Induction of Epigenetic Alterations by Dietary and Other Environmental Factors

Abstract

Dietary and other environmental factors induce epigenetic alterations which may have important consequences for cancer development. This chapter summarizes current knowledge of the impact of dietary, lifestyle, and environmental determinants of cancer risk and proposes that effects of these exposures might be mediated, at least in part, via epigenetic mechanisms. Evidence is presented to support the hypothesis that all recognized epigenetic marks (including DNA methylation, histone modification, and microRNA (miRNA) expression) are influenced by environmental exposures, including diet, tobacco, alcohol, physical activity, stress, environmental carcinogens, genetic factors, and infectious agents which play important roles in the etiology of cancer. Some of these epigenetic modifications change the expression of tumor suppressor genes and oncogenes and, therefore, may be causal for tumorigenesis. Further work is required to understand the mechanisms through which specific environmental factors produce epigenetic changes and to identify those changes which are likely to be causal in the pathogenesis of cancer and those which are secondary, or bystander, effects.

Given the plasticity of epigenetic marks in response to cancer-related exposures, such epigenetic marks are attractive candidates for the development of surrogate endpoints which could be used in dietary or lifestyle intervention studies for cancer prevention. Future research should focus on identifying epigenetic marks which are (i) validated as biomarkers for the cancer under study; (ii) readily measured in easily accessible tissues, for example, blood, buccal cells, or stool; and (iii) altered in response to dietary or lifestyle interventions for which there is convincing evidence for a relationship with cancer risk.

Introduction

Diet and lifestyle are major determinants of chronic disease and premature mortality. A recent quantitative analysis of the impact of 12 modifiable dietary, lifestyle, and metabolic risk factors on mortality in the United States concluded that smoking and high blood pressure are responsible for largest number of deaths with each accounting for about one in five or six deaths in adults (Danaei et al., 2009). Obesity, physical inactivity, alcohol consumption, and poor diet were also major risk factors for premature death, and cardiovascular diseases, cancers, respiratory diseases, and injuries were the most prominent causes of mortality (Danaei et al., 2009). From an etiological perspective, it is difficult to separate the effects of a number of diet and lifestyle factors since they may be interdependent, for example, obesity results from a sustained positive imbalance between dietary energy intake (poor diet) and energy expenditure (physical inactivity; Mathers, 2010), and dietary factors, including high salt intake, positive energy imbalance (obesity), and heavy alcohol consumption, contribute to risk of hypertension (Stanner, 2005). In addition, it may be helpful to consider clusters of risk factors since some factors contribute to risk of several diseases, for example, obesity is a risk factor for cancer at a number of sites and for cardiovascular diseases. A prospective study of 20,244 middle aged and older people in the United Kingdom found that the combined effects of just four health behaviors (current nonsmoking, not physically inactive, moderate alcohol intake, and plasma vitamin C concentration > 50 μM (a biomarker indicative of intakes of at least five servings of vegetables and fruits daily)) predicted a fourfold difference in total mortality with an estimated impact equivalent to 14 years in chronological age (Khaw et al., 2008). These studies reinforce the importance of understanding the dietary and lifestyle determinants of cancers and other age-related chronic diseases especially those neoplasms for which there is a striking relationship between ageing and cancer incidence (DePinho, 2000). Equally important is the need to use that understanding to develop, and to implement, effective interventions such as those proposed by the World Cancer Research Fund/American Institute for Cancer Research (2007) in the domains of food, nutrition, and physical activity.

For more than half a century, it has been evident that tobacco consumption causes lung cancer. Tobacco also causes tumors of the larynx, pancreas, kidney, and bladder and, in conjunction with alcohol consumption, tobacco use is associated with a high incidence of carcinomas of the oral cavity and esophagus (Stewart and Kleihues, 2003). In most economically developed countries, where tobacco consumption accounts for up to 30% of malignant tumors (Stewart and Kleihues, 2003), there are sustained and progressive disincentives to tobacco smoking. With the corresponding reduction in cigarette-induced lung cancer, lung cancer in individuals who have never smoked tobacco products is an increasing medical and public health issue and research is underway to identify the genetic (and other) causes of this type of lung cancer (Li et al., 2010).

In their recent report, the World Cancer Research Fund/American Institute for Cancer Research (2007) recommended that, for those who consume alcoholic drinks, consumption should be limited to no more than two drinks per day for men and one per day for women. This recommendation is based on the convincing evidence that alcoholic drinks are a cause of cancers of the mouth, pharynx, larynx, and esophagus in both sexes and breast (in women) and colorectum (in men). Such drinks are also a probable cause of hepatocellular cancer (both sexes) and of colorectal cancer in women (World Cancer Research Fund/American Institute for Cancer Research, 2007). Even low to moderate alcohol consumption is associated with increased cancer risk in women, especially among current smokers (Allen et al., 2009). In addition, obesity and alcohol consumption combine to increase the risk of morbidity and mortality from liver cirrhosis (Hart et al., 2010, Liu et al., 2010) and these adverse exposures contribute to the development of primary hepatocellular carcinoma (Siegel and Zhu, 2009).

Studies in rodents have shown that both voluntary and imposed exercise reduces bowel cancer (Basterfield et al., 2005) and there is now convincing evidence from human epidemiological studies that physical activity protects against colon cancer (World Cancer Research Fund/American Institute for Cancer Research, 2007). In addition, physical activity probably protects against postmenopausal breast cancer and endometrial cancer (World Cancer Research Fund/American Institute for Cancer Research, 2007). As noted below, obesity is also a risk factor for the same cancers (and others) so that it is difficult to separate, etiologically, the adverse effects of sedentary behavior per se from the contribution which the latter makes to positive energy balance and, therefore, to obesity-related cancers. However, observations from a population-based, case-control study of colon cancer in Shanghai, China, indicated that commuting physical activity modified significantly the adverse effects of increased adiposity (measured as body mass index (BMI, Hou et al., 2004). The highest colon cancer risk was experienced by those in the highest quintile of BMI and with the lowest physical activity level (Hou et al., 2004).

Over the past 10–15 years, the evidence that overweight and obesity increase cancer risk at several sites (including several sites within the gastrointestinal tract plus the endometrium, kidney, gallbladder, and breast; postmenopausal) has strengthened considerably (World Cancer Research Fund/American Institute for Cancer Research, 2007). As with cardiovascular disease and type 2 diabetes, abdominal (visceral) adiposity appears to be particularly pathogenic and increases the risk of large bowel cancer (Larsson and Wolk, 2007). Central obesity is also a probable cause of cancers of the pancreas, endometrium, and breast (postmenopausal; World Cancer Research Fund/American Institute for Cancer Research, 2007). These observations begin to connect, etiologically, several cancers with the metabolic syndrome and may suggest that there are common underlying mechanisms with chronic systemic inflammation as a key potential candidate (Hotamisligil, 2006). Support for the concept that metabolic dysregulation is central to the development of cancer, diabetes, and cardiovascular disease is provided by the recent observation of a common transcriptional signature shared by cancer and inflammatory and metabolic diseases (Hirsch et al., 2010). Of genes showing aberrant expression in cancer, 54 were also implicated in aspects of the metabolic syndrome (Hirsch et al., 2010). Importantly, 11 out of 13 drugs used to treat noncancer diseases also inhibited cellular transformation (Hirsch et al., 2010). These results suggest that interventions which reduce the risk of the metabolic syndrome may also be effective in cancer prevention.

Three decades ago, the pioneering studies of Doll and Peto (1981) (Jenab et al., 2009, Johansson et al., 2010, Spencer et al., 2008) provided strong evidence that dietary factors may be as important as smoking behavior in explaining variation in cancer risk—each explaining about 30% of the variance. However, diet is a very complex exposure and conventional epidemiological approaches are relatively blunt instruments when attempting to dissect out which dietary components, in what doses and over what time periods, enhance risk or protect against cancer development. Such difficulties underlie current attempts to estimate dietary exposure using biomarker approaches (Jenab et al., 2009, Johansson et al., 2010, Spencer et al., 2008). Nevertheless, there is good evidence that diets which contain a high proportion of plant foods (fruits, vegetables, relatively unprocessed cereals, and pulses) are associated with lower risk of several common cancers whereas higher intakes of red, and especially processed, meats appear to be causal for colorectal cancer (World Cancer Research Fund/American Institute for Cancer Research, 2007). In contrast, there is an inverse correlation between higher intakes of fish and colorectal cancer risk (Norat et al., 2005). Very recent data from the large European Prospective Investigation of Cancer and Nutrition (EPIC) study suggest rather weak negative associations between fruit and vegetable intake and overall cancer risk (Boffetta et al., 2010). Recent evidence that above median circulating concentrations of methionine, B₆, and folate several years prior to disease onset combine to reduce dramatically risk of lung cancer (Johansson et al., 2010) point to this area as a high priority for future research into the influence of methyl donors on cancer etiology.

Although there is strong evidence from observational studies that higher intakes of specific micronutrients, for example, vitamin A, folate, selenium, and vitamin D are associated with lower cancer risk, the outcomes of intervention studies testing the antineoplastic effects of such nutrients have often been disappointing (see, e.g., Mathers, 2009 for review of folate and bowel cancer risk). In general, dietary intervention studies for cancer chemoprevention have been carried out in higher risk subjects using relatively large (sometimes unphysiological) doses of individual micronutrients and such designs may have limited relevance in attempts to understand the impact of diet on cancer risk.

In contrast with the many detailed mechanistic studies of the effects of physical stressors such as oxidative stress and inflammation on cancer risk, there has been limited research which attempts to understand how psychosocial stressors modulate cancer biology. Much of the available data relate to associations between psychosocial stress (or ability to cope with such stress) and breast cancer risk. When this topic was reviewed a decade ago, the authors concluded that the evidence for a relationship between psychosocial factors and risk of breast cancer was weak and called for research with a theoretical grounding and greater methodological rigor (Butow et al., 2000). A more recent review (Nielsen and Gronbaek, 2006) concluded that stress does not seem to increase the risk of breast cancer and that there was inadequate evidence on which to make a judgment about the effects of stress on breast cancer progression. Again, these authors drew attention to key methodological issues including the way in which stress was conceptualized and measured (Nielsen and Gronbaek, 2006), issues which will need to be addressed to allow this potentially important field to progress.

Since the industrial era, certain occupations (and sites of habitation) have been important source of exposure to cancer-causing chemicals such as polycyclic aromatic hydrocarbons (PAHs; Bosetti et al., 2007). While such exposures have been eliminated from most workplaces, risks remain in some newly industrialized countries where regulations are less restrictive or less rigorously enforced (Stewart and Kleihues, 2003). In addition, some heavy metals notably lead, cadmium, mercury, and arsenic pose threats to human health (Jarup, 2003). The pollution of drinking water by arsenic in the Ganges delta alluvium in Bangladesh has resulted in an estimated 40 million people being at risk of arsenic poisoning-related diseases, including skin cancer (Alam et al., 2002).

In summary, there is now good knowledge of many of the environmental factors which modulate cancer risk. The challenge is to understand how these exposures affect cancer biology, that is, the accumulation of genetic damage and aberrant gene expression which are fundamental hallmarks of tumor development. Given the impact of a wide variety of environmental exposures on epigenetic marks (see below) and the key role which such marks play in modulating gene expression, an a priori case can be made for the importance of understanding the effects of dietary, lifestyle, and other environmental factors on epigenetic marks as a step toward the development of interventions for cancer risk reduction. It has been argued that epigenetic changes are an early event in tumor development (Feinberg et al., 2006) and that changes induced by environmental exposures may only alter risk of cancer initiation but also increase the likelihood that further genetic insults will be acquired (Baylin and Ohm, 2006). Experimental evidence is accumulating that DNA methylation patterns (Belshaw et al., 2008) and gene expression patterns (Polley et al., 2006) are altered even in apparently normal tissues of those at higher cancer risk.