Stuart H. Yuspa and Peter G. Shields
INTRODUCTION
As early as the 1800s, initial observations of unusual cancer incidences in occupational groups provided the first indications that chemicals were a cause of human cancer, which was then confirmed in experimental animal studies during the early and mid 1900s. However, the extent to which chemical exposures contribute to cancer incidence was not fully appreciated until population-based studies documented differing organ-specific cancer rates in geographically distinct populations and in cohort studies such as those that linked smoking to lung cancer.1 The most commonly occurring chemical exposures that increase cancer risk are tobacco, alcoholic beverages, diet, and reproductive factors (e.g., hormones). Today, it is recognized that cancer results not solely from chemical exposure (e.g., in the workplace or at home), but that a variety of biologic, social, and physical factors contribute to cancer pathogenesis.2,3 For some common cancers, it also has been recognized that heritable factors also contribute to cancer risk from chemical exposure (e.g., genes involved in carcinogen metabolism, DNA repair, a variety of cancer pathways).4 Twin studies show that for common cancers, nongenetic risk factors are dominant, and the best associations for genetic risks of sporadic cancers indicate that the risks for specific genetic traits are typically less than 1.5-fold.5–7 The role of the tumor microenvironment, the cancer stem cells, and feedback signaling to and from the tumor also have been recently recognized as important contributors to carcinogenesis, although how chemicals affect these have yet been clearly demonstrated.8–10
The experimental induction of tumors in animals, the neoplastic transformation of cultured cells by chemicals, and the molecular analysis of human tumors have revealed important concepts regarding the pathogenesis of cancer and how laboratory studies can be used to better understand human cancer pathogenesis.7,11,12 Chemical carcinogens usually affect specific organs, targeting the epithelial cells (or other susceptible cells within an organ) and causing genetic damage (genotoxic) or epigenetic effects regulating DNA transcription and translation. Chemically related DNA damage and consequent somatic mutations relevant to human cancer can occur either directly from exogenous exposures or indirectly by activation of endogenous mutagenic pathways (e.g., nitric oxide, oxyradicals).13,14 The risk of developing a chemically induced tumor may be modified by nongenotoxic exogenous and endogenous exposures and factors (e.g., hormones, immunosuppression triggered by the tumor), and by accumulated exposure to the same or different genotoxic carcinogens.7,15
Analyses of how chemicals induce cancer in animal models and human populations has had a major impact on human health. Experimental studies have been instrumental in replicating hypotheses generated from human studies and identifying pathobiologic mechanisms. For example, animal experiments confirmed the carcinogenic and cocarcinogenic properties of cigarette smoke and identified bioactive chemical and gaseous components.1 The transplacental carcinogenicity of diethylstilbestrol and the hazards of specific occupational carcinogens such as vinyl chloride, benzene, aromatic amines, and bis(chloromethyl)ether led to a reduction in allowable exposures of suspected human carcinogens from the workplace and a reduction in cancer rates. Dietary factors that enhance or inhibit cancer development and the contribution of obesity to specific organ sites have been identified in models of chemical carcinogenesis, and alterations in diet and obesity are expected to result in reduced cancer risk. Experimental animal studies are the mainstay of risk assessment as a screening tool to identify potential carcinogens in the workplace and the environment, although these studies do not prove specific chemical etiologies as a cause of human cancer because of interspecies differences and the use of maximally tolerated doses that do not replicate human exposure.
THE NATURE OF CHEMICAL CARCINOGENS: CHEMISTRY AND METABOLISM
The National Toxicology Program, based mostly on experimental animal studies and supported by epidemiology studies when available, lists 45 chemical, physical, and infectious agents as known human carcinogens and about 175 that are reasonably anticipated to be human carcinogens (http://ntp.niehs.nih.gov/?objectid=035E57E7-BDD9-2D9B-AFB9D1CADC8D09C1), whereas the International Agency for Research on Cancer (IARC) lists 113 agents as carcinogenic to humans and 66 that are probably carcinogenic to humans (http://monographs.iarc.fr/ENG/Classification/index.php). Table 7.1 provides a selected list of known human carcinogens, as indicated by the IARC, which are continuously updated.16 Most chemical carcinogens first undergo metabolic activation by cytochrome P450s or other metabolic pathways so that they react with DNA and/or alter epigenetic mechanisms.11,17 This process, evolutionarily presumed to have been developed to rid the body of foreign chemicals for excretion, inadvertently generates reactive carcinogenic intermediates that can bind cellular molecules, including DNA, and cause mutations or other alterations.18 Recent data indicate that metabolizing enzymes also have the ability to cross-talk with transcription factors involved in the regulation of other metabolizing and antioxidant enzymes.19 DNA is considered the ultimate target for most carcinogens to cause either mutations or gross chromosomal changes, but epigenetic effects, such as altered DNA methylation and gene transcription, also promote carcinogenesis.20 The formation of DNA adducts, where chemicals bind directly to DNA to promote mutations, is likely necessary but not sufficient to cause cancer.
