inhibition of carcinogenesis by minor dietary constituents1€¦ · a protection by vegetables...

(CANCER RESEARCH (SUPPL.) 52. 2085s-2091s. April I. 1992)

Inhibition of Carcinogenesis by Minor Dietary Constituents1

Lee W. Wattenberg2

Laboratory of Medicine and Pathology, I'niversity of Minnesota, Minneapolis, Minnesota 55455

Abstract

A major foundation of the field of cancer chemoprevention has residedin an impressive number of animal studies showing that cancer can beprevented by a variety of chemical compounds. In the search for increasingly effective inhibitors, both synthetic and naturally occurring compounds are being investigated. One group of naturally occurring compounds of particular interest consists of minor dietary nonnutrients. Foodsof plant origin frequently contain as much as several per cent of thesecompounds. In recent years there has been a growing awareness that theymay have important effects on the consequences of exposure to carcinogenic agents. This information comes from studies of whole diets; individual dietary constituents, particularly those from plants; and finallyfrom investigations of pure compounds. The nonnutrient inhibitors ofcarcinogenesis have several different mechanisms of action. Some areblocking agents; i.e.. they prevent carcinogens from reaching or reactingwith critical target sites. Others are suppressing agents; i.e., they preventevolution of the neoplastic process in cells that otherwise would becomemalignant. Occasionally, one compound will show both mechanisms. Thenonnutrient compounds have the important attribute of there being dataon their consumption by humans. Thus, it may be possible to evaluatetheir impact on cancer risk. Ultimately, information pertaining to thesecompounds may be useful in terms of diet selection or their use as dietarysupplements.

Cancer can be prevented by a surprisingly large number ofchemical compounds. This type of prevention is referred to aschemoprevention. Two forms exist: one, general populationchemoprevention, entails efforts at providing measures applicable to large population groups; the other, targeted chemoprevention, is directed toward individuals at increased risk ofcancer in specific tissues. The two differ in terms of allowabletoxicity and specificity of agent targeting. Driving forces forgeneral population chemoprevention have been both epidemi-ological and experimental data indicating protective effects ofdietary constituents against the occurrence of cancer. Epide-miological studies have shown that diets containing large quantities of vegetables and, to a lesser extent, fruits are associatedwith relatively low risks of cancer. The assignment of whichconstituents of these foods of plant origin are responsible forprotection has been indecisive. In the search for inhibitorycompounds, a group currently under investigation is that of theminor nonnutrient constituents of the diet. This presentationprovides an overview of experimental studies in which membersof this very diverse group of compounds have been shown toinhibit carcinogenesis. The implications of these data could bequite profound. Ultimately, they could be relevant to decisionson food selection and composition for the purpose of providingincreased protection against cancer as well as for possible useof individual compounds as specific dietary adjuncts.

Foods contain major and minor constituents. The major onesare protein, fat, carbohydrate, and fiber. The minor ones arenutrients such as vitamins and minerals and a large number ofnonnutrients, i.e., compounds with no known nutritional value.Nonnutrient compounds will be the principle focus of this

' Presented at "Nutrition and Cancer," the first conference of the International

Conference Series on Nutrition and Health Promotion, April 17-19, 1991,Atlanta, GA. The research described was supported by Grant SIG 5 from TheAmerican Cancer Society.

2To whom requests for reprints should be addressed.

paper; many are unfamiliar to individuals in the biomÃ©dicalsciences, including specialists in nutrition. They have beenignored because they seem to be biologically quite inert. However, it is becoming apparent that some of these compoundscan have rather profound effects in cancer prevention.

The role of nonnutrients in the diet in preventing cancercomes from three levels of data, each of which is successivelymore defined. The first level consists of data derived fromstudies in which comparisons of carcinogenic responses weremade between animals fed crude diets composed of naturalconstituents containing both nutrients and nonnutrients andsemipurified diets composed mainly of purified materials andalmost completely lacking nonnutrients. Several of these experiments show a lower tumor response in animals fed the crudediets (1-4). The results are important because they suggest aprotective role of nonnutrients in some carcinogenic processesand because the concentration of the nonnutrients in conventional diets is adequate to bring about this preventive effect.The second level of information comes from studies of singlecomponents, such as vegetables, fruits, or crude products derived from plant sources. In several instances, inhibition ofcarcinogenesis was found. Finally, experiments with a largenumber of specific dietary nonnutrient compounds, particularlyfrom vegetables and fruits, show protection against neoplasia.

Comparisons between the Effects of Crude and SemipurifiedDiets on the Carcinogenic Response of Experimental Animals

Studies from the early work of Silverstone et al. (ÃŒ)demonstrated that in some experiments animals fed a crude diet wereat lesser risk of developing cancer than were those fed a semi-purified diet. The crude diets (Purina chow) contain components such as grain, beet pulp, alfalfa meal, cane molasses, fishmeal, and other crude materials. In contrast, semipurified dietsconsist of defined dietary constituents, such as casein, starch,corn oil, minerals, and vitamins. The term semipurified meansthat some of these constituents, principally vegetable oils, arenot pure. The two types of diets are nutritionally comparablebut differ profoundly in amounts of nonnutrients. In one investigation, male DBA or C3H mice placed on the two differenttypes of diets developed hepatomas spontaneously. In a seriesof experiments, 2 to 5 times as many of the mice fed thesemipurified diet developed hepatomas compared with groupsfed the crude diet. In some experiments of cancers of other sitessuch as skin cancers, mammary cancers, and leukemias, theresults were indecisive or no effects were found.

More recently, the effects of feeding AIN-76, a semipurifieddiet, and NIH-07, a crude diet, were compared for carcinogen-induced neoplasia in mice given the tobacco-specific carcinogen,NNK-' (2). Feeding of the diets was begun 2 weeks before

carcinogen challenge and continued for the duration of theprotocol. The animals fed the semipurified diet had 3 times asmany pulmonary adenomas as did those fed the crude diet.

'The abbreviations used are: NNK, 4-(methylnitrosatnino)-l,3-(pyridyl)-l-butanone; BP, benzo(a)pyrene; DMBA, 7,12-dimethylbenz(a)anthracene; AMT,ally! methyl trisulfide; AMD, allyl methyl disulfide; DAT. diallyl trisulfide; DAS,diallyl sulfide.

