Fd Chem. Toxic. Vol. 30, No. 4, pp. 289-295, 1992 Printed in Great Britain.All rights reserved
0278-6915/92 $5.00 + 0.00 Copyright © 1992Pergamon Press Ltd
A N T I M U T A G E N I C I T Y OF ELLAGIC ACID TOWARDS THE FOOD M U T A G E N IQ: INVESTIGATION INTO POSSIBLE MECHANISMS OF ACTION A. D. AYRTON*, D. F. V. LEWIS, R. WALKERand C. IOANNIDES'[" Molecular Toxicology Research Group, Division of Toxicology, School of Biological Sciences, University of Surrey, Guildford, Surrey GU2 5XH, UK (Accepted 18 December 1991)
Abstract--The ability of the plant phenol ellagic acid to inhibit the mutagenicity of the food mutagen IQ was evaluated using Salmonella typhimurium strain TA98 in the Ames mutagenicity test. Ellagic acid caused a concentration-dependent decrease in the S-9- and microsome-mediated mutagenicity of IQ. The plant phenol did not interact directly with the IQ-derived mutagenic species and did not modify the cytosol-mediated activation of the promutagen. At the concentrations used in the mutagenicity studies, ellagic acid failed to inhibit microsomal mixed-function oxidase activity, including that mediated by the P450I family responsible for the bioactivation of IQ, despite being an essentially planar molecule as indicated by computer-graphic analysis. The inhibitory effect of ellagic acid was independent of its ability to chelate Mg2+ . However, pre-incubation of ellagic acid with the bacteria, followed by removal of the plant phenol, did not completely prevent the inhibitory effect of the phenol on the mutagenicity of IQ. Intraperitoneal administration of ellagic acid to rats caused a decrease in total cytochrome P-450 levels and related activities as well as in cytosolic glutathione S-transferase activity. Finally, the possibility that the reported anticarcinogenic action of ellagic acid reflects nothing more than non-selective destruction of hepatic cytochromes P-450, and thus reduced chemical activation, is considered.
INTRODUCTION Naturally occurring plant polyphenols, such as ellagic acid (2,3,7,8-tetrahydroxy [1] benzopyrano[5,4,3-cde] [1] benzopyran-5,10-dione), have been claimed to possess antimutagenic and anticarcinogenic activities towards major groups of environmental carcinogens, including polycyclic aromatic hydrocarbons, mycotoxins and nitrosamines (Dixit and Gold, 1986; Mandal et al., 1987; Mukhtar et al., 1986). These studies were provoked by initial reports that ellagic acid was a potent inhibitor of the mutagenicity of the bay-region diol epoxides, the ultimate carcinogenic intermediates of polycyclic aromatic hydrocarbons, in the Ames test (Sayer et aL, 1982; Wood et al., 1982) and inhibited the binding of 3-methylcholanthrene to D N A (Mukhtar et al., 1984). In human bronchial epithelial cells, however, ellagic acid increased the toxicity of the 7,8-diol of benzo[a]pyrene, the precursor of the ultimate carcinogen (Teel et al., 1986). Other workers failed to detect any effect of ellagic acid on the metabolism of benzo[a]pyrene by mouse skin in organ culture (Shugart and Kao, 1984). Finally, studies conducted *Present address: Rh6ne-Poulenc Agrochimie, Division Protection des Cultures, 14-20 rue Pierre Baizet, 69263 Lyon, Cedex 09, France. tTo whom all correspondence should be addressed. Abbreviations: DCNB ffi 1,2-dichloronitrobenzene; Glu-P-I = 2-amino-6 methyldipyrido[l,2-a:Y,2'-d]imidazole; IQ ffi 2-amino-3-methylimidazole[4,5-f]quinoline.