Genotoxic carcinogens may transfer simple alkyl or complexed (aryl) alkyl groups to specific sites on DNA bases.18,21 These alkylating and aryl-alkylating agents include, but are not limited to, N-nitroso compounds, aliphatic epoxides, aflatoxins, mustards, polycyclic aromatic hydrocarbons, and other combustion products of fossil fuels and vegetable matter. Others transfer arylamine residues to DNA, as exemplified by aryl aromatic amines, aminoazo dyes, and heterocyclic aromatic amines. For genotoxic carcinogens, the interaction with DNA is not random, and each class of agents reacts selectively with purine and pyrimidine targets.7,18,21 Furthermore, targeting carcinogens to particular sites in DNA is determined by nucleotide sequence, by host cell, and by selective DNA repair processes (see later discussion), making some genetic material at risk over others. As expected from this chemistry, genotoxic carcinogens can be potent mutagens and particularly adept at causing nucleotide base mispairing or small deletions, leading to missense or nonsense mutations. Others may cause macrogenetic damage, such as chromosome breaks and large deletions. In some cases, such genotoxic damage may result in changes in transcription and translation that affect protein levels or function, which in turn alter the behavior of the specific host cell type. For example, there may be effects on cell proliferation, programmed cell death, or DNA repair. This is best typified by the signature mutations detected in the p53 gene caused by ingested aflatoxin in human liver cancer22 and by polycyclic aromatic hydrocarbons human lung cancer caused by the inhalation of cigarette smoke.15,23,24 Similarly, a distinct pattern of mutations is detected in pancreatic cancers from smokers when compared with pancreatic cancers from nonsmokers.25
Some chemicals that cause cancers in laboratory rodents are not demonstrably genotoxic. In general, these agents are carcinogenic in laboratory animals at high doses and require prolonged exposure. Synthetic pesticides and herbicides fall within this group, as do a number of natural products that are ingested. The mechanism of action by nongenotoxic carcinogens is not well understood, and may be related in some cases to toxic cell death and regenerative hyperplasia. They may also induce endogenous mutagenic mechanisms through the production of free radicals, increasing rates of depurination, and the deamination of 5-methylcytosine. In other cases, nongenotoxic carcinogens may have hormonal effects on hormone-dependent tissues. For example, some pesticides, herbicides, and fungicides have endocrine-disrupting properties in experimental models, although the relation to human cancer risk is unknown.
ANIMAL MODEL SYSTEMS AND CHEMICAL CARCINOGENESIS
Most human chemical carcinogens can induce tumors in experimental animals; however, the tumors may not be in the same organ, the exposure pathways may differ from human exposure, and the causative mechanisms may not exist in humans. In many cases, however, the cell of origin, morphogenesis, phenotypic markers, and genetic alterations are qualitatively identical to corresponding human cancers. Furthermore, animal models have revealed the constancy of carcinogen–host interaction among mammalian species by reproducing organ-specific cancers in animals with chemicals identified as human carcinogens, such as coal tar and squamous cell carcinomas, vinyl chloride and hepatic angiosarcomas, aflatoxin and hepatocellular carcinoma, and aromatic amines and bladder cancer. The introduction of genetically modified mice designed to reproduce specific human cancer syndromes and precancer models has accelerated both the understanding of the contributions of chemicals to cancer causation and the identification of potential exogenous carcinogens.26,27 Furthermore, construction of mouse strains genetically altered to express human drug–metabolizing enzymes has added both to the relevance of mouse studies for understanding human carcinogen metabolism and the prediction of genotoxicity from suspected human carcinogens and other chemical exposures.28 Together, these studies have indicated that carcinogenic agents can directly activate oncogenes, inactivate tumor suppressor genes, and cause the genomic changes that are associated with autonomous growth, enhanced survival, and modified gene expression profiles that are required for the malignant phenotype.29
Genetic Susceptibility to Chemical Carcinogenesis in Experimental Animal Models
The use of inbred strains of rodents and spontaneous or genetically modified mutant strains have led to the identification and characterization of genes that modify risks for cancer development.30–32 For a variety of tissue sites, including the lungs, the liver, the breast, and the skin, pairs of inbred mice can differ by 100-fold in the risk for tumor development after carcinogen exposure. Genetically determined differences in the affinity for the aryl hydrocarbon hydroxylase (Ah) receptor or other differences in metabolic processing of carcinogens is one modifier that has a major impact on experimental and presumed human cancer risk.33–35 The development of mice reconstituted with components of the human carcinogen–metabolizing genome should facilitate the extrapolation of metabolic activity by human enzymes and cancer risk.27,28,36 Such mice also show that other loci regulate the growth of premalignant foci, the response to tumor promoters, the immune response to metastatic cells, and the basal proliferation rate of target cells.30 In mice susceptible to colon cancer due to a carcinogen-induced constitutive mutation in the APC gene, a locus on mouse chromosome 4 confers resistance to colon cancer.31 The identification of the phospholipase A2 gene at this locus and subsequent functional testing in transgenic mice revealed an interesting paracrine protective influence on tumor development.31 This gene, and several other genes mapped for susceptibility to chemically induced mouse tumors (PTPRJ, a receptor type tyrosine phosphatase, and STK6/STK15, an aurora kinase), have now been shown to influence susceptibility to organ-specific cancer induction in humans.30,31
