Cancer and the Environment: Gene-Environment Interaction (2002)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

3 The Links Between Environmental Factors, Genetics, and the Development of Cancer The past decade has witnessed important advances in the understanding of factors that influence cancer risk. Several environmental factors continue to sur- face as potentially instrumental in explaining the wide global variation in the incidence and biological behavior of various tumors. For example, discoveries that both essential and nonessential dietary nutrients can markedly influence several key biological events—including cell cycle regulation, processes involved in replication or transcription, immunocompetence, and factors involved with apoptosis, or programmed cell death—have strengthened convictions that spe- cific foods or components may markedly influence cancer risk. Analyses of the incidence of cancer in twin pairs and in families are tradi- tional methods for answering questions about the relationships between cancer etiology, genes, and the environment. Sorting out the relative roles of each in the initiation and progression of cancer can lead to clearer elucidation of how shared environmental influences can disparately affect the health of individual members of a community, that is, why some people exposed to a specific agent develop cancer when others do not. Finally, although environmental, occupational, and recreational exposures to carcinogens contribute to cancer risk in humans, variation in incidence and progression of cancers among individuals can be attributed to interindividual variation in genetic makeup. Recent research has identified functional polymor- phisms that influence an individual’s cancer risk and has focused on gene prod- ucts involved in activation and detoxification of carcinogens and DNA repair. Gene polymorphisms that are important in apoptosis will increasingly be recog- nized as clues to individual susceptibility to cancer, explaining why individuals with shared environmental exposures do not always share cancer morbidity and mortality. 25 

26 CANCER AND THE ENVIRONMENT DIET AS A MODIFIER OF CANCER RISK There are unprecedented opportunities for using the food supply to achieve genetic potential, that is, to optimize our performance and reduce the risks of diseases, said John Milner, Department of Nutrition, Pennsylvania State Univer- The influence of diet in the sity. Although 80 percent of cancers are development of cancer is somewhat related to environmental factors, the in- uncertain. However, the general fluence of diet in the development of consensus is that approximately 35 to cancer is somewhat uncertain. However, 40 percent of cancers relate to dietary the general consensus is that approxi- habits mately 35 to 40 percent of cancers relate John Milner to dietary habits, although the range might be quite large. Even though science has come a long way in understanding what factors are important in controlling cancer risks or modifying health in general, we still do not really know who is going to benefit, and under what circumstances, said Milner. In fact, we do not yet know if there are some people who would be placed at risk because of exaggerated intakes of certain types of foods or food components. The whole issue of the role of diet in health is exceedingly complex when trying to assess the relative roles of individual foods as they relate to overall cancer risk. There are some areas of agreement, however, said Milner. More than 80 percent of the studies that have been published reveal a reduction in cancer risk with an increase in fruit and vegetable consumption. However, there is considerable variability among popu- lations, suggesting that a person’s genetics may be important in determining the response. He added that we need to have a better understanding of how genes are involved in the cancer process and how individual nutrients can modify these genes and ultimately influence the probability of developing cancer. Some of the strongest evidence linking diet and cancer comes from the epidemiological observation that increased vegetable and fruit consumption is associated with a reduction in the risk for cancers of the mouth and pharynx, esophagus, lung, stomach, colon, and rectum. Likewise considerable evidence points to a host of essential and nonessential nutrients as modifiers of cancer risk at a variety of sites. Milner noted that part of this variation in cancer risk may arise from variation in the intake of one or more essential nutrients supplied by either plant or animal food sources. Vegetables derived from various parts of plants including roots (e.g., carrots, parsnips), leaves (e.g., spinach, lettuce), flowers (e.g., artichoke, broccoli), stalks (e.g., celery, rhubarb), and seeds (e.g., corn, peas), as well as a host of fruits, provide thousands of chemically diverse phytonutrients that may contribute to these observations. Some of these phytonutrients—including flavonoids, carotenoids, organosulfides, and 

