part 2. mechanisms of carcinogenesis chapter 11. Mechanisms of carcinogenesis: from initiation and promotion to the hallmarks Bernard W. Stewart PART 2 CH A PTER 11 Introduction malignancy by describing a tumour The types of biological agents and of as “an abnormal mass of tissue, the radiation now recognized by IARC For many decades, a corollary to growth of which exceeds and is un- as carcinogenic to humans (Group 1) the contemporary understanding of coordinated with that of the surround- are few compared with the number the nature of cancer and of carcino- ing tissue, and that continues to grow of chemicals in this category (IARC genesis has been the recognition of in the same excessive manner after 2012a, b, c, d, e, f); there is a much causative agents. Since the 1950s, cessation of the stimulus that caused larger number of chemicals for which many agents that contribute to the it”. According to the same textbook, at least some evidence of carcino- development of cancer have been development of tumours of the skin, genicity is available (see Volumes categorized as initiators or promot- the alimentary canal, or the respira- 1–105 of the IARC Monographs, ers, on the basis of studies of chem- tory tract was to be expected among available from http://publications. ical carcinogenesis in mouse skin individuals exposed “to various iarc.fr). (Berenblum and Shubik, 1947). noxious agents in the environment”. Research has established how Cancer was described with ref- Causation of cancer in humans or many carcinogenic chemicals cause, erence to causative agents. Thus, animals by certain chemicals, radia- or are likely to cause, malignant a 1970s pathology text (Cappell tion, and biological agents was rec- transformation, but the biological and Anderson, 1974) introduced ognized by early in the 20th century. processes involved are diverse, and Part 2 • Chapter 11. Mechanisms of carcinogenesis: from initiation and promotion to the hallmarks 93 Table 11.1. A selection of proposals for the categorization of chemical carcinogensa Mode of action Exposure context Chemistry Human relevance Agent type of bioassay data Genotoxic Tobacco smoke PAHs DNA binding Atmospheric pollutants Direct-acting Alcoholic N-nitroso PPARα activation Pesticides beverages compounds Pro-carcinogen Occupation Aromatic amines α2u-Globulin Organic solvents nephropathy Inorganic carcinogen Pollution Halogenated Urinary tract calculi Endocrine disruptors organic compounds Non-genotoxic Diet Naturally Disinfection occurring by-products compounds Solid-state carcinogen Pharmaceutical Inorganic Pharmacological drugs compounds steroids Hormone Exogenous hormones Immunosuppressant Promoter PAHs, polycyclic aromatic hydrocarbons; PPAR, peroxisome proliferator-activated receptor. a Knowledge about chemical carcinogens is presented from a variety of perspectives apart from that of mechanism of action. The listings indicate those used in particular publications (e.g. Searle, 1984; Tomatis et al., 1990; Vainio et al., 1992; Vainio and Hietanen, 2003; Hsu and Stedeford, 2010) as ways of ordering data, as indicated by chapter headings in many cases, and are not necessarily comprehensive. Categories shown in bold involve or include at least one Volume 100 (Group 1) agent. there is no generally accepted mech- categorization of chemical carcin- “hallmark” or “hallmarks” have been anistic basis for classifying chemi- ogens on the basis of the organ af- published. These papers typically cal carcinogens (Loeb and Harris, fected (Warshawsky and Landolph, describe signal transduction path- 2008), beyond categorization ac- 2006). ways and their therapeutic implica- cording to genotoxicity (Weisburger Currently, the most widely rec- tions. Although the characterization and Williams, 1981). There is no ognized description of the nature of by Hanahan and Weinberg (2011) single comprehensive basis for cat- cancer is that presented by Hanahan of the hallmarks of cancer did not egorization; chemical carcinogens and Weinberg in two reviews – pub- refer to chemical carcinogens or are sometimes ordered according lished more than a decade apart – causative agents in general, recent- to the context in which information is that identify the “hallmarks” of can- ly the hallmarks have been used to presented, with genotoxicity ordered cer (Hanahan and Weinberg, 2000, characterize chemical carcinogens according to mutational signatures, 2011). These papers have been so (Kleinstreuer et al., 2013). or agents categorized in relation to influential that others refer to “the These considerations give rise to differing classes of receptors. There hallmarks” without further qualifica- two questions: (i) whether previous- have been many proposals for the tion, for example in the title of a re- ly