MOLECULAR EPIDEMIOLOGY, CHEMICAL CARCINOGENESIS, AND CANCER RISK IN HUMAN POPULATIONS
Molecular epidemiology is the application of biologically based hypotheses using molecular and epidemiologic methods and measures. New technologies continue to allow epidemiologic studies to improve the testing of biologically based hypotheses and to develop large datasets for hypothesis generation, most notably the application of various –omics technologies via next-generation sequencing (e.g., genomics, epigenomics, transcriptomics), proteomics, and metabolomics. The greatest challenge now is to develop methods that allow for analysis cutting across various technologies.37–43 Recent advances now include the role of microRNA and long noncoding RNAs in tumor development and progression because of their impact on the regulation of gene expression.44,45 Chemical effects on microRNAs and the resultant gene expression is currently being identified.46 Using such technologies, emerging evidence is noting the importance of the microbiome and associated infections as a risk of human cancer.47–50 The complexity of environmental exposure and how it interacts with humans to affect numerous biologic pathways has been characterized as the exposome, also expressed as a multidimensional complex dataset.51Therefore, the important goal remains: to characterize cancer risk based on gene–environment interactions. However, we remain challenged because cancer is a complex disease of diverse etiologies by multiple exposures causing damage in different genes; for example, genen–environmentn interactions, for which the variable n is not known.
Two fundamental principles underlie current studies of molecular epidemiology. First, carcinogenesis is a multistage process, and behind each stage are numerous genetic events that occur either due to an exogenous insult such as a chemical exposure or an endogenous insult, such as from free radicals generated via cellular processes or errors in DNA replication. Therefore, identifying a cancer risk factor can be challenging because of the multifactorial nature of carcinogenesis, given that any one risk factor occurs within a background of many risk factors. Second, wide interindividual variation in response to carcinogen exposure and other carcinogenic processes indicate that the human response is not homogeneous, so that experimental models and epidemiology (e.g., the use of a single cell clone to study a gene’s effect experimentally or the assumption that the population responds similarly to the mean in epidemiology studies), might not be representative of susceptible and resistant groups within a population.
Genetic Susceptibility
In humans, the determination of genetic susceptibility can be assessed by phenotyping or genotyping methods. Phenotypes generally represent complex genotypes. Examples of phenotypes include the assessment of DNA repair capacity in cultured blood cells, mammographic breast density, or the quantitation of carcinogen-DNA adducts in a target organ. Phenotypes now also include profiles of methylation that affect gene expression, a so-called epigenetic effect, for example, identified though next-generation sequencing or other methods.52 The contribution of genetics to cancer risk from chemical carcinogens can range from small to large, depending on its penetrance.4 Highly penetrant cancer-susceptibility genes cause familial cancers, but account for less than 5% of all cancers. Low-penetrant genes cause common sporadic cancers, which have large public health consequences.
A genetic polymorphism (e.g., single nucleotide polymorphisms) is defined as a genetic variant present in at least 1% of the population. Because of the advent of improved genotyping methods that have reduced cost and increased high throughput, haplotyping and whole genomewide association studies are ongoing. Although haplotyping studies, facilitated through the International HapMap Project (www.hapmap.org), have not proven useful for predicting human cancers; high-density, whole genomewide, single nucleotide polymorphism association studies have shown remarkable consistency for many gene loci, although the risk estimates are only 1.0 to 1.4, which are not useful in the clinic for individual risk assessment.6 For example, the contribution of genetic polymorphisms to cancer risk, at least for breast cancer, appears to improve risk modeling by only a few percent; known breast cancer risk factors account for about 58% of risk, and adding 10 genetic variants increases the risk prediction only to 62%.53 Genes under study are from pathways that affect behavior, activate and detoxify carcinogens, affect DNA repair, govern cell-cycle control, trigger apoptosis, effect cell signaling, and so forth.