2085s

on June 23, 2020. © 1992 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

NONNUTRIENT INHIBITORS OF CARCINOGENESIS

These data agree with epidemiological investigations in whicha protection by vegetables against lung cancer in the humanswas observed (5).

In other experiments, the comparisons focused on either theinitiation or postinitiation phases of carcinogenesis. One studycompared a crude diet (Purina rat chow) with a semipurifieddiet (Teklad) in mice during the initiation phase. Two experiments were performed, one with /3-propiolactone and one withBP (3). Animals were placed on the experimental diets 2 weeksbefore carcinogen challenge and were maintained on the dietsuntil 1 day after the last dose of carcinogen. The target organwas the forestomach. In the first experiment, mice treated with/3-propiolactone and fed the crude diet had approximately one-third as many forestomach tumors as did those fed the semi-purified diet. In the second experiment, with BP-induced neoplasia, no difference in carcinogenic response was observed.This experiment shows that the dietary effects on initiation bydifferent carcinogens can vary.

A study aimed at investigating the effects of feeding a crudewith a semipurified diet on the postinitiation phase of carcinogenesis was carried out by Longnecker et al. (4, 6). The experimental model was azaserine-induced neoplasia of the pancreasin rats. These investigators showed that feeding a crude dietduring the postinitiation phase resulted in a lower incidence ofpancreatic neoplasia compared with feeding a semipurified diet.In summary, a number of experiments show that the carcinogenic response of animals fed crude diets was less than in thosefed a semipurified diet. The crude diets contain an abundanceof nonnutrient dietary constituents whereas the semipurifieddiets contain relatively small quantities. Thus, these studiesprovide an initial indication that nonnutrient dietary constituents are inhibitors of carcinogenesis.

Effects of Single Crude Dietary Constituents on Carcinogen-induced Neoplasia in Experimental Animals

Single crude dietary constituents have been added to semi-purified diets so that their effects on the carcinogenic responsecould be studied. The results of adding cruciferous vegetableswere studied with DMBA-induced neoplasia in Sprague-Dawleyrats. In this experimental model, the animals develop mammarytumors after a single administration of DMBA by p.o. intubation. Frozen-thawed pieces of vegetable were placed in the cagesonce a day (7). The animals consume this type of vegetablesupplement very rapidly thereby mimicking human meal consumption. The vegetable supplements were provided beginning1 week after carcinogen administration only during the postinitiation phase of carcinogenesis. Under these conditions, cabbage and broccoli inhibited mammary tumor formation. Inanother study with this tumor model, orange oil was added toa semipurified diet. A dose-dependent inhibition of mammarytumor formation resulted (8). In other experiments, the effectsof onion and garlic oils on epidermal carcinogenesis in mousewere studied. Topical applications of these two oils producedinhibition in the postinitiation phase in this animal model (9).In other work, extracts of plant materials containing proteaseinhibitors inhibited the postinitiation phase of carcinogenesisof the skin, breast, large bowel, and lung (10). Green teaadministered in the drinking water was recently shown to inhibitcarcinogenesis in mice (II).4

' A. Conney, unpublished observations.

Inhibition of Carcinogenesis by Specific NonnutrientConstituents of the Diet

The finding of specific compounds of plant origin with inhibitory effects on carcinogenesis has resulted from two lines ofinvestigation. One of these is the identification of a constituentin crude plant materials. The plant material used had carcinogen-inhibitory effects or produced a biological response, suchas enhancement of detoxification systems, likely to result incancer prevention. In these studies, o-limonene, a major constituent of citrus fruit oils that inhibits carcinogenesis, wasfound to be preventive (12). Several groups of inhibitors alsowere identified in cruciferous vegetables. These include Ãndoles,aromatic isothiocyanates, dithiolethiones, and phenols. Studiesto identify the active constituents of garlic and onion oils thatinhibit carcinogenesis showed a number of organosulfur compounds in these oils that have inhibitory properties (13, 14).

The second type of investigation that identified nonnutrientdietary inhibitors focused initially on synthetic compounds.Subsequently, these structurally related compounds were foundin natural products. Initial studies of synthetic chemicals resulted in identification of flavone as an inhibitor. Flavone, itself,is a synthetic compound. However, a large number of naturallyoccurring flavone derivatives were found to inhibit carcinogenesis (15-17).

The mechanisms of action of many of the minor nonnutrientinhibitors of carcinogenesis are poorly understood, making itdifficult to organize them into a precise pattern. One organizational framework is classification of inhibitors by the time incarcinogenesis when they are effective. In this framework, inhibitors can be divided into three categories. The first consistsof compounds that prevent the formation of carcinogens fromprecursor substances. The second contains compounds thatinhibit carcinogenesis by preventing carcinogenic agents fromreaching or reacting with critical target sites in the tissues.These inhibitors are called blocking agents, which describestheir mechanism of action (they exert a barrier function). Inhibitors of the third category act subsequent to exposures tocarcinogenic agents. These are termed suppressing agents because they act by suppressing the expression of neoplasia incells previously exposed to doses of carcinogens that otherwisewould cause cancer (15). Studies of various inhibitors havemade it apparent that some inhibitors act at more than onetime in the carcinogenic process (7).

Blocking Agents

Blocking agents prevent carcinogens from reaching or reacting with critical target sites by several mechanisms. One mechanism is inhibiting of reactions requiring metabolic activation.A second entails inducing of activities of enzyme systems thatdetoxify carcinogens. A third is the trapping of reactive carcinogenic species. Some blocking agents prevent the action oftumor promoters by preventing the promoter from reaching orreacting with its cellular target or by blocking subsequent cellular events required for tumor promotion to occur. A list ofnonnutrient blocking agents is shown in Table 1.