ENVIRONMENTAL, GENETICS, AND CANCER 27 isothiocyanates—have been the focus of recent research to determine both their effects on risk and their mechanisms of action. Despite the clear linkages that have been found between the risk of develop- ing some types of cancers and dietary patterns, inconsistencies have been de- tected, which might reflect the multifactorial and complex nature of cancer, the specificity that individual dietary constituents have in modifying specific genetic pathways, and the temporal relationship between dietary intervention and pheno- typic changes in tumor incidence or behavior. The chemical and biological di- versity of dietary components in combination with a range of molecular targets makes pinpointing the importance of diet in various cancers a challenge, empha- sized Milner. It is likely that this challenge will be augmented by advances in cell biology and epidemiology. For instance, when limonene (found in citrus fruits) is added to tumor cells it has been found to enhance several genes while suppressing others. Since several of the identified genes are involved in the pathways leading to apoptosis, it is possible that agents such as limonene could play a role in the cell signaling involved in programmed cell death. Similarly, studies with a variety of other nutrients, including selenium, isothiocyanates, and allyl sulfide, have been reported to modify at least 20 different gene products associated with cancer prevention. In addition, knockout and transgenic animals can provide important clues about the specific site of action of dietary components. The use of these genomic technologies to evaluate the effects of nutrients offers exciting opportunities for determining which cellular change is most important in bringing about a change in the incidence or behavior of a tumor. Preclinical evidence suggests that diverse dietary constituents including se- lenium, allyl sulfur, genistein, and A reductionist approach to diet and resveratrol can influence the same ge- cancer prevention may produce netic pathways associated with tumor oversimplifications and confusion. We cell proliferation and apoptosis. Such clearly need to know what the common effects raise concerns about mechanisms are that account for potential interactive and cumulative ef- specific bioactive food components but fects among nutrients, said Milner. In must also recognize that we eat whole addition, compounds such as diallyl di- foods. sulfide, which is found in crushed gar- John Milner lic, can actually suppress the growth rate of cells, and indole-3-carbinol, found in cabbage, can shift estradiol metabolism, which can affect tumor formation. The only problem, said Milner, is that we may have to consume about three-quarters of a pound of cabbage a day and several cloves of garlic to bring about a re- sponse. We know of a few examples where isolated food components and intact foods do not bring about the same biological response. Thus, a reductionist 

28 CANCER AND THE ENVIRONMENT approach to diet and cancer prevention may produce oversimplifications and confusion. We clearly need to know what the mechanisms are that account for specific bioactive food components but must also recognize that we eat whole foods. Astonishing strides have been made in understanding how molecules and genetic pathways differ in precancerous and malignant cells and from their nor- mal counterparts. Capitalizing on the differences in cellular signatures that are characterized by active and inactive genes and cellular products could assist in determining who should and should not benefit from intervention strategies. Clearly, added Milner, such information will help clarify the reason for discrep- ancies among preclinical, epidemiological, and intervention studies. At least part of the variation in response to dietary components can probably be explained by the consumer’s genetic profile. It is now becoming apparent that the prevalence of polymorphisms is variable among studied populations, and these differences could influence the response to diet. Evidence exists that ge- netic polymorphisms may modulate cancer risk through their influence on folate metabolism. For example, epidemiologic studies have reported that the relation- ship between dietary folate and colorectal cancer risk is influenced by polymor- phism in methylenetetrahydrofolate reductase activity. Variation in the response to folate metabolism is not unique since other studies suggest that variation in receptors for vitamin D may also be linked to cancer risk. Considerably more information is needed about how genetic polymorphisms influence the response to dietary components and ultimately cancer risk, added Milner. Unquestionably, cancer is intertwined with environmental factors including diet. Strategies to prevent cancer through modification of either diet or specific dietary patterns will probably not be uniformly effective for all individuals, said Milner. He stressed that a better understanding of gene–nutrient interactions will be needed to determine those who might benefit most from dietary intervention and those who might be placed at risk. For example, there are data suggesting that some women who consume large amounts of fruits and vegetables may be at increased risk of giving birth to children with infantile leukemia. These women appear to have a reduced ability to remove some of the flavonoids from their system, which thus accumulate and become toxic to the developing fetus. Al- though in most cases there likely will be benefits from increased consumption of fruits and vegetables during pregnancy, in a small subset of the population an opposite response may occur. Future research in nutrition and cancer prevention must give top priority to studies that seek to understand the basic molecular and genetic mechanisms by which nutrients influence the various steps in carcino- genesis. “By understanding the importance of the genetic profile, we can identify who is going to benefit and who is not going to benefit from dietary interven- tion,” concluded Milner. 