in vivo demonstrated that ellagic acid, topically applied to the skin of mice, afforded protection against the induction of skin tumours by 3-methylcholanthrene and 7,12-dimethylbenzanthracene (Lesca, 1983; Mukhtar et al., 1984). In the same animal species, the plant phenol was reported to inhibit the benzo[a]pyrene-induced pulmonary tumorigenesis (Lesca, 1983). It has been proposed that ellagic acid could exert its antimutagenic and anticarcinogenic actions through a combination of one or more of the following distinct mechanisms: (1) inhibition of the enzymes responsible for bioactivation of the carcinogen, such as the cytochrome P-450-dependent mixed-function oxidases; (2) stimulation of the enzymes involved in the detoxication of reactive intermediates, such as the glutathione S-transferases; and (3) through direct interaction of the phenol with the reactive intermediates to form inactive adducts, as previously described (Das et al., 1985; Wood et al., 1982). A fourth possible mechanism is that ellagic acid may interact with D N A in such a way as to decrease the number of binding sites available for interaction with the ultimate mutagens/carcinogens (Teel, 1986). It is now widely recognized that a major group of genotoxic carcinogens are the imidazo-quinolines and -quinoxalines and carbolines generated during the cooking of protein-rich foods (Overvik and Gustaffson, 1990). These chemicals are potent bacterial mutagens in the presence of appropriate activation systems (Kato et al., 1983) and are carcinogenic in
289
290
A.D. AYRTONet aL
mice and rats, giving rise to tumours in multiple sites including the liver (Ohgaki et aL, 1991). Since these compounds are likely to be ingested simultaneously with plant polyphenols, they provide an appropriate model for investigating the possible mechanisms of action of ellagic acid following dietary exposure, and this is the purpose of the present study. MATERIALSAND METHODS IQ (2-amino-3-methylimidazo [4,5-f]quinoline), GIu-P-I (2-amino-6-methyldipyrido [1,2-a:3,2'-d]imidazole) (Wako Fine Chemicals, 4040 Neuss 1, Germany), ethoxyresorufin, pentoxyresorufin, resorufin (Molecular Probes, Eugene, OR, USA), Aroclor 1254 (Robens Institute of Health and Safety, Guildford, Surrey, UK), 1,2-dichloronitrobenzene (Alldrich Chemicals Co. Ltd, Gillingham, Dorset, UK), abscissic acid (Fluka AG, Buchs, Switzerland), ellagic acid, menadione, chlorogenic acid, ethoxycoumarin, cytochrome c, bovine serum albumin and all cofactors (Sigma Co., Poole, Dorset, UK) were purchased. The Salmonella typhimurium strain TA98 was a kind gift from Professor B. N. Ames, University of California, Berkeley, CA, USA. Male Wistar albino rats (150-200 g) were purchased from the Experimental Biology Unit, University of Surrey. Induction of the hepatic mixedfunction oxidase system was achieved by a single ip injection of Aroclor 1254 (500 mg/kg), dissolved in corn oil (200 mg/ml), the animals being killed on the fifth day following administration. The effect of ellagic acid on the hepatic mixed-function oxidase system was investigated following single daily ip administrations of the plant phenol (100 mg/kg) for 5 days, whereas controls received the corresponding volume of vehicle, 0.9% (w/v) NaCI in 0.1 mM-phosphate buffer pH 7.4, all animals being killed 24 hr after the last dose. Hepatic post-mitochondrial supernatant (9000g supernatant, S-9 fraction), microsomal fraction (105,000g pellet resuspended in homogenizing medium) and cytosolic fraction (105,000g supernatant) were prepared as previously described (loannides and Parke, 1975). The following assays were carried out on the microsomal fraction: the O-dealkylations of ethoxyresorufin (Burke and Mayer, 1974), pentoxyresorufin (Lubet et al., 1985) and ethoxycoumarin (Ullrich and Weber, 1972), the NADPH-dependent reduction of cytochrome c Table 1. Plant phenol