Biomarkers of Cancer Risk
The evaluation of dose and risk estimates in epidemiologic studies can include four components: namely, external exposure measurements, internal exposure measurements, biomarkers estimating the biologically effective dose, and biomarkers of effect or harm. The latter three measurements are biomarkers that improve on the first by quantifying exposure inside the individual and at the cellular level to characterize low-dose exposures in low-risk populations, providing a relative contribution of individual chemical carcinogens from complex mixtures, and/or estimating total burden of a particular exposure where there are many sources.54
Chemicals cause genetic damage in different ways, namely in the formation of carcinogen-DNA adducts leading to base mutations or gross chromosomal changes. Adducts are formed when a mutagen, or part of it, irreversibly binds to DNA so that it can cause a base substitution, insertion, or deletion during DNA replication. Gross chromosomal mutations are chromosome breaks, gaps, or translocations. The level of DNA damage is the biologically effective dose in a target organ, and reflects the net result of carcinogen exposure, activation, lack of detoxification, lack of effective DNA repair, and lack of programmed cell death. A variety of assays have been used for determining carcinogen-macromolecular adducts in human tissues; for example, for assessing risk from tobacco smoking for lung cancer and aflatoxin and liver cancer.55,56 Important considerations for the assessment of biomarkers include sensitivity, specificity, reproducibility, accessibility for human use, and whether it represents a risk measured in a target organ or surrogate tissue. No single biomarker has been considered to be sufficiently validated for use as a cancer risk marker in an individual as it relates to chemical carcinogenesis.57 However, there is some evidence that DNA adducts are cancer risk factors in both cohort and case-control studies.58
People are commonly exposed to N-nitrosamine and other N-nitroso compounds from dietary and tobacco exposures, which are associated with DNA adduct formation and cancer. Exposure can occur through endogenous formation of N-nitrosamines from nitrates in food or directly from dietary sources, cosmetics, drugs, household commodities, and tobacco smoke. Endogenous formation occurs in the stomach from the reaction of nitrosatable amines and nitrate (used as a preservative), which is converted to nitrites by bacteria. The N-nitrosamines undergo metabolic activation by cytochrome P450s (CYP2E1, CYP2A6, and CYP2D6) and form DNA adducts. Biomarkers are available to assess N-nitrosamine exposure from tobacco smoke (e.g., urinary tobacco-specific nitrosamine levels) or DNA, including in target organs such as the lungs. Recent data indicate that increasing levels of tobacco-specific nitrosamine metabolites are associated with increased lung cancer risk.55
Heterocyclic amines are formed from the overheating of food with creatine, such as meat, chicken, and fish.59 Heterocyclic amines, estimated based on consumption of well-done meat, have been associated with breast and colon cancer, presumably through metabolic activation mechanisms and DNA damage.59 Aflatoxins, another food contaminant, are considered to be a major contributor to liver cancer in China and parts of Africa, especially interacting with hepatitis viruses, and urinary aflatoxin adduct levels are predictors of liver cancer risk.56
Aromatic amines are another class of human carcinogens. Aryl aromatic amines have been implicated in bladder carcinogenesis, especially in occupationally exposed cohorts (e.g., dye workers) and tobacco smokers.60 These compounds are activated by cytochrome P4501A2 and excreted via the N-acetyltransferase 2 gene. They are genotoxic, and the quantitative assessment using biomarkers has been more difficult, but some persons have studied DNA adducts as well.61
Polycyclic aromatic hydrocarbons (PAH) are large, aromatic (three or more fused benzene rings) compounds that are a class of more than 200 chemicals. These compounds are ubiquitous in the environment and present in the ambient air. They are formed from overcooking foods, fireplaces, charcoal barbeques, burning of coal and crude oil, tobacco smoke, and can be found in various occupational settings. In order for PAHs to exert their toxic effect, they must undergo metabolic activation via cytochromes P4501A1 and P4503A4 to form DNA adducts, or are excreted via pathways involving the glutathione-S-transferase genes. PAHs are associated with an increased risk of lung and skin cancer in the occupational setting, although risk varies by type of industry and the individual being exposed.62,63 Benzo(a)pyrene (BaP), the most frequently studied PAH, serves as a model for chemical carcinogens. The bay region diol epoxide binds to DNA, mostly as the N2-deoxyguanosine adduct. The evidence linking BaP-deoxyguanosine adducts with a carcinogenic effect in lung cancer is very strong, including site-specific hotspot mutations in the p53 tumor suppressor gene.64–68 Various biomarkers of exposure have been developed for assessing PAH exposure. These include measuring DNA adducts, protein adducts, and urinary 1-hydroxypyrene; only the latter is a validated biomarker of exposure and no adducts have been validated as biomarkers of cancer risk. However, recent data indicate that PAH metabolites might be risk factors for lung cancer.58
Air pollution has been recently classified by the IARC as a known human lung carcinogen.69 Studies that support the conclusion include cohort studies that use biomarkers of exposure.70 Such markers include measurements of 1-hydroxypyrene, DNA adducts, chromosomal aberrations, micronuclei, oxidative damage to nucleobases, and methylation changes.71
Epidemiologic and experimental studies have linked benzene to hematologic toxicity, including aplastic anemia, myelodysplastic syndrome, and acute myeloid leukemia.72–74 Benzene is metabolized by hepatic P4502E1 (CYP2E1), yielding benzene oxide and hydroquinone, among other reactive metabolites. Circulating hydroquinones may be further metabolized to reactive benzoquinones by myeloperoxidase in bone marrow white blood cell precursors and stroma. Benzene metabolites are reported to have a variety of biologic consequences on bone marrow cells, including covalent binding to DNA and protein, alterations in gene expression, cytokine and chemokine abnormalities, and chromosomal aberrations.75 There are well-established biomarkers of exposure to benzene, but to date, biomarkers of toxicity have not been validated (except for high-level exposure workplaces and effects of peripheral blood counts).