ENVIRONMENTAL, GENETICS, AND CANCER 29 GENETIC EPIDEMIOLOGY AS A TOOL FOR STUDYING GENE– ENVIRONMENT INTERACTIONS Mounting evidence supports the concept that cancer is generally a polygenic multifactorial disease, which makes environment an important modifier in the risk of cancer, stated Kari Hemminki, Karolinska Institute. It is estimated that only 1 percent of cancers are caused by “cancer syndromes” and up to 5 percent result from highly penetrant single-gene mutations; thus, the majority are poly- genic. Studies with various animal and in vitro models, initiation and promotion models, adenoma carcinoma models, and immortalized human cells provide evi- dence that polygenic mechanisms are important in cancer, at least in experimen- tal systems. Almost all of the known cancer syndromes are monogenic and conform to a two-stage model of development; that is, they require inactivation of two copies of a tumor suppressor gene in order to initiate. These syndromes tend to be dominant Mendelian conditions, which can be assessed in family studies cover- ing two or more generations. However, such studies provide no data on recessive Mendelian conditions and have a limited resolving power in polygenic condi- tions. Consequently, apart from highly penetrant single-gene mutations, the risks posed by low-penetrance single-gene mutations, polygenes, and recessive genes are poorly understood. Hemminki described a study of data obtained from 44,000 same-sex twin pairs to assess cancer risks for co-twins of twins with cancer. There were almost 10,000 pairs in which one of the members had cancer. The analysis of environ- mental and inherited contributions was based on correlations between monozy- gotic twins who share the genome completely, that is, 100 percent concordance in their genomes. A similar concordance was carried out with dizygotic twins, the difference being the assumption that only 50 percent of the genes are com- mon. The assumption is that the environment is affecting monozygotic and dizy- gotic twins similarly. Some of these different effects will then be 100 percent, or Twin studies as tools for understanding genes, the environment, and cancer Genetic: if monozygotic twins are more similar for a given trait than dizygotic twins Shared Environment (e.g., diet and childhood experiences): if there is twin simi- larity not accounted for by genetic effects Nonshared Environment: anything that is not hereditary and not shared between relatives, that is, sporadic causes of cancer 

30 CANCER AND THE ENVIRONMENT 1. The nonshared random environmental effect was the largest factor for all cancers, accounting for 58 to 82 percent of the total variation (Table 3-1) (Lichtenstein et al., 2000). Statistically significant heritability estimates were detected for cancers of the colorectum (35 percent), breast (27 percent), and prostate (42 percent). The estimates for shared environmental effects ranged from 0 to 20 percent, but none were statistically significant. A Swedish family cancer database, containing 10 million people, is the largest population-based data set ever used for studies on familial cancer, said Hemminki. The data are used to develop estimates for the environmental and inherited components in cancer, using the genetic relationships among family members to calculate the effects of genotype, shared environment, and nonshared environment. The database has been used in modeling cancer causation and has revealed that environmental causes explained most of the total variation for all neoplasms except thyroid cancer, for which heritable causes were largest. There also appears to be a subgroup of cancer patients who develop a second cancer to which there is a strong genetic predisposition, that often cannot be predicted by a family history. This phenomenon is typical of polygenic disease. Hemminki reported that the twin and family data quantified nonshared envi- ronmental effects as ranging from 40 to 90 percent for different cancers. It is of interest to note that this effect was large for some cancers of identified environ- mental causes, such as lung and cervical cancers. In contrast, shared environ- ment—common family experiences and habits—accounted for 0 to 30 percent of cancer etiology. For all cancer, the genetic effect was estimated to be 26 percent; however, there is evidence supporting heritability for all cancers. TABLE 3-1 Heritable and Environmental Effects from Twin Studies Proportion of Variance Attributed to Heritable Shared Nonshared Cancer Effects Environmental Effects Environmental Effects Stomach 0.28 0.10 0.62 a Colorectum 0.35a 0.05 0.60 Lung 0.26 0.12 0.62 a Breast 0.27a 0.06 0.67 a Cervix uteri 0 0.20 0.80 a Ovary 0.22 0 0.78 a Prostate 0.42 0 0.58 Bladder 0.31 0 0.69 Leukemia 0.21 0.12 0.66 a a95% does not include 0.0. SOURCE: Lichtrnstein et al. (2000). Reprinted with permission. 

