Author's personal copy

CHAPTER 5

Carcinogenesis: Mechanisms and Manifestations David E. Malarkey1, Mark Hoenerhoff1, Robert R. Maronpot2 1National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina, USA, 2Maronpot Consulting, LLC, Raleigh, North Carolina, USA

OUTLINE

1. Introduction 107 2.5. Gene Regulation 127 1.1. Overview 107 2.6. Cancer Stem Cell Theory 129 1.2. Background 109 2.7. Tumor Regression 133 1.3. Initiation, Promotion, and Progression 110 3. Identifying Carcinogens 134 1.4. Preneoplasia 115 3.1. Animal Bioassays 134 1.5. Hypertrophy 116 3.2. Data Evaluation and Interpretation 136 1.6. Metaplasia and Anaplasia (Atypia) 117 3.3. Genotoxic and Non-Genotoxic Carcinogens 139 1.7. Benign vs malignant neoplasms 117 3.4. Risk Assessment 140 2. Cancer is a Genetic Disease 117 3.5. Molecular Epidemiology 141 2.1. Overview 117 4. Conclusions 143 2.2. Oncogenes, Tumor Suppressors, Apoptosis and Repair Genes 120 Acknowledgments 144 2.3. Cell Proliferation and Apoptosis 124 Suggested Reading 144 2.4. Somatic Mutation Theory 125

We foresee cancer research as an increasingly exponentially increased risk for developing logical science, in which complexities are manifesta- cancer. There is a similar window of increased tions of a small set of underlying organizing princi- susceptibility to chemically-induced neoplasms ples. (Hanahan and Weinberg, 2011) in rodents treated with either a single or chronic exposure(s) to carcinogen(s). Identifying poten- 1. INTRODUCTION tial human carcinogens in rodent bioassays has been a major focus of the field of toxicologic 1.1. Overview pathology. It is estimated that 5% of human cancers are Cancer is a major cause of debilitation and caused by viruses, 5% by radiation, and the death in humans and animals. Cancer develops remaining 90% by chemicals. Of these, an esti- as a function of age, environment, diet, and mated 30% are caused by the use of tobacco genetic makeup, whether in man or animals. As products and the rest by chemicals associated humans reach their sixth decade, they face an with diet, lifestyle, and the environment. The

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition. http://dx.doi.org/10.1016/B978-0-12-415759-0.00005-4 107 Copyright Ó 2013 Elsevier Inc. All rights reserved.

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 108 5. CARCINOGENESIS importance of chemical products in the etiology of than 100 known human carcinogens (Table 5.1) cancer is reflected in the fact that up to 8% of all and virtually all of them also cause cancer in human cancers are related to occupational chem- animals. Animal and human cancers are funda- ical exposure. All chemical carcinogens, or their mentally similar and frequently share morpho- derivatives, are highly reactive electrophiles, logical, biological, and molecular biological which have electron-deficient atoms that can react features. In fact, approximately 30% of human with nucleophilic, electron-rich sites in the cell. carcinogens were first identified in animal Deoxyribonucleic acid (DNA), in particular, is studies. Two Japanese pathologists, Yamigawa made up of an array of nucleophilic centers at and Ichikawa, are credited with the original which these DNA-damaging agents can form demonstration that a chemical could produce adducts through one or more covalent bonds. cancer in animals. With chronic exposure of the To date, approximately 6 million chemicals skin (pinna) to coal tars, rabbits developed squa- have been identified and registered with the mous cell carcinomas, some of which metasta- chemical abstracts services. Of these, more than sized. These findings in 1918 confirmed Percival an estimated 50 000 are used regularly in Pott’s strong epidemiological observations in commerce and industry. Less than 2000, however, 1775 of increased rates of cutaneous scrotal have been examined for their carcinogenic cancer in chimney sweeps and demonstrated potential. that chronic exposures were necessary for the According to the International Agency of induction of some cancers. Research on Cancer (IARC), the United States Chronic inflammation and infectious agents Environmental Protection Agency (EPA), and have also been implicated in the development the United States National Toxicology Program’s of a number of human and animal cancers. Report on Carcinogens (RoC), there are more Nobel laureates Drs Robin Warren and Barry

TABLE 5.1 Classification of Carcinogens by the International Agency for Research on Cancer (IARC), United States (US) National Toxicology Program’s Report on Carcinogens (RoC), and the US Environmental Protection Agency (EPA)

Classification No. of agents IARC’s 2012 classification of carcinogens Group 1 Carcinogenic to humans 108 Group 2A Probably carcinogenic to humans 64 Group 2B Possibly carcinogenic to humans 271 Group 3 Not classifiable as to its carcinogenicity to humans 508 Group 4 Probably not carcinogenic to humans 1 NTP’s 2012 Report on Carcinogens (RoC) classification “Known to be human carcinogens” 56 “Reasonably anticipated to be human carcinogens” 189 US EPA classification [Integrated Risk Information System (IRIS)]a Group A Carcinogenic to humans 14 Group B Likely to be carcinogenic to humans 91 Group C Suggestive evidence of carcinogenic potential 9 Group D Inadequate information to assess carcinogenic potential 147 Group E Not likely to be carcinogenic to humans 7 a Includes guidelines from 1986 to 2005.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 1. INTRODUCTION 109

Marshall were credited with determining 5. inducing angiogenesis; that infection with Helicobacter pylori was 6. activating invasion and metastasis; a common cause of gastric inflammation and 7. reprogramming energy metabolism; and ulcers in man. Some skeptical scientists were 8. evading immune destruction. not convinced until Dr Marshall developed gastritis soon after drinking a Petri dish of the bacteria. 1.2. Background In subsequent years it was shown that chronic helicobacter gastritis is associated with the It is still accepted today that cancer originates development of gastric lymphomas and carci- in a single cell and develops through the clonal nomas, and thereby Helicobacter pylori is listed proliferation of their progeny. However, the as a human carcinogen. Some of the lymphomas search for a comprehensive theory of carcinogen- appear reversible and regress after treatment esis has not been forthcoming. Investigation into with antibiotics. Certain strains of mice the two predominant and antagonistic theories: (including A/JCr and B6C3F1/N) with chronic humoralist and cellular, dominated most of the Helicobacter hepaticus hepatitis develop signifi- second half of the 19th century. cantly higher rates of liver cancer compared to The humoralist view regarded cancer as origi- uninfected controls. Helicobacter infections may nating from certain hereditary characteristics of confound interpretation of the results of cancer the individual associated with susceptibility to occurrence in animal bioassays or human epide- contract the disease. Cellular pathologists such miological studies. as Muller and Virchow, however, argued that Our understanding of cancer biology is cancer was related to a form of chronic irritation. evolving at a rapid rate. Conceptual views of This latter view was supported by experimental carcinogenesis are formed by the piece-by-piece studies in mouse skin where wounding seemed discovery of key elements of the complex biolog- to play a tumorigenic role. Over the next century, ical puzzle that this disease entails. This piece- thousands of chemicals have been shown to by-piece discovery includes early findings of transform cells in vitro and to be carcinogenic the evidence of clonal evolution of cancer, the in animals. Those that are known to be carcino- Knudsen two-hit hypothesis progression from genic in humans are many times fewer (see benign to malignant growth, discovery of onco- Tables 5.1 and 5.2). genes and tumor suppressor genes, the somatic Some of the most potent carcinogens have mutation theory, the Fearon-Vogelstein multistep been extracted from fossil fuels or are synthetic colon cancer mutation model, mutator pheno- products created by industry. However, a variety type, and the cancer stem-cell theory. of occupational causes of cancer had been docu- With the advent of new technologies in molec- mented prior to the Industrial Revolution. In ular analysis, such as gene expression profiling, 1531, Paracelsus had described the “mala metal- networks, microRNAs, gene discovery, and lorum” among miners for silver and other pathway analysis, carcinogenesis is proving to metals, including uranium. This observation be much more complex than being simply a clonal was later interpreted as radiation-induced lung evolution of a cell that sustained two genetic cancer. In 1775, Pott attributed scrotal skin “hits” by a carcinogen. The current multistep cancers to prolonged exposure to soot in model of carcinogenesis involves at least 80 chimney sweeps. A few years later, on the basis cancer gene mutations or alterations, about of this observation, the Danish Chimney Sweeps a dozen of which are “drivers” of the cancer Guild ruled that its members must bathe daily. growth processes. The hallmarks of carcinogen- No public health measure since that time has so esis (Hanahan and Weinberg, 2011) include successfully controlled a form of cancer. genetic alterations involved in: Cancers related to industrial activity were reported early in the industrial age in several 1. sustaining proliferative signaling; locations with increasing frequency. Examples 2. evading growth suppressors; include the “aniline” bladder cancer related to 3. resisting cell death; the systemic effects of naphthylamines and the 4. enabling replicative immortality; “paraffin cancer” caused by exposure to shale

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 110 5. CARCINOGENESIS

TABLE 5.2 Selected Agents (of more than 100) Classified in the environment as a result of cigarette smoke as Human Carcinogens by the International and automobile exhaust fumes. It became the Agency for Research on Cancer (IARC) and most intensely studied chemical carcinogen the National Toxicology Program’s Report on because of the belief that its chemical structure Carcinogens (RoC) was related to that of steroid hormones. Steroidal hormones were at the time the only Acetylaldehyde endogenous compounds reported to induce Aflatoxin tumors in mice and humans. It was shown later Arsenic that compounds with many different chemical structures could induce neoplasia. It also Asbestos became clear that age and individual differences Benzene were contributing factors in the susceptibility to cancer, and these differences, presumably of Benzo[a]pyrene genetic origin, could be inherited. 1,3 Butadiene Cadmium 1.3. Initiation, Promotion, and Progression Chromium VI In early studies, it was observed that a long Coal tar latent period could elapse from exposure to carcinogens to the development of cancer. In Dioxins 1941, Rous and Kidd painted the skin of rabbits Estrogens and found that if this painting were interrupted, tumors would disappear only to reappear if the Ethanol application of tar was re-established. It seemed, Helicobacter pylori therefore, that a reversible process was taking place in those cells that did not attain the Ionizing radiation complete neoplastic state. Formaldehyde These cells had undergone what Rous called initiation. Further development of tumors would Radon then require what was termed promotion, the Vinyl chloride process by which the initiated cell expands clon- ally into a detectable cell mass that is either benign or preneoplastic. Finally, cells must undergo additional changes in their progression oil. The interest in industrial carcinogens coex- to a malignant neoplasm. Today it is known isted with the realization that a definitive that for a normal cell to evolve into a malignant cause–effect relationship existed between tobacco one, heritable changes involving multiple, inde- and cancer. pendent genes are required. This “multi-hit” Early reports by Hill in 1761, which called model is consistent with the incidence rates of attention to the association of “immoderate use cancer that increase exponentially with age. of snuff” and the development of “polypusses,” The concepts of initiation and promotion were later received experimental confirmation by first described during experiments on mouse Roffo in 1931 when he induced skin cancer in skin carcinogenesis, but have since been applied rabbits by painting with tobacco-derived tar. to a variety of other tissues and species. During However, tobacco tar was shown to be only the initiation phase of chemical carcinogenesis, weakly carcinogenic in experimental models. a normal cell undergoes an irreversible change Thus, a serious effort was undertaken to identify characterized by an intrinsic capacity for autono- potential carcinogens using the mouse skin mous growth. This capacity for autonomous experimental model. growth remains latent for weeks, months, or Polycyclic aromatic hydrocarbons were the years, during which time the initiated cell may initial target of research. Benzo[a]pyrene, the be phenotypically indistinguishable from other most potent carcinogenic agent of tar, is present parenchymal cells in that tissue.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 1. INTRODUCTION 111

Spontaneous initiation refers to a situation in such as identification of cell surface immuno- which the exact agent responsible for the initi- globulin markers and glucose-6-phosphatase ating event is unknown. One possibility that dehydrogenase variants, restriction fragment can lead to spontaneous initiation would be length polymorphisms, cytogenetic studies, a situation wherein there is infidelity in the single-cell transplantation studies, and identifi- action of DNA polymerase during normal cation of chromosome inactivation mosaics has cellular division or during DNA repair. Initiation permitted identification of the monoclonal and operationally implies that there is alteration to polyclonal nature of individual neoplasms. cellular DNA at one or more sites in the genome. The vast majority of human and animal Such alteration represents a mutational event, neoplasms studied to date are of monoclonal which can be hereditary. origin. There are several salient characteristics Metabolic activation of a carcinogen to its of initiation. It can occur following a single expo- chemically reactive products and their subse- sure to a known carcinogen. Changes produced quent reaction with cellular targets (e.g., DNA by the initiator may be latent for weeks or bases) occurs within a few hours of exposure. months, and are considered irreversible. The Most tissues have the ability to repair this interval between initiation and promotion damage over a period of days or weeks. may be as long as 1 year in mouse skin-painting Currently accepted dogma suggests that the studies. chemically damaged DNA, if not first repaired Several lines of evidence indicate that initia- by normal cellular processes, is converted to tion is additive and that the yield of neoplasms a stable biological lesion (mutation, chromo- is dose-dependent. Increasing the dose of initi- somal rearrangement, etc.) during DNA replica- ator increases the incidence and multiplicity of tion. Thus, if a round of cell replication occurs resulting neoplasms and shortens the latency to before the DNA damage is repaired, the lesion manifestation of neoplasms. Because the initi- in the DNA is regarded as “fixed.” This phenom- ating event must be “fixed” by a round of cell enon may explain the high frequency of proliferation, it becomes obvious that initiation neoplasms in proliferating tissue where there is is dependent on the cell cycle. an intrinsically high rate of cell turnover coinci- Finally,there is no readily measurable threshold dent with exposure to a DNA-damaging agent. dose for maximum and minimum responses In contrast to the step of initiation, the to initiators. Unrealistically large numbers of conversion of an initiated cell to a fully malig- animals would be required to demonstrate nant neoplasm is usually a prolonged process, minimum responses, and confluence of multiple lasting months in animals and years in neoplasms following high doses precludes accu- humans. Based on the hypothesis that most rate quantitation of the neoplastic response. Prop- initiators are mutagenic or genotoxic, a battery erties of initiators are summarized in Table 5.3. of short-term in vitro and in vivo mutagenicity Initiators interact with host cellular macro- tests have been developed to permit the detec- molecules and nucleic acids in specific patterns, tion of chemicals with potential initiating typically involving the generation of reactive activity. Identification of initiating agents is electrophiles, esters, or free radicals that bind especially important because of the irreversible covalently to nucleophilic sites in critical cellular and hereditary nature of the alterations that macromolecules. The majority of known carcino- occur during initiation. While useful when posi- gens have both initiating and promoting activity tive results are obtained, the predictive ability of and can thus induce neoplasms rapidly and in short-term mutagenicity tests for the ultimate high yield when given repeatedly. When given carcinogenic potential of xenobiotics is not at sufficiently low single doses, even a complete absolute. carcinogen may act as a “pure” initiator Exposure of experimental animals to chemi- requiring subsequent promotion for the detec- cals with initiating activity may ultimately result tion of any neoplasms. Under such circum- in the induction of multiple neoplasms in a given stances, the agent can be regarded operationally tissue. Each individual neoplasm is often found as an “incomplete carcinogen.” A cell that has to be monoclonal in origin, having arisen from undergone the irreversible change that permits a single initiated cell. Application of techniques its ultimate neoplastic transformation may be

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 112 5. CARCINOGENESIS

TABLE 5.3 General Characteristics of Initiators and agents themselves are so classified only in an Promoters of Neoplasia operational sense. INITIATORS/INITIATION A promoter is typically given at some time after chemically-induced or fortuitous initiation, Irreversible and the doses of agent used are insufficient to Additive produce cancer without prior initiation. It should be appreciated that when classical promoters are Cannot identify cells administered at sufficiently high doses and for “Pure” initiation does not result in neoplasia unless prolonged intervals, neoplasia can occur without promoter is subsequently applied evidence of prior initiation. Under these condi- tions, a promoting agent must be considered Number of initiated cells dependent on dose a complete carcinogen unless fortuitous initiation No measurable threshold dose from background radiation, dietary contami- nants, environmental toxins, and so on is Agents are considered carcinogens believed to have occurred. However, under Must be administered before the promoter typical experimental conditions commonly Only one exposure may suffice employed in short- and medium-term initia- tion–promotion experiments, neoplasia does Electrophile production and covalent binding to DNA not typically occur in animals that have not Agents usually mutagenic been previously initiated. The temporal PROMOTERS/PROMOTION sequence of promoter administration is critical to the operational definition of promotion. The Reversible agent must be administered after initiation and Non-additive cause enhancement of the neoplastic process to be considered a promoter. If an agent is given Agents not capable of initiation simultaneously with an initiator and results in Modulated by diet, hormonal, environmental, and enhancement of the development of neoplasms, related factors it is regarded as a co-carcinogen rather than a promoter. Measurable threshold dose While some promoters such as phorbol esters Agents not considered carcinogens but co-carcinogens are co-carcinogenic, not all promoters (e.g., phenobarbital, phenol) possess co-carcinogenic- Must be administered after the initiator ity and, conversely, not all co-carcinogens are Prolonged exposure is usually required promoters. Under these same conditions of simultaneous administration, a diminution in No electrophile production and no covalent binding to DNA the neoplasm response is considered evidence of anticarcinogenic activity. Thus, several rodent Agents usually not mutagenic liver tumor promoters, which are active when Reproduced from Handbook of Toxicologic Pathology, 2nd Ed. given after a variety of initiators, prevent or W. M Haschek, C. G. Rousseaux and M. A. Wallig, eds. (2002) delay the development of liver neoplasms Academic Press, Vol. I, Table V, p. 94, with permission. when added to diets along with an active carcinogen. Finally, reversing the order of administration phenotypically indistinguishable from adjacent by giving a known promoter prior to an initiator normal parenchymal cells. However, when stim- may prevent the expression of carcinogenic ulated properly, it has an intrinsic capacity for activity on the part of the initiator. While upper autonomous growth. and lower thresholds have been demonstrated Promotion is classically considered that experimentally for promoters, some consider portion of the multistep carcinogenic process that, in an absolute sense, it is statistically impos- where specific agents, known as promoters, sible unequivocally to prove or disprove the exis- enhance the development of neoplasms from tence of thresholds for promoters for much the a background of initiated cells. The promoting same reasons that this cannot be done for