ARISTOLOCHIC ACID AND UROTHELIAL CANCERS AS A MODEL FOR IDENTIFYING HUMAN CARCINOGENS
Aristolochic acids come from the Aristolochia genus of plants, which have been used for herbal remedies (e.g., birthwort, Dutchman’s pipe). The case of the carcinogen aristolochic acid, which is identified as a Class 1 human carcinogen by the IARC (http://monographs.iarc.fr/ENG/Monographs/vol100A/mono100A-23.pdf), presents a powerful example of how the forces of epidemiology, classical chemical carcinogenesis, and genomics collaborate to unravel the pathogenesis and prevention of a specific human cancer.76 In the 1990s, epidemiologists independently reported on three distinct unrelated population groups that developed nephrotoxicity (interstitial fibrosis) and an extraordinary high incidence of urothelial cancer of the upper urinary track after exposure for different reasons and in different parts of the world (Belgium, the Balkans, and China). In Belgian women ingesting an extract from plants of the Aristolochia species for weight reduction, which was provided to them in a weight loss clinic, nearly 50% developed this unusual syndrome. A similar clinical picture (so-called Balkan endemic nephropathy) was reported for residents farming around the Danube River and eating home-baked bread from wheat contaminated with seeds from Aristolochia weeds grown in the same fields. In China, the Aristolochia herbs have been used for centuries in Chinese medicine and are prominently prescribed in Taiwan, a nation with the highest incidence of urothelial cancer in the world, as remedies for ailments of the heart, liver, snake bites, arthritis, gout, childbirth, and others.
Common to all Aristolochia species are one of two major nitrophenanthrene carboxylic acid toxicants, namely, aristolochic acid I and II (http://monographs.iarc.fr/ENG/Monographs/vol100A/mono100A-23.pdf).77,78 The oral administration of aristolochic acid to rodents is highly carcinogenic, producing predominantly forestomach cancers and lymphomas, along with cancers of the lung, kidney, and urothelium (http://monographs.iarc.fr/ENG/Monographs/vol100A/mono100A-23.pdf). The major route of excretion of aristolochic acid is through the kidneys. These clinical and experimental observations inspired further analyses of the mechanism of action of these potent human carcinogens. Studies in intact mice and mice reconstituted with humanized P450 revealed that CYP1a and CYP2a were responsible for both the activation and the detoxification of aristolochic acid I and II, and that NAD(P)H:quinone oxidoreductase produced the ultimate reactive aristolactam I nitrenium species.78 The molecular action of the ultimate carcinogen is remarkably specific, targeting purine nucleotides in DNA to form DNA adducts and binding at the exocyclic amino group of deoxyadenosine and deoxyguanosine with a far greater affinity for dA over dG (Fig. 7.1). DNA adducts from aristolochic acids have been found in both experimental animals and humans. Furthermore, unlike any other human carcinogen, the predominant mutagenic outcome is an A:T transversion with a marked preference for the nontranscribed strand of DNA, notably in the p53 gene.77,79 The A:T to T:A transversions are extremely uncommon among the mutation spectrum in all eukaryotes. These unique properties of aristolochic acid DNA adducts appear to elude DNA repair mechanisms that commonly focus on transcribing DNA, resulting in persistent carcinogen-DNA adducts in human tissues and surgical tumor specimens, thus confirming the association of exposure with a biologic effect.80 In experimental models in mice where human p53 is substituted for the mouse gene, multiple sites on p53 are mutated, almost all of which are those unusual A:T transversions.81 Modern genomic techniques have unraveled other selective properties of this unusual but potent human chemical carcinogen. Whole genome and exome sequencing of multiple aristolochic-associated kidney cancers from patients confirmed the high frequency of the unusual A:T to T:A transversion mutations. Furthermore, an unusual pattern emerges where there is selectivity for mutations at splice sites with a preferable consensus sequence of T/CAG. Among the many mutations detected, certain targets stand out, particularly in p53, MLL2, and other genes the products of which function in regulating gene expression through higher chromosome order.82,83 This cancer story covers the gamut of all elements of chemical carcinogenesis, and its illumination has opened a door for cancer prevention.
REFERENCES
1. U.S. Department of Health and Human Services. The Health Consequences of Smoking: 50 Years of Progress. A Report of the Surgeon General. Atlanta: Author; 2014.
