochrome P450 enzymes. Humans are known to have at least 20 distinct cytochrome P450s (and may have as many as 100), which generally result ultimately in the addition of hydroxyl groups (OH) to different kinds of chemicals. Cytochrome P450s are also responsible for other types of metabolism, including steps in the biosynthesis of hormones (such as oestrogen) and cholesterol. While P450 enzymes usually generate less toxic species, they can also form reactive intermediates, notably epoxides, such as the 7,8-epoxide of benzo[a]pyrene. These epoxides are very reactive and spontaneously hydrolyse to the corresponding diols, in a reaction that is also catalysed by the enzyme epoxide hydrolase. Subsequent epoxidation of the 7,8-diol leads to the formation of the 7,8-diol-9,10epoxide, which is the ultimate carcinogen of benzo[a]pyrene (see below). Other metabolites may also be important for carcinogenesis.

Mutagen/carcinogen reaction with DNA (adduction) Although the majority of the benzo[a]pyrene diol-epoxide reacts with water to give the corresponding benzo[a]pyrene tetraol, a significant fraction can covalently react with DNA, which occurs principally at the N 2-position of the guanine base of DNA to give the adduct ( 1 )-trans-antiB[a]P-N 2-dG (Figure 2) (Harvey, 1997; Singer and Grunberger, 1983; Conney, 1982). Other carcinogens may react at different sites in DNA: for example, following metabolism, aflatoxin B1 reacts at N7 of guanine; most

aromatic amines (e.g. benzidine and PhIP, Figure 1) preferentially react at C8 of guanine; and DMN and NNK (Figure 1) add methyl groups to many sites in DNA, the most important in terms of mutagenesis (and toxicity) being O6 of guanine. Many carcinogens have a natural affinity for DNA that enhances covalent adduction and thus their mutagenic/ carcinogenic potency. In a general sense this is attributable to the hydrophobic portion of the mutagen/carcinogen, which has an affinity for the hydrophobic bases in DNA. In many cases, potent carcinogens have flat, planar, aromatic regions (e.g. the fused benzene rings in benzo[a]pyrene) that insert (intercalate) between the base pairs of DNA in a step that directly precedes covalent reaction.

DNA repair In most cases, DNA adducts are removed from DNA by a process called DNA repair, which restores the sequence and integrity of DNA. DNA repair is a multifaceted process, and – as with metabolism – involves a large number of diverse enzymes (Nickoloff and Hoekstra, 1998; Friedberg et al., 1995; Chaney and Sancar, 1996). Some DNA repair proteins are able to reverse the adduct-forming step and, thereby, directly restore the integrity of the DNA base to which the carcinogen was attached. An example is alkylguanine alkyltransferase, which transforms the premutagenic lesion, O6-methylguanine, which is formed from alkylating agents like dimethylnitrosamine (Figure 1), directly back to guanine in DNA. Other strategies of DNA repair involve the removal of a stretch of the DNA strand surrounding the adduct by a complex of proteins. This stretch of DNA is then resynthesized by a DNA polymerase using the complementary DNA strand as a template, which restores the integrity of the DNA. Base excision repair and nucleotide excision repair are the two most prominent examples. In some cases DNA repair is not completed before a DNA polymerase encounters an adduct in the strand it is attempting to copy (during DNA replication), in which case the polymerase is blocked (at least temporarily). This can trigger a cell to induce a variety of additional strategies to avoid the accumulation of deleterious mutations. One prominent example is called post-replication repair, which involves ‘borrowing’ one strand of DNA for the identical, sister chromosome (by a process called homologous recombination) and placing it opposite the DNA adduct to accomplish bypass. Since the adduct is not bypassed by a DNA polymerase, this process is error free.

Cell cycle arrest and apoptosis Cells have at least two additional ways to minimize the impact of DNA damage (Stein et al., 1999; Nurse, 2000;


7


Earnshaw et al., 1999). First, cells actively dividing are at greatest risk for mutation, since they are actively involved in DNA replication. Dividing cells follow an orderly process called the cell cycle, involving four stages: cell division (or mitosis, M); gap 1 (G1); DNA synthesis (S); and gap 2 (G2). Cells have DNA surveillance mechanisms, and if excessive DNA damage is detected, then cell cycle arrest (or delay) occurs, which allows more time for DNA repair to rectify this DNA damage. Second, a cell also has a way of assessing whether it has accumulated too much DNA damage and, if it has, it initiates a process termed apoptosis (pronounced: ah-po-to-sis), or ‘programmed cell death’, whereby it purposely destroys itself rather than risk the accumulation of too many mutations. One common example is very sunburned skin. The peeling of the exposed skin cells is actually the sloughing off of dead cells from the upper layer of skin. These dead cells have received a very high dose of mutagenic ultraviolet light from the sun, and have ‘decided’ to undergo apoptosis. The dead cells are replaced by cells from farther down in the dermis that were shielded from and relatively unaffected by the dose of ultraviolet light.

Mutagenesis Mutagenesis is defined as any heritable alteration in the primary sequence of bases in the genetic substance DNA (Hall and Hainaut, 2000; Loechler, 1994; Miller, 1983). When an adduct is not repaired, a DNA polymerase may attempt to read past this adduct (e.g., ( 1 )-trans-antiB[a]P-N2-dG; Figure 2), in which case it has a much greater chance of incorporating the wrong base (e.g. A instead of C), than when reading a normal base (e.g. G in this case). This will eventually lead to a mutation, when T is incorporated opposite this A in the next round of DNA replication. There are three general types of mutations. (1) A base substitution involves the exchange of one base pair for another; e.g. a G:C base pair was transformed into a T:A base pair in the example given in the previous paragraph. (2) A deletion involves the loss of one or more base pairs from a DNA sequence. (3) An insertion involves the addition of one or more base pairs from a DNA sequence. All of these mutations have been shown to contribute to carcinogenesis on basis of the kinds of mutations observed in the genes associated with human cancers.