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 1. INTRODUCTION 113 initiators. One can never be certain that an practical definition of a promoter must include apparent no observable effect level (NOEL) the designation of the susceptible tissue. would indeed be without effect if a sufficiently Tumor promotion may be modulated by several large enough number of animals were used. factors, such as age, sex, diet, and hormone Promoters include agents such as drugs, balance. The correlation of increased rates of plant products, and hormones that do not breast cancer in women following a “western” interact directly with host cellular DNA (are life style has implicated meat and fat consumption not genotoxic) but somehow influence the as playing an important role in breast cancer expression of genetic information encoded in development. Experimental demonstration of the the cellular DNA (e.g., epigenetics). It has role of a high-fat diet in the promotion of been suggested that promoting agents may mammary cancer in rats exposed to the mammary cause gene repression and derepression in cells. carcinogen dimethylbenzanthracene (DMBA) has Some experimental evidence suggests that the been documented. regulation of gene expression is unique to the Similarly, bile acids, as modulated by fat nature of the promoting agent administered. consumption, are known promoters of rat liver Some promoters are believed to produce their carcinogenesis. Age- and sex-associated modula- effect by interaction with receptors in the cell tions in hormonal levels of estrogens, proges- membrane, cytoplasm, or nucleus (e.g., terone, and androgens have been implicated as hormones, dioxin, phorbol ester, polychlori- potential promoters of breast cancer on the basis nated biphenyls). Alternatively, some hydro- of epidemiology studies in humans. Experi- philic and hydrophobic promoting agents exert mental studies have shown repeatedly that these their effect through their molecular orientation hormones, in addition to pituitary prolactin, at cellular interfaces. Other promoters are mito- serve to promote mammary cancer in rats initi- genic, stimulating DNA synthesis and enhanced ated with mammary carcinogens. cell proliferation. This may occur directly or, Some promoters cause hyperplasia and/or alternatively, indirectly by targeting cells with inflammation. This is particularly true in skin a shortened G1 phase, thereby giving them initiation–promotion studies using phorbol esters a selective proliferative advantage. At least in as promoters, but is also seen in hyperplasia of tissue culture, some promoters have been hepatocytes following treatment with mitogenic shown to inhibit intercellular communication agents such as phenobarbital. In the rodent liver, (metabolic cooperation). In some situations, phenobarbital causes a transient hyperplasia of metabolism appears to play little role in the hepatocytes. It should be remembered that action of promoters. some materials could cause hyperplasia and Experimental evidence suggests that the mole- inflammation but be without promoting effects. cule as a whole may exert the promotional effect This observation has led some investigators to and that the molecular configuration determines consider that the ability to stimulate DNA the activity of the agent. When promoter metabo- synthesis and cell division and the ability to lism does occur, it typically results in inactivation induce inflammation are essential but not suffi- of the agent. Possible exceptions include d-limo- cient properties of promoters. Properties of nene and trimethyl pentane (unleaded gasoline), promoters are summarized in Table 5.3. It has which are promoters of renal tumors in rats. been suggested and confirmed experimentally Promoters appear to have relatively high tissue that the process of promotion can be divided specificity. Thus, phenobarbital functions as into at least two stages in the mouse skin- a promoter for rodent liver neoplasia but not in painting initiation–promotion model. The the urinary bladder. In contrast, 12-0-tetradeca- two-stage promotion model is based on the stim- noylphorbol-13-acetate is a potent skin and ulation of basal cell hyperplasia (dark basal forestomach neoplasm promoter, but has no keratinocytes) in stage 1 of promotion followed appreciable activity in the liver. Antioxidants by enhancement of cell proliferation in stage 2. such as 3-tert-butyl-4-methoxyphenol and 2,6-di- In the classical mouse skin carcinogenesis tert-butyl-4-methoxyphenol may act as promoters model, a phorbol ester binds to a membrane in one organ, anti-promoters in another organ, receptor, induces protein kinase C (PKC), and is and have no effect in a third organ. Thus, the effective with only one application, the process

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 114 5. CARCINOGENESIS being at least partially irreversible, in the first matter of time before the emergence of a subpop- stage of promotion. Weak or non-promoting ulation with more malignant biological charac- agents such as mezerein are effective as second- teristics or at least a differential growth stage promoters, but require multiple applica- advantage. This can occasionally be observed tions and do not always have receptor-binding during early stages of experimental hepatocarci- properties. Similar multistage promotion has nogenesis when phenotypically distinguishable not yet been demonstrated in other experimental neoplasms can arise within existing foci of carcinogenesis model systems. altered hepatocytes. Progression is that part of the multistep In addition, serial transplantation of tumors neoplastic process associated with the develop- frequently shows an enhancement of malignant ment of an initiated cell into a biologically malig- properties associated with increased numbers nant cell population. In common usage, of passages. Here there is a presumed selection progression is used frequently to signify the for more rapidly growing subpopulations stages whereby a benign proliferation becomes because the subsequent serial passage is typi- malignant or where a neoplasm develops from cally dictated by the size of the transplant growth a low grade to a high grade of malignancy. in the recipient. During progression, neoplasms show progres- Associated with progression is the develop- sively increased invasiveness, develop the ment of an increased degree of karyotypic ability to metastasize, and have alterations instability and of aneuploidy. This latter in biochemical, metabolic, and morphologic phenomenon may be related to the not infre- characteristics. quent observation of abnormal mitoses in malig- Tumor cell heterogeneity is an important char- nant neoplasms. Finally, chromosomal acteristic of tumor progression. Expression of rearrangement is associated with several clini- this heterogeneity includes antigenic and protein cally malignant neoplasms, especially leukemias. product variants, ability to elaborate angiogen- It is probable that such rearrangements are esis factors, emergence of chromosomal variants, a consequence of karyotypic instability and that development of metastatic capability, altered apposition of critical portions of genes upstream metabolism, and decreased sensitivity to radia- or downstream from genomic enhancers or dere- tion. The development of intraneoplastic diver- pressors imparts a proliferative advantage and sity may come about as a consequence of a metastatic capability to affected cells within genetic change such as loss of polymorphic an evolving neoplastic proliferation. restriction fragments in DNA of malignant Distinction between tumor promotion and tumors or similar random processes such as tumor progression is not readily discernible in additional genomic “hits” by genotoxic agents. the routine histopathologic evaluation of neop- Alternatively, the heterogeneity observed in lasms. In fact, the distinction may be somewhat tumor progression may be generated by epige- academic in that promotion may be considered netic, regulatory mechanisms operative as part of the process of progression. In both situa- a continuation of the process of promotion. tions the critical event is accentuated growth. More than likely, genetic and epigenetic events Distinguishing progression from promotion is subsequent to initiation operate in a non- the presence of structural genomic alterations in mutually exclusive manner during progression, the former and the absence of definable struc- possibly in an ordered cascade of latter events tural changes in the genome of the latter. Both superimposed on earlier events. structural genomic changes and biochemical The most plausible mechanism of progression changes associated with tumor progression invokes the notion that during the process of cannot be defined by conventional histopa- tumor growth there is a selection that favors thology. Emerging technologies centered on enhanced growth of a subpopulation of the histochemistry, immunocytochemistry, and in neoplastic cells. In support of this mechanism is situ hybridization to identify products of proto- the observation of increased phenotypic hetero- oncogenes and activated oncogenes offer geneity that is observed in malignant versus promise to help distinguish various stages of benign neoplastic proliferations. Presumably progression in the evolution from benign to a variety of subpopulations arise and it is only a malignant neoplasms.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 1. INTRODUCTION 115

1.4. Preneoplasia TABLE 5.4 Examples of Presumptive Preneoplastic Lesions

Most neoplasms are believed to be derived a from the clonal proliferation of a single initiated Tissue Presumptive preneoplastic lesion cell. Usually at some point early in the clonal Mammary gland Hyperplastic alveolar nodules expansion, the differentially proliferating cells may become phenotypically distinguishable Atypical epithelial proliferation from the surrounding normal parenchyma. Lobular hyperplasia Although such lesions may not as yet have suffi- Intraductal hyperplasia cient characteristics to qualify as neoplasms, their recognition has led many to regard them Hyperplastic terminal duct as “preneoplastic.” According to the multistage Liver Focus of altered hepatocytes model of carcinogenesis, there is a morphological continuum from hyperplasia/preneoplasia to Hepatocellular hyperplasia adenoma and carcinoma, which is reflected in Oval cell proliferation the pattern observed in tumor types. Patients or animals with preneoplastic lesions Cholangiofibrosis are at increased risk of developing neoplasia at Kidney Karyocytomegaly the tissue site in which the preneoplasia is present. The preneoplastic lesions themselves are believed Atypical tubular dilation to progress to neoplasia, although unequivocal Atypical tubular hyperplasia proof for this is difficult to document. Examples Skin Increase in dark basal keratinocytes of preneoplasia from human medicine include leukoplakia of the oral cavity or vulva, senile kera- Focal hyperplasia/hyperkeratosis tosis, and xeroderma pigmentosum. Numerous Pancreas (exocrine) Focus of acinar cell alteration putative preneoplastic lesions have been identi- fied in laboratory animals (Table 5.4). Hyperplastic nodules Awell-studied example of preneoplasia is seen Colon Atypical acinar cell nodules in experimental studies of liver neoplasia using Aberrant crypt foci rats exposed to potent hepatocarcinogens. The initial change detected in liver tissue consists of Hyperplasia foci of cellular alteration. These foci consist of a Many of these presumptive preneoplastic lesions are seen in nests or islands of altered hepatocytes that differ carcinogenicity studies utilizing specific animal model systems. phenotypically from adjacent normal hepato- Generalizations about these presumptive preneoplastic lesions are cytes due to the tinctorial quality of the cyto- inappropriate outside the context of the specific animal model system being used. plasm. The foci of altered hepatocytes (FAH) Reproduced from Handbook of Toxicologic Pathology, 2nd Ed. often precede the development of hepatocellular W. M Haschek, C. G. Rousseaux and M. A. Wallig, eds. (2002) neoplasms. Academic Press, Vol. I, Table II, p. 88, with permission. There are several phenotypic types of FAH, including eosinophilic, basophilic, clear cell, The exact significance of FAH is unknown. and mixed, and the ultimate neoplasms may Empirical observations speak both for and resemble them phenotypically. Because of their against their role in oncogenesis. On the one consistent production by known hepatocarcino- hand, FAH occur spontaneously in aging gens and their temporal relationship with ulti- untreated rats that have a negligible background mate neoplasia, FAH are regarded operationally incidence of hepatic neoplasia. Also, in some as preneoplastic lesions. Because the number of experimental models, the stability or persistence neoplasms eventually generated represents of foci is dependent on continued exposure to a very small proportion of the number of foci a xenobiotic agent or feed ingredient. On the produced (estimates range from 1 neoplasm other hand, the consistent induction of foci by for every 1000 to 10 000 foci), conservative known hepatocarcinogens and the occasional pathologists regard the foci as “putatively observation of “neoplasia within a focus” preneoplastic.” suggest that these putatively preneoplastic

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 116 5. CARCINOGENESIS lesions are related to hepatic neoplasia. This Exposure of animals to liver enzyme inducers finding may simply reflect that additional critical results in a signature of toxicological changes genetic perturbations to undergo neoplastic characterized by an increase in liver weight, transformation have occurred. Despite the hepatocellular hypertrophy, cell proliferation, absence of definitive proof that hepatic foci of and (frequently in chronic [2-year] studies) hep- cellular alteration are precursor lesions for atocarcinogenesis (Liver and Gallbladder, Chapter hepatic neoplasms, several investigators have 45). Hepatic enzyme induction is generally an identified chemicals as having hepatocarcino- adaptive response associated with increases in genic activity based only on induction of foci liver weight, induction of gene expression, and by these chemicals. It has been proposed that morphological changes in hepatocytes. The regulatory decisions regarding the production, additive growth and functional demands that distribution, and use of chemicals can be based initiated the response to hepatic enzyme induc- on the observation that they induce FAH in rat tion cover a wide range of stimuli, including carcinogenicity studies. pregnancy and lactation, hormonal fluctuations, dietary constituents, infections associated with acute phase proteins, as well as responses to 1.5. Hypertrophy exposure to xenobiotics. Common xenobiotic While not technically a proliferative change, enzyme inducers trigger pathways involving hypertrophy deserves mention because it is the constitutive androstane receptor (CAR), sometimes diagnosed incorrectly as hyperplasia the peroxisome proliferator-activated receptor (Morphologic Manifestations of Toxic Cell Injury, (PPAR), the aryl hydrocarbon receptor (AhR), Chapter 4). Hypertrophy and hyperplasia may and the pregnane-X-receptor (PXR). Liver occur together, and several factors link hyper- enlargement in response to hepatic enzyme trophy to rodent hepatocarcinogenesis (Table induction is typically associated with hepatocel- 5.5). Hypertrophy is under various regulatory lular hypertrophy and often with transient controls, and thus is limited in amount and hepatocyte hyperplasia. The hypertrophy may duration. Hypertrophy may be classified in show a lobular distribution, with the pattern a manner similar to how hyperplasia is classi- of lobular zonation and severity reflecting fied. Compensatory or adaptive hypertrophy species, strain, and gender differences in addi- represents a physiological response to a stim- tion to effects from specific xenobiotics. Toxicity ulus, such as is seen with muscle hypertrophy and hepatocarcinogenicity may occur when subsequent to prolonged exercise or in enzyme liver responses exceed adaptive changes or induction in the liver following exposure to induced enzymes generate toxic metabolites chemical inducers such as phenobarbital. (Table 5.5). These undesirable consequences Combined with degenerative changes, such as are influenced by the type and dose of xenobi- necrosis and vacuolization, hepatocyte hyper- otics, and show considerable species differences trophy is associated with the development of in susceptibility and severity that need to be hepatocellular neoplasms in rats and mice. understood for assessing the potential effects on human health from similar exposures to TABLE 5.5 Factors Linking Hepatocyte Hypertrophy to specific xenobiotics. Liver Cancer Induction in the Rat and Mouse Hormones can also induce hypertrophy. For example, injection of growth hormone from the Enzyme induction of > 140% anterior pituitary induces hypertrophy of liver Liver weight increase of > 150% at 1 year cells, which have an increase in their RNA con- tent. Whether the various types of hypertrophy Induction of cytochrome p450s that elevate oxygen are considered physiological, adaptive, or patho- radicals and reactive electrophiles capable of causing logical depends on the philosophy of the person DNA damage making the judgment. For example, because Hepatocyte hyperplasia and degeneration (necrosis hypertension is a disease, then cardiac hyper- and/or vacuolization) trophy that occurs secondary to the hypertension can be considered pathologic. However, one could Polyploidy and aneuploidy argue that cardiac hypertrophy is an adaptive

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 2. CANCER IS A GENETIC DISEASE 117 physiological response to an increased demand evidence exists that the benign lesions progress for work, regardless of the proximate cause. to malignancy in some studies. “Malignant” comes from the Latin word 1.6. Metaplasia and Anaplasia (Atypia) malignus, and means malicious. Malignant neoplasms grow rapidly and are characterized Qualitative changes such as metaplasia and by local invasiveness. Areas of necrosis seen in anaplasia (or atypia) can occur in hyperplastic some malignant neoplasms presumably result cells and represent one of the hallmarks of when growth is so rapid that the neoplastic tissue pathological hyperplasia. However, qualitative outgrows the existing blood supply. Malignant changes in the phenotype or in the association growth is disorganized, and such neoplasms and organization of groups of cells can also occur may spread by extension into adjacent tissues or in the absence of clear-cut evidence of increases in by metastasis to distant sites via blood and/or cell number. lymphatic circulation (Figure 5.1). Characteristics Metaplasia is the reversible substitution of one of malignant neoplasms are listed in Table 5.6. type of fully differentiated cell for another within Some malignant neoplasms, particularly carci- a given tissue, and is seen most commonly in nomas, at some time in their evolution are at an epithelial tissues. The notion that metaplasia intraepithelial or in situ stage. In situ carcinomas usually involves the substitution of a less special- are microscopic lesions that have cytological ized cell type should be avoided in that it criteria of malignancy, but are localized and involves a value judgment regarding whether have not gone beyond the basement membrane a cell expressing a different set of genes is indeed (Table 5.6). less specialized. For example, squamous meta- plasia occurring in an area normally populated by ciliated respiratory epithelium represents a situation where fully specialized squamous 2. CANCER IS A GENETIC DISEASE epithelium provides protection against irritation. Removal of the irritating stimulus results in 2.1. Overview replacement by fully specialized ciliated The carcinogenic process involves the alter- epithelium. ations of four broad categories of cancer genes, namely the activation of oncogenes, inactivation 1.7. Benign vs malignant neoplasms of tumor suppressors, evasion of apoptosis genes, and defective DNA repair genes (Table 5.7). Of “Benign” comes from the Latin word benignus, the approximately 20 000 genes in the mamma- and means innocuous. A benign neoplasm is lian genome, hundreds are known or proposed a localized growth of well-differentiated, non- oncogenes while relatively fewer are tumor invasive tissue (Figure 5.1). Its growth is by suppressors, apoptosis evaders, or repair genes. expansion, and it may produce compression of There are other fundamental cellular alter- adjacent normal tissues. Benign neoplasms ations, including limitless replicative potential, ordinarily grow very slowly and are usually sustained angiogenesis, and ability to invade not life-threatening unless they interfere with and metastasize. During the multistep process vital functions, such as in a cardiac Schwannoma each neoplasm, arising from an individual cell, or a so-called “benign” meningioma severely accumulates at least 80 genetic alterations in compressing vital areas in the brain or spinal cancer genes, a dozen of which are considered cord. Characteristic features of benign neoplasia “driver” mutations that drive the cancer’s are listed in Table 5.6. uncontrolled growth. Multistep models of carci- Controversy regarding the significance of nogenesis have proven useful for defining events benign neoplasia with respect to the develop- in the neoplastic process and form the corner- ment of malignancy is similar to that associated stone of current hypotheses of the biological with preneoplastic lesions. In chemical carcino- mechanisms of carcinogenesis. Multistage genicity tests using rodents, carcinogens models of carcinogenic processes have been frequently produce both benign and malignant demonstrated in a variety of organ systems, neoplasms of a given tissue, and morphologic such as skin, liver, urinary bladder, lung, kidney,