2. Colditz GA, Wei EK. Preventability of cancer: the relative contributions of biologic and social and physical environmental determinants of cancer mortality. Annu Rev Public Health 2012;33:137–156.
3. Lynch SM, Rebbeck TR. Bridging the gap between biologic, individual, and macroenvironmental factors in cancer: a multilevel approach. Cancer Epidemiol Biomarkers Prev 2013;22:485–495.
4. Rahman N. Realizing the promise of cancer predisposition genes. Nature 2014;505:302–308.
5. Lichtenstein P, Holm NV, Verkasalo PK, et al. Environmental and heritable factors in the causation of cancer—analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med 2000;343:78–85.
6. Hunter DJ, Chanock SJ. Genome-wide association studies and “the art of the soluble.” J Natl Cancer Inst 2010;102:836–837.
7. Luch A. Nature and nurture - lessons from chemical carcinogenesis. Nat Rev Cancer 2005;5:113–125.
8. Taddei ML, Giannoni E, Comito G, et al. Microenvironment and tumor cell plasticity: an easy way out. Cancer Lett 2013;341:80–96.
9. Fessler E, Dijkgraaf FE, De Sousa E Melo, et al. Cancer stem cell dynamics in tumor progression and metastasis: is the microenvironment to blame? Cancer Lett 2013;341:97–104.
10. Hanahan D, Coussens LM. Accessories to the crime: functions of cells recruited to the tumor microenvironment. Cancer Cell 2012;21:309–322.
11. Irigaray P, Belpomme D. Basic properties and molecular mechanisms of exogenous chemical carcinogens. Carcinogenesis 2010;31:135–148.
12. Xia HJ, Chen CS. Progress of non-human primate animal models of cancers. Dongwuxue Yanjiu 2011;32:70–80.
13. Yi C, He C. DNA repair by reversal of DNA damage. Cold Spring Harb Perspect Biol 2013;5:a012575.
14. Dizdaroglu M. Oxidatively induced DNA damage: mechanisms, repair and disease. Cancer Lett 2012;327:26–47.
15. Wogan GN, Hecht SS, Felton JS, et al. Environmental and chemical carcinogenesis. Semin Cancer Biol 2004;14:473–486.
16. Baan R, Grosse Y, Straif K, et al. A review of human carcinogens—Part F: chemical agents and related occupations. Lancet Oncol 2009;10:1143–1144.
17. Rendic S, Guengerich FP. Contributions of human enzymes in carcinogen metabolism. Chem Res Toxicol 2012;25:1316–1383.
18. Luch A. The mode of action of organic carcinogens on cellular structures. EXS 2006;65–95.
19. Anttila S, Raunio H, Hakkola J. Cytochrome P450-mediated pulmonary metabolism of carcinogens: regulation and cross-talk in lung carcinogenesis. Am J Respir Cell Mol Biol 2011;44:583–590.
20. Pogribny IP, Beland FA. DNA methylome alterations in chemical carcinogenesis. Cancer Lett 2012 [Epub ahead of print].
21. Shrivastav N, Li D, Essigmann JM. Chemical biology of mutagenesis and DNA repair: cellular responses to DNA alkylation. Carcinogenesis 2010;31:59–70.
22. Kew MC. Aflatoxins as a cause of hepatocellular carcinoma. J Gastrointestin Liver Dis 2013;22:305–310.
23. Feng Z, Hu W, Hu Y, et al. Acrolein is a major cigarette-related lung cancer agent: preferential binding at p53 mutational hotspots and inhibition of DNA repair. Proc Natl Acad Sci U S A 2006;103:15404–15409.
24. Porta M, Crous-Bou M, Wark PA, et al. Cigarette smoking and K-ras mutations in pancreas, lung and colorectal adenocarcinomas: etiopathogenic similarities, differences and paradoxes. Mutat Res 2009;682:83–93.
25. Blackford A, Parmigiani G, Kensler TW, et al. Genetic mutations associated with cigarette smoking in pancreatic cancer. Cancer Res 2009;69:3681–3688.
26. Eastmond DA, Vulimiri SV, French JE, et al. The use of genetically modified mice in cancer risk assessment: challenges and limitations. Crit Rev Toxicol 2013;43:611–631.
27. Boverhof DR, Chamberlain MP, Elcombe CR, et al. Transgenic animal models in toxicology: historical perspectives and future outlook. Toxicol Sci 2011;121:207–233.
28. Cheung C, Gonzalez FJ. Humanized mouse lines and their application for prediction of human drug metabolism and toxicological risk assessment. J Pharmacol Exp Ther 2008;327:288–299.
29. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 2011;144:646–674.
30. Demant P. Cancer susceptibility in the mouse: genetics, biology and implications for human cancer. Nat Rev Genet 2003;4:721–734.
31. Klatt P, Serrano M. Engineering cancer resistance in mice. Carcinogenesis 2003;24:817–826.
32. Lynch D, Svoboda J, Putta S, et al. Mouse skin models for carcinogenic hazard identification: utilities and challenges. Toxicol Pathol 2007;35:853–864.