Tumorigenesis In most cases, cells in an organism are in a state of stasis, and new cells are only produced (by cell division) to replace old cells that have been lost (e.g. by injury) (Balmain et al., 2000; Cooper, 1995; Varmus and Weinberg, 1993). (There are numerous cases when this is not true, most notably during development.) Cell division is tightly regulated and 8

is overseen by the protein products of a set of growth control genes, such as those that control the cell cycle (see above). If these genes become mutated, the cell loses the checks and balances necessary to ensure that it divides only when it is supposed to. Normal cells have so-called tumour suppressor genes, which permit cell division to occur only when a cell receives a proper growth signal (e.g. to repair damaged tissue). The loss of a tumour suppressor gene by mutation may contribute to uncontrolled cell growth (cancer). A second class of genes called protooncogenes are active in the signalling pathway for cell growth; if these genes are mutated to oncogenes, they can potentially send their signals to grow continually rather than only when it is proper. The best-studied tumour suppressor gene and oncogene are p53 and ras, respectively, and they have been found to be mutated in respectively 50% and 20% of all human tumours. Human tumours are also associated with mutations in genes that encode proteins involved in DNA repair and in apoptosis. Cells do not become tumorigenic following the mutation of a single cancer gene, but rather seem to require a progression of mutations. This has been most extensively studied in the case of colon cancer, where approximately six steps or stages appear to be required. Progression proceeds from relatively early stages, such as benign polyps, through noninvasive adenomas and, finally, to fully malignant carcinomas. It appears that the cells remain at a particular stage until one cell in the mass acquires a mutation in another protooncogene or tumour suppressor gene, following which it is poised to progress to the next stage of the tumorigenic process.

References Balmain A, Brown R and Harris CC (eds) (2000) Twentieth anniversary special issue [which contains reviews of various carcinogenesis topics]. Carcinogenesis 21: 339–531. Chaney SG and Sancar A (1996) DNA repair: enzymatic mechanisms and relevance to drug response. Journal of the National Cancer Institute 88: 1346–1360. Conney AH (1982) Induction of microsomal enzymes by foreign chemicals and carcinogens by polycyclic aromatic hydrocarbons. Cancer Research 42: 4875–4917. Cooper GM (1995) Oncogenes, 2nd edn. London: Jones and Bartlett. Earnshaw WC, Martins LM and Kaufmann SH (1999) Mammalian caspases: structure, activation, substrates and functions during apoptosis. Annual Review of Biochemistry 68: 383–424. Eaton DL and Groopman JD (eds) (1994) The Toxicology of Aflatoxins. New York: Academic Press. Friedberg EC, Walker GW and Seide W (1995) DNA Repair and Mutagenesis. Washington, DC: ASM Press. Greenblatt MS, Bennett WP, Hollstein M and Harris CC (1994) Perspectives in cancer research: mutations in the p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis. Cancer Research 54: 4855–4878. Hainaut P, Hernandez T, Robinson A et al. (1998) IARC database of p53 gene mutations in human tumors and cell lines: updated compilation, revised formats and new visualization tools. Nucleic Acids Research 26: 205–213.



Hall J and Hainaut R (eds) (2000) Reviews in mutation research. Mutation Research 462: 49–430. Harvey RG (1997) Polycyclic Aromatic Hydrocarbons: Chemistry and Cancer. New York: Wiley-VCH. Hoffmann D and Hoffmann I (1997) The changing cigarette, 1950–1995. Journal of Toxicology and Environmental Health 50: 307–364. Kitchen KT (1998) Carcinogenicity: Testing, Predicting and Interpreting Chemical Effects. New York: Marcel Dekker. Loechler EL (1994) Mechanism by which aflatoxins and other bulky carcinogens induce mutations. In: Eaton DL and Groopman JD (eds) The Toxicology of Aflatoxins: Human Health, Veterinary, and Agricultural Significance, pp. 149–178. Orlando, FL: Academic Press. Loechler EL, Henry B and Seo K-Y (2000) Biological responses to chemical carcinogens. In: The Molecular Basis of Human Cancer. Totowa, NJ: Humana Press. Miller JH (1983) Mutational specificity in bacteria. Annual Review of Genetics 17: 215–238.

Nickoloff JA and Hoekstra MF (eds) (1998) DNA Damage and Repair, vols. 1 and 2. Totowa, NJ: Humana Press. Nurse P (2000) A long twentieth century of the cell cycle and beyond. Cell 2000: 71–78. Searle CE (ed.) (1984) Chemical Carcinogens, 2nd edn, ACS Monograph 182. Washington DC: American Chemical Society. Selikoff I and Hammond EC (1978) Asbestos-associated diseases in the United States shipyards. Cancer 28: 67–99. Singer B and Grunberger D (1983) Molecular Biology of Mutagens and Carcinogens. New York: Plenum Press. Stein GS, Baserga R, Giordano A and Denhardt DT (eds) (1999) The Molecular Basis of Cell Cycle and Growth Control. New York: WileyLiss. Varmus H and Weinberg RA (1993) Genes and the Biology of Cancer. New York: Scientific American Library.


9