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 118 5. CARCINOGENESIS

FIGURE 5.1 Photomicrographs of an adenoma (A and B) and adenocarcinoma (C and D) arising within the lung (arrowheads) from two different mice. The adenoma has features similar to alveolar type II pneumocyte hyper- plasia, such as well-differentiated cells (C) forming papillary structures and lining alveoli (arrows). The solid areas and compression are the diagnostic features of adenomas. Pulmonary adenocarcinomas are characterized by larger size (generally), solid areas with moderate to marked cellular dysplasia (D), poor differentiation, increased and often abnormal mitotic figures (arrows), and/or vascular or stromal invasion. Adenocarcinomas tend to metas- tasize to other parts of the lung (intrapulmonary) or to distant sites (extrapulmonary). Interestingly, some meta- static cells grow as well-differentiated ciliated respiratory cells, thereby recapitulating their cell of origin. intestine, mammary gland, and pancreas, by operationally defined phases of carcinogenesis – utilizing various experimental animal models. initiation, promotion, and progression – are From use of these model systems, various useful for discussion and understanding of carci- agents have been categorized as initiators, nogenesis, but in fact each of these phases in the promoters, and complete carcinogens. The process may consist of multiple stages.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 2. CANCER IS A GENETIC DISEASE 119

TABLE 5.6 Comparative Features of Benign and Malignant Neoplasms

Benign Malignant General effect on Little; usually does not cause death Will almost always kill the host if the host untreated Rate of growth Slow; may stop or regress More rapid (but slower than “repair tissue”) Autonomous; never stops or regresses Histologic features Encapsulated; remains localized at Infiltrates or invades; metastasizes primary site Mode of growth Usually grows by expansion, displacing Invades, destroys and replaces surrounding normal tissue surrounding normal tissue Metastasis Does not metastasize Most can metastasize Architecture Encapsulated; has complex stroma and Not encapsulated; usually has poorly adequate blood supply developed stroma; may become necrotic at center Danger to host Most are without lethal significance Always ultimately lethal unless removed or destroyed in situ Injury to host Usually negligible but may become very Can kill host directly by destruction of large and compress or obstruct vital vital tissue tissue Radiation Radiation sensitivity near that of normal Radiation sensitivity increased in rough sensitivity parent cell; rarely treated with radiation proportion to malignancy; often treated with radiation Behavior in tissue Cells are cohesive and inhibited by Cells do not cohere; frequently not mutual contact inhibited by mutual contact Resemblance in Cells and architecture resemble tissue of Cells atypical and pleomorphic; tissue origin disorganized bizarre architecture Mitotic figures Mitotic figures are rare and normal Mitotic figures may be numerous and abnormal in polarity and configuration Shape of nucleus Normal and regular; show usual stain Irregular; nucleus frequently affinity hyperchromatic Size of nucleus Normal; ratio of nucleus to cytoplasm Frequently large; nucleus-to-cytoplasm near normal ratio increased Nucleolus Not conspicuous Hyperchromatic and larger than normal

Reproduced from Handbook of Toxicologic Pathology, 2nd Ed. W. M Haschek, C. G. Rousseaux and M. A. Wallig, eds. (2002) Academic Press, Vol. I, Table III, p. 89, with permission.

A large body of experimental data supports the Since the mid-1980s, oncogenes have been contention that malignant cancer is generally irre- identified in the tumor DNA of many human versible. Even in initiation–promotion studies, neoplasms as well as in spontaneous and chemi- the initial mutational changes constituting initia- cally-induced neoplasms in animals (Table 5.8). tion may remain latent for weeks or months These oncogenes are frequently in the ras onco- before being expressed by administration of gene family, and the activated ras oncogenes a promoting agent. This “memory effect” is frequently differ from their homologous proto- compatible with a stable somatic mutation. oncogene by virtue of a single point mutation.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 120 5. CARCINOGENESIS

TABLE 5.7 Selected Examples of Cancer Genes Classified Oncogenes were first identified as the transform- by Mechanism(s) of Action ing genes of retroviruses. These oncogenes were not necessary for the life cycle of the virus but Oncogenes: were responsible for transforming virally Growth factors (Sis) infected cells and producing cancer in the host. Growth factor receptors (EGFR, PDGFR, HER2/neu) It was later learned that the transforming onco- genes were not intrinsic viral genes but rather Signal transduction (ras, abl, b-catenin) were normal cellular structural genes captured Cell cycle regulators (cyclin D, Cdk4) from eukaryotic organisms previously infected by a retrovirus. The virus integrates in to the Transcription factors (myc) host genome and after rearrangement a portion Tumor suppressors: of the normal cellular genes assimilates in to the b viral genome, a process called transduction. The Cell surface (TGF- R) transduced oncogene in the viral genome is Cytoskeleton (NF2) referred to as a viral oncogene (v-onc). The homol- Cytosol (APC/b-catenin, PTEN, Smad 2) ogous gene in the host genome is called a cellular oncogene (c-onc) or a proto-oncogene. Transcription factors (Rb1, p53, Brca1) In capturing the cellular proto-oncogene, the Cell cycle regulators virus lost some of its own structural genes and consequently often became incapable of repli- Apoptosis evasion (p53, bcl-2, bcl-x, bax) cating in the absence of helper viruses. As the Defective DNA repair (XP, HNPCC, MLH1, ATM, genomes of various mammalian and submam- Brca1, Brca2) malian species were examined, it was found that homologs to the retroviral oncogenes were present in species as diverse as yeast, fruit flies, In some experimental situations, carcinogen- amphibians, birds, and mammals. The high induced activation of some oncogenes appears degree of evolutionary conservation of these to be an early event in the carcinogenic process. proto-oncogenes suggested that they served While it can be argued that activation of proto- important normal functions in the cell. oncogenes may be a necessary event in the It is now known that proto-oncogenes encode genesis of some cancers, it is unlikely that for proteins that are important in cell growth, a single point mutation associated with an acti- development, and differentiation. Because vated oncogene is sufficient in and of itself for cancer is a perturbation of normal cell growth the development of all cancers. and differentiation, the potential significance of alterations in proto-oncogenes became apparent. 2.2. Oncogenes, Tumor Suppressors, It is thus not surprising that examination of DNA Apoptosis and Repair Genes from human and animal tumors has demon- strated the existence of dominant transforming Oncogenes are dominant-acting structural oncogenes, some of which correspond to those genes that encode for protein products responsible for the carcinogenicity of acute trans- capable of transforming the phenotype of a cell. forming retroviruses.

TABLE 5.8 Chemical-Specific ras Mutations in Hepatocellular Neoplasms in B6C3F1 Mice

Codon 61 mutation spectrum (normal [ CAA) Frequency (%) of H-ras codon 61 Treatment mutations in hepatocellular tumors AAA CGA CTA Controla 183/333 (56%) 106 50 21 DENa,b 63/239 (26%) 16 32 15 a Maronpot et al. (1995). Mutations in the ras proto-oncogene: Clues to etiology and molecular pathogenesis of mouse liver tumors. Toxicology 101, 125–156. b DEN, diethylnitrosourea from various dosages and dosing regimens. Data combined from multiple studies.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 2. CANCER IS A GENETIC DISEASE 121

An activated proto-oncogene has come to be such, when intact, they have a function opposite called an oncogene, and is a proto-oncogene that to that of oncogenes and might effectively has been altered quantitatively or qualitatively, oppose the action of an oncogene. While proto- resulting in inappropriate or overexpression. oncogenes have to be activated to influence Activation can occur in several ways. Retroviral carcinogenesis, suppressor genes have to be transduction had been shown to result in the inactivated for the transformed phenotype to acquisition of point mutations, deletions, or gene be expressed. Inactivation can be achieved by fusions within the coding sequence of the trans- chromosome loss, gene deletion, recombination, duced proto-oncogene. This leads to abnormal gene conversion, or point mutation. functioning and changes in levels and schedules Proto-oncogene activation is generally the of expression of encoded protein products. result of a somatic mutation. Mutant forms of Retroviruses can also affect the expression of tumor suppressor genes might be present in proto-oncogenes by a process called insertional germ cells and may thus be hereditary. It is mutagenesis. In this situation, the retroviral generally necessary for both alleles of a tumor DNA integrates into the host-cell DNA adjacent suppressor gene to be inactivated to ultimately to or within the coding sequence of a proto-onco- result in cancer. Exceptions include inactivation gene. The powerful retroviral promoters (regula- of a normal allele by genomic imprinting or by tory genes) then drive transcription of the normal dominant-negative mechanisms, as is proposed or truncated gene product of the proto-oncogene. for the p53 tumor suppressor gene. Activation of proto-oncogenes can also occur by In common human cancers, multiple tumor mechanisms independent of retroviral involve- suppressor genes may be affected, supporting ment. Point mutations and DNA rearrangements, the notion that cancer development involves such as translocations or gene amplifications, can perturbation of several levels of growth control. result in proto-oncogene activation by leading to Inactivation of the tumor suppressor gene p53 is altered levels or schedules of expression of the a frequent occurrence in a variety of human normal protein product, or in normal or altered cancers. Thus, inactivation or loss of tumor levels of expression of abnormal protein. suppressor genes, working in concert with the acti- The activation of proto-oncogenes in sponta- vation of oncogenes and with a variety of endoge- neous and chemically-induced neoplasia has nous and exogenous stimuli, plays an important received considerable attention over the years. part in the complex process of carcinogenesis. A variety of activated oncogenes have been The pivotal role of cell proliferation in all documented in rodent neoplasms. From some phases (e.g., initiation, promotion, progression) experimental studies it appears that certain types of the multistep process of carcinogenesis is inex- of oncogenes are activated by carcinogen treat- tricably linked to positive and negative cell cycle ment, and that this activation is sometimes an control mechanisms as influenced by oncogenes, early event in tumor induction. Other studies tumor suppressor genes, growth factors and with human and rodent neoplasms suggest that their cognate receptors, hormones and their oncogene activation is involved later in the carci- receptors, and the action of exogenous agents nogenic process, specifically during tumor (e.g., chemicals and viruses) on cell cycle control. progression. Uncontrolled cellular proliferation is the hall- In vitro neoplastic transformation and somatic mark of neoplasia, and many cancer cells demon- cell fusion studies provide evidence that malig- strate damage to genes that regulate their cell nant transformation represents a balance cycles directly. between genes for expression and for suppres- The prevailing model of the cell cycle is that of sion of malignancy. There is also convincing a series of transitions at which certain criteria evidence that growth suppressor genes, gener- must be met before the cell proceeds to the next ally referred to as tumor suppressor genes, phase. The cell cycle is composed of an S (DNA play a critical role in in vivo carcinogenesis. synthesis) and an M (mitotic) phase, separated Growth suppressor genes are regulatory genes by two gap phases (G1 and G2). Progression that normally function to limit or suppress through the cell cycle is tightly controlled by normal growth by inhibiting the activity of a group of heterodimeric protein kinases structural genes responsible for growth. As comprising a cyclin as a regulatory element and

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 122 5. CARCINOGENESIS a catalytic subunit known as a cyclin-dependent is maintained through a series of checkpoints kinase (Cdk). There are many combinations of (Figure 5.2). cyclin/Cdk complexes, and each phase of the The presence of these checkpoints allows DNA cycle is characterized by a specific pattern of repair before further progression into the cycle. expression and activity. The components of checkpoint control may not Five major classes of mammalian cyclins necessarily be essential to the workings of the (termed A–E) have been described. Cyclins C, cycle; instead, their role is to “brake” the cycle D1–3, and E reach their peak of synthesis in the face of stress or damage. Abrogation of and activity during the G1 phase and regulate cell cycle checkpoints with agents such as meth- the transition from G1 to S phase. However, ylxanthine analogs or pentoxifylline increases cyclins A and B1–2 achieve their maximal the cytotoxicity of DNA-damaging agents. levels later in the cycle, during the S and G2 The importance of DNA damage in triggering phases, and are regarded as regulators of the a cell cycle shutdown is obvious. Replication of transition to mitosis. Association with cyclins a damaged template would certainly result in not only activates cyclin-dependent kinases irreversible chromosomal aberrations and a high but also determines their substrate specificity. mutation rate. Two major checkpoints are thought Depending on the cyclin partner and therefore to be particularly important following DNA the cell cycle stage, different key target mole- damage and have been established at the middle cules are phosphorylated. These events occur to end of G1 (preceding DNA replication) and in a highly regulated temporal sequence that G2 (preceding chromosome segregation).

FIGURE 5.2 The cell cycle. After mitotic (M) cell division, cells either enter an indefinite resting phase called Gap 0 (G0), or immediately or eventually enter the Gap 1 (G1) phase to prepare the cell for duplicating its DNA. The DNA is copied during the synthesis (S) phase and then enters Gap 2 (G2) prior to dividing into two cells during mitosis (M). The cell cycle is controlled by positive and negative feedback signals involving p53, cyclins, cyclin- dependent kinases (Cdks), Rb, E2F, and control over apoptotic processes (see text). Cells must successfully proceed through so-called “checkpoints” (stop signs), the major ones being before S and M, to reduce the likelihood of introducing transforming mutational events in the progeny cells. p53 is one of the main suppressor proteins exerting control by suppressing progression through the cell cycle in DNA-damaged cells as well as aiding in DNA repair. Because of its major roles in surveillance of cell integrity, p53 has been considered the “guardian of the genome.” In severely damaged cells p53 plays a role in active cell death. p53 mediates its effects through various pathways, including exerting control of the cell cycle via proteins such as p21, p16, and p27. When p53 is mutated or altered in function, an individual’s cancer risk is markedly elevated. Furthermore, p53 mutation is the most frequent cancer gene alteration in human cancers, occurring in up to 60% of cancer cases.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 2. CANCER IS A GENETIC DISEASE 123

Loss of the G1 checkpoint triggers genomic binding core. Lack of p53 permits synthesis of instability at the time of interaction of unrepaired damaged DNA and increases the incidence of DNA with the DNA replication machinery, selected types of mutations. This increased inci- leading to deletion-type mutations and aberrant dence has been shown after a variety of DNA gene amplification. Inactivation of the G1 check- damage mechanisms, such as ionizing radiation point serves as an initiation step that makes (strand breaks), alkylation by methyl-methane the cell susceptible to unregulated growth (initia- sulfonate (MMS), ultraviolet irradiation (photo- tion), increasing the probability of subsequent dimers), and a variety of environmental carcino- genetic alterations and establishing the fully devel- gens. Thus, one of the major roles of p53 is to oped neoplastic phenotype. Control at the G1 ensure that, in response to genotoxic damage, checkpoint is dependent on cyclin D1 (degraded cells arrest in G1 and attempt to repair their at the G1/S transition) and cyclin E (degraded in DNA before it is replicated. mid-S phase). Overexpression of either cyclin D1 The wild-type p53 protein is normally kept at or cyclin E and subsequent activation of the cyclin very low steady-state cellular levels by its rela- D1 and cyclin E/Cdk complexes results in entry tively short half-life. However, it is stabilized into S phase and decreased G1 time. and accumulates in cells undergoing DNA Cyclin D1 is overexpressed in many human damage or in those responding to certain forms cancers, including breast and non-small cell of stress. After DNA damage, p53 binds lung carcinomas, sarcomas, melanomas, B cell a consensus-binding site and activates the tran- lymphomas, and squamous cell carcinomas of scription of several “downstream” genes. the head and neck. Cyclin D1/Cdk4 complexes One of these genes codes for the p21 protein. act to phosphorylate pRB, the product of the reti- The p21 gene belongs to a family of negative cell noblastoma susceptibility gene. pRB does not cycle regulators, which function as cyclin-depen- seem to possess sequence-specific binding dent kinase inhibitory molecules. Genes that activity, but instead exerts a negative regulatory encode these proteins are designated CKI genes. effect on gene expression through complex These negative regulators form stable complexes formation with DNA-binding proteins, including with cyclin/Cdk units and inactivate them. p21 members of the E2F family. In non-dividing or G0 inactivates cyclin E-Cdk2, cyclin A-Cdk2, and (arrested) cells, underphosphorylated pRB is cyclins D1-, D2-, and D3-Cdk4 complexes, thereby bound to E2F family members, leading to repres- inhibiting pRB phosphorylation and preventing sion of E2F-mediated transcription. progression of the cell cycle beyond G1. Upon phosphorylation by cyclin/Cdk The p53 protein also activates the BAX gene, complexes, pRB dissociates from E2F proteins, involved in the regulation of apoptosis. leading to transcription of genes promoting S Apoptosis is a cell suicide mechanism that leads phase entry. Thus, underphosphorylated pRB to “programmed cell death.” Apoptotic cells maintains cells in G1, whereas phosphorylation undergo cell shrinkage and chromosomal inactivates pRB and allows exit from G1. In condensation in response to DNA damage. These humans, inactivation of pRB is observed most changes prevent the replication of cells that have commonly in retinoblastomas, osteosarcomas, sustained a degree of genetic damage beyond carcinoid tumors, and non-small cell lung cancers. repair. The p53 protein regulates its own function Another tumor suppressor gene, p53, is neces- through the activation of the MDM2 gene. The sary for G1 phase arrest after DNA damage. product encoded by MDM2 is a 90-kDa zinc Mutations at the p53 locus are the most frequent finger protein (mdm2), which also contains genetic alterations associated with cancer in a p53-binding site. The mdm2 protein binds to humans. The p53 protein can be divided into p53 and acts as a negative regulator, inhibiting three main regions: the amino-terminal transacti- wild-type p53 transcriptional activity and vation domain, the sequence-specific DNA- creating an autoregulatory feedback loop. binding central core, and the multifunctional A pRB-binding site has also been identified at carboxy-terminal domain, which includes tetra- the carboxy-terminal domain of mdm2 that inter- merization and nuclear localization domains. acts with pRB and restrains its functions. Thus, The majority of p53 mutations involve several overexpression of mdm2 inactivates both p53 highly conserved regions within the DNA and pRB in a fashion similar to some viral