Blocking Agents That Act by Inhibiting Carcinogen Activation.Several groups of nonnutrient compounds prevent carcinogenesis by inhibiting carcinogen activation reactions (Table 2).Almost all carcinogens occurring in food require metabolicactivation. Examples of such carcinogens are include aflatoxin,nitrosamines, polycyclic aromatic hydrocarbons, hydrazines,

2086s




Table 1 Some nonnutrient blocking agents in food

Terpenes(12. 15, 18)Organosuindes(14. 18-21)Aromatic isothiocyanates

(15, 22, 23)Ãndoles(13, 15)Dithiolethiones (24, 25)Phenols (15)Flavones(15-17, 26)Tannins (11, 27,28)

Ellagic acid (29-31)Curcumin (25, 32)Coumarins (15)Conjugated dienoic linoleic acids (25, 33)Nucleophiles(15, 25)tÃ-Carotene(25)18-/3-Glycyrrhetinic acid (34)Glucarates (35-37)

and heterocyclic amines. Thus, blocking agents inhibiting activation of these and other carcinogens may play a role inpreventing neoplasia resulting from this type of exposure. Benzyl isothiocyanate is one such naturally occurring blockingagent. Benzyl isothiocyanate, found in cruciferous vegetables,is formed by the hydrolysis of the glucosinolate glucotropaeolin.The glucosinolate is the storage form in plant material. Whenplants are processed or damaged, the glucosinolate undergoeshydrolysis with resultant formation of the free aromatic isothiocyanate (39). Early experiments showed that benzyl isothiocyanate inhibited DMBA-induced mammary tumor formationin rats when given by p.o. intubation 2 h before the carcinogen(38). More recently, the effects of benzyl isothiocyanate onnitrosamine metabolism in rat liver microsomes and culturedesophagus were studied by Chung et al. (40, 41). These investigators found that administration of benzyl isothiocyanate 2 hbefore killing markedly inhibited activation of yV-nitrosodimeth-ylamine and the tobacco-specific carcinogen, NNK. Subsequently, benzyl isothiocyanate administered by gavage 15 minbefore p.o. carcinogen challenge inhibited the neoplastic effectsof JV-nitrosodiethylamine and BP (22, 38).

Organosulfur compounds found in Allium species, includinggarlic, onions, leeks, and shallots (42-44) are a second groupof naturally occurring compounds that can inhibit carcinogen-esis by preventing carcinogen activation. In one series of experiments, they inhibited 7V-nitrosodiethylamine-induced carci-nogenesis of the forestomach and, to a lesser extent, the lungsin female A/J mice (18). The most potent of the naturallyoccurring organosulfur compounds was diallyl disulfide. Inother studies Wargovich et al. (19) showed that diallyl sulfideinhibits yV-nitrosomethylbenzylamine-induced esophageal cancer in rats when the organosulfur compound was administeredp.o. 3 h before nitrosamine. In related work, Brady et al. (45)demonstrated that diallyl sulfide inhibits microsomal metabolism of nitrosamines by rats when administered 3 h before therats are killed. Inhibition of 1,2-dimethylhydrazine-inducedneoplasia of the large bowel in mice by diallyl sulfide was alsoobserved when this compound was administered p.o. 3 h beforethe carcinogen (14).

A third group of naturally occurring compounds that inhibitcarcinogen activation and carcinogenesis when given shortlybefore carcinogen exposure are monoterpenes. Two such compounds have been investigated, D-limonene and o-carvone. D-Limonene is a major constituent of citrus fruit oils (46). Forexample, orange oil contains more than 90% D-limonene. D-Carvone is a major constituent of caraway seed oil. This oilcontains approximately 50% o-carvone (47). o-Limonene, D-carvone, orange oil, and caraway seed oil all have been foundto inhibit activation of dimethylnitrosamine.5 Carcinogenesisexperiments have been performed in which D-limonene, D-carvone, or caraway seed oil were administered p.o. to femaleA/J mice l h before NDEA (18). All three substances pro-

5L. W. Wattenberg and J. L. Leong, unpublished observations.

foundly inhibited forestomach tumor formation and the occurrence of pulmonary adenomas in these animals. More recently,D-limonene and citrus fruit oils were found to inhibit thecarcinogenicity of NNK (48).

Glucobrassicin and glucotropaeolin, two glucosinolates occurring in cruciferous vegetables, were also found to inhibitDMBA-induced mammary neoplasia in rats when administeredby gavage 4 h before the carcinogen. Thus, data now show thatfour widely occurring groups of naturally occurring compounds,i.e., aromatic isothiocyanates, organosulfur compounds foundin Allium species, monoterpenes, and glucosinolates, can inhibitcarcinogenesis if administered shortly before carcinogen challenge (Table 2).

Blocking Agents That Act by Increasing Carcinogen Detoxification. Blocking agents can prevent the occurrence of neoplasiaby increasing the detoxification of carcinogens. Two generalcategories of blocking agents acting by this mechanism havebeen identified. They are designated type A and type B inhibitors (15). Type A inhibitors induce an increase in activity ofphase 2 enzymes, which perform conjugation and other reactions that detoxify many carcinogenic agents. Prominent amongthese is the marked increase in glutathione 5-transferase activity. UDP-glucuronosyltransferase, epoxide hydrolase, andNAD(P)H-quinone reducÃaseactivities are also enhanced. Increases in tissue glutathione levels are found. Type B inhibitorsenhance both phase I and phase 2 enzyme activities. The phase1 enzymes include the cytochrome P-450 systems and are verycomplex. Reviews of the effects of type A and type B inhibitorshave been published (15).

Because type A inhibitors are effective, there have been acontinuing effort to identify compounds in this category ofchemopreventive agents. Recent work has focused on the organosulfur compounds found in Allium species. In initial studies, the capacity of AMT to induce increased glutathione 5-transferase activity in rodent tissues was investigated. AMTinduces increased activity of this enzyme system when given 4and 2 days before assay. With this administration schedule,AMT was also found to inhibit BP-induced neoplasia of theforestomach of female A/J mice (20). Experimental protocolswith the test compound being given 4 and 2 days before carcinogen challenge are commonly used in studying the inhibitoryeffects of agents that act by inducing increased activity ofdetoxification systems.

Other organosulfur compounds were studied under thesesame conditions. Ultimately, the experiments have includedfour allyl group-containing compounds with one, two, or threelinearly connected sulfur atoms: AMT, AMD, DAT, and DAS.Four corresponding saturated compounds in which propylgroups are substituted for the allyl groups also were studied.All four allylic compounds inhibited BP-induced neoplasia ofthe forestomach. Their saturated analogues showed almost noinhibitory activity, indicating the importance of allyl groups.DAT, with two allyl groups, was more potent than AMT, withone, providing evidence for the role of allyl groups in theinhibitory effects observed. DAS and AMD, but not DAT orAMT, inhibited pulmonary adenoma formation (20). The factthat monosulfide and disulfide inhibit tumor formation in thelung but that trisulfide does not indicates that the number ofsulfur atoms in the molecule can control the organ sites atwhich protection against carcinogenesis occurs. All four allyliccompounds induce increased glutathione 5-transferase activityin the forestomach but vary in capacity to induce glutathione5-transferase activity in the lung, liver, and small bowel. Their

2087s




Table 2 Blocking agents preventing neoplasia by inhibiting carcinogen activation

GroupAromatic

isothiocyanatesOrganosulfur

compoundsMonoterpenesGlucosinolatesCompoundsBenzyl

isothiocyanatePhenethylisothiocyanateDiallyl

suindeDiallyl disulfideAIM mercaptanAllyl methyldisulfiden-Limonene

r>-CarvoneGlucobrassicin

GlucotropaeolinSourceCruciferous

vegetablesCruciferousvegetablesAllium

sp.Allium sp.Allium sp.Alliumsp.Citrus

fruit oilsCaraway seedoilCruciferous

vegetablesCruciferous vegetablesCarcinogens

inhibitedNDEA,"BP, DMBA

NDEA. BP. DMBA,NNKDMH,

NMBANDEANDEANDEANDEA,

NNKNDEADMBA

DMBARef.22,3822,2314,

19181818Â«.*1313

" NDEA, jV-nitrosodiethylamine; DMH, 1,2-dimethylhydrazine; NMBA, W-nitrosomethylbenzylamine.* L. W. Wattenberg and J. Coccia, unpublished.

saturated analogues produced little or no induction (20). Likethe organosulfur compounds from Allium species, aromaticisothiocyanates from cruciferous vegetables, in particular, benzyl isothiocyanate, induce increased glutathione S-transferaseactivity and have other properties of type A inhibitors.

Blocking Agents Effective against Tumor Promoters. A number of naturally occurring nonnutrient inhibitors of tumor promotion have been found. A property for identifying inhibitorsof this type is their capacity to prevent an increased ornithinedecarboxylase activity after exposure to the promoter. Thisassay has been particularly effective in mouse epidermis. However, in other instances, biological responses such as inhibitionof increased cellular proliferation or tumor formation in response to promoter administration has been the means by whichthe inhibitor was identified. Table 1 lists compounds that inhibittumor promotion, including phenols, flavones, tannins, curcu-min, glycyrrhetinic acid, and glucarates. Mechanisms by whichthese inhibitors work include inhibition of components of thearachidonic acid cascade, antioxidant activity, and modulationof hormonal responses.

Suppressing Agents

Suppressing agents prevent the evolution of the neoplasticprocess in cells previously exposed to doses of carcinogenicagents that otherwise would cause cancer (15). Table 3 listsminor nonnutrient constituents of the diet that can act assuppressing agents in animal models. The number of suppressing agents identified is considerably smaller than the numberof blocking agents. The mechanisms by which blocking agentsact are relatively simple, although the systems themselves canbe very complex. The mechanism of action for suppressingagents is not well defined, as might be expected because theevents that they counteract are those basic to cancer, also apoorly understood process. Suppressing agents act during theearliest stages of the neoplastic process, when the number ofpathological alterations is considerably smaller than at laterstages (55). As more information becomes available concerningthe multiple events leading to malignancy, targeting of suppressing agents to counteract defined early changes may be

Table 3 Some nonnutrient suppressing agents in food

Protease inhibitors (10)Inhibitors of the arachidonic acid cascade (49-51)Terpenes(12)Aromatic isothiocyanates (IS)Inositol hexaphosphate (52, 53)Nerolidol (57)

feasible. The vast amount of data indicates that alterations ingenetic material occur during carcinogenesis. Correcting genetic alterations during neoplastic changes in solid tissues isdifficult. Suppressing agents can modulate consequences of thegenetic changes but are not likely to correct the genetic defectsthemselves. Accordingly, it would be anticipated that suppressing agents may have reversible effects. Consequently, it is likelythat continuous administration of suppressing agents will benecessary to sustain inhibition of carcinogenesis. Retinoids, themost extensively investigated of the suppressing agents, demonstrate this characteristic both in vitro and in vivo. Suppressionis maintained as long as the agent is present; on removal, theneoplastic process recurs.

Studies of crude diets for suppressing effects are relativelysparse. One such study cited previously was carried out byLongnecker et al. (4, 6), who demonstrated a suppressing effectof crude diets on the occurrence of pancreatic cancer. This studyis not definitive, because animals fed the crude diet gained lessweight than did those fed the semipurified diet. It is wellestablished that retarding weight gain also can diminish thecarcinogenic response. However, the difference in the incidenceof pancreatic neoplasms between the two groups of animals wasgreater than expected from the modest difference in weightgain.

Suppressing agents have been found in individual dietaryconstituents. Studies with cruciferous vegetables and with orange oil demonstrated suppressing effects (8). In the initialstudies with cruciferous vegetables, dehydrated powders prepared from Chieftain Savoy cabbage inhibited DMBA-inducedmammary-tumor formation. In further investigations, it wasfound that frozen-thawed segments of cabbage leaf placed directly in cages produced a comparable reduction of neoplasiain the animals (56). The amount of cabbage leaf fed was of theorder of 5 g/kg body weight/day. More recently, similar studieswith frozen-thawed segments of broccoli and cabbage weregiven beginning 1 week after DMBA and continued for 18weeks after DMBA. The rats were fed a semipurified diet, andthe vegetable additions were given 6 days a week. The resultsshowed that fewer animals receiving cabbage had mammarytumors compared with nonsupplemented controls (100% versus56%) and that the average number of tumors per rat wasdiminished markedly (2.5 versus 0.8). Similar findings wereobtained with broccoli (56).

Possibly related to the suppressive effects of cruciferousvegetables is the finding that benzyl isothiocyanate, a constituent of these plants, will inhibit DMBA-induced mammarycarcinogenesis when administered beginning 1 week afterDMBA challenge (15, 38). Suppressive effects also were pro-

2088s



NONNt'TRIENT INHIBITORS OF CARC1NOGENF.SIS

duced by orange oil or its principal constituent, D-limonene (8,12). In recent studies, nerolidol, a sesquiterpene structurallyrelated to farnesol, suppressed azoxymethane-induced neoplasia of the large bowel in rats (57). The observation that suppressing agents occur in commonly consumed foods indicatesthat further searches should be made for this category of chem-

opreventive agents.Table 3 lists some defined minor nonnutrient dietary constit

uents that have suppressing activity. Included are proteaseinhibitors, a very interesting group of suppressing agents. Thein vitro data showing their suppressive effect are striking. Somein vivo studies have also shown inhibition, but not many of suchinvestigations have been performed (10). Another group ofsuppressing agents with a common characteristic are inhibitorsof components of the arachidonic acid cascade. This groupincludes various polyphenolic compounds. Studies with synthetic inhibitors of the arachidonic acid cascade, such as indo-methacin and piroxicam, demonstrated a suppressive effect,thus suggesting that the naturally occurring compounds mayact similarly (49-51). Inhibitors of the arachidonic acid cascadecan be potent inhibitors of tumor promotion (5). The relationship between their mechanisms of suppression and inhibitionof tumor promotion is not clear. Two additional groups ofnaturally occurring suppressing agents are the terpenes andisothiocyanates (12, 15). Further work is required to more fullyevaluate their potential role as suppressing agents.

Most of the research with suppressing agents has focused onsynthetic compounds and a few minor dietary compounds thatare nutrients. As mentioned previously, the retinoids, syntheticanalogues of vitamin A, are the most extensively studied suppressing agents. Vitamin A has only limited suppressing activity(54). It is relatively toxic at the large amounts generally usedfor suppression. However, in a careful study by Zile et al. (58),moderate supplementation of the diet with vitamin A inhibitedcarcinogen-induced mammary neoplasia in rats. ÃŸ-Carotenehasalso been studied extensively for its suppressing activity inanimal systems. Many of these studies have given negativeresults. A few studies giving positive results were published(59). Selenium salts and inositol hexaphosphate were alsoshown to have suppressing effects (60-63).

Inhibitors with Both Blocking and Suppressing Activity

During studies of inhibitors of carcinogenesis, some substances showed both blocking and suppressing effects. The firstwas sodium cyanate, where both blocking and suppressingeffects against DMBA-induced mammary carcinogenesis in rats( 15). Subsequently, cruciferous vegetables were shown to havethese characteristics, as was one of their constituents, benzylisothiocyanate. In experiments with cabbage, consumption of asemipurified diet with added dehydrated powders prepared fromChieftain Savoy cabbages showed blocking effects againstDMBA-induced mammary carcinogenesis in rats. It was alsofound that consumption of the cabbage powders beginning 1week after DMBA administration suppressed mammary tumorformation. Subsequently, benzyl isothiocyanate was shown toblock DMBA-induced mammary carcinogenesis and to suppress formation of mammary tumors under similar conditionsto those used with crude plant materials (15).

Citrus fruit oils were also found to have both blocking andsuppressing capacities. The most extensively studied is orangeoil, which has a blocking effect against DMBA-induced mammary neoplasia and a suppressive effect in this experimental

model (8). o-Limonene, found in orange oil, was shown to haveboth blocking and suppressing effects against DMBA-inducedmammary tumor formation (12). Inhibitors having both blocking and suppressing capacities, such as cruciferous vegetablepowders, benzyl isothiocyanate, orange oil, and D-limonene,have another property in common; They all induce increasedglutathione 5-transferase activity in mouse tissues. The significance of this relationship remains to be explained.

The finding of inhibitors with both blocking and suppressingactivity appears important. However, mechanisms for thesedual inhibitory actions are not known. Two distinctive possibilities merit consideration. One is that a coordinated protectivesystem against potentially toxic compounds exists that includesboth a detoxification component and a suppressive component.The latter would be a fail-safe means of protection should theformer be not completely effective. On exposure to toxic compounds, the detoxification system constitutes the first line ofdefense. Phase 2 enzymes, as well as glutathione, are components of this detoxification (blocking) defense. Because blockingeffects may not always be completely successful, a second lineof defense might prevent manifestations of damage that occurs,i.e., suppression.

Whereas a coordinated protective mechanism entailing increased activity of detoxification systems and suppression ofmanifestations of toxicity appears to be a reasonable possibility,the inclusion of a decrease in cytochrome P-450 enzyme activities might appear incongruous. However, an insight into howsuch enzyme inhibitions might fit into an overall protectivesystem was provided by experiments that show that somecompounds that reduce cytochrome P-450 activity increaseinterferon levels in tissues (64, 65). Such findings suggest thata decrease in activity of one mechanism that is involved inprotection can be accompanied by an enhanced activity of adifferent protective system.

A recent publication by Nebert (66) conceptualize how blocking and suppressing activity might be brought about by thesame compound. This investigator discusses a possible relationship between components of the nuclear receptor superfamily,with foreign chemicals as ligands, and cell proliferation. Themouse peroxizome proliferator-activated receptor, a novelmember of the nuclear receptor superfamily, has foreign chemicals as its only ligands (67). Numerous studies show thatperoxizome proliferation is associated with liver carcinogenesis.Relationships are proposed between the level of activity of drug-metabolizing (foreign body-metabolizing) enzymes, the steady-state level of small oxygenated molecules (including steroids)that act as signals for growth, differentiation, and possiblytumor promotion. If such interactions exist, opposing systemsmight exist as well. One would entail inhibition of the activityof the drug-metabolizing systems and a decrease in signals forcell proliferation. Although these conjectures are highly speculative, they form the conceptual basis for relating the activityof systems that detoxify chemical carcinogens to nuclear controlmechanisms of cell proliferation. If a defense mechanism withboth blocking and suppressing components exists, it could betriggered by a receptor mechanism.

A second possibility that might account for a compoundhaving both blocking and suppressing activity is high reactivitywith multiple effects due to this attribute. In such cases, blocking and suppressing effects would be triggered separately. Aromatic isothiocyanates are highly reactive compounds and theircapacity to act as both blocking and suppressing agents mightbe due to their high chemical reactivity. In contrast, D-limonene

2089s



NONNUTRIENT INHIBITORS OF CARC1NOGENESIS

is not a highly reactive molecule. However, it is metabolized toepoxides that are highly reactive. The paucity of data makes itdifficult to evaluate the two major mechanisms of action considered. If a receptor activating both blocking and suppressingaction were identified, it would be a major achievement. Evenif such a unified defense system does not exist, knowledge aboutthe characteristics of inhibitors having both blocking and suppressing effects might enhance the possibility of obtainingchemopreventive agents with maximum protective capacities.

Summary

EpidemiolÃ³gica! studies have shown that diets containinglarge quantities of vegetables and, to a lesser extent, fruits areassociated with relatively low risks from cancer. The information about which constituents of these foods are responsible forprotection has been indecisive. One group of inhibitory compounds currently under investigation consists of minor nonnu-trients in diets. Foods of plant origin frequently containamounts of these substances as high as several per cent. Inrecent years there has been a growing awareness that thesenonnutrient compounds may have important effects on theconsequences of exposure to carcinogenic agents. This information comes from studies of whole diets; studies of individualdietary constituents, particularly those of plant origin; and alarge number of investigations with pure compounds. The non-nutrient inhibitors of carcinogenesis have several differentmechanisms of action. Some are blocking agents; i.e., theyprevent carcinogens from reaching or reacting with criticaltarget sites. Others are suppressing agents; i.e., they preventevolution of the neoplastic process in cells which otherwisewould become malignant. Occasionally the same compoundwill show both mechanisms. As studies of the inhibitory capacities of nonnutrient dietary inhibitors of carcinogenesis continue, a clearer picture of their potential impact in cancerprevention will emerge. Pending definitive data and as a practical consequence of the current state of knowledge, the basicrecommendation of the National Research Council, the National Cancer Institute, and the American Cancer Society thatthe daily diet should contain generous amounts of vegetablesand fruits should be followed. For individuals of sufficientsocioeconomic status, following this recommendation is a matter of education and choice. For the economically deprived,making available diets that decrease the risk of cancer involvesa public health concern.

References

1. Silverstone, II.. Solomon. R. !>.. and Tannenbaum, A. Relative influences ofnatural and semipurified diets on tumor formation in mice. Cancer Res., 12:750-756, 1952.

2. Hecht, S. S., Morse, M. A., Shantu, A., Stoner, G. D., Jordan, K. G., Choi,C-I., and Chung, F-L. Rapid single-dose model for lung tumor induction inA/J mice by 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone and the effectof diet. Carcinogenesis (Lond.), 10: 1901-1904. 1989.

3. Wattenberg, L. W., Borchert, P., Destafney, C. M., and Coccia, J. B. Effectsofp-methoxyphenol and diet on carcinogen-induced neoplasia of the mouseforestomach. Cancer Res., 43: 4747-4751, 1983.

4. Longnccker. D. S., Roebuck, B. D.. Yager, J. D.. Lilja, H. S., and Siegmund.B. Pancreatic carcinoma in azaserine-treated rats: induction, classification,and dietary modulation of incidence. Cancer (Phila.), 47: 1562-1572. 1981.

5. Committee on Diet, Nutrition and Cancer. Diet, Nutrition and Cancer, pp.358-370. Washington, DC: National Academy Press, 1982.

6. Longnecker, D. S., Kuhlman. E. T., Roebuck, B. D., and Curphey T. J.Pancreatic carcinogenesis in azaserine-treated rats: inhibition by a solventmixture in the diet. Food Chem. Toxicol., 25: 823-828, 1987.

7. Wattenberg, L. W. Inhibition of carcinogenesis by naturally-occurring andsynthetic compounds. In: Y. Kuroda, D. M. Shankel, and M. D. Waters

12.

13.

14.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

(eds.), Antimutagenesis and Anticarcinogenesis Mechanisms II, pp. 155-166. New York: Plenum Publishing Corp., 1990.Wattenberg, L. W. Inhibition of neoplasia by minor dietary constituents.Cancer Res., 43: 2448s-2453s, 1983.Belman, S. Onion and garlic oils inhibit tumor promotion. Carcinogenesis(Lond.), 4: 1063-1065, 1983.Troll, W., Wiesner, R., and Frankel K. Anticarcinogenic action of proteaseinhibitors. Adv. Cancer Res., 49: 265-283, 1987.Fujiki, H., Suganuma, M., Suguri H., Takagi, K., Yoshizawa, S., Ootsuyama,A., Tanooka, H., Okuda, T., Kobayashi, M., and Sugimura, T. New anti-tumor promoters: (-)-epigallocatechin gallate, and sarcophytols A and B. In:Y. Kuroda, M. Shankel, and M. D. Waters (eds.), Antimutagens and Anticarcinogenesis Mechanisms II, pp. 205-212. New York: Plenum PublishingCorp., 1990.Elson, C. E., Maltzman, J. L., Boston, J. L., and Tabber, M. A. Anticarcinogenic activity of </ limonelle during initiation and promotion/progressionstages of DMBA-induced mammary carcinogenesis. Carcinogenesis (Lond.),9:331-332, 1988.Wattenberg, L. W., Hanley, A. B., Barany, G., Sparnins, V. L., Lam, L. K.T., and Fenwick, G. R. Inhibition of carcinogenesis by some minor dietaryconstituents. In: Y. Hayashi et al. (eds.), Diet, Nutrition and Cancer, pp.193-203. Tokyo: Japan Scientific Societies Press, 1985.Siimi)oslii, H., and Wargovich, M. J. Chemoprevention of 1,2-dimethylhy-drazine induced colon cancer in mice by naturally occurring organosulfurcompounds. Cancer Res., 50: 5084-5087, 1990.Wattenberg, L. W. Chemoprevention of cancer. Cancer Res., 45: 1-8, 1985.Wei, H., Tye, L.. Bresnick, E., and Bin, D. F. Inhibitory effect of apigenin,a plant flavonoid, on epidermal ornithine decarboxylase and skin tumorpromotion in mice. Cancer Res., 40: 499-502, 1990.Verma, A. K., Johnson, J. A., Gould, M. N., and Tanner, M. A. Inhibitionof 7,12-dimethylbenz(u)anthracene- and A'-nitrosomethylurea-induced ratmammary cancer by dietary flavonol quercetin. Cancer Res., 48:5754-5758,1988.Wattenberg, L. W., Sparnins, V. L., and Barany, G. Inhibition of A'-nitro-sodiethylamine carcinogenesis by naturally occurring organosulfur compounds and monoterpenes. Cancer Res., 49: 2689-2692, 1989.Wargovich, M. J., Woods, C., Eng, V. W. S., Stephens, L. C, and Gray, K.Chemoprevention of A'-nitrosomethylbenzylamine-induced esophageal can