33. Lash LH, Hines RN, Gonzalez FJ, et al. Genetics and susceptibility to toxic chemicals: do you (or should you) know your genetic profile? J Pharmacol Exp Ther 2003;305:403–409.
34. Di PG, Magno LA, Rios-Santos F. Glutathione S-transferases: an overview in cancer research. Expert Opin Drug Metab Toxicol 2010;6:153–170.
35. Feng S, Cao Z, Wang X. Role of aryl hydrocarbon receptor in cancer. Biochim Biophys Acta 2013;1836:197–210.
36. Jiang XL, Gonzalez FJ, Yu AM. Drug-metabolizing enzyme, transporter, and nuclear receptor genetically modified mouse models. Drug Metab Rev 2011;43:27–40.
37. Tuna M, Amos CI. Genomic sequencing in cancer. Cancer Lett 2013;340:161–170.
38. MacConaill LE. Existing and emerging technologies for tumor genomic profiling. J Clin Oncol 2013;31:1815–1824.
39. Watson IR, Takahashi K, Futreal PA, et al. Emerging patterns of somatic mutations in cancer. Nat Rev Genet 2013;14:703–718.
40. Dumas ME. Metabolome 2.0: quantitative genetics and network biology of metabolic phenotypes. Mol Biosyst 2012;8:2494–2502.
41. Adamski J, Suhre K. Metabolomics platforms for genome wide association studies—linking the genome to the metabolome. Curr Opin Biotechnol 2013;24:39–47.
42. Verma M, Khoury MJ, Ioannidis JP. Opportunities and challenges for selected emerging technologies in cancer epidemiology: mitochondrial, epigenomic, metabolomic, and telomerase profiling. Cancer Epidemiol Biomarkers Prev 2013;22:189–200.
43. Edwards SL, Beesley J, French JD, et al. Beyond GWASs: illuminating the dark road from association to function. Am J Hum Genet 2013;93:779–797.
44. Di LG, Garofalo M, Croce CM. MicroRNAs in cancer. Annu Rev Pathol 2014;9:287–314.
45. Cheetham SW, Gruhl F, Mattick JS, et al. Long noncoding RNAs and the genetics of cancer. Br J Cancer 2013;108:2419–2425.
46. Izzotti A, Pulliero A. The effects of environmental chemical carcinogens on the microRNA machinery. Int J Hyg Environ Health 2014 [Epub ahead of print].
47. Kostic AD, Gevers D, Pedamallu CS, et al. Genomic analysis identifies association of Fusobacterium with colorectal carcinoma. Genome Res 2012;22:292–298.
48. Compare D, Nardone G. Contribution of gut microbiota to colonic and extracolonic cancer development. Dig Dis 2011;29:554–561.
49. Ahn J, Chen CY, Hayes RB. Oral microbiome and oral and gastrointestinal cancer risk. Cancer Causes Control 2012;23:399–404.
50. Schwabe RF, Jobin C. The microbiome and cancer. Nat Rev Cancer 2013;13:800–812.
51. Wild CP, Scalbert A, Herceg Z. Measuring the exposome: a powerful basis for evaluating environmental exposures and cancer risk. Environ Mol Mutagen 2013;54:480–499.
52. Brennan K, Flanagan JM. Epigenetic epidemiology for cancer risk: harnessing germline epigenetic variation. Methods Mol Biol 2012;863:439–465.
53. Wacholder S, Hartge P, Prentice R, et al. Performance of common genetic variants in breast-cancer risk models. N Engl J Med 2010;362:986–993.
54. Boffetta P, van der Hel O, Norppa H, et al. Chromosomal aberrations and cancer risk: results of a cohort study from Central Europe. Am J Epidemiol 2007;165:36–43.
55. Yuan JM, Gao YT, Wang R, et al. Urinary levels of volatile organic carcinogen and toxicant biomarkers in relation to lung cancer development in smokers. Carcinogenesis 2012;33:804–809.
56. Wogan GN, Kensler TW, Groopman JD. Present and future directions of translational research on aflatoxin and hepatocellular carcinoma. A review. Food Addit Contam Part A Chem Anal Control Expo Risk Assess 2012;29:249–257.
57. Hatsukami DK, Benowitz NL, Rennard SI, et al. Biomarkers to assess the utility of potential reduced exposure tobacco products. Nicotine Tob Res 2006;8:599–622.
58. Yuan JM, Gao YT, Murphy SE, et al. Urinary levels of cigarette smoke constituent metabolites are prospectively associated with lung cancer development in smokers. Cancer Res 2011;71:6749–6757.
59. Turesky RJ, Le ML. Metabolism and biomarkers of heterocyclic aromatic amines in molecular epidemiology studies: lessons learned from aromatic amines. Chem Res Toxicol 2011;24:1169–1214.