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 124 5. CARCINOGENESIS oncoprotein products, demonstrating a potential both of two ways. First, the enhanced cell turn- link between p53 and pRB in cell cycle regula- over could lead to “fixation” of spurious or spon- tion, apoptosis, and tumor progression. taneous genotoxic damage. DNA damage is The cell cycle and its regulatory proteins are believed to occur continually from cellular expo- altered to the benefit of many viral agents. sure to endogenous or exogenous genotoxic Progression of the cell cycle is advantageous for insults or simply from endogenous errors in many viruses, as they require activation of host DNA replication. Fortunately, the vast majority replication machinery to replicate their own of such DNA damage is repaired prior to cell genome. Several viral proteins have been shown division by efficient cellular enzymatic systems. to interact with and alter p53 and pRB. The faster the cells are dividing, the greater the The SV 40 large T antigen, the adenovirus EIB chance that genotoxic damage would not be 55-kDa protein, and the human papillomavirus repaired prior to cell division. Once cell division E6 protein each bind to p53 and inactivate it. has taken place, the genomic alteration is “fixed” Similarly, complex formation of underphos- and is hereditary. If such a genomic alteration phorylated pRB with the SV 40 large T antigen, confers a selective growth advantage to the cell the adenovirus EIA protein, and the papilloma- and its progeny, there will be clonal expansion virus E7 protein leads to pRB inactivation and and progressive development of neoplasia. cellular immortalization. High-transformation A second way in which enhanced mitogenesis strains of tumor viruses encode proteins that could contribute to neoplasia is to stimulate cell bind and inactivate both p53 and pRB. In the division in an already initiated cell, thereby normal cell, the growth deregulation caused by providing the growth stimulus whereby it can pRB inhibition can be counteracted by apoptotic expand clonally. In situations of continued expo- cell death produced by normal p53. sure to a non-genotoxic mitogen, it is likely that With the loss of both p53 and pRB, E2F activa- both potential mechanisms could act in a comple- tion stimulates unchecked cellular proliferation, mentary fashion. In either case, the mechanism of leading to the emergence of neoplastic cell cancer “induction” would be considered growth. The high rate and mutation pattern of secondary. A variety of situations exist in animal p53 and pRB in primary tumors have rendered carcinogenicity models where non-genotoxic them prototype tumor suppressor genes. chemicals given at sufficiently high doses may Furthermore, detection of p53 and pRB muta- play a causative role in the development of tions and altered expression of their encoded neoplasia simply by enhancing cell proliferation. products appear to be of clinical prognostic Examples of agents that may operate through significance when identified in specific cancers. this potential mechanism are given in Table 5.9. Additional gene products activated in response Confirmation of whether enhanced cell turn- to DNA damage include transcription factors, over is a realistic explanation for the observed growth factors, growth factor receptors, protec- carcinogenicity of non-genotoxic carcinogens tive enzymes, and proteins associated with will require carefully conducted studies that inflammation and tissue injury and repair. These demonstrate enhanced cell turnover at doses findings are consistent with the current under- that result in cancer, and no increase in cell turn- standing of the molecular basis of carcinogenesis over at non-carcinogenic doses. If enhanced as a multistep process. Therefore, it is essential to neoplasia in an animal test system is a conse- further advance our understanding of the intri- quence of increased cell proliferation and if the cate molecular mechanisms that govern chemical increased cell proliferation is related to exposure carcinogenesis. This will allow us to improve to excessive amounts of certain chemicals, then it strategies for assessing human cancer risk and becomes important to place the animal test to design effective treatment regimens. results into the context of expected environ- mental exposure in assessing potential risk to 2.3. Cell Proliferation and Apoptosis human health. Current data strongly suggest that cell death Increased cell proliferation can make an may be as essential as cell proliferation in carcino- important contribution to the process of carcino- genesis. The ratio between cell birth and counter- genesis. This increase can come about in either or balancing cell death determines tumor growth.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 2. CANCER IS A GENETIC DISEASE 125

TABLE 5.9 Non-Genotoxic Carcinogens that May Act Indirectly through Enhanced Cell Proliferation

Agent Target Potential mode of action Carbon tetrachloride Liver Necrosis and cytotoxicity a-Hexachlorocyclohexane Liver “Additive” hyperplasia Phenobarbital Liver “Additive” hyperplasia Polybrominated biphenyls Liver “Additive” hyperplasia Di (2-ethylhexyl)phthalate Liver “Additive” hyperplasia and peroxisome proliferation Methyl clofenapate Liver “Additive” hyperplasia and peroxisome proliferation a Unleaded gasoline Kidney Toxicity from 2u-globulin and resulting regenerative hyperplasia in male rats a Pentachloroethane Kidney 2u-globulin induction in male rats a d-Limonene Kidney 2u-globulin induction in male rats Sodium saccharin Urinary bladder Hyperplasia Biphenyl Urinary bladder Hyperplasia secondary to bladder calculi Butylated hydroxytoluene Urinary bladder ?

Reproduced from Handbook of Toxicologic Pathology, 2nd Ed. W. M Haschek, C. G. Rousseaux and M. A. Wallig, eds. (2002) Academic Press, Vol. I, Table VII, p. 104, with permission. Two forms of cell death may be seen in cancer has a viral etiology. According to the mutational development: necrosis and apoptosis. Necrosis hypothesis, one or more point mutations are typically occurs when a developing cancer responsible for initial and/or critical steps in outgrows its blood supply. Apoptosis is an the neoplastic process. In 1914, the somatic muta- energy-dependent process that involves active tion theory for cancer was put forward. gene transcription and translation. In preneoplas- According to the somatic mutation theory, tic lesions, apoptosis is the predominant form of cancer originates when an otherwise ordinary cell death that is observed. Chemicals, food depri- cell undergoes a mutation. In 1953, it was postu- vation, certain cytokines, growth factors, tumor lated that if a large enough population of somatic suppressor genes, and withdrawal of mitogenic cells lives for a sufficient length of time, gene agents in experimental in vivo carcinogenesis mutations will occur in some of them. As the may trigger enhanced apoptosis. It has been mutated cells proliferate, there is a finite proba- reported that the growth of dioxin-promoted pre- bility that some of them will sustain at least neoplastic liver foci in rat hepatocarcinogenesis is a second mutation (Figure 5.3). due to inhibition of apoptosis rather than to As the process of successive mutation and enhanced cell proliferation. An underlying prin- proliferation continues, cells eventually sustain ciple of cancer chemotherapy is the selective enough genetic alteration to become autonomous induction of apoptosis in neoplastic cells. and result in cancer. The accumulation of succes- sive mutations would be expected to increase as 2.4. Somatic Mutation Theory a function of age and of the degree of cell prolif- eration. Early occurrence of cancer might be There is strong evidence that a critical step in expected to result from exposure to mutagens carcinogenesis is a structural alteration occurring as well as to agents that increase the rate of cell in the genetic machinery of a somatic cell. This proliferation. The relevance of the somatic muta- appears to be true whether the active agent is tion theory to multistep carcinogenesis has a chemical or ionizing radiation, or if the cancer been shown repeatedly to be the most probable

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 126 5. CARCINOGENESIS

FIGURE 5.3 Depiction of how neoplasms evolve through a multistep genetic and morphologic continuum and how somatic cells are prone to acquisition of genetic or epigenetic alterations influenced by predisposing alterations, acute or chronic environ- mental exposures, and with age. The molecular environment may permit events that foster the development and progression of cancer.

explanation for the initiation of carcinogenesis the association of the Philadelphia chromosome (Figure 5.3). with the lymphoma/leukemia group of human In more recent times, the possibility that diseases. However, in light of the relatively few malignant cancer can result from multiple constant karyotypic changes associated with genetic insults has been championed by several specific histogenic types of neoplasms and the investigators and is gaining widespread accep- random nature of the preponderance of chromo- tance. There seems little doubt that genetic muta- somal aberrations observed in neoplasms, it is tion plays a causative role in the genesis of some difficult to establish a cause–effect relationship cancers. However, it has not been demonstrated between specific chromosomal aberrations and unequivocally that mutation is a universal, suffi- cancer. Non-random chromosomal aberrations cient, or necessary prerequisite for all cancers. in animal neoplasms include trisomy 15 in T cell It has been proposed that neoplastic cells likely lymphoma of the mouse, trisomy 13 in mouse have a higher mutation rate than normal cells mammary adenocarcinoma, and trisomy 4 in eth- and thereby increase the likelihood of neoplastic ylnitrosurea-induced neurogenic neoplasms in cells acquiring further mutations conducive to the rat. neoplastic growth features. This was referred to Reciprocal translocation is a form of gene rear- as the “mutator phenotype.” It suggests that rangement where portions of two chromosomes early mutation in stability genes (i.e. DNA repair, are simply exchanged with no net loss of genetic mismatch repair, DNA replication, or chromo- information. This can result in an alteration of the some maintenance) will lead to the mutator structure of the genes by virtue of their new loca- phenotype and further mutations contribute to tion and/or in abnormal expression of the translo- the subsequent invasive and metastatic proper- cated gene(s). The Philadelphia chromosome ties of the neoplastic growth. associated with human lymphoma/leukemia Numerical and/or structural chromosomal represents a prototypical example of this phenom- abnormalities and alterations are virtually enon. This specific chromosomal abnormality universally present in neoplasms, particularly consists of a translocation between the long arms malignant neoplasms. Chromosomal aberrations of chromosomes 9 and 22, and is seen in 85% of are found in apparently spontaneous neoplasms patients with chronic myelogenous leukemia. as well as in those induced by chemical carcino- Translocations between other chromosomes are gens or by oncogenic viruses. Against a backdrop associated with different forms of leukemia. of apparently random chromosomal aberrations Gene amplification represents a situation observed in neoplasms, it is also apparent that where there is an increase in the amount of there are a few constant karyotypic changes asso- DNA present in a specific region of a chromo- ciated with specific neoplasms. Most notable is some. Chromosomal abnormalities observed in

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 2. CANCER IS A GENETIC DISEASE 127 karyotype preparations such as homogeneously Through the action of the RNA-induced silencing stained regions, abnormal banding patterns, and complex they “silence” gene expression via double minutes are the result of gene amplifica- a post-transcriptional event. MiRNAs act to tion. Assuming that the amplified genes are tran- control various pathways, including cell growth, scriptionally active, an excess of the product differentiation, and survival. MiRNAs have been encoded by the amplified genes would be antici- shown to undergo changes in expression in pated. Drug-induced gene amplification, possibly cancer cells, and frequent amplifications and with an associated specific karyotypic change, is deletions of miRNA loci have been identified in known to result in drug resistance due to an many cancers. They act by increasing the expres- increase in the amount of gene product. A number sion of oncogenes or by reducing the expression of observations have been made with respect to of tumor suppressor genes. gene amplification and oncogenesis. The classical MiRNAs can act like a tumor suppressor when skin tumor promoter, phorbol ester, causes ampli- the miRNA enhances expression of oncogenes fication of specific genes in cultured cells. DNA that lead to overproduction of the oncogene amplification has been shown to occur at the site product. Conversely, miRNA can reduce the of insertion of oncogenic retroviruses. tumor suppressor proteins’ expression and It has been shown that increased production of thereby act as an oncogene. MiRNA profiling of normal protein products of proto-oncogenes several human tumors has led to the downregu- (e.g., c-ras) that have been amplified may lation or deletion of certain miRNAs in leuke- contribute to the malignant phenotype. Amplifi- mias and lymphomas, resulting in increased cation of c-myc is seen in small cell cancer of the expression of BCL2, the anti-apoptotic protein. lung in humans, and the degree of amplification MiRNA-mediated up regulation of ras and myc is correlated with clinical aggressiveness of this oncogenes has been detected in some lung cancer. The causative role of DNA transposition, tumors and B cell leukemias, respectively. Over- other types of DNA rearrangement, and even expression of specific miRNAs has been amplification of endogenous genes in either documented in a number of cancers in some the origin or the development of neoplasia is human brain and breast tumors, but a critical difficult to ascertain at present. Even the mecha- oncogene or tumor suppressor gene has yet to nisms causing such somatic chromosomal be identified. Look for more about miRNAs abnormalities remain obscure. and role in cancer in the coming years of exciting However, the sites for chromosomal rearrange- new understanding. ment seem to correspond to specific fragile loca- tions in the genome that may be uniquely Epigenetics susceptible to breakage. Further documentation Epigenetics refers to reversible, heritable of chromosomal abnormalities in various changes in gene expression that occur without neoplasms may provide important clues about mutation. Such changes involve post-transla- basic genetic mechanisms associated with carci- tional modifications of histones and DNA meth- nogenesis. In some cases, specific chromosomal ylation, both of which affect gene expression. In abnormalities have been shown to have progno- normal, differentiated cells, the majority of the stic significance. Also, modulation of gene prod- genome is not expressed. Some portions of the ucts may foster malignant behavior by causing genome are silenced by DNA methylation and uncontrolled cell proliferation and metastasis. histone modifications that lead to the compaction of DNA into heterochromatin. On the other hand, cancer cells are characterized by a global 2.5. Gene Regulation DNA hypomethylation and selective promoter- localized hypermethylation. Indeed, it has MicroRNAs become evident during the past few years that MicroRNAs (or MiRNAs) are small non- tumor suppressor genes are sometimes silenced coding, single-stranded RNAs, approximately by hypermethylation of promoter sequences 20 nucleotides in length, that control gene expres- rather than mutation. sion. They can incorporate into what is referred The non-mutational genetic hypothesis postu- to as the RNA-induced silencing complex. lates that a genomic perturbation other than

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 128 5. CARCINOGENESIS a mutation may lead to cancer. One of the mech- of differentiation conferring a selective growth anistic explanations for this hypothesis relates to advantage to the altered phenotype. changes in the pattern of DNA methylation. Altered patterns of DNA methylation are associ- Intercellular Communication ated with exposure to chemicals and are Among the potential regulatory factors associ- tissue-specific. Consistent with a role of altered ated with the phenotypic expression of trans- methylation in carcinogenesis are the observa- formed cells are those associated with tions that DNAs from specific tumors have intercellular communication. Intercellular evidence of hypomethylation; carcinogens are communication (sometimes referred to as meta- known to interfere with DNA methylation; and bolic cooperation, based on a popular assay for the drug 5-azacytidine, which causes under its measurement) is known to play an important methylation of genes, has been shown to enhance role in the phenotypic expression of some trans- cell transformation. In general, genes from formed cells in vitro. In co-cultivation studies, cancer cells are hypomethylated as compared to the presence of normal cells is sufficient to their normal counterparts, although important prevent the phenotypic expression of trans- qualifications that relate to the test system used formed cells. should be noted. Apparently there is sufficient communication Hypomethylation is believed to influence the between cells, most probably occurring via gap regulation of transcription and gene expression junctions, such that biochemical substances elab- and may be associated with cellular differentia- orated by the normal cells block the expression of tion. Unfortunately, as with many abnormalities an abnormal phenotype by the transformed cells. that are associated with cancer cells, we are faced Although not a ubiquitous characteristic of non- with a dilemma in deciding if hypomethylation genotoxic carcinogens, it has been suggested that of genes is a cause of malignancy or a conse- the inhibition of cellular communication may be quence of the altered metabolism of malignant the mode of action of many, but not all, chemicals cells. that induce neoplasia but do not affect the genome directly. Blocked Differentiation Current studies have shown that many An apparent characteristic of transformed tumor-promoting agents inhibit gap junction cells is a partial or total block in terminal cell intercellular communication associated with differentiation. A block of normal differentia- the aberrant expression of connexin and the tionimpliesanalterationinnormalgene loss of function of cell adhesion molecules. It expression. While it has been postulated that has also been suggested that the inhibition of neoplasia may represent blocked or abnormal intercellular communication may be one cellular differentiation, the paucity of knowl- unifying mechanism for all forms of neoplasm edge relative to normal mechanisms that regu- promotion. Non-cytotoxic levels of chemicals late differentiation makes it difficult to known to be promoters in lung, liver, skin, elucidate the potential role of abnormal cell colon, breast, or esophagus have been shown differentiation in the genesis or development to inhibit intercellular communication in cell of cancer. culture systems. In fact, experiments exploring the possibility An in vivo phenomenon that might be consid- that neoplasia is a disease of blocked differentia- ered a different form of intercellular communi- tion have shown that, under specific environ- cation relates to the postulation that epidermal mental conditions, malignant cells have an chalones are instrumental in the promotion of intrinsic ability to revert to a normal phenotype. skin cancer. Chalones are tissue-specific This suggests that if blocked differentiation plays substances that inhibit the proliferation of imma- a role in the development of neoplasia, the ture cells and probably operate through a cell endogenous and/or exogenous factors regu- membrane receptor mechanism rather than via lating differentiation must be present continually gap junctions. to maintain the neoplastic phenotype. It has been Chalones are believed to be produced by argued, however, that cancer is not blocked neighboring cells, and may operate as a type of differentiation but rather represents a new state paracrine control mechanism. Skin tumor