ENVIRONMENTAL, GENETICS, AND CANCER 31 The data presented by Hemminki on twins, families, and second cancers provide additional support to the multistage theory of carcinogenesis. If most cancers are indeed polygenic, this should be adequately considered in study designs for gene mapping approaches. Linkage analysis in families of multiple affected individuals is not sufficient to identify cancer-related genes, said Hemminki. Instead, what are needed are large case-control studies with stringent clinical criteria so that the different types of cancer can be distinguished and there is a large enough sample size to enable even the rare homozygotes to be scored, emphasized Hemminki. In addition, it will be important to study people with multiple cancers or second cancers, because they can provide a good indi- cation of whether polygenic effects are operating. MOLECULAR CARCINOGENESIS, MOLECULAR EPIDEMIOLOGY, AND HUMAN RISK ASSESSMENT The macroenvironment—our lifestyle, the air we breathe, the food we eat, the chemicals we are exposed to, as well as viruses, radiation, and physical agents we come in contact with—that combines with the microenvironment of our cells to either prevent or enhance carcinogenesis was described by Curtis Harris, National Cancer Institute (see Figure 3-1; Wang et al., 1997). In addition, a great deal of interindividual variation in our genetic makeup plays a role in the incidence and variability of cancer. Much of the work on identifying functional polymorphisms that influence an individual’s cancer risk has focused on gene products involved in the activation and detoxification of carcinogens and, more recently, on DNA repair. The idea that genomic instability might play a role in cancer is also an old one. Aneuploidy was recognized in the nineteenth century and was postulated to play a role in some cancers. More recently, tripolar spindles of DNA have been associated with the overexpression of an oncoprotein from the hepatitis B virus, which could explain how the virus contributes to hepatocellular carcinoma. In the 1970s and 1980s a set of genes, called tumor suppressor genes, was elucidated, one of which was called p53. These genes are so named because they prevent cancer by recognizing defective cell programming. The p53 gene recog- nizes the signal created by a precancerous condition and responds by killing the cell by a process called programmed cell death, or apoptosis. It has subsequently been shown that p53 mutations are common in diverse types of human cancer, where they are involved in genomic instability. The gene is involved in some pathways of apoptosis and cell cycle control, and among its many functions, it is a transcription factor. It suppresses some genes and upregulates others. It is at the crossroads of multiple cellular stress response pathways, DNA damage of varying kinds, hypoxia, and oncogene activation. (Figure 3-2). In 1979, the p53 tumor suppressor gene was identified. It has been the subject of intense research in the past 20 years and is involved in many cellular 

32 CANCER AND THE ENVIRONMENT Macro Environment Caretaker Genes • Chemicals • DNA repair • Viruses • Carcinogen Metabolic Activation • Radiation • Carcinogen Detoxification • Physical Agents Interindividual GeneN EnvironmentN Variation Micro Environment Caretaker Genes • Oxyradicals • Cell Cycle Control • Nitric Oxide • Programmed Cell Death • Hormones • Growth Factors Cancer FIGURE 3-1 Multiple genes (GeneN) can interaction with a number of environments (EnvironmentN) in the development of cancer. SOURCE: Harris (1997). Reprinted with permission. responses including differentiation, DNA repair, inhibition of angiogenesis, and apoptosis—programmed cell death. It is suggested that p53 plays a role in the development of cancer through mutation at various sites. Knockout mice missing p53 can develop normally, but they are highly tumor prone. Molecular studies estimate that approximately half of all human cancers, including some forms of skin, lung, and liver cancers, carry p53 mutations. Interestingly, the mutational sites in radon-associated lung cancer differ from lung cancer caused by tobacco smoking alone. These differ- ences may have implications for cancer diagnosis and treatment in the future. Cancer formation is a multistage process involving the activation of proto- oncogenes and the inactivation of tumor suppressor genes. Harris explained how carcinogens could affect any of these stages through genetic and epigenetic mechanisms. 

ENVIRONMENTAL, GENETICS, AND CANCER 33 Examining the mutational spectra of cancer-related genes (e.g., p53, BRCA1, and p16INK4) may provide a molecular link between etiological agents and hu- man cancer. For example, mutations in the evolutionarily conserved codons of the p53 tumor suppressor gene are common in diverse types of human cancer, and the p53 mutational spectra differ among cancers of the colon, lung, esopha- gus, breast, liver, brain, reticuloendothelial tissues, and hemopoietic tissues. Analysis of these mutations can provide clues to the mutagenic mechanisms and the function of specific regions of p53. Genetic polymorphisms are likely to play a role in the risk of lung cancer in smokers, ex-smokers, and individuals exposed to secondhand smoke. For example, women with a GST-null (glutathione S- transferase) genotype have about a twofold increased risk of developing lung cancer, and if they are exposed to high levels of environmental tobacco smoke, the risks are five- or sixfold higher. The hypothesis that women are more suscep- tible than men to tobacco-smoke-induced lung cancer is another controversial area deserving study, said Harris. Studies show more carcinogen-induced DNA damage in lungs of women than men who smoke, possibly due to a relative decrease in DNA repair capacity in women versus men. Much of the work in the field right now is investigating the mechanisms that lead to the activation of p53, largely through the kinase pathways. Studies are concentrating on the type and location of mutations found in the p53 gene in a variety of cancers. For example, in liver tumors from persons living in geo- graphic areas where aflatoxin B1 and hepatitis B virus are cancer risk factors, DNA Oncogene Activation Damage Hypoxia (myc, E1A, ras, E2F1) HIF-1A ATM, ATR, CHK2 N0 S2* P14ARF p53 Mdm2 Cell Checkpoints DNA Repair Apoptosis P21WAF1 Gadd45, p48, p53R2 Bax, Bcl-2 14-3-3 XPB*, XPD* PIGs Gadd 45* ref1 (APE1), Pol XPB*, XPD*, WRN*, BLM* RAD51, Rad54, BRC A1 FIGURE 3-2 The p53 gene: at the crossroads of the cellular response. SOURCE: Harris (2000). Reprinted with permission. 