used mechanism-based descrip- categorization of chemical carcin- cent perspective on tumour metabol- tions of chemical carcinogens may ogens according to various crite- ism (Cantor and Sabatini, 2012). be recast in relation to the hallmarks; ria. A selection of these is shown Since 2000, about 200 cancer re- and (ii) whether, and to what extent, in Table 11.1; others include the search papers with a title including the hallmarks provide opportunities 94 to characterize agents apart from Morphological and genetic biological macromolecules had been currently known carcinogens as changes variously demonstrated over dec- contributing to the development of ades, carcinogen adducts in DNA cancer. Both of these matters are Within 20 years of the publica- were crucial. addressed in this chapter. tions cited above, the identification Alkylation of DNA by N-nitroso of multistage carcinogenesis with compounds was shown by Magee Multistage carcinogenesis particular carcinogens or other exog- and Farber (1962), with tumorigene- enous agents had become irrelevant sis attributable to the pro-mutagenic Exogenous agents to an understanding of cancer de- O6-methylguanine product, which The widely accepted paradigm of velopment. Over the same decades, mispairs with thymine. In rats, acti- carcinogenesis as involving a mul- the context in which carcinogenesis vation of H-Ras in mammary gland tistage process is generally recog- was best understood changed from tumours induced by N-methyl-N′- nized to have been developed from rodents to humans. Critical to this nitrosourea was correlated with the two-stage model of carcino- transition was the identification of H-Ras mutation at codons 12, 13, genesis in mouse skin (Berenblum multistage carcinogenesis with alter- and 61 (Sukumar et al., 1983). and Shubik, 1947), which typically ations in gene structure or expres- However, although this insight had involves a polycyclic aromatic hydro- sion rather than with the impact of been gained, it was clear that the eti- exogenous agents. carbon (PAH) and a phorbol ester ology of some types of cancer, such PART 2 (identified as the active agent in the A key development was the as breast cancer in humans, did not CH A PTER 11 irritant croton oil). Because tumor- correlation by Vogelstein et al. (1988) primarily involve alkylating agents. igenesis in animals is amenable to of morphological change during the Thus, in human cancer RAS activa- histological examination at all stag- development of colon cancer in hu- tion is a relevant genetic change in es, morphological criteria can be mans with particular genetic change. tumour tissue, without reference to used to characterize the process. The concept was applicable to all exogenous agents (Bos et al., 1987). With the production of malignant tu- tumour types. Thus, in a diagram il- Although the concept of multi- mours as the end-point, two-stage or lustrating multistage carcinogenesis stage carcinogenesis was estab- multistage carcinogenesis was read- with respect to human lung cancer, lished through the use of exogenous ily described in various organ sites in Harris (1992) made no reference to agents that target particular organ animals, including the liver and the any particular exogenous agents as sites in animals, by 1990 multistage bladder (Slaga et al., 1978). mediating specific stages in tumori- carcinogenesis was primarily iden- Thus, in relation to hepatocarcino- genesis, and showed the transitions tified with altered structure or ex- genesis, agents such as phenobarbi- between stages as being mediated pression of genes associated with tal, dichlorodiphenyltrichloroethane, by alterations in the structure or ex- cell proliferation, specifically as de- polychlorinated biphenyls, butylated pression of oncogenes and tumour scribed in human tumours. However, hydroxytoluene, and estradiol ben- suppressor genes. the focus of that research has not zoate were identified as promoters Oncogenes and tumour suppres- involved the specification of genetic (Dohi et al., 1996). The relevant sor genes mediate altered prolifera- change over time in a manner that experimental observations, in addi- tive activity in a positive and nega- might account for the emergence of a tion to indicating the possible risk to tive sense, respectively. Classically, metastatic cell population from with- humans presented by the relevant increased proliferative activity due in normal tissue. Rather, the relevant chemicals, also led to the contempo- to oncogene expression accounted research has involved the identifica- rary understanding of the nature of for the transformation of NIH 3T3 tion of disordered signal transduction malignancy itself. That understand- cells by DNA isolated from tumours pathways, with a view to developing