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 2. CANCER IS A GENETIC DISEASE 129 promoters, such as phorbol esters, possibly inac- synthetic estrogens, and some hormones do tivate the chalone receptor site, thereby switch- bind to protein. ing off the mechanism of growth control in the Most endogenous hormones are believed to affected cell(s). The existence of chalones has act through a specific cellular receptor mecha- been debated. Some regard them as hormone- nism. Once there is binding to a specific like substances with inhibitory activity. A membrane, cytosolic, or nuclear receptor, the number of growth-inhibitory substances have receptor becomes activated and intracellular been described (e.g., TGF-13). Many of these do activity occurs that affects transcriptional or not meet the tissue-specific definition of a chalone translational processes, resulting in the synthesis but nonetheless appear to regulate normal but of specific proteins. Just how hormonal interac- not tumor cell growth. tion with a cell receptor leads to cancer is unknown, but one potential mechanism relates Hormones to an increase in cell turnover among cells that Hormones are chemical messengers that bind already possess latent genetic change. to specific cellular receptors and form a Thus, endogenous hormones may serve to hormone–receptor complex that triggers a cellular promote spontaneously or otherwise initiated response (see Endocrine System, Chapter 58). The cells. Alternatively, hormonal imbalance could cellular response is specific for both the hormone lead to increased proliferation in a sensitive cell and the target cell. The target-cell response to population with secondary genotoxic damage hormone stimulation is typically an increase or from any one of several environmentally preva- decrease in cell division, or an acceleration or lent genotoxic agents. This certainly is consistent deceleration in differentiation. There is an almost with the frequent observation of endocrine continual discovery of endogenous messenger tumors developing against a previously existing substances and growth factors that fit the broad background of hyperplasia. Furthermore, the definition of “hormone.” two possibilities are not mutually exclusive. Like hormones, hormone-related peptides, The precise signals and mechanisms by which commonly referred to as growth factors, play hormones and related growth factors regulate an important role in the control of growth and cell proliferation and differentiation remain to differentiation of cells, tissues, and organs. A be identified, and are areas of active research. major difference between hormones and growth Evidence suggests that the mechanism of action factors is that hormones are produced in an of hormones and growth factors may be closely endocrine tissue and act on cells at a distant associated with and related to the action of onco- site, whereas growth factors are secreted by gene protein products. a variety of normal and abnormal tissues and act on nearby responsive cells. Endogenous hormones have long been known 2.6. Cancer Stem Cell Theory to be associated with the development of specific For more than 100 years cancers have been neoplasms and in some cases with the inhibition recognized as morphologically heterogeneous of carcinogenesis (e.g., diethylstilbestrol, DES). populations of cells that arise from a single cell, Hormones or hormone imbalances undoubtedly but it has become clear in the past decade that play a major causative role in cancers of certain functional heterogeneity also exists. Now, many hormone-sensitive tissues (ovary, uterus, pros- studies suggest that a distinct subpopulation of tate, testes, endocrine organs). However, it is cancer cells designated as “cancer stem cells” unlikely that they play a major causative role in (CSC) sustain cancer growth. Specifically, only the development of neoplasia in non-hormone- a minority of tumor cells (CSCs) from some of sensitive tissues. Through the use of two-stage these cancers have the capacity to regenerate animal models for mammary and thyroid cancer, a tumor and sustain its growth when injected hormones have been demonstrated to function as into immune-compromised mouse. tumor promoters of neoplasia. There is little There is extensive evidence that multiple types evidence that there is direct interaction of endog- of cancers arise from transformation of stem enous hormones with DNA, although minimal and/or multipotent progenitor cells during car- DNA binding has been shown with some cinogenesis; therefore, it is reasonable to consider

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 130 5. CARCINOGENESIS that cancer in some cases may be a primary the numerous genetic mutations needed for disorder of stem cell differentiation. Cancer transformation. stem cells, as these transformed stem cells are Many similarities between normal stem cells defined, are neoplastic cells that have stem cell and tumor cells exist, and tumorigenic cells properties, including the ability to self-renew in display functional and phenotypic qualities of order to propagate additional malignant stem the normal cells from which they arise. There is cells, and differentiate to give rise to pheno- significant evidence that neoplastic cells undergo typically diverse non-carcinogenic cells. self-renewal and differentiation similar to that of Given these characteristics, these transformed normal stem cells. Both cancer cells and stem stem cells therefore have an indefinite ability to cells have indefinite proliferative potential, as self-renew, which is the driving force behind well as the ability to undergo organogenesis carcinogenesis. This is the “cancer stem cell and give rise to new tissues, although disorga- hypothesis.” This hypothesis is not new to the nized and atypical in the case of tumorigenesis. field of cancer research. The formal cancer stem Both cancerous and normal tissue is composed cell hypothesis was first proposed over 150 years of heterogeneous cell populations with various ago: that these dormant embryonic components morphologies and proliferative ability. exist in adult tissues and may be activated to Stem cells and various neoplastic cells express later become tumorigenic. Over the decades, telomerase, and have anti-apoptotic mechanisms, the idea that tissue-specific stem cells may be and increased membrane transporter activity the cell of origin of various cancers and that (thus being able to efflux chemicals and chemo- cancers may represent a maturation arrest of therapeutics from the intracellular compartment). stem cells advanced the study of the cancer Telomerase is a type of reverse transcriptase that stem cell hypothesis. Indeed, the cancer stem preserves the length of telomeres. It is active cell hypothesis continues today as an enormous during cell replication, and plays a role in prolif- focus of the field of cancer research. eration and immortality of cancer cells. The telo- Through the ability to self-renew, a small pop- meres of chromosomes of normal somatic cells ulation of carcinogenic stem cells within shorten during each cell division and are a part a neoplasm would have the ability to maintain of cell senescence. In cancer cells, telomerase malignant clones. Self-renewal is a process by may be expressed and lengthen telomeres, which a stem cell pool maintains its numbers contributing to cell immortality and autonomous through symmetric and asymmetric division. In neoplastic growth. Both stem cells and neoplastic symmetric cell division, the progeny are identical cells have the ability to migrate (or rather metas- to the initial stem cell; in asymmetric stem cell tasize in the case of neoplastic cells), and both self-renewal, one of the two progeny are identical share embryonic markers such as b-HCG and to the initial stem cell, whereas the other cell is AFP, and several commonly recognized stem a committed progenitor cell, which undergoes cell markers such as c-kit and CD34 are found cellular differentiation. The traditional clonal in a variety of tumors. origin of carcinogenesis states that transforma- The process of tumorigenesis in many types of tion occurs through a series of sequential muta- cancer involves dysregulation of embryonic and tions resulting in preneoplastic lesions that developmental pathways involved in regulation progress to neoplasia and, finally, metastasis. of embryogenesis, cell fate determination, and However, in many tissue sites in which cancers stem cell maintenance, including Wnt, Notch, arise, cell populations have a high rate of cell and Sonic Hedgehog pathways. In many ways, turnover and thus do not persist for significantly neoplastic cells differ from stem cells only in long enough time to accumulate the necessary terms of their unrestricted and disordered number of mutations sufficient for cellular trans- growth and genetic alterations. formation. On the other hand, through the tightly Significant evidence exists for the existence of regulated process of self-renewal, normal stem tissue-specific cancer stem cells in various cells are able to function over the lifespan of the organs, consistent with the theory that cancers host, and thus their longevity and continued arising in various tissues originate from transfor- mitotic activity makes them a significant target, mation of tissue-specific stem or progenitor cells. and potential reservoir for, the accumulation of Identification of these cells can be problematic,

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 2. CANCER IS A GENETIC DISEASE 131 because stem cells represent a very small propor- that could be serially transplanted and was tion of the overall cell population in an organ. morphologically identical to the original In order to identify cancer stem cells, investiga- neoplasm. Alternatively, injection of 105 CD133- tors have used a stem cell immunophenotype negative cells failed to produce a tumor when based on immunocytologic and immunohisto- transplanted into mice, indicating that only chemical expression of various stem cell markers brain tumor cells possessing a stem cell pheno- and cell surface markers, followed by cell sorting type were able to survive and proliferate to repo- to identify a cancer stem cell population. Stem pulate the neoplastic population, and that the cell populations have been identified using panels non-stem cell tumor population failed to of a number of immunocytochemical markers, promote neural cancer. Identification of these including CD133, CD34, CD38, CD24, CD44, brain tumor-initiating cells gives strong support SOX2, OCT4, BMI1, ESA, Nanog, Nestin, and c-kit. for the cancer stem cell hypothesis as the basis Significant biologic evidence for the role of for many solid tumors. stem cells as oncogenic targets of mutagenesis Studies on human lung cancer provided was initially reported in hematologic malignan- evidence that bronchioalveolar stem cells were cies such as leukemia and multiple myeloma. In putative stem cell targets of oncogenic KRAS, these diseases, it was documented that only and when transformed were the precursors to a small subset of tumor cells was actually capable lung tumors in vivo in mice. These stem cells of extensively proliferating, and that the progres- normally maintain Clara cell populations in the sion of the neoplastic disease was driven by bronchioles and alveolar epithelial cells within this population that possessed stem or progenitor the lung. When transformed, these multipotent cell characteristics. In human acute myelogenous progenitor cells gave rise to pulmonary adeno- leukemia (AML), using the hematopoietic cell carcinomas, contrary to previous belief that þ þ surface phenotype CD34 /CD38 /Thy1 ,inves- differentiated pulmonary alveolar or bronchiolar tigators found that the small population of AML cells were the cell of origin for this type of cells possessing this phenotype could repopulate cancer. It has been shown that, similar to brain and induce leukemia in immunocompromised tumors, tumorigenic cells in small cell and mice, with the same morphology and cell surface non-small cell lung cancer in humans are marker expression. Those cells that did not a rare population of CD133-positive, undifferen- exhibit such a phenotype were unable to propa- tiated cancer stem cells that are able to self- gate, grow, and reproduce tumors in immuno- renew, grow indefinitely, and repopulate compromised mice. secondary transplanted hosts, whereas CD133- While the earliest evidence of cancer stem cells negative tumor cells are not. comes from studies in hematologic malignancies The CD133 cell surface marker has also been such as human acute and chronic leukemia and used to identify colon cancer-initiating cells myeloma, multiple researchers have provided using a similar xenograft model, in which only evidence for the existence of cancer stem cells the small proportion of CD133þ colonic adeno- in solid tumors, including brain, lung, colon, carcinoma cells were able to maintain growth ovarian, pancreas and prostate tumors, as well and recapitulate the original tumor morphologi- as in breast cancer and melanoma, among others. cally in secondary hosts. Through molecular investigation, the use of stem There is mounting evidence supporting cell markers, developmental pathways, and a transformed stem cell in malignant mela- xenograft transplantation experiments in mouse noma. Melanomas are comprised of a heteroge- models, numerous studies have provided signif- neous cell population, and stem and icant evidence for the relevance of cancer stem progenitor cell markers have been detected in cells as initiators of various types of cancer in malignant melanoma cells in humans. Several humans. studies have examined the presence of tumor- A mouse xenograft model has been used to igenic subpopulations of human malignant identify neural cancer stem cells through the melanoma cancer stem cells expressing use of a stem cell marker, CD133, and provided markers of stem cells (including CD133, evidence that injection of as few as 100 of these CD166, and Nestin, CD20, and the ABCG2 cancer-initiating cells produced a tumor in mice transporter) associated with high tumorigenic

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 132 5. CARCINOGENESIS potential, and found that melanoma cells functional and phenotypic aspects of the stem expressing these markers had higher tumori- cells they originate from. genic potential when transplanted into immu- The idea that the growth of a solid tumor is nocompromised mice. driven by proliferation of cancer stem cells has The existence of mammary epithelial stem profound implications for cancer therapy. Treat- cells was demonstrated over 40 years ago ment of cancer as an homogeneous population through the use of transplantation studies which of proliferative clones with unlimited growth showed that the entire repertoire of the gland, potential is problematic, because many tumors including fully functional mammary epithelial may initially respond to therapy only to recur, outgrowths containing ducts, lobules, and myoe- often months to years later in life. If, in the pithelial cells, could be generated from epithelial view of the cancer stem cell hypothesis, solid fragments taken from various regions of the tumors are considered to be composed predom- mammary gland at different stages of postnatal inantly of an heterogeneous population of development in the mouse. non-proliferative tumor cells and a small In 1998, it was shown that multipotent stem subpopulation of transformed stem cells that cells present in the mature mouse mammary predominantly drive carcinogenesis, this would gland had the capability of self-renewal, and account for the recurrence of tumors following that mammary gland outgrowths in secondary an initial regression period, since this subpopu- recipients of transplants were derived from lation of cancer stem cells is more resistant to self-renewing stem cells. In 2006, researchers current chemotherapeutics. demonstrated that an entire functioning The current dogma is that cancer cells acquire mammary gland could be recapitulated from resistance to chemotherapeutics as they evolve injection of a single mammary stem cell possess- and undergo additional mutations; an alterna- ing the immunophenotype Lin-/CD29high/ tive explanation to account for drug resistance þ CD24 . Interestingly, using the stem cell pheno- is that the cancer stem cell population inherently þ þ type (CD44 /CD24 , ESA ), breast cancer stem possesses the ability to resist current therapies, cells reconstituted secondary xenograft hosts and thus is able to contribute to disease progres- with as few as 200 tumor cells from primary sion or tumor reoccurrence. human breast carcinomas, whereas tumor cells A major cause of cancer-related death, metas- without this phenotype failed to produce tumors tasis, is currently considered to occur through despite transplantation of thousands of cells. sequential acquisition of mutations leading to The cancer stem cell origin hypothesis does selection of clones with metastatic potential; not discount the somatic mutation theory, and however, it has been shown for decades that in it is possible that each paradigm could be certain cancers a small number of individual contributing to transformation and ultimate cancer cells disseminate and lie dormant at formation of a neoplasm. The cancer stem cell distant sites in a state of quiescence for up to hypothesis accepts the contribution of genetic several years, until triggered to outbreak. Alter- instability, mutation, and epigenetic factors that natively, some disseminated tumor cells never influence the evolution of heterogeneity in result in the formation of metastatic tumors. a malignancy; however, in this case, the target One well-accepted explanation is that such cells of the genetic or epigenetic alterations is the are destroyed by immune surveillance. Alterna- stem cell, rather than the somatic cell. tively, one may consider that under the cancer Furthermore, the concept of initiation and stem cell hypothesis a tumor is composed largely promotion that is a significant part of the somatic of transformed but poorly proliferative tumor mutation theory applies to the cancer stem cell cells, and that only a small proportion, the highly hypothesis, in that mutations that lead to clonal carcinogenic cancer stem cell subset, is able to expansion of a neoplasm may initially occur in form a new tumor or metastasize. a stem cell. In further support of stem cells as If the growth and metastasis of tumors are targets of transforming mutations and progres- driven by a small fraction of highly tumorigenic sion towards malignancy, there are several simi- cancer stem cells within a heterogeneous popula- larities between normal stem cells and tumor tion of tumor cells, this may explain the failure to cells, and neoplastic cells indeed portray both develop therapies that are able to consistently

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 2. CANCER IS A GENETIC DISEASE 133 eradicate solid tumors. Therefore, a paradigm TABLE 5.10 Non-Genotoxic Agents Associated with shift in cancer therapy should be considered, in Liver Tumor Regression which therapy is directed towards killing cancer Chlordane stem cells, which may result in much more effec- tive therapies. Phenobarbital Clofibrate 2.7. Tumor Regression Nafenopin Regression of foci of altered hepatocytes and Y-14,643 hepatocellular neoplasms has been occasionally described in rats and only rarely in mice. Mech- anisms for tumor regression are unclear, but this by mimicking sex steroids, or by altering the phenomenon suggests that some histologically- synthesis and elimination of sex steroids or their appearing benign and malignant “neoplasms” receptors. Since some cases of tumor regression depend on the sustained presence of the inciting appear to be related to hormone dependency, chemical. In a study of chlordane there was such as the androgen- or contraceptive-induced evidence of regression of benign and malignant liver tumors described in humans, perhaps hepatocellular neoplasms after cessation of expo- chlordane acts by a similar mechanism. sure in B6C3F1 mice. H-ras mutations, which Liver tumor regression has also been observed commonly occur in up to 60% of neoplasms in after cessation of chronic exposure to a variety of this strain of mouse, were not present. non-genotoxic rodent hepatocarcinogens, Other non-genotoxic murine hepatocarcino- including chlordane, nafenopin, phenobarbital, gens that induce tumors with a lower frequency and Wyeth 14,643 (a peroxisome proliferator) of H-ras mutation than occurs in spontaneous (Table 5.10). In contrast, some other non-geno- tumors include hexachlorobenzene, Aroclor, toxic murine hepatocarcinogens, such as methy- phenobarbital, ciprofibrate, dieldrin, chloroform, lene chloride and dieldrin, induce hepatocellular and methyl clofenapate. The absence of ras neoplasms that persist and progress after cessa- mutations and evidence for regression implies tion of prolonged exposures. Perhaps chlordane- that the reversible “neoplasms” may lack the induced neoplasms develop by a series of appropriate genomic mutations necessary for changes that progress from a dependent to an autonomous proliferation. Alternatively, chlor- autonomous state. For example, some histologi- dane may provide a selective growth advantage cally benign and malignant liver neoplasms are to the hepatocellular lesions without ras muta- sustained by the continual presence of chlor- tions, eliminate ras mutated cells through hepa- dane, and with cessation of exposure only tocyte necrosis or apoptosis, or induce the autonomous neoplasms persist and some tumorigenic pathways that do not involve ras eventually metastasize. It also appears that activation. There was no histological evidence pathways of chlordane-induced hepatocarcino- of immune-mediated rejection of any neoplasms genesis progress independently of ras activation. in the study. Interestingly, bcl2 and bclx, proteins that inhibit apoptotic pathways, were elevated TABLE 5.11 Non-Genotoxic Mechanisms of in hypertrophied centrilobular hepatocytes and Carcinogenesis liver neoplasms from rodents treated with chlor- dane, phenobarbital, or clofibrate. Additionally, Alteration of endogenous events a gene array profiling study supported the DNA repair involvement of growth-promoting factors and oncogenes in chlordane hepatocarcinogenesis. Oxidative DNA damage Complete regression of hepatocellular carci- Methylation nomas occurs, rarely, in humans with cirrhosis and in children after cessation of androgen Selective growth advantage therapy. Benign liver tumors have regressed after Cell death/proliferation withdrawal of oral contraceptives in women. Induction of a hypermutable state Recent evidence suggests that chlordane can act