60. Burger M, Catto JW, Dalbagni G, et al. Epidemiology and risk factors of urothelial bladder cancer. Eur Urol 2013;63:234–241.
61. Besaratinia A, Tommasi S. Genotoxicity of tobacco smoke-derived aromatic amines and bladder cancer: current state of knowledge and future research directions. FASEB J 2013;27:2090–2100.
62. International Agency for Research on Cancer. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans: Some Non-Heterocyclic Polycyclic Aromatic Hydrocarbons and Some Related Exposures. Volume 92. Lyon, France: World Health Organization; 2010.
63. Boffetta P, Autier P, Boniol M, et al. An estimate of cancers attributable to occupational exposures in France. J Occup Environ Med 2010;52:399–406.
64. Mordukhovich I, Rossner P Jr, Terry MB, et al. Associations between polycyclic aromatic hydrocarbon-related exposures and p53 mutations in breast tumors. Environ Health Perspect 2010;118:511–518.
65. Pfeifer GP, Denissenko MF, Olivier M, et al. Tobacco smoke carcinogens, DNA damage and p53 mutations in smoking-associated cancers. Oncogene 2002;21:7435–7451.
66. Pfeifer GP, Hainaut P. On the origin of G →. T transversions in lung cancer. Mutat Res 2003;526:39–43.
67. Sjaastad AK, Jorgensen RB, Svendsen K. Exposure to polycyclic aromatic hydrocarbons (PAHs), mutagenic aldehydes and particulate matter during pan frying of beefsteak. Occup Environ Med 2010;67:228–232.
68. Hussain SP, Amstad P, Raja K, et al. Mutability of p53 hotspot codons to benzo(a)pyrene diol epoxide (BPDE) and the frequency of p53 mutations in nontumorous human lung. Cancer Res 2001;61:6350–6355.
69. Loomis D, Grosse Y, Lauby-Secretan B, et al. The carcinogenicity of outdoor air pollution. Lancet Oncol 2013;14:1262–1263.
70. Raaschou-Nielsen O, Andersen ZJ, Beelen R, et al. Air pollution and lung cancer incidence in 17 European cohorts: prospective analyses from the European Study of Cohorts for Air Pollution Effects (ESCAPE). Lancet Oncol 2013;14:813–822.
71. Demetriou CA, Raaschou-Nielsen O, Loft S, et al. Biomarkers of ambient air pollution and lung cancer: a systematic review. Occup Environ Med 2012;69:619–627.
72. Galbraith D, Gross SA, Paustenbach D. Benzene and human health: A historical review and appraisal of associations with various diseases. Crit Rev Toxicol 2010;40:1–46.
73. Vlaanderen J, Portengen L, Rothman N, et al. Flexible meta-regression to assess the shape of the benzene-leukemia exposure-response curve. Environ Health Perspect 2010;118:526–532.
74. Vlaanderen J, Lan Q, Kromhout H, et al. Occupational benzene exposure and the risk of chronic myeloid leukemia: a meta-analysis of cohort studies incorporating study quality dimensions. Am J Ind Med 2012;55:779–785.
75. Snyder R. Leukemia and benzene. Int J Environ Res Public Health 2012;9:2875–2893.
76. Grollman AP. Aristolochic acid nephropathy: harbinger of a global iatrogenic disease. Environ Mol Mutagen 2013;54:1–7.
77. Hollstein M, Moriya M, Grollman AP, et al. Analysis of TP53 mutation spectra reveals the fingerprint of the potent environmental carcinogen, aristolochic acid. Mutat Res 2013;753:41–49.
78. Stiborova M, Martinek V, Frei E, et al. Enzymes metabolizing aristolochic acid and their contribution to the development of aristolochic acid nephropathy and urothelial cancer. Curr Drug Metab 2013;14:695–705.
79. Attaluri S, Bonala RR, Yang IY, et al. DNA adducts of aristolochic acid II: total synthesis and site-specific mutagenesis studies in mammalian cells. Nucleic Acids Res 2010;38:339–352.
80. Sidorenko VS, Yeo JE, Bonala RR, et al. Lack of recognition by global-genome nucleotide excision repair accounts for the high mutagenicity and persistence of aristolactam-DNA adducts. Nucleic Acids Res 2012;40:2494–2505.
81. Nedelko T, Arlt VM, Phillips DH, et al. TP53 mutation signature supports involvement of aristolochic acid in the aetiology of endemic nephropathy-associated tumours. Int J Cancer 2009;124:987–990.
82. Hoang ML, Chen CH, Sidorenko VS, et al. Mutational signature of aristolochic acid exposure as revealed by whole-exome sequencing. Sci Transl Med 2013;5:197ra102.
83. Poon SL, Pang ST, McPherson JR, et al. Genome-wide mutational signatures of aristolochic acid and its application as a screening tool. Sci Transl Med 2013;5:197ra101.