cer in rats by the naturally occurring thioether, diallyl sulfide. Cancer Res.,48:6872-6875, 1988.Sparnins, V. L.. Barany, G., and Wattenberg, L. W. Effects of organo-sulfurcompounds from garlic and onions on benzo(a)pyrene-induced neoplasia andglutathione 5-transferase activity. Carcinogenesis (Lond.), 9:131-134,1988.Sparnins, V. L., Mott, A. W., Barany, G., and Wattenberg, L. W. Effects ofallyl methyl trisulfide on glutathione S-transferase activity and BP-inducedneoplasia in the mouse. Nutr. Cancer, 8: 211-215, 1986.Wattenberg, L. W. Inhibitory effects of benzyl isothiocyanate administeredshortly before diethylnitrosamine on pulmonary and forestomach neoplasiain A/J mice. Carcinogenesis (Lond.), 12: 1971-1973. 1987.Morse, M. A., Wang, C-X., Stoner. G. D., Mandai, S., Conran, P. B., Shantu,G. A., Hecht, S., and Chung, F-L. Inhibition of 4-(methylnitrosamino)-l(3-pyridyl)-l-butanone-induced DNA adduci formation and tumorigenicity inthe lung of F344 rats by dietary phenethyl isothiocyanate. Cancer Res., 49:549-553, 1989.Wattenberg, L. W., and Bueding, E. Inhibitory effects of 5-(2-pyrazinly)-4-methyl-l,2-dithiol-3-thione (Oltipraz) on carcinogenesis induced bybenzo(a)pyrene, diethylnitrosamine and uracil mustard. Carcinogenesis(Lond.), 7: 1379-1381, 1986.Hartman, P. E., and Shankel, D. M. Antimutagens and anticarcinogens: asurvey of putative interceptor molecules. Environ. Mol. Mutagen., IS: 145-182, 1990.Kato. R., Nakadate. T., Yamamoto, S., and Sugimura, T. Inhibition of 12-0-tetra-decanoylphorbol-13-acetate-induced tumor promotion and ornithinedecarboxylase activity by quercetin: possible involvement of lipoxygenaseinhibition. Carcinogenesis (Lond.), 4: 1301-1305, 1983.Mukhtar, H., Das, M., Khan. W. A., Wang, Z. Y., Bik, D. P.. and Bickers,D. R. Exceptional activity of tannic acid among naturally occurringplant phenols in protecting against 7,12-dimethylbenz(a)anthracene-,benzo(a)pyrene-, 3-methylcholanthrene-, and A'-methyl-A'-nitrosourea-in-duced skin tumorigenesis in mice. Cancer Res., 48: 2361-2365, 1988.Athar, M., Khan. W. A., and Mukhtar, H. Effect of dietary tannic acid onepidermal, lung, and forestomach polycyclic aromatic hydrocarbon metabolism and tumorigenicity in Sencar mice. Cancer Res., 49: 5784-5788, 1989.Mandel, S., and Stoner, G. D. Inhibition of /V-nitrosobenzylmethylamine-induced esophageal tumorigenesis by ellagic acid. Carcinogenesis (Lond.),//: 55-61, 1990.Chang, R. L., Huang, M. T., Wood, A. W., Wong C. Q., Newmark, H. L.,Yagi, H., Sayer, J. M., Jerind, D. M., and Conney, A. H. Effect of ellagicacid and hydroxylalcd flavonoids on the tumorigenicity of benzo(a)pyreneand (Â±)-7tf,8-dihydroxy-9,10-epoxy-7,8,9,10-tetrahydrobenzo(a)pyrene onmouse skin and in the newborn mouse. Carcinogenesis (Lond.), <>:1127-1133, 1985.Tanaka. T., Iwala, H., Niwa, K., Mori, Y., and Mori, H. Inhibitory effect ofellagic acid on iV-2-fluorenylacetamidc induced liver carcinogenesis in maleAC1/N rats. Jpn. J. Cancer Res., 79: 1297-1303, 1988.Huang, M-T., Smart, R. C., and Conney, A. D. Inhibition of tumor promotion by curcumin, a major constituent of the food additive, tumeric. Proc.

2090s



NONNIITRIENT INHIBITORS OF CARCÃ•NOGESESIS

Am. Assoc. Cancer Res., 2Â«:173. 1987.33. Ha, Y. L., Stockson, J., and Pariza. M. Inhibition of benzo(o)pyrene-induced

mammary tumor formation by conjugated dienoic derivatives of linoleic acid.Cancer Res., 50: 1097-1101. 1990.

34. Nishino, 11.. Kitagawa, k . and Iwashima, A. Antitumor-promoting activityof glycyrrhetinic acid in mouse skin tumor formation by 7.12-dimethylbenz-(a)-anthracene plus teleocidin. Carcinogenesis (Lond.), 5: 1529-1530. 1984.

35. Walaszek. Z.. Hanavsek-Walaszek. M.. and Webb. T. E. Dietary glucarate-mediated reduction of sensitivity of murine strains to chemical carcinogene-sis. Cancer Lett.. 33: 25-32. 1986.

36. Walaszek. A., and Hanavsek-Walaszek, M. Dietary glucarate inhibits ratmammary tumorigenesis induced by A'-methyl-A'-nitrosourea. Proc. Am.

Assoc. Cancer Res., 28: 153, 1987.37. Walaszek, Z. Potential use of i>-glucaric acid derivatives in cancer prevention.

Cancer Lett., 54: 1-8. 1990.38. Wattenberg. L. W. Inhibition of carcinogenic effects of polycyclic hydrocar

bons by benzyl isothiocyanate and related compounds. J. Nail. Cancer Inst.,5Â«:395-398, 1977.

39. Fenwick. G. R., Heaney. R. K.. and Mullin. W. J. Glucosinolates and theirbreakdown products in food and food plants. CRC Crit. Rev. Food Sci.Nutr., 18: 123-201. 1983.

40. Chung. F-L., Juchatz, A., and Hecht, S. S. Effects of dietary compounds ona-hydroxylation of A'-nitrosopyrrolidine and A"-nitrosonornicotine in rattarget tissues. Cancer Res., 44: 2924-2928. 1984.

41. Chung, F-L., Wang, M., and Hecht, S. S. Effects of dietary' Ã•ndolesandisothiocyanateson A'-nÂ¡trosodÂ¡methylamÂ¡neand4-(methylnitrosamine)-l-(3-pyTidyl)-l-bu(anone-Â«-hyd^oxylationand DNA methylation in rat liver. Cancer Res., 45: 539-543, 1985.

42. Fenwick, G. R.. and Hanley, A. B. The genus Allium. CRC Crit. Rev. FoodSci. Nutr.. 22: 199-271, 1985.

43. Fenwick, G. R., and Hanley. A. B. The genus Allium. Part 2. CRC Crit. Rev.Food Sci. Nutr.. 22: 273-377. 1985.

44. Fenwick. G. R.. and Hanley. A. B. The genus Allium. Part 3. CRC Crit. Rev.Food Sci. Nutr., 23: 1-73. 1985.

45. Brady, J. F., Li, D., Ishizaki, H.. and Yang, C. S. Effect of diallyl sulfide onrat liver microsomal nitrosamine metabolism and other monooxygenaseactivities. Cancer Res., 48: 5937-5940. 1988.

46. Shaw, P. E. Review of quantitative analyses of citrus essential oils. Agrie.Food Chem., 27: 246-257, 1979.

47. Solzin, 1 . J. The analysis of essential oils and extracts (oleoresins) fromseasoningsâ€”a critical review. CRC Crit. Rev. Food Sci. Nutr., 9: 345-373,1977.

48. Wattenberg, L. W., and Coccia, J. Inhibition of 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone Carcinogenesis in mice by i>limonene and citrus fruitoils. Carcinogenesis (Lond.), 12: 115-117, 1991.

49. Narisawa, T.. Sato, M., Tahi, M., Kudo. T., Takahashi, T., and Goto, A.Inhibition of the development of methylnitrosourea-induced rat colon tumorsby indomethacin treatment. Cancer Res., 4l: 1954-1957, 1981.

50. Rubio. C. A. Antitumoral activity of indomethacin on experimental esopha-geal tumors. J. Nati. Cancer Inst., 72: 705-707. 1984.

51. Kelloff, G. J., Malone. W. F.. Boone. C. W., Sigman. C. C.. and Fay, J. R.Progress applied chemoprevention research. Semin. Oncol., 17: 438-455,1990.

52. Shamsuddin. A., Elsayed, P. M., and Ullah, A. Suppression of large intestinalcancer in F344 rats by inositol hexaphosphate. Carcinogenesis (Lond.), 9:577-580. 1988.

53. Shamsuddin. A., and Ullah, A. Inositol hexaphosphate inhibits large bowelintestinal cancer in F344 rats 5 months after induction by azoxymethane.Carcinogenesis (Lond.). IO: 625-626, 1989.

54. Committee on Diet, Nutrition, and Cancer. Diet, Nutrition and Cancer, pp.141-144. Washington. DC: National Academy Press, 1982.

55. Pearson, E. R., and Vogelstein, B. A genetic model of colorectal tumorigenesis. Cell, 61: 759-767. 1990.

56. Wattenberg, L. W., Shafer, H. W., Waters, L., and Davis, D. W. Inhibitionof mammary tumor formation by broccoli and cabbage. Proc. Am. Soc.Cancer Res.. JO: 181. 1989.

57. Wattenberg. L. W. Inhibition of azoxymethane-induced neoplasia of the largebowel by 3-hydroxy-3.7.l l-trimethyl-l.6,10-dodecatriene. Carcinogenesis(Lond.), 12: 151-152, 1991.

58. Zile, M. H.. Cullum, M. E., Roltsch, I. A., DeHoog, J. V., and Welsch, C.W. Effects of moderate vitamin A supplementation and lack of dietaryvitamin A on the development of mammary tumors in female rats treatedwith low carcinogenic dose levels of 7,12-dimethylbenz(a)anthracene. CancerRes., 46: 3495-3503, 1986.

59. Krinksy, N. I. Carotenoids and cancer in animal models. J. Nutr., II9: 123-126, 1989.

60. Medina, D., and Shephard. F. Selenium-mediated inhibition of 7,12-di-methyl-benz(a)anthracene-induced mammary tumorigenesis. Carcinogenesis(Lond.), 2:451-455, 1981.

61. Thompson, H. J., Meeker. L. D., and Kokosha, S. Effect of inorganic andorganic form of dietary selenium on the promotional stage of mammaryCarcinogenesis in the rat. Cancer Res.. 44: 2803-2806. 1984.

62. Shamsuddin. A. M., Elsayed. A. M., and Ullah, A. Suppression of largeintestinal cancer in F344 rats by inositol hexaphosphate. Carcinogenesis(Lond.). 9: 577-580, 1988.

63. Shamsuddin, A. M.. and Ullah, A. Inositol hexaphosphate inhibits largeintestinal cancer in F344 rats 5 months following induction by azoxymethane.Carcinogenesis (Lond.), 10:625-626, 1989.

64. Deloria. L., Abbott, N.. Gooderham, N.. and Mannering, G. J. Induction ofxanthine oxidase and depression of cytochrome P-450 by interferon inducers:Genetic differences in the responses of mice. Biochem. Biophys. Res. Commun., 131: 109-114. 1985.

65. Gooderham, N. J., and Mannering, G. J. Depression of cytochrome P-450and alterations of protein metabolism in mice treated with the interferoninducer polyriboinosinic acid, polyribocytidylic acid. Arch. Biochem. Biophys., 250: 418-425. 1986.

66. Nebert, D. W. Growth signal pathways. Nature (Lond.), 347:709-710,1990.67. Isseman. I., and Green. S. Activation of a member of the steroid hormone

receptor superfamily by peroxisome proliferators. Nature (Lond.), 347: 645-649. 1990.

2091s



1992;52:2085s-2091s. Cancer Res Lee W. Wattenberg Inhibition of Carcinogenesis by Minor Dietary Constituents

Updated version

http://cancerres.aacrjournals.org/content/52/7_Supplement/2085s

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/52/7_Supplement/2085sTo request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



inhibition of carcinogenesis by minor dietary constituents1€¦ · a protection by vegetables...

Documents