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 134 5. CARCINOGENESIS

The findings suggest that chlordane appears to The current strategy for identifying the poten- act through enzyme induction as well as non- tial carcinogenicity of chemicals involves genotoxic, hormone-like mechanisms that systemic (including dermal) exposure of male promote cell survival and inhibit cell death. If and female rats and mice to high doses and frac- such mechanisms (Table 5.11) can be further tions thereof of chemicals over a 2-year period. elucidated, then we can more accurately assess At the end of 2 years, the carcinogenicity of the the human risks of chemicals such as chlordane. chemical being studied is assessed by measuring the excess in neoplasm production above back- ground levels of incidence and multiplicity, doc- 3. IDENTIFYING CARCINOGENS umenting the occurrence of rare neoplasms with negligible background levels, or demonstrating 3.1. Animal Bioassays a reduced latency in neoplasm development in exposed animals versus controls. Widespread and routine evaluation of chemi- If an agent is observed to produce neoplasia in cals for their carcinogenic potential began in the experimental animals, the agent is regarded earnest in the mid-1960s with the use of a stan- as a potential human carcinogen. In countries dardized protocol for the “bioassay” program throughout the world, legal requirements of the National Cancer Institute (NCI). The orig- mandate that all new chemical agents and drugs inal NCI “bioassay” has evolved to the present- be tested in animal bioassays to determine if they day National Toxicology Program (NTP) 2-year cause cancer in the test animals. Additionally, and perinatal carcinogenicity study. since the mid-1960s in the United States, the Throughout this period a variety of alternative National Cancer Institute (NCI) and currently in vivo and in vitro assay schemes have been the National Toxicology Program (NTP) have introduced in the hopes of identifying potential collectively conducted animal bioassays on carcinogens in less time and for less money. more than 600 chemical agents to assess their Some of these alternative testing models have potential to cause cancer. been highly beneficial tools to help explore the Of those 600 agents, about half (290) are iden- processes of carcinogenesis and the types of tified as rodent carcinogens. A number of them carcinogens, to identify the genotoxicity of are considered known or suspected human chemicals, and to identify the specific mecha- carcinogens. Fifty-seven percent of the 290 NTP nisms whereby given agents produce a carcino- rodent carcinogens (Table 5.12) induced at least genic response in rodents. Testing batteries, tier an increase in liver neoplasms (primarily in approaches, and decision-point analyses have B6C3F1 mice), 22% an increase in lung tumors, been proposed to provide a basis for prioritizing and 14% in renal (primarily in F344/N rat), chemicals in the testing queue and to provide 13% in mammary gland, 13% in hematopoietic, a larger database for risk assessment. 12% in forestomach, 10% in thyroid, and 9% in These efforts serve to emphasize that, in the last endothelial neoplasms. analysis, there is no ideal test for identifying The original NCI “bioassay” was intended as potential human carcinogens. In general, efforts a screen for carcinogenicity in the rodent. It was to supplant the long-term carcinogen “bioassay” generally not intended to be used for risk assess- with less costly short-term tests have been frus- ment, determining carcinogenic potency, identi- trated by poor concordance of the latter with the fying subtleties of chronic toxicity, determining traditional long-term test results. In the mean- the mechanism of observed carcinogenic time, increasing costs for conducting long-term responses, or establishing pathogenesis of lesions carcinogenicity studies have resulted in fewer (see Carcinogenicity Assessment, Chapter 27). The chemicals being tested annually for potential car- implication behind the original NCI “bioassay” cinogenicity. Currently, much thought is being was that if a chemical was found to cause given to developing medium-term tests, such as neoplasms in treated animals, further studies 6- to 9-month studies in genetically modified would need to be designed and conducted to mice, for carcinogenicity concomitantly with anal- specifically determine parameters such as yses of reasonable ways to reduce the cost of the the mechanism of cancer induction, the nature standard long term rodent carcinogenicity test. of the dose–response relationship, chemical

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 3. IDENTIFYING CARCINOGENS 135 metabolism, and target organ dosimetry. The to the high dose or EMTD, at least two additional more definitive testing of positive chemicals lower doses are employed in NTP rodent carci- has generally not occurred. nogenicity studies with a conscious effort to However, in the absence of such additional utilize human and environmental-exposure relevant data, “bioassay” results have been information to set the low doses. This permits used as a basis for human risk assessment and better definition of dose–response and allows regulation. In an effort to consciously expand better utilization of data for risk assessment the original NCI “bioassay” and to make the (see Risk Assessment, Chapter 31). test results more relevant for the interpretations To fully characterize the toxicity of chemicals, being applied, considerable evolution has modifications to the chronic rodent carcinogenic- occurred in the design, conduct, and interpreta- ity study protocol have been selectively intro- tion of long-term rodent carcinogenesis tests con- duced by the NTP. Interim sacrifices and “stop ducted under the auspices of the NTP. studies” are incorporated into the design of While the maximum tolerated dose (MTD) many carcinogenicity studies to better define concept has been retained in principle, it is the pathogenesis and biological relevance applied more conservatively in setting the high of the anticipated response. The route of admin- dose. However, a fundamental difference still istration of chemicals is chosen to mimic natural exists in how the NTP selects the high dose in routes of human exposure whenever possible. a chronic rodent carcinogenicity study and how Consequently, corn-oil gavage of chemicals is industrial organizations determine the high rarely used in contemporary study design. Ancil- dose in their otherwise similar chronic rodent lary studies, such as chemical disposition and carcinogenicity studies. High-dose selection metabolism, toxicokinetics, reproductive toxicity, relates to philosophic approaches to testing for teratology, behavioral testing, and immunotoxic- toxicity and carcinogenicity. Governmental and ity, frequently complement contemporary 2-year similar organizations have traditionally con- carcinogenicity studies. In addition, in vitro and ducted their studies to determine if exposure to in vivo genotoxicity tests are conducted for a chemical can be harmful under any circum- each chemical. stances, whereas industry most frequently The NTP and others who assess carcinogenic conducts its tests to determine if the chemical is activity of environmental or pharmaceutical reasonably safe. agents have utilized a number of animal models. The first philosophic approach would tend to The majority of NTP studies have been con- lead to situations where the animals might be ducted with the Fischer344/N (F344) rat and overdosed, whereas the second approach would the B6C3F1/N mouse (a hybrid of the C57BL/6 tend toward underdosing. There is no right or and C3H parental strains). The F344 rat is prone wrong to either philosophy, so long as the study to developing nephropathy, mononuclear cell results are interpreted appropriately. In addition leukemia, seizures, and/or testicular interstitial (Leydig) cell tumors, whereas the B6C3F1 mouse TABLE 5.12 Frequency of Positive Cancer Response in is genetically susceptible to the spontaneous 290 Positive NTP 2-Year Rat or Mouse development of liver and lung cancer. These Cancer Studies spontaneous lesions can sometimes interfere Liver 57% with the interpretation of the bioassay. More recently, the NTP has examined the utility of Lung 22% the Wistar Han and the Sprague-Dawley rat to Kidney 22% replace the F344 rat, and has completed analysis of genetically modified mice – p53 deficient, Mammary gland 14% Tg.AC v-HA-ras transgenic, and p16(INK4a)/ Hematopoietic 13% p19(ARF) – for screening carcinogens. Additional carcinogenicity models have been Forestomach 12% developed, including a rat urinary bladder Thyroid 10% model, a rat pancreatic cancer model, a gastric cancer model, a rat thyroid cancer model, a fish Vascular system 9% liver neoplasm model, and rat and mouse colon

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 136 5. CARCINOGENESIS cancer models. Each has its own limitations. A results obtained using the “gold standard” short-term multi-organ test system for carcino- chronic rodent carcinogenicity test. Few individ- gens has been proposed in which rats are treated uals seem to have considered simple questions by multiple carcinogens to cause initiation in such as: Why would a short-duration lung several target tissues followed by administration tumor assay system be expected to identify of the test chemical for 12 weeks. Endpoints that chemicals that produce renal cancer in a chronic lend themselves to quantitation include preneo- rodent study? Despite this unfortunate short- plastic lesions in the liver, thyroid, lung, fore- coming, these alternative in vivo carcinogenicity stomach, urinary bladder, and esophagus. models can play an important role in defining Over the years it has become apparent that chemical carcinogenesis. there is no panacea for identifying potential First, in vivo “short-term” tests are useful in human carcinogens, there is no absolute refer- defining the nature of the carcinogenic response ence for carcinogenicity, and virtually no human observed in a chronic rodent carcinogenicity surrogate is ideal or immune to criticism. Equiv- study. For specific target tissues, they can iden- ocal results can occur in any animal model no tify whether a chemical is an initiator or matter how well a carcinogenicity study is promoter, and they can help define the relative designed. In the last analysis, carcinogenicity potency of the carcinogen. Second, they can data are most relevant to the test model help elucidate the mechanistic basis of carcino- employed. The greatest potential of the various genesis. Third, when used as part of a battery “short-term” organ-specific animal carcinogen- or tier approach to carcinogen testing, they esis models is in understanding the underpin- help set priorities for more extensive carcinoge- nings of the carcinogenic process. nicity testing of chemicals. When used with potent carcinogens, a preneo- plastic or neoplastic response is often observed 3.2. Data Evaluation and Interpretation in a few weeks, hence the designation “short term.” However, for less potent carcinogens and Interpretation of results from human epidemi- for non-carcinogens, tests are typically conducted ological studies and animal bioassays to identify for several months. In the case of the mouse skin- carcinogenic agents has often proven difficult painting model, lifetime studies may sometimes and controversial. Humans are rarely exposed be conducted to assess carcinogenicity. Because to only one potential cancer-causing agent in of the sharp contrast in the study duration for their lifetime, and the amount and duration of these so-called “short-term” in vivo test models that exposure may be difficult or impossible to relative to truly “short-term” (days or weeks) in quantify rigorously. Many years may intervene vivo test systems, they have more recently been between exposure to a potential carcinogen and referred to as “medium-term” bioassays for ultimate development of neoplasia, making carcinogens, indicating a study shorter than the accurate assessment of cause and effect almost conventional 2-year rodent carcinogenicity study. impossible. Relative to man, rodents infrequently “Short-term” or “medium-term” in vivo carci- develop prostate, colon, pancreatic, cervical, and nogenesis animal models consist of an intact uterine carcinomas, but develop lung, mammary, animal capable of chemical metabolism (activa- hematopoietic, and skin tumors at similar rates tion and detoxification), complex tissue and (Table 5.13). Neoplasms of the liver, kidney, hormonal interactions, and possessing repair and thyroid are much more frequent in rodents mechanisms, all of which could influence the than in humans. ultimate expression of carcinogenic responses Despite such limitations, epidemiological to chemical exposure. studies that clearly show an association between Various investigators have proposed specific a given chemical exposure or lifestyle habit and “short-term” or “medium-term” in vivo rodent an enhanced rate of a specific cancer are models to supplant the more costly chronic regarded as the most relevant method for identi- rodent carcinogenicity studies (bioassays). In fication of human carcinogens. While animal those instances where appropriate validation bioassays have proven useful for identification studies have been conducted, these models of agents that can cause cancer in the laboratory have shown unacceptable concordance with rodent, they only identify an agent as potentially

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 3. IDENTIFYING CARCINOGENS 137 hazardous to human health. Additional facts and induction of renal tumors in the male rat by factors must be considered in classifying such an some chemicals appears dependent on a path- a agent as a likely human carcinogen (see Statistical way involving the deposition of 2u-globulin in a Assessment of Toxicologic Pathology Studies, the tubular epithelium; 2u-globulin is not found Chapter 30). in humans. Thereby, there is less concern about The current approach for assessing the scien- potential carcinogenic effects of such chemicals tific relevance of either epidemiological or in humans when they induce only kidney tumors animal bioassay results to human health risk in male rats by that specific mode of action. involves a “weight-of-evidence” procedure Another example is the occurrence of thyroid in which national and international panels neoplasms in rats treated with non-genotoxic of expert scientists from several disciplines enzyme-inducing agents that in turn increase examine all available information on the the amount of UDP-glucuronyl transferase, suspect agent and reach a consensus opinion. which binds to thyroxin (T4). Because of the Included in this analysis are the strength of excreted and lower circulating levels of T4, the epidemiological evidence, the dose– thyroid stimulating hormone (TSH) is released response curve of the animal response, compar- and this causes secondary proliferation (hyper- ative species metabolism and ability to extrapo- plasia and neoplasia of the thyroid). late between species, likely mechanism of The high quality of the pathology evaluation cancer induction for the agent in question, the for both neoplastic and non-neoplastic lesions is genotoxicity of the agent, the amount of the a result of rigorous peer-review (see Peer Review agent in the environment, and the number of and Pathology Working Groups, Chapter 17). The people potentially exposed to the agent. Based final interpretation of study results is likewise on this type of analysis, so far more than 100 subjected to intensive peer review, and any agents have been classified as known human carcinogenic response is categorized according carcinogens by the International Agency for to NTP levels of evidence (Table 5.14) of carcino- Research on Cancer (IARC) and the US Health genic activity. All of the above has resulted in and Human Services Annual Report on Carcin- markedly improved data, but has required ogens (ROC); virtually all are carcinogenic in a substantial expenditure of time and money. animals (Table 5.1). There are positive and negative attributes of Due to limitations in the animal studies, there the contemporary NTP 2-year rodent carcinoge- remains a lack of certainty in determining if nicity study. On the positive side, the test as per- a chemical is not carcinogenic to either humans formed currently is a thoroughly conducted or animals. The interpretation of animal carcino- and peer-reviewed assessment of carcinogenicity genicity results has proven difficult and contro- and chronic toxicity for the chemical under versial because there are concerns about study. The study results are reasonably inter- adequacy of the rodent bioassay to represent preted, and provide a better basis for risk assess- the potential human risk. ment than previously designed rodent The central issues for interpretation are those “bioassays” permitted. All data are published data dealing with inter-species and low-dose to allow for close scrutiny of the scientific basis extrapolation. For example, the chemical for each interpretation. To date, the 2-year rodent carcinogenicity TABLE 5.13 A Comparison of Frequent and Infrequent study appears to be the single best system for Tumor Types in the Human and Rodent identifying carcinogenic potential. Not only has the 2-year rodent carcinogenicity test been suffi- High in treated rodents but Liver, kidney, thyroid ciently sensitive in identifying known human not in humans: carcinogens, it has also identified carcinogens High in both treated rodents Lung, hematopoietic, such as aflatoxin, 4-aminobiphenyl, bis(chlo- and humans: mammary, skin romethyl)ether, diethylstilbestrol, melphalan, mustard gas, and vinyl chloride prior to High in humans but not in Prostate, colon, discovery of their carcinogenicity in humans. treated rodents: pancreas, brain, Disadvantages of the chronic rodent carcino- cervix, uterus genicity test relate primarily to the high cost of

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 138 5. CARCINOGENESIS each study and to the long time interval required cause cancer in mice or rats. Using judgment to obtain results. The high costs make the studies and scaling in assessing the significance of the too expensive to repeat, thereby obviating the carcinogenicity findings should qualify even principle of reproducibility demanded by good this endpoint. For example, a chemical that science. However, in 70 “near-replicate” compar- causes an unequivocal increase in malignant isons, good reproducibility of positivity, target neoplasia in multiple sites and in both sexes of site, and carcinogenic potency was found by rats and mice should be given different consider- comparing published studies done in rats, ation than a chemical that causes a marginal mice, or hamsters. increase of a benign neoplasm whose back- The identification of rodent carcinogens on ground incidence is normally quite variable in the basis of repeated exposure to high doses of one sex of one species. This approach is referred chemical has been criticized as not realistic to as the “weight-of-evidence” method of inter- with respect to human exposure situations in pretation, and it goes further than just evaluating all cases, although clear exceptions exist. neoplasm incidence data. Another criticism of the current 2-year rodent Also included in the “weight-of-evidence” carcinogenicity study is that it does not define method for assessing potential carcinogenic risk the mechanisms of any observed carcinogenesis. to humans is consideration of the nature of the However, this criticism seems unfounded in that dose–response curve, the pharmacokinetics of the “standard” chronic rodent carcinogenicity the chemical, whether the maximum tolerated test is not designed to define mechanisms of dose (MTD) was exceeded in the carcinogenicity carcinogenesis. While cost and lack of evidence study, comparative species metabolism, route of reproducibility continue to be valid criticisms of administration, genotoxicity, cytotoxicity, and of the chronic rodent carcinogenicity test, it still many other factors (see Pharmacokinetics and Tox- remains the best test system for definitively icokinetics, Chapter 2). identifying potential human carcinogens. The judgment of whether a chemical poses An important issue relative to the chronic a carcinogenic hazard should be made in light of rodent carcinogenicity test is how the results the total evidence provided by all sources of are interpreted and used. The studies determine available relevant information. The MTD concept if a particular chemical under specific conditions continues to be one of the most debated and controversial issues in toxicology. Studies that are negative for carcinogenicity and were conducted TABLE 5.14 National Toxicology Program’s Levels of with doses below the MTD leave doubt about the Evidence (LOE) of Carcinogenic Activity adequacy of the test. One never knows if the Clear evidence of carcinogenic activity animals were sufficiently exposed to the chemical. Studies conducted at or above the MTD may Dose-related (i) increase of malignant neoplasms, compromise host homeostasis to such a degree (ii) increase of a combination of malignant and benign that a positive result has little relevance to human neoplasms, or (iii) marked increase of benign exposure situations. Because it is not always neoplasms if there is an indication from this or other possible to determine what the MTD will be prior studies of the ability of such tumors to progress to malignancy to exposure of the animals over much of their lifespan, the real problem comes with how the Some evidence of carcinogenic activity study results are used in risk assessment. Thus, Chemical-related increased incidence of neoplasms in conclusions regarding carcinogenicity or lack which the strength of the response is less than that thereof as determined in chronic rodent tests required for clear evidence must be qualified as occurring under conditions of the specific study in question. The limitations Equivocal evidence of carcinogenic activity of the particular study should be pointed out as Marginal increase of neoplasms that may be chemical they relate to MTD, the original design and related purpose of the study, study conduct, confound- ing toxicity, and other study conditions. No evidence of carcinogenic activity In addition, some standardized approach for Reproduced courtesy of the NTP. describing the reasoning used to arrive at