(Williams and Kamin, 1962) and total c y t o c h r o m e P-450 levels (Omura and Sato, 1964); on the cytosolic fraction glutathione S-transferase activity was assayed using 1,2-dichloronitrobenzene (DCNB) as the acceptor substrate (Habig et al., 1974). Protein was determined on both hepatic fractions (Lowry et al., 1951) using bovine serum albumin as standard. The mutagenicities of IQ and GIu-P-I were determined in the Ames test (Maron and Ames, 1983), using S. typhimurium strain TA98. In experiments where microsomes only were used for the activation, the activation system was supplemented with glucose6-phosphate dehydrogenase (1 unit/plate). In studies where the role of the cytosol was evaluated, t h e bacteria, IQ and microsomal activation system were pre-incubated for 20 min, at 37°C in a shaking water-bath, before termination of the microsomal reaction by the addition (100/d/plate) of menadione (900/zM). This was followed by the addition of either buffer or the cytosolic fraction. IQ, GIu-P-1 and the plant antimutagens were always dissolved in dimethylsulphoxide so that the total amount of t h e solvent in the activation system was constant and never exceeded 100/zl. At the concentrations used, none of the phenols was toxic to the bacteria (results not shown). The molecular geometry of ellagic acid was determined using the COSMIC package (written by J. G. Vinter, A. Davies and M. R. Saunders of SKF Ltd). The molecular plot, which is viewed in, and perpendicular to, the molecular plane, was generated using the PLUTO crystallographic package. The following van der Waals radii were utilized: C, 1.6 A; O, 1.4 A and H, 1.2 A. Statistical evaluation was carried out using Student's t-test RESULTS
Of the plant compounds studied, only ellagie acid produced a concentration-dependent decrease in t h e Aroclor 1254-induced mutagenicity of IQ (Table 1) and was therefore investigated further. A more detailed study showed that, when post-mitochondrial fractions from Aroclor 1254-induced rats were used as activation system, the inhibition of the IQ-mediated mutagenic response by ellagic acid was related to the concentration both of the plant phenol and of t h e mutagen (Fig. IA). When isolated microsomes (i.e. in the absence of cytosol) were used as the activation system, the mutagenic response of IQ was considerably lower than that following activation by t h e
Amestest* Histidine revertantsper platet at a plant phenol concentration (M) of:
Effect of various phenols on the mutagcnicity of IQ in the 0
Chlorogenic acid 6879 + 1467 Ellagic acid 10352 ± 860 Abscissicacid 4537± 69
10 -7
10 -6
10 -5
10 -4
8410 + 1105 9087 ± 292 6578 ± 330
7681 ± 1460 8520 ± 370
7452 ± 440 7656 ± 495
6579 ± 674 1424 ± 201 5649± 490
4990± 1072
5538± 588
*The test was carried out using Salmonella typhimurium strain TA98, IQ (I00 ng/plate) and Aroclor 1254-induced hepatic post-mitochonddal activation systems (I0%, v/v). tResults are presented as mean + SD for triplicates. The spontaneous reversion rate of 25 ± 6 has already been subtracted.
Ellagic acid inhibits the mutagenicity of IQ post-mitochondrial (S-9) fraction (Fig. 1B). The mutagenic activity of IQ was once again inhibited by ellagic acid in a concentration-dependent fashion. Finally, neither ellagic acid nor the other plant cornpounds, at the concentrations used, displayed any mutagenic effect (results not shown). When the microsomal activation of IQ was terminated by the addition of menadione, addition of the cytosol, as expected, enhanced the mutagenic potency of the carcinogen. When incorporated into the cytosol, ellagic acid did not influence the mutagenicity of IQ (Fig. 2). When ellagic acid was incorporated into the buffer, instead of the cytosol, the mutagenicity of IQ was not significantly altered, except for a decrease at the highest concentration of the phenol, 5 x 10 -s M (Fig. 2). At the concentrations used in the mutagenicity studies, ellagic acid had no effect on the O-dealkylations of ethoxyresorufin or pentoxyresorufin, or the NADPH-dependent reduction of cytochrome c. B
10,000
(A) 8000
c
4000
2000
o
', 0
20
40
60
80
I
I
1 O0
120
IQ concentration (ng/plate) 1200
_=
-
(B)
ca- ~ 1000
--
4=# C
800
-
>
600
C
400
.=
I
:6 .m
-I-
--
500
--
400
--
300
--
200
--
100
-
~...........---t
x
O. .~
~, ~>
',~
= -l-
0
10-7
,
, ,=~,ld 10-6
= , =l,,,I
i
, ~,,=,,I
10-5
10-4
Ellagic acid c o n c e n t r a t i o n (M)
Fig. 2. Effect of ellagic acid on the cytosolic potentiation of the microsome-mediated mutagenicity of IQ. IQ (10 ng) was preincubated with Salmonella typhimurium strain TA98 and an Aroclor 1254-induced hepatic microsomal activation system (10%] for 20 rain at 37°C in a shaking water-bath. The reaction was terminated by addition of 100/tl menadione (900#M). Cytosol (&) or buffer ( 1 ) containing various concentrations of ellagic acid were then added and a further 20-min pre-incubation was carried out at 37°C. Results are presented as mean + SD for triplicates. The spontaneous reversion rate of 25 + 7 has already been subtracted.