34 CANCER AND THE ENVIRONMENT most p53 mutations are at the third nucleotide pair of codon 249. A dose-depen- dent relationship between dietary aflatoxin B1 intake and codon 249ser p53 muta- tions is observed in hepatocellular carcinoma. Exposure of human liver cells to aflatoxin B1 in vitro produces specific p53 mutants; the mutation load in vivo is positively correlated with dietary aflatoxin B1 exposure. These results demon- strate that the expression of a specific mutant p53 protein provides a specific growth or survival advantage to liver cells. Other associations between the p53 mutational spectra and exposures to carcinogens have been observed. For example, the induction of skin cancer by ultraviolet light is accompanied by specific p53 mutations. In another example, the p53 mutational spectrum in radon-associated lung cancer from uranium min- ers differs from that in lung cancer caused by tobacco smoking alone, noted Harris. These genetic changes in the tumor suppressor genes have implications for cancer diagnosis, prognosis, and therapy, according to Harris. The association of a suspected carcinogenic exposure and cancer risk in populations can be studied with classic epidemiologic techniques. However, these techniques are not applicable to the assessment of risk in individuals, stressed Harris. A goal of molecular epidemiology is to integrate molecular biology, in vitro and in vivo laboratory models, biochemistry, and epidemiology to infer individual cancer risk. Carcinogen–macromolecular adduct levels, somatic cell mutations, and DNA adducts can be measured to determine the biologically effective dose of a carcinogen. Molecular epidemiology also explores host can- cer susceptibilities, such as carcinogen metabolic activation, DNA repair, endog- enous mutation rates, and inheritance of mutated tumor suppressor genes. Substantial interindividual variation for each of these biologic end points has been shown, highlighting the need to assess cancer risk on an individual basis. As Ernst Mayr (1982) wrote in The Growth of Biological Thought, “In biology one rarely deals with classes of identical entities, but nearly always studies of populations consisting of unique individuals.” There is a wide varia- tion from one individual to another in the ability to metabolically activate and damage DNA. Given the pace of the past decade, said Harris, it is feasible that future advances will allow molecular epidemiologists to develop a cancer risk profile for an individual that includes assessment of a number of exposure and host factors. Cancer-related genes, such as p53, provide a useful molecular link between environmental agents and cancer itself. This will help focus preventive strategies and strengthen quantitative risk assessments. SUMMARY Cancer is the second leading cause of mortality in the United States today, resulting in more than half a million deaths each year. Although recent data show a downward trend in mortality rates due to cancer—mostly as a result of early detection and improved therapies—the incidence of some cancers is in- 

ENVIRONMENTAL, GENETICS, AND CANCER 35 creasing. Important advances have increased our understanding of the factors that influence cancer risk, explaining in part why some cancer rates are increas- ing while others are decreasing, and guiding us closer to the means to reverse adverse trends. A growing body of knowledge dramatically illustrates the influence of the environment, genes, and their interactions in the international variation reported in cancer incidence. A variety of linkages clearly exist between environmental exposures, diet, lifestyle factors, and cancer. Genetic factors also are known to be involved in the predisposition to and development of some cancers. Recent progress in identifying and characterizing highly penetrant susceptibility genes in familial cancer has revolutionized our understanding of the critical genetic mechanisms in cancer etiology. Studies that combine genetic analysis with as- sessment of exposures and diet can explain why not everyone exposed to a particular cancer-causing chemical will develop cancer. Genetic research also is shedding light on why some cancer patients respond to therapy and others do not. The interactions of multiple modifier genes with various environmental fac- tors—that is, gene–environment interactions—explain why cancer rates vary across populations, among exposed groups, and even within families. The re- search community is now studying cancer with an expanded and enhanced view of environmental health and exposures that include factors such as diet, lifestyle, metabolic alterations, socioeconomic status, and various environmental expo- sures.