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 3. IDENTIFYING CARCINOGENS 139 a conclusion about the carcinogenicity of a chem- substitution occurs either through oxidative ical is desirable conceptually and for consistent mechanisms or by direct electrophilic attack, interpretation. For this purpose, categorization and results in a change of base pairing because of the observed response using levels of evidence of deamination or a shift in tautomeric (struc- (see Table 5.14) is helpful. tural isomer) equilibrium. Examples include nitrous oxide (NO), which causes oxidative 3.3. Genotoxic and Non-Genotoxic deamination, and formaldehyde, which forms Carcinogens crosslinks within DNA. Formaldehyde also causes hydroxymethyl adducts, and thus may Supportive of the arguments for the mutational also act as an alkylating agent. hypotheses of cancer causation are the correla- tion between mutagenicity and carcinogenicity, Alkylating Genotoxic Agents the correlation between faulty DNA repair mech- Alkylating chemical carcinogens either anisms and some cancers, and the heritable nature interact directly with cellular genomic material of neoplastic transformation. While the correla- (direct acting carcinogens) or must first be tion between mutagenicity and carcinogenicity is metabolized by the host to a reactive species not perfect, many chemical carcinogens are muta- (indirect acting carcinogens). Examples of direct genic either alone or after metabolic activation. alkylating agents include methylnitrosurea, eth- Because it is well known that some mutagens ylnitrosourea, methylmethane sulfonate, and are not carcinogens and some carcinogens are ethylmethane sulfonate. Indirectly acting alkylat- not mutagens, invocation of the association ing agents require metabolic activation to an elec- between mutagenicity and carcinogenicity neces- trophilic intermediate and include chemicals sarily must be qualified. Practical considerations such as diethylnitrosamine, benzo[a]pyrene, logically compel us to admit that no single muta- nitrofurans, nitroquinoline oxide, ethylene genicity test or even battery of mutagenicity tests dibromide, ethylene dichloride, N-acetyl-2-ami- would be expected to reflect the complexity and nofluorene, and dimethylhydrazine. diverse mechanisms of carcinogenesis. Nonethe- Pathways for activation involve cytochrome less, the association between mutagenicity and P450, reduction with NADPH diaphorase, reac- carcinogenicity, where it occurs, is intuitively tion with glutathione, or oxidation mediated by appealing as a potential mechanism for the peroxidases or via prostaglandin synthetase. For genesis of some cancers. specifics regarding metabolic activation of differ- Endogenous hereditary defects in mutational ent types of chemical carcinogens, the reader is repair in the face of exposure to environmental encouraged to consult cited reference material; mutagenic factors are known to lead to some see also Biochemical and Molecular Basis of Toxicity cancers. Most notable is the occurrence of carci- (Chapter 1). For both direct and indirect acting nomas and melanomas of the skin in xeroderma pig- agents, the reactive form of the carcinogen is mentosum patients who sustain mutations caused an electron-deficient species (electrophilic form) by ultraviolet radiation. Because a constellation that interacts non-enzymatically with electron- of biochemical abnormalities may be present in rich or nucleophilic molecular sites in the cell. the same cells that have a defective excision These nucleophilic sites are not limited to repair of pyrimidine dimers, it cannot be DNA but also include RNA and cellular protein. adamantly stated that the observed skin cancers Thus, the reaction is not specific for genomic result exclusively from the accumulation of material or for nucleic acids. The interaction of mutations resulting from deficient DNA repair. electrophilic forms of the carcinogen with host However, the association between defective cellular material results in the formation of cova- DNA repair and carcinogenesis in xeroderma pig- lent adducts (addition products). These interac- mentosum patients lends support to the somatic tions take place primarily between the mutation hypothesis of cancer formation. electrophilic form of the carcinogen and the nitrogen, oxygen, and sulfur atoms in the cellular Non-Alkylating Genotoxic Agents macromolecules. This class of carcinogen directly substitutes for Adducts formed may be small, as is seen with exocyclic amino groups of nucleosides. The simple alkylating agents, or large, so-called

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 140 5. CARCINOGENESIS

“bulky adducts,” as occur with polycyclic Cancer risk assessment historically has relied aromatic hydrocarbons. In either case, configura- upon classic epidemiology, results from chronic tional or conformational changes occur in the exposures of rodents to potential carcinogens, DNA, which can lead to steric hindrance and and mathematical models of these findings. The result in infidelity of DNA replication. process of carcinogenesis in rodents is a temporal The electrophilic forms of carcinogens react at sequence that represents months in rodents, as numerous macromolecular cellular sites and opposed to many years in humans. The field of produce adducts and mutations throughout the risk assessment has been forced to be cautious genome. Some of these mutations may be lethal. mainly because of the limited knowledge of the Non-lethal reactions with cellular targets are the complex pathogenesis of cancer. Public health most relevant for carcinogenesis in that initiation strategies are now based upon the premise that occurs when one of these is in a critical genomic reduction or prevention of exposure to cancer- site. Non-lethal alteration in the somatic cell causing factors, whether endogenous or exoge- genotype is consistent with the heritable change nous, will decrease the incidence of cancer. that is observed in cancer cells. Risk assessment requires numerical estimates Evidence of alteration of the cellular genome is of human health risks, and numerous mathe- supported by the observation of chromosomal matical models have been developed based on abnormalities and altered gene expression in animal bioassays, human exposure data, cancer cells. Furthermore, adduct formation at and epidemiology studies to predict a tumor res- DNA nucleophilic sites such as phosphate resi- ponse in a population. These models are based dues and hydrogen-bonding sites between base on the theories of multistage carcinogenesis, and pairs can also result in miscoding or in sufficient take into account initiation, proliferation, multi- distortion of DNA structure to result in infidelity ple genetic alterations, preneoplastic lesions, and of DNA replication. Because DNA damage the monoclonality of cancer development. The (adduct formation) can be repaired efficiently models are only as good as the input data, and by cellular enzymes, it is necessary for cell prolif- often have uncertainties in their predictive value eration to occur prior to repair of the DNA because of factors such as variability of response damage in order for a heritable change to occur. in outbred populations, sensitivity in measuring Once a round of cell proliferation takes place, tissue dose, and for low-dose extrapolations. the genomic alteration is said to be “fixed” and These models are useful in determining allowable the affected cell(s) is initiated. The round or levels of human exposures, but it should be rounds of cell replication that serve to fix the remembered that they are often based on our molecular lesions may occur de novo, may be imprecise knowledge. As our scientific under- induced by the inherent toxicity of the chemical standing increases, these “provisional” models carcinogen, or may be induced by the promoting will be adjusted to better represent the risks. activity of the chemical carcinogen. Epidemiologic studies that clearly show an association between chemical exposure and an 3.4. Risk Assessment increased risk for the development of cancer provide the strongest evidence for identifying Considering a risk assessor is required to esti- human carcinogens. For most suspected carcino- mate the dose of an environmental carcinogen gens, it is often difficult or impossible to quantify that will lead to an increased lifetime probability the amount and duration of exposure in humans for cancer of one in a million, continued efforts to or even to associate the exposure with a disease accumulate and utilize relevant information are that may not appear for decades. The animal necessary to make scientifically sound estimates. bioassay identifies agents that are potentially These estimates must also reflect the many hazardous to human health and must be consid- endogenous and exogenous factors that influ- ered along with as many scientifically-based ence cancer development, including metabolism analyses as are known, such as the dose– of carcinogens, DNA repair capacities, variable response relationship, comparative species genomic stability among animal species, metabolism, ability to extrapolate between inherited predispositions, immune status, species, mechanisms of cancer induction, geno- hormones, and our environment. toxicity of the agent, amount of agent in the

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 3. IDENTIFYING CARCINOGENS 141 environment, number of people exposed, and may lead to the identification of more human strength of the epidemiologic evidence in order carcinogens. Moreover, animal studies, such as to estimate the likelihood that a chemical is carci- in the mouse liver tumor models, may help nogenic in humans (see Pathology Issues in the further define these associations and elucidate Design of Toxicology Studies, Chapter 19). the effects of dose. Retrospective or prospective epidemiological It has been proposed that the carcinogen- methods to study human populations using specific patterns of ras gene mutation observed existing historical records associated with known in tumors reflect the mechanisms of carcinogen- cases of neoplasia have demonstrated the associ- esis, and some patterns of H-ras mutations in ation of cigarette smoking with lung cancer, and mouse liver tumors suggest that they occur exposure to asbestos with mesotheliomas. from direct chemical–ras gene (genotoxic) interactions. 3.5. Molecular Epidemiology The B6C3Fl mouse liver is a model in which many genotoxic and non-genotoxic chemicals Proto-oncogene and tumor suppressor gene induce hepatocellular tumors. H-ras activation mutation assays have become a popular tool is a common event in tumorigenesis, and the for investigating tumor etiology in humans frequency and pattern of activating mutations and rodents mainly because mutations tend to are compared between tumors that are chemi- be chemical specific (see The Application of Toxi- cally-induced and spontaneous to help in cogenomics to the Interpretation of Toxicologic determining possible mechanisms by which Pathology, Chapter 11). Because the inactivating chemicals induce tumors. Some genotoxic mutations of the p53 gene are primarily compounds cause mutations that can be missense mutations (90%) and there are many explained by the type of DNA adducts formed “hotspots,” it is “well suited for this form of and some non-genotoxic carcinogens induce molecular archaeology” in examining human tumors that have the same mutational spectra and animal tumors. Study of the patterns of as spontaneous tumors, suggesting that these oncogene activation in spontaneous versus chemicals act by promoting the spontaneously chemically-induced rodent neoplasms has initiated hepatocytes. provided data that suggest that the molecular Molecular epidemiology of cancer is the lesions associated with chemically-induced study of molecular alterations, primarily muta- cancer are sometimes different from those docu- tions, in investigating the causative agents of mented in spontaneous cancer. Furthermore, cancer, as well as identifying individual cancer the patterns of oncogene activation in several risk. The objective to identify cancer-causing rodent model systems appear to be carcin- agents based on the occurrence of predictable ogen-specific, are consistent with known or molecular alterations that are found in the expected DNA adduct formation, and in some neoplasm is intriguing. It is based on the cases are similar to patterns of oncogene activa- hypothesis that there are carcinogen-specific tion documented in human neoplasms. patterns of mutations that reflect direct interac- Mutations in the ras gene from skin tumors in tions of carcinogens with cancer genes. For sun-exposed areas appear to be induced by UV example, lung and colon cancers from people irradiation; K-ras mutations in human lung and who smoke tend to have a specific mutation colon tumors may be the result of adduct forma- in the ras oncogene or p53 tumor suppressor tion due to carcinogens from cigarette smoke; gene (i.e., mostly a G to T nucleotide base and aflatoxin B1-associated human liver tumors substitution) and that this mutation is likely appear to have a specific mutation at codon due to direct interaction of the carcinogen in 249 of the p53 gene. There is epidemiological smoke benzo[a]pyrene with DNA. evidence suggesting that for some solvent- Such chemical-specific mutational profiles associated human leukemias, the exposure was (or “molecular signatures”) have been used to associated with an increased risk of developing establish a causal nature between particular a tumor with ras activation. genetic events in tumors and carcinogen, such The analysis of subpopulations of humans at as neoplasms associated with exposure to risk with unique genetic alterations in tumors radon, aflatoxin B1, vinyl chloride, and the

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 142 5. CARCINOGENESIS nitrosamines. The strongest evidence for linkage in tumor DNA include those caused by indirect between a cancer-causing agent and a specific oxidative DNA damage via oxygen or lipid radi- type of neoplasm is that of the CC to TT double cals produced during carcinogen metabolism or base changes observed in skin neoplasms of man cytotoxic injury; a selective growth promotion or and animals. This mutation is consistent with death of tumor cells with a specific ras genotype; the predicted UV-induced damage of dipyrimi- or a hypermutable state induced by the carcin- dine dimers. In liver tumors from persons living ogen (Table 5.11). Other factors that can affect in geographic areas with a high exposure to afla- the profile of ras mutations found in mouse liver toxin B1 there is a frequent mutation at the third tumors, and thus add an additional level of nucleotide pair of codon 249 in the p53 gene, complexity in the evaluation of chemical geno- suggesting the mutation is chemical-specific toxicity in tumor DNA, include genetic back- and imparts a specific growth or survival advan- ground (strain) of mouse, dose, dosing tage to the mutated liver cells. regimen, and time of sampling. In some cases Animal studies have confirmed that there are when humans and rodents are exposed to the certainly chemical-specific mutational profiles same mutagenic carcinogen, such as vinyl chlo- in neoplasms; however, there are many examples ride, the specific cells or tissue at risk and where the mutational profile varies by strain specific mutations differ. (Table 5.15), species, dose, or dosing regimen. Using tumor DNA from the National Toxi- For example, diethylnitrosamine, a strong, cology Program (NTP) carcinogenesis testing cross-species hepatocarcinogen, will induce liver program to detect dominant transforming genes, neoplasms in mice, rats, and rainbow trout, but patterns of oncogene activation have been docu- the frequency and type of ras mutation in the mented for spontaneous and chemically-induced neoplasm varies widely, and the mutations are neoplasms in Fischer 344 rats and in B6C3Fl not simply a reflection of direct DNA mice. Results to date show a low frequency interaction. (about 3%) of activated oncogenes in sponta- In some studies, in vitro mutation assays neous neoplasms of the Fischer 344 rats. are a poor predictor of liver tumor mutation In contrast, a variety of epithelial neoplasms profiles in the mouse. In this complex process, induced by benzidine congeners as well as lung carcinogens might also be influencing events tumors induced by tetranitromethane inhalation such as DNA repair, oxidative DNA damage, have shown a high frequency of oncogene activa- methylation, cell death, proliferation, and/or tion. In the B6C3F1 mouse there is a high a hypermutable state. frequency (56%) of oncogene activation in spon- Other possible mechanisms by which carcin- taneous liver tumors and a lower frequency (8%) ogen treatment could affect ras mutation profile in spontaneous non-liver tumors. The oncogenes

TABLE 5.15 Species and Strain Comparison of the Variety of Frequencies and Spectra of ras Mutations in Hepatocellular Neoplasms Induced by Diethylnitrosamine (DEN)a

Codon 61 mutation (normal [ CAA) H-ras codon 61 Strain mutations AAA CGA CTA CCA Other ras C3H 54/114 (47%) 28 24 2 0 0 B6C3F1 63/239 (26%) 16 32 15 0 0 CD-1 9/25 (36%) 7 1 0 0 N-ras (4/25; 16%) C57BL 2/59 (3%) 0 1 0 1 0 Rainbow trout eeeeeK-ras (6/7; 85%) F344 rat 0/10 eeee0/10 a DEN, diethylnitrosourea from various dosages and dosing regimens. Data combined from multiple studies. Mouse data mainly from Maronpot et al. (1995). Mutations in the ras proto-oncogene: Clues to etiology and molecular patho- genesis of mouse liver tumors. Toxicology 101, 125–156.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 4. CONCLUSIONS 143 in spontaneous liver tumors in the B6C3F1 (MNU), and such mutations would be expected mouse most frequently involve activation of the to lead to nucleotide mispairing during DNA H-ras proto-oncogene by point mutations in the replication. The altered DNA would be expected sixty-first codon. to give rise to an abnormal protein product that In chemically-induced hepatic neoplasia in could theoretically alter cell growth or this mouse hybrid, there is also a high frequency differentiation. of oncogene activation, but the pattern of onco- MNU is very labile, with an estimated biolog- gene activation differs from that observed in ical half-life of a few minutes. Thus, it is likely spontaneous liver tumors. In addition to a “back- that the effect produced by this chemical carcin- ground” level of mutations in the sixty-first ogen occurred as an early event in the carcino- codon identical to those observed in spontaneous genic process. Because hormone-stimulated cell hepatocellular neoplasms, additional point division in the mammary tissue is also necessary mutations were documented in codons 12 or for the development of mammary neoplasia in 117 of the H-ras proto-oncogene in addition to this model, it is unlikely that activation of the the activation of K-ras and N-ras oncogenes in H-ras gene is sufficient in and of itself for the chemically treated mice that had an unequivocal production of mammary neoplasia. Another increase in hepatocellular tumors. example of association of an activated H-ras Molecular epidemiologic studies aimed at gene with neoplasia is seen in the carcinomas identifying an individual’s risk of developing produced in mouse skin by initiation with 7,12- cancer have found that persons with germ- dimethylbenz[a]anthracene (DMBA) followed line mutations in cancer genes (i.e. BRCA1 by promotion with phorbol ester. In this instance or BRCA2) or variations (polymorphisms) of there is an A–T transversion (mutation) in the carcinogen-metabolizing enzyme activities (i.e., sixty-first codon of the H-ras gene. Once again, cytochrome p450s or glutathione-S-transferases) the oncogene activation produced by the or DNA repair capacities can be at increased risk DMBA treatment was not sufficient to cause of developing neoplasia in their lifetime. High- carcinoma development but required promotion throughput analyses to examine single nucleo- by phorbol ester to obtain cancer. Rodent test tide polymorphisms (SNPs) are being used to systems used for the in vivo detection of chemi- search for biomarkers of cancer-risk individuals, cal carcinogens often depend on demonstration and some of this information is being used to of an increased incidence of common neoplasms help people to take preventive measures to or on the induction of novel neoplasms in chron- decrease their risk of developing cancer. ically treated animals. The most frequently identified activated onco- genes detected in chemically-induced neoplasms belong to the ras gene family. Ras oncogenes were 4. CONCLUSIONS originally detected and isolated by transfection of NIH3T3 cells (immortalized mouse fibro- Research is contributing to our better under- blasts) using DNA from tumors; this method of standing and interpretation of in vivo and in vitro detection has been replaced by newer molecular findings of environmentally related carcinogen- techniques. Cloning of these genes from the esis. Studies are identifying the many similari- tumor DNA revealed that many were H-ras, K- ties in the molecular pathogenesis of cancer ras,orN-ras genes that differed from their between rodents and humans. The new technol- proto-oncogene homologs by a single point ogies will continue to contribute towards our mutation located in a specific codon. Thus, improved understanding of human risks of when adolescent rats are given a single exposure disease and the betterment of human health. to nitrosomethylurea (NMU), mammary tumors Although cancer research appears to progress result that have an activated H-ras oncogene at a snail’s pace, take heart that overall we are with a G–A transition (mutation) in the twelfth making great strides in preventing and treating codon. cancer. The animal models are at least in part G–A transitions are consistent with the Q6- responsible for the steady increases in the methyl guanine adduct that is formed by meth- average American’s lifespan seen over the past ylating agents such as methyl-nitrosourea three decades.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 144 5. CARCINOGENESIS