6000
c
t~
600
291
200
0 o
20
40
60
80
100
120
IQ concentration (ng/plate)
Fig l. Inhibition of the mutagenicity of IQ by ellagic acid. IQ, Salmonella typhimurium strain TA98 and hepatic S-9 (A) or microsomal (B) activation system (10%) derived from Aroclor 1254-induced rats were incubated in the absence (&) and presence of ellagic acid at concentrations of 5 X 10-TM ( I ) , 10-6 M (rq), 5 x 10-6M (O), I0-SM (A) and 5 x 10-~M (O). Each point represents the mean of triplicates. The spontaneous reversion rate of 22 + 5 has already been subtracted.
When Glu-P- 1 was used as the premutagen (20-80 ng! plate) in the Ames test and microsomes from Aroclor 1254-rats as the activation system, ellagic acid (5 x 10-7-5 x 10 -5 M) had no significant influence on the mutagenic response (results not shown). Presence of magnesium ions did not modify the mutagenic effect of IQ; similarly, the inhibitory effect of ellagic acid was independent of the presence of magnesium ions (Table 2). Pre-incubation of the bacteria with ellagic acid, followed by the removal of the plant phenol by successive washings with nutrient broth, did not completely prevent the inhibitory effect of the phenol on the mutagenicity of IQ (Table 3). When administered to rats by the ip route, precipitated ellagic acid was present in the abdominal cavity and body-weight loss was observed in the animals treated with the phenol. The same treatment gave rise to a decrease in the microsomal levels of total cytochrome P-450, N A D P H cytochrome c reductase activity and microsomal and cytosolic protein levels, but not all effects, were statistically significant (Table 4). Similarly, the O-deethylation ofethoxycoumarin and cytosolic glutathione S-transferase activity were significantly inhibited; the O-dealkylations of ethoxyresorufin and pentoxyresorufin were also reduced by the ellagic acid treatment but the effect was not statistically significant (Table 5). Microsomes from ellagic acid-treated rats did not differ significantly from preparations from control animals in their ability to convert GIu-P-I to mutagens in the Ames test (Table 5). The computer-optimized dimensions of ellagic acid were: length, 12.1 A.; width 10.2A and depth 3.2A,
A . D . AYRTON et al.
292
Table 2. Effect of Mg 2+ ions and ellagic acid on the mutagenicity of IQ* Histidine revertants per platet Supplementation of activation system
+ MgCI 2
- MgCI2
-Ellagic acid
2711 + 474 779 _+ 140
2638 ± 284 736 __ 160
*The study was carried out using Salmonella typhimurium strain TA98. The bacteria, IQ (20 ng) and microsomes (10%, v/v) from Aroclor 1254-treated rats, in the presence and absence of MgCI 2 (8 raM) and/or ellagic acid (5 x l0 -5 M) were pre-incubated for 20 minutes at 37°C in a shaking water-bath. ?Results are presented as mean + SD for three plates. The spontaneous reversion rate of 25 _+ 5 has already been subtracted.
giving rise to an area/depth 2 of 12.0, showing that it is essentially a planar molecule (Fig. 3). DISCUSSION
Consumption of diets with a high vegetable content has been repeatedly correlated with reduced human cancer incidence at a number of sites (Graham et al., 1978). This observation has led to the evaluation of various plant compounds as possible anticarcinogens that may contribute to reduced cancer risk. The most extensively studied is ellagic acid, which was shown to inhibit the mutagenicity of a number of promutagens and ultimate mutagens in the Ames test (Dixit and Gold, 1986; Mandal et al., 1987; Wood et al., 1982) and to diminish the carcinogenicity of chemical carcinogens such as polycyclic aromatic hydrocarbons (Mukhtar et al., 1984 and 1986; Lesca, 1983); however, other workers failed to demonstrate an anticarcinogenic effect of ellagic acid towards benzo[a]pyrene and 3-methylcholanthrene (Chang et al., 1985; Smart et al., 1986a). In the present study we investigated the possible antimutagenic effect of ellagic acid towards IQ because both compounds can be present in food. Ellagic acid decreased the S-9-mediated mutagenicity of IQ, which was virtually abolished at a phenol concentration of 10 -s M. One or more of the following mechanisms may be responsible for the antimutagenic effect of ellagic acid: (1) ellagic acid or one of its metabolites may inhibit the microsomal Table 3. Mutagenicity of IQ to bacteria treated with ellaglc acid* Ellagic acid concentration (M)
Histidine revertants per plate
0 5 x 10 -7 5 x 10 -6 5 x 1 0 -s 5 x 10 -4
792 + 60 781+38 734+47 728+50 659 ± 78
*Salmonella typhimurium strain TA98 and ellaglc acid were pre-incubated for I hr at 37°C in a shaking water-bath. The bacteria were then washed three-times with nutrient broth and were then mixed with IQ (10ng) and Aroclor 1254-induced microseines (10%, v/v). Results are presented as mean + SD of tripficates. The spontaneous reversion rate of 29 + 9 has already been subtracted.
Table 4. Effect of treatment with ellagicacid on some rat hepatic parameterst Parameter Body-weight gain (g) Liver/body wt x 100 NADPH-cytochrome c reductase (nmol/min/mg protein) Total cytochrome P-450 (nmol/mg protein) Cytosolic protein (mg/g liver) Microsomal protein (mg/g liver)
Control
Ellagic acid
6.6 ± 5.5 4.7 -t- 0.1 18.5 ± 1.6
--8.0 + 3.0* 5.3 ± 0.2* 11.9 ±0.3
0.47 ± 0.01
0.25 ± 0.01"
88.9 +_4.6 17.8 -k 1.8
66.5 ± 2.2* 13.8 ± 0.4
*P < 0.05. tRats were treated with single daily ip injections of ellaglc acid (100 mg/kg) for 5 days whereas controls received the corresponding dose of the vehicle, phosphate buffered saline; all animals were killed 24 hr after last administration. Results are presented as mean ± SEM for five animals.
N-hydroxylation of IQ and/or its further metabolism to the ultimate mutagen by cytosolic enzymes; (2) the phenol may interact directly and non-enzymically with the proximate and/or ultimate mutagens, rendering them inactive; (3) ellagic acid, by virtue of its affinity for divalent ions, may complex with magnesium ions, thus depriving the activation system of an important component; and (4) the plant phenol may interact with DNA in such a way as to protect it from the ultimate mutagen. All these mechanisms were investigated. The reduction in the mutagenicity of IQ by ellagic acid was maintained even when isolated microsomes, rather than the post-mitochondrial supernatant (S-9), were used as the activation system, indicating that the phenol impairs the microsomal activation of IQ. The activation of this carcinogen is selectively catalysed by the P-450I family, and especially the A2 isoenzyme (Kate, 1986) and it is conceivable that ellagic acid is a potent inhibitor of this isoenzyme. Initially, a computer-graphic analysis was carried out to establish whether ellagic acid has the necessary molecular shape to interact with this cytochrome P-450 isoenzyme (Lewis et al., 1986). With an area/depth 2 ratio of 12 it is apparent that ellagic acid is an essentially planar molecule with a high potential for interaction with the P-450I family. However, in contrast to the predicted interaction, ellagic acid at Table 5. Rat hepatic mixed-function oxidase and glutathione Stransferase activities following treatment with ellagic acid Parameter Ethoxycoumarin O-deethylase (pmol/min/mg protein) (pmol/min/nmol P-450) Ethoxyresorufin O -deethylase (pmol/min/mg protein) (pmol/min/nmol P-450) Pentoxyresorufin O-depentylase (pmol/min/mg protein) (pmol/min/nmol P-450) Glutathione S-transferase (nmol/min/mg protein) Microsomal activationof Glu-P-I (2/*g) (histidine revertants/nmol P-450)
Control
Ellagic acid
41.8 + 7.4 89.0 + 12.8
20.3 -J- 3.4* 88.9 + 23.9
92 -6 1 200 + 10
68 -6 10 270 + 40
15 4- 2 33 + 6
II + I 46 -t- 4
78 ± 5
54 + 4*
1580 -I-210
2290 + 330
*P < 0.05. ?Treatment of animals and presentation of data are as in the footnotes to Table 4.
Ellagic acid inhibits the mutagenicity of IQ
293
Fig 3. Space-filled model of ellagic acid. Model was drawn using the PLUTO computer program and utilizing the following van der Waals radii to generate the computer-graphicplot of the molecular geometry: carbon, 1.6 A; oxygen, 1.4 A and hydrogen, 1.2 A. the concentrations studied failed to inhibit the Odeethylation of ethoxyresorufin and the bioactivation of GIu-P-I to mutagens, two reactions catalysed exclusively by the P-450I proteins, A1 and A2 respectively (Guengerich et al., 1982; Phillipson et al., 1984; Yamazoe et al., 1984). This lack of inhibition may be attributed to the relatively poor lipophilicity of the polyhydroxy phenol (also compatible with its poor absorption when orally administered; