Table 8.8 lists a number of miscellaneous dietary factors that act as exogenous modifiers of car- cinogenesis. As noted, these are quite heterogeneous and some are as yet undefined, such as fiber, intestinal flora, and cruciferous vegetables. However, ethanol, and to a lesser extent diets deficient in methyl groups, are better defined in a chemical sense. In particular, ethanol as a dietary factor in carcinogenesis plays a very significant role not only in animals but also in hu- mans (Chapter 11).
Ethanol. As further discussed in Chapter 11, the excessive intake of alcoholic beverages by humans has long been considered a significant risk factor for various types of cancer (Thomas, 1995; Longnecker and Enger, 1996), although a direct causal relationship has been questioned by some (cf. Rubin, 1996). Furthermore, epidemiological studies in the human have virtually all been carried out with dietary intake of ethanolic mixtures containing a variety of other substances, a number of which are known to be at least mutagenic if not carcinogenic (cf. Blot, 1992). In animal studies, however, it is possible to utilize the major organic constituent of alcoholic beverages, ethanol itself. Even in experimental studies, however, the demonstration of ethanol as an exogenous modifier of carcinogenesis has not been uniform (cf. Seitz and Simanowski, 1988; Garro and Lieber, 1990). Yet, there is sufficient evidence from a number of animal studies to indicate that, under some conditions, ethanol may act as an enhancer of exper- imental carcinogenesis by either a cocarcinogenic or promoting effect. Examples of such effects on carcinogenesis in several tissues of rats and mice may be seen in Table 8.11. In the rat, several different concentrations of ethanol in the drinking water enhance hepatocarcinogenesis induced by several different chemical agents. This fact—together with the effects noted in esophageal, gastric, and mammary carcinogenesis (Table 8.11)—tends to present experimental situations that may be analogous to the genesis of neoplasia in the same tissues in the human (Longnecker and Enger, 1996; Thomas, 1995). Singletary (1997) has reviewed the effect of ethanol in experi- mental mammary carcinogenesis and demonstrated that, in the mouse, spontaneous mammary carcinogenesis is little if at all affected by dietary ethanol. In the rat, however, most studies have indicated an enhancement of mammary carcinogenesis, with a few examples exhibiting no ef- fect. Even the study noted in the table (Singletary et al., 1995) indicated that high concentrations of ethanol (30% of calories) exhibited no enhancing effect on mammary carcinogenesis induced by MNU, while ethanol intakes of 15% or 20% of calories exhibited a significant enhancement, primarily during the promotion stage. An interesting example of an exogenous modification of carcinogenesis by ethanol is that of the administration of DMN to C57BL mice, wherein the hepatocarcinogenic effect of the agent was essentially not altered by ethanol, but animals receiv- ing the chemical carcinogen plus ethanol exhibited a dramatic incidence of olfactory neuroblas- tomas (Griciute et al., 1987).
In view of the widespread exposure to dietary ethanol, a variety of studies on the potential mechanisms of its toxicity have been carried out. A general survey of the genotoxicity of alcohol and its primary metabolite acetaldehyde (Bridges et al., 1987) concluded that ethanol is weakly mutagenic to bacteria and clastogenic to eukaryotic cells both in vivo and in vitro. The effect of
ethanol on germ cells was not clarified. However, ethanol with some degree of specificity in- duces cytochrome P450 2E1 in mammals (Tsutsumi et al., 1993). This cytochrome also activates several nitrosamines, including DMN and DEN, which may be important in its inhibitory and enhancing effects on carcinogenesis induced by these agents (Swann et al., 1984; Carlson,1990). Further in line with this suggestion is the fact that lowering of the carboydrate intake, which would be expected to enhance xenobiotic metabolism in the liver (Hartshorn et al., 1979), together with ethanol ingestion enhanced the initiation of DEN hepatocarcinogenesis in rats (Yonekura et al., 1992). In addition to these factors, ethanol enhances lipid peroxidation and in- hibits excisional DNA repair (cf. Garro and Lieber, 1990).
Dietary Deficiency of Methyl Groups and Carcinogenesis. As discussed previously (Chap- ter 3), diets deficient in components involved in the utilization and production of methyl groups may, on prolonged administration, induce a development of liver neoplasms in rats (Mikol et al.,1983; Ghoshal and Farber, 1984; Nakae et al., 1992) as well as in mice (Yoshida et al., 1993). However, even prior to the demonstration of this direct dietary-deficient induction of neoplasia, the importance of methyl donors in the diet and the responses to chemical carcinogens was well known (cf. Rogers, 1995). Dietary methyl deficiency enhances the induction of hepatic neoplasia in rats by aflatoxin B1, DEN, AAF, and several other chemicals. As noted above, folate deficiency can contribute substantially to the lack of normal methyl group metabolism (cf. Rogers, 1995).
The most obvious mechanism for the carcinogenic effect of methyl group deficiency is a lack of methylation of DNA or abnormalities in this process. This phenomenon has been shown in animals fed such methyl-deficient diets (Wainfan et al., 1989; Zapisek et al., 1992). These changes in nucleic acid methylation, which are accompanied by alterations in gene expression, are quite reversible on reintroduction of a normal diet (Christman et al., 1993). Correlations of alterations in the expression of specific proto-oncogenes have been noted in livers of rats fed a methyl-deficient diet (Dizik et al., 1991). In addition, methyl group–deficient diets induce alter- ations in nucleotide pools (James et al., 1992) and an increase in the level of uracil in hepatic DNA (Pogribny et al., 1997). An alternative or complementary mechanism of the effect of me- thyl-deficient diets is the enhanced lipid peroxidation (Yoshida et al., 1993) and deoxyguanosine hydroxylation of DNA occurring in hepatocytes of rats fed a methyl-deficient diet (Nakae et al.,1990). That such an oxidant effect of the diet is significant is also shown by the inhibitory effects of derivatives of vitamins C and E on the induction of preneoplastic lesions in livers of animals fed such diets (Mizumoto et al., 1994). Lombardi and associates (Lombardi and Smith, 1994) have studied the effect of choline deficiency and a more intense methyl-deficient diet on hepato- carcinogenesis, demonstrating a promoting action of such diets that may or may not be enhanced by phenobarbital (Shinozuka and Lombardi, 1980; Saito et al., 1990). Lombardi and associates have further argued that the promotion of endogenous initiated cells present within the livers of normal untreated rats can explain the promoting and carcinogenic effects of choline deficiency (Lombardi and Smith, 1994). Such a mechanism was also proposed by Sawada et al. (1990). Thus, while the mechanistic effects of methyl-deficient diets are quite varied, their effects on altering the stages of carcinogenesis appear to occur primarily at the stage of promotion through their en- hancement of cell replication (Lombardi and Smith, 1994) and effects on DNA structure and gene expression. However, it is also reasonable to assume that the promutagenic and potentially clasto- genic effects of the diets may enhance the transition from the stage of promotion to progression.
Minor Components of Fruits and Vegetables as External Modifiers of Carcinogenesis. That a high consumption of fruits and vegetables leads to a statistically significant protective effect on cancer incidence in a variety of sites in humans has been shown in a large number of studies (Block et al., 1992; Chapter 11). Wattenberg and associates (cf. Wattenberg and Loub,1978) were among the first to investigate specific chemicals occurring within vegetables, nota- bly cruciferous plants (cabbage, cauliflower, broccoli, etc.), that exerted a preventive effect on carcinogenesis. Since that time, with the advancement of knowledge in this area, Wattenberg (1985; 1992) has proposed a classification of minor nonnutrient inhibitors of carcinogenesis found in fruits and vegetables according to their action in relation to the carcinogenic process. The first category consists of compounds that prevent the formation of carcinogens from precur- sor substances either in vivo—e.g., ascorbate’s prevention of the formation of nitroso com- pounds from the reactions of nitrites with appropriate amines—or prior to consumption of such compounds by the organism. Examples of the latter would be compounds preventing the forma- tion of carcinogenic heterocyclic amines in meat products (Chapter 3) or those preventing the formation of urethane in alcoholic beverages and inhibiting the destruction of hydrazines in mushrooms (cf. Wattenberg, 1990). The second category of nonnutrient dietary inhibitors of car- cinogenesis is that comprising so-called blocking agents. These agents inhibit metabolic activa- tion of carcinogens and/or increase carcinogen detoxification by enhancing phase I and phase II pathways of xenobiotic metabolism (Chapter 3) in vivo. Those in the third category, suppressing agents, inhibit the development of the neoplastic process in cells previously exposed to carcino- genic agents. This category includes protease inhibitors (St. Clair et al., 1990; Hocman, 1992) and other dietary proteins (Papenburg et al., 1990; Messina et al., 1994), inhibitors of the arachi- donic acid cascade (cf. Rao et al., 1993a), inositol hexaphosphate, and other chemicals (cf. Wat- tenberg, 1992). The structures of some of these blocking and suppressing agents are seen in Figure 8.11. Several of the phenolic chemicals seen in the figure act as blocking agents by en- hancing the level of glutathione S-transferase activities. Benzyl isothiocyanate acts in a similar manner, while a large class of chemicals termed glucosinolates, occurring in cruciferous vegeta- bles, induce these phase II activities and also act on phase I enzyme activities. The glucosino- lates share a common basic molecular skeleton, containing a β-D-thioglucose grouping, a side chain, and a sulfonated oxime moiety. The various members of this group differ in the side chain
Figure 8.11 Chemical structures of some nonnutrient minor constituents of fruits and vegetables that can serve as exogenous modifiers of carcinogenesis. β-naphthoflavone is a synthetic congener, but the re- maining structures are of naturally occurring substances.
(Verhoeven et al., 1997). Ellagic acid is, like several other chemicals noted in Figure 8.11, anti- mutagenic, occurring naturally in grapes, nuts, and strawberries (cf. Hayatsu et al., 1988). A number of other phenolic compounds found in tea and coffee also exert inhibitory effects on carcinogenesis (Dragsted et al., 1993). From this discussion, therefore, it becomes obvious that these nonnutrient minor constituents of fruits and vegetables use a number of different mecha- nisms for their inhibitory effects on carcinogenesis. However, these may be grouped primarily into effects on initiation (carcinogen activation) and the stage of promotion as emphasized by Wattenberg (1992).
Dietary Fiber as an Exogenous Modifier of Carcinogenesis. Unlike most of the other exogenous modifiers considered in this chapter, studies on the effect of dietary fiber on carcino- genesis originated largely from human epidemiological investigations (cf. Howe et al., 1992). Since such investigations in the human did indicate a significant protective effect of dietary fiber from colon cancer, investigative studies to reproduce such effects in animals have been under- taken. Specific studies on the effect of various types of dietary fiber have indicated that poorly fermentable fibers such as wheat bran and cellulose are protective, while pectin and some other fermentable fiber supplements may have no effect or enhance chemically induced colon carcino- genesis in rodents (Jacobs, 1986; Bingham, 1990). In a recent study, Pierre et al. (1997) demon- strated that the administration of short-chain fructo-oligosaccharides to mice carrying the min mutation reduces the occurrence of colon tumors. Potential mechanisms for this inhibitory effect of fiber on colon carcinogenesis in animals (and by analogy in humans) include a reduction in fecal bile acid concentrations and an increase in the presence of short-chain fatty acids, particu- larly butyrate, as a result of fermentation of the fiber in the colon by bacteria (cf. Klurfeld,
1992). Cellulose fiber in particular has been demonstrated to prevent DNA damage by bile acids in an in vitro system (Cheah and Bernstein, 1990). As part of these mechanisms, it is now well known that changing the fiber content of the diet may also change the colonic microflora—e.g., germ-free rats exhibited a higher incidence of colon tumorigenesis than conventional rats when both were treated with azoxymethane by the intrarectal route (Reddy et al., 1975). In earlier studies, Popp and associates (Rickert et al., 1984) demonstrated that alteration in the fiber com- position of the diet as well as in the microflora of the large intestine dramatically affected the carcinogenicity of 2,6-dinitrotoluene. In addition, lactic acid–fermenting bacteria, even when fed to animals, decreased the incidence of both preneoplastic and neoplastic lesions in mice ad- ministered dimethylhydrazine and other carcinogenic compounds (cf. Rafter, 1995).