Acknowledgments Maronpot, R.R., Boorman, G.A., 1996. The contribution of the mouse in hazard identification studies. Toxicol. Pathol. 24, We are grateful to Dr Stephen Mastorides for his contribu- 726–731. tions to previous editions of the Handbook of Toxicologic Pritchard, J.B., French, J.E., Davis, B.J., Haseman, J.K., 2003. Pathology, some of which remain with permission in this The role of transgenic mouse models in carcinogen iden- updated chapter. We also thank Elizabeth Ney of NIEHS for tification. Environ. Health Perspect. 111, 444–454. her assistance generating the tables and figures. Tennant, R.W., Margolin, B.H., Shelby, M.D., Zeiger, E., This research was supported (in part) by the Division of the Haseman, J.K., Spalding, J., Caspary, W., Resnick, M., National Toxicology Program of the NIH, National Institute of Environmental Health Sciences. Stasiewwicz, S., Anderson, B., Minor, R., 1987. Prediction of This chapter may be the work product of an employee or chemical carcinogenicity in rodents from in vitro genetic group of employees of the National Institute of Environmental toxicity assays. Science 236, 933–941. Health Sciences (NIEHS), National Institutes of Health (NIH); however, the statements, opinions, or conclusions contained herein do not necessarily represent the statements, opinions, Molecular Carcinogenesis or conclusions of NIEHS, NIH, or the United States government. Fearon, E.R., Vogelstein, B., 1990. A genetic model for colo- rectal tumorigenesis. Cell 61, 759–767. Hanahan, D., Weinberg, R.A., 2011. Hallmarks of cancer: the next generation. Cell 144, 652–674. SUGGESTED READING Hussain, S.P., Hofseth, L.J., Harris, C.C., 2001. Tumor suppressor genes: at the crossroads of molecular carcino- Cancer Bioassays genesis, molecular epidemiology and human risk assess- Beyer, L.A., Beck, B.D., Lewandowski, T.A., 2011. Historical ment. Lung. Cancer 34, S7–S15. perspective on the use of animal bioassays to predict car- Kinzler, K.W., Vogelstein, B., 1996. Lessons from hereditary cinogenicity: evolution in design and recognition of utility. colorectal cancer. Cell 87, 159–170. Crit. Rev. Toxicol. 41, 321–338. Loeb, L.A., 1994. Microsatellite instability: marker of a mutator Boorman, G.A., Maronpot, R.R., Eustis, S.L., 1994. Rodent phenotype in cancer. Cancer Res. 54, 5059–5063. carcinogenicity bioassay: past, present, and future. Toxicol. Malarkey, D.E., Parker, J.S., Turman, C.A., Scott, A.M., Pathol. 22, 105–111. Paules, R.S., Maronpot, R.R., 2005. Microarray data analysis Borowsky, A.D., Munn, R.J., Galvez, J.J., Cardiff, R.D., of mouse neoplasia. Toxicol. Pathol. 33, 127–135. Ward, J.M., Morse 3rd, H.C., Kogan, S.C., Aldape, K.D., Vogelstein, B., Kinzler, K.W., 2004. Cancer genes and the Louis, D.N., Bosenberg, M.W., 2004. Mouse models of pathways they control. Nat. Med. 10, 789–799. human cancers (part 3). Comp. Med. 54, 258–270. Wood, L.D., Parsons, D.W., Jones, S., Lin, J., Sjo¨blom, T., Boyland, E., 1969. The correlation of experimental carcinogenesis Leary, R.J., Shen, D., Boca, S.M., Barber, T., Ptak, J., and cancer in man (review). Prog. Exp. Tumor Res. 11, 222–234. Silliman, N., Szabo, S., Dezso, Z., Ustyanksky, V., Cohen, S.M., Arnold, L.L., 2011. Chemical Carcinogenesis. Nikolskaya, T., Nikolsky, Y., Karchin, R., Wilson, P.A., Tox. Sci. 120 (S1), S76–S92. Kaminker, J.S., Zhang, Z., Croshaw, R., Willis, J., Farber, E., 1984. Cellular biochemistry of the stepwise devel- Dawson, D., Shipitsin, M., Willson, J.K., Sukumar, S., opment of cancer with chemicals. Cancer Res. 44, 5463–5474. Polyak, K., Park, B.H., Pethiyagoda, C.L., Pant, P.V., Hoenerhoff, M.J., Hong, H.H., Ton, T.V., Lahousse, S.A., Ballinger, D.G., Sparks, A.B., Hartigan, J., Smith, D.R., Sills, R.C., 2009. A review of the molecular mechanisms of Suh, E., Papadopoulos, N., Buckhaults, P., Markowitz, S.D., chemically induced neoplasia in rat and mouse models in Parmigiani, G., Kinzler, K.W., Velculescu, V.E., National Toxicology Program bioassays and their relevance Vogelstein, B., 2007. The genomic landscapes of human to human cancer. Toxicol. Pathol. 37, 835–848. breast and colorectal cancers. Science 318, 1108–1113. Huff, J., Haseman, J., 1991. Long-term chemical carcinogenesis experiments for identifying potential human cancer Hepatocarcinogenesis hazards: collective database of the National Cancer Insti- tute and National Toxicology Program (1976–1991). Allen, D.G., Pearse, G., Haseman, J.K., Maronpot, R.R., 2004. Environ. Health Perspect. 96, 23–31. Prediction of rodent carcinogenesis: an evaluation of pre- Huff, J., Cirvello, J., Haseman, J., Bucher, J., 1991. Chemicals chronic liver lesions as forecasters of liver tumors in NTP associated with site-specific neoplasia in 1394 long-term carcinogenicity studies. Toxicol. Pathol. 32, 393–401. carcinogenesis experiments in laboratory rodents. Environ. Christensen, J., Romach, E.H., Healy, L.N., Gonzales, A.J., Health Perspect. 93, 247–270. Anderson, S.P., Malarkey, D.E., Cattley, R.C., Huff, J., Jacobson, M.F., Davis, D.L., 2008. The limits of two- Goldsworthy, T.L., 1999. Altered Bcl-2 family expression year bioassay exposure regimens for identifying chemical during nongenotoxic hepatocarcinogenesis in mice. Carci- carcinogens. Environ. Health Perspect. 116, 1439–1442. nogenesis 20, 1583–1590.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy SUGGESTED READING 145

Hailey, J.R., Haseman, J.K., Bucher, J.R., Radovsky, A.E., Initiation, Promotion, and Progression Malarkey, D.E., Miller, R.T., Nyska, A., Maronpot, R.R., 1998. Impact of Helicobacter hepaticus infection in B6C3F1 Argyris, T.S., 1985. Promotion of epidermal carcinogenesis by mice from twelve National Toxicology Program two-year repeated damage to mouse skin. Am. J. Ind. Med. 8, 329–337. carcinogenesis studies. Toxicol. Pathol. 26, 602–611. Borzsonyi, M., Day, N.E., Lapis, K., Yamasaki, H. (Eds.), 1984. Hall, A.P., Elcombe, C.R., Foster, J.R., Harada, T., “Models, Mechanisms, and Etiology of Tumour Promo- Kaufmann, W., Knipple, A., Kuttler, K., Malarkey, D.E., tion.” IARC. France, Lyon. Maronpot, R.R., Nishikawa, A., Nolte, T., Schulte, A., Hennings, H., Yuspa, S.H., 1985. Two–stage tumor promotion Strauss, V., York, M.J., 2012. Liver Hypertrophy: A Review in mouse skin: An alternative interpretation. J. Natl. of Adaptive (Adverse and Non-adverse) Changes – Cancer. Inst. 74, 735–740. Conclusions from the 3rd International ESTP Expert Workshop. Toxicol. Pathol. 2012 Jun 21 [Epub]. Hoenerhoff, M.J., Pandiri, A.R., Lahousse, S.A., Hong, H.H., Clonal Origins Ton, T.V., Masinde, T., Auerbach, S.S., Gerrish, K., Fialkow, P.J., 1979. Clonal origin of human tumors. Annu. Rev. Bushel, P.R., Shockley, K.R., Peddada, S.D., Sills, R.C., 2011. Med. 30, 135–143. Global gene profiling of spontaneous hepatocellular Nowell, P., 1978. Tumors as clonal proliferations. Arch. Cell carcinoma in B6C3F1 mice: similarities in the molecular Pathol. 29, 145–150. landscape with human liver cancer. Toxicol. Pathol. 39, 678–699. Malarkey, D.E., Devereux, T.R., Dinse, G.E., Mann, P.C., Cancer Stem Cell Theory (Extended Maronpot, R.R., 1995. Hepatocarcinogenicity of chlordane Bibliography) in B6C3F1 and B6D2F1 male mice: evidence for regression in B6C3F1 mice and carcinogenesis indepen- Aguirre-Ghiso, J.A., 2007. Models, mechanisms and clinical dent of ras proto-oncogene activation. Carcinogenesis 16, evidence for cancer dormancy. Nat. Rev. Cancer 7, 834–846. 2617–2625. Al-Hajj, M., Wicha, M.S., Benito-Hernandez, A., Morrison, S.J., Maronpot, R.R., Fox, T., Malarkey, D.E., Goldsworthy, T., 1995. Clarke, M.F., 2003. Prospective identification of tumori- Mutations in the ras proto-oncogene: Clues to etiology and genic breast cancer cells. Proc. Natl. Acad. Sci. U.S.A. 100, molecular pathogenesis of mouse liver tumors. Toxicology 3983–3988. 101, 125–156. Al-Hajj, M., Becker, M.W., Wicha, M., Weissman, I., Maronpot, R.R., Yoshizawa, K., Nyska, A., Harada, T., Clarke, M.F., 2004. Therapeutic implications of cancer stem Flake, G., Mueller, G., Singh, B., Ward, J.M., 2010. cells. Curr. Opin. Genet. Dev. 14, 43–47. Hepatic Enzyme Induction: Histopathology. Toxicol. Bonnet, D., Dick, J.E., 1997. Human acute myeloid leukemia is Pathol. 38, 776. organized as a hierarchy that originates from a primitive Stout, M.D., Kissling, G.E., Sua´rez, F.A., Malarkey, D.E., hematopoietic cell. Nat. Med. 3, 730–737. Herbert, R.A., Bucher, J.R., 2008. Influence of Helicobacter Conheim, J., 1867. Ueber entzundung und eiterung. Path. hepaticus infection on the chronic toxicity and carcinoge- Anat. Physiol. Klin. Med. 40, 1–79. nicity of triethanolamine in B6C3F1 Mice. Toxicol. Pathol. Conheim, J., 1875. Congenitales, Quergestreiftes Musk- 36, 783–794. elsarkon der Nireren. Virchows. Arch. 65, 64. Donnenberg, V.S., Donnenberg, A.D., 2005. Multiple drug Spectrum of Proliferative Lesions resistance in cancer revisited: the cancer stem cell hypothesis. J. Clin. Pharmacol. 45, 872–877. Bannasch, P., 1986. Preneoplastic lesions as end points in Dontu, G., Wicha, M.S., 2005. Survival of mammary stem cells carcinogenicity testing. I. Hepatic preneoplasia. Carcino- in suspension culture: implications for stem cell biology genesis 7, 689–695. and neoplasia. J. Mammary Gland. Biol. Neoplasia. 10, Bannasch, P., 1986. Preneoplastic lesions as end points in 75–86. carcinogenicity testing. II. Preneoplasia in various non- Dontu, G., Al-Hajj, M., Abdallah, W.M., Clarke, M.F., hepatic tissues. Carcinogenesis 7, 849–852. Wicha, M.S., 2003. Stem cells in normal breast development Cardiff, R.D., Anver, M.R., Boivin, G.P., Bosenberg, M.W., and breast cancer. Cell Prolif. 36 (Suppl. 1), 59–72. Maronpot, R.R., Molinolo, A.A., Nikitin, A.Y., Rehg, J.E., Durante, F., 1874. Nesso fisio-pathologico tra la struttura dei nei Thomas, G.V., Russell, R.G., Ward, J.M., 2006. Precancer in materni e las genesi di alcuni tumori maligni. Arch. Mem- mice: animal models used to understand, prevent, and ori. ed. Osservazioni. di. Chirugia. Practica. 11, 217–226. treat human precancers. Toxicol. Pathol. 34, 699–707. Eramo, A., Lotti, F., Sette, G., Pilozzi, E., Biffoni, M., Di Medina, D., 1988. The preneoplastic state in mouse mammary Virgilio, A., Conticello, C., Ruco, L., Peschle, C., De tumorigenesis. Carcinogenesis 9, 1113–1119. Maria, R., 2008. Identification and expansion of the Nowell, P.C., 1986. Mechanisms of tumor progression. Cancer tumorigenic lung cancer stem cell population. Cell Death Res. 46, 2203–2207. Diff. 15, 504–514.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146 Author's personal copy 146 5. CARCINOGENESIS

Fang, D., Nguyen, T.K., Leishear, K., Finko, R., Kulp, A.N., Pierce, G.B., 1967. Teratocarcinoma: model for a develop- Hotz, S., Van Belle, P.A., Xu, X., Elder, D.E., Herlyn, M., mentalconceptofcancer.Curr.Top.Dev.Biol.2, 2005. A tumorigenic subpopulation with stem cell proper- 223–246. ties in melanomas. Cancer Res. 65, 9328–9337. Reya, T., Morrison, S.J., Clarke, M.F., Weissman, I.L., 2001. Stem Farrar, J.D., Katz, K.H., Windsor, J., Thrush, G., cells, cancer, and cancer stem cells. Nature 414, Scheuermann, R.H., Uhr, J.W., Street, N.E., 1999. Cancer 105–111. dormancy. VII. A regulatory role for CD8þ T cells and IFN- Ricci-Vitiani, L., Lombardi, D.G., Pilozzi, E., Biffoni, M., gamma in establishing and maintaining the tumor- Todaro, M., Peschle, C., De Maria, R., 2007. Identification dormant state. J. Immunol. 162, 2842–2849. and expansion of human colon-cancer-initiating cells. Kim, C.F., Jackson, E.L., Woolfenden, A.E., Lawrence, S., Nature 445, 111–115. Babar, I., Vogel, S., Crowley, D., Bronson, R.T., Jacks, T., Schatton, T., Murphy, G.F., Frank, N.Y., Yamaura, K., Waaga- 2005. Identification of bronchioalveolar stem cells in Gasser, A.M., Gasser, M., Zhan, Q., Jordan, S., Duncan, L.M., normal lung and lung cancer. Cell 121, 823–835. Weishaupt, C., Fuhlbrigge, R.C., Kupper, T.S., Sayegh, M.H., Klein, W.M., Wu, B.P., Zhao, S., Wu, H., Klein-Szanto, A.J., Frank, M.H., 2008. Identification of cells initiating human Tahan, S.R., 2007. Increased expression of stem cell markers melanomas. Nature 451, 345–349. in malignant melanoma. Mod. Pathol. 20, 102–107. Sell, S., 2004. Stem cell origin of cancer and differentiation Kordon, E.C., Smith, G.H., 1998. An entire functional therapy. Crit. Rev. Oncol. Hematol. 51, 1–28. mammary gland may comprise the progeny from a single Sell, S., Pierce, G.B., 1994. Maturation arrest of stem cell cell. Development 125, 1921–1930. differentiation is a common pathway for the cellular origin Lang, S.H., Frame, F.M., Collins, A.T., 2009. Prostate cancer of teratocarcinomas and epithelial cancers. Lab. Invest. 70, stem cells. J. Pathol. 217, 299–306. 6–22. Lessard, J., Sauvageau, G., 2003. Bmi-1 determines the prolif- Shackleton, M., Vaillant, F., Simpson, K.J., Stingl, J., erative capacity of normal and leukaemic stem cells Nature Smyth, G.K., Asselin-Labat, M.L., Wu, L., Lindeman, G.J., 423, 255–260. Visvader, J.E., 2006. Generation of a functional Li, C., Heidt, D.G., Dalerba, P., Burant, C.F., Zhang, L., Adsay, V., mammary gland from a single stem cell. Nature 439, Wicha, M., Clarke, M.F., Simeone, D.M., 2007. Identification 84–88. of pancreatic cancer stem cells. Cancer Res. 67, 1030–1037. Singh, S.K., Hawkins, C., Clarke, I.D., Squire, J.A., Bayani, J., Liu, S., Dontu, G., Wicha, M.S., 2005. Mammary stem cells, Hide, T., Henkelman, R.M., Cusimano, M.D., Dirks, P.B., self-renewal pathways, and carcinogenesis. Breast. Cancer 2004. Identification of human brain tumour initiating cells. Res. 7, 86–95. Nature 432, 396–401. Lowenberg, B., Terpstra, W., 1998. Maturation hierarchy of Soltysova, A., Altanerova, V., Altaner, C., 2005. Cancer stem leukemic stem cells. Stem cells 16 (Suppl. 1), 85–89. cells. Neoplasma. 52, 435–440. Monzani, E., Facchetti, F., Galmozzi, E., Corsini, E., Benetti, A., Spillane, J.B., Henderson, M.A., 2007. Cancer stem cells: Cavazzin, C., Gritti, A., Piccinini, A., Porro, D., a review. ANZ. J. Surg. 77, 464–468. Santinami, M., Invernici, G., Parati, E., Alessandri, G., La Taipale, J., Beachy, P.A., 2001. The Hedgehog and Wnt sig- Porta, C.A., 2007. Melanoma contains CD133 and ABCG2 nalling pathways in cancer. Nature 411, 349–354. positive cells with enhanced tumorigenic potential. Eur. J. Till, J.E., Mc, C.E., 1961. A direct measurement of the radiation Cancer 43, 935–946. sensitivity of normal mouse bone marrow cells. Radiat. O’Brien, C.A., Pollett, A., Gallinger, S., Dick, J.E., 2007. A human Res. 14, 213–222. colon cancer cell capable of initiating tumour growth in Tu, S.M., Lin, S.H., Logothetis, C.J., 2002. Stem-cell origin of immunodeficient mice. Nature 445, 106–110. metastasis and heterogeneity in solid tumours. Lancet. Pardal, R., Molofsky, A.V., He, S., Morrison, S.J., 2005. Stem Oncol. 3, 508–513. cell self-renewal and cancer cell proliferation are regulated Welm, B., Behbod, F., Goodell, M.A., Rosen, J.M., 2003. Isola- by common networks that balance the activation of proto- tion and characterization of functional mammary gland oncogenes and tumor suppressors. Cold Spring Harb. stem cells. Cell Prolif. 36 (Suppl. 1), 17–32. Symp. Quant. Biol. 70, 177–185. Yoo, M., Hatfield, D.L., 2008. The cancer stem cell theory: Is it Park, I.K., Qian, D., Kiel, M., Becker, M.W., Pihalja, M., correct? Mol. Cells 26, 514–516. Weissman, I.L., Morrison, S.J., Clarke, M.F., 2003. Bmi-1 is Zhang, S., Balch, C., Chan, M.W., Lai, H.C., Matei, D., required for maintenance of adult self-renewing haemato- Schilder, J.M., Yan, P.S., Huang, T.H., Nephew, K.P., 2008. poietic stem cells. Nature 423, 302–305. Identification and characterization of ovarian cancer-initi- Pathak, S., 2002. Organ- and tissue-specific stem cells and ating cells from primary human tumors. Cancer Res. 68, carcinogenesis. Anticancer Res. 22, 1353–1356. 4311–4320.

I. PRACTICE OF TOXICOLOGIC PATHOLOGY

Haschek and Rousseaux’s Handbook of Toxicologic Pathology, Third Edition, 2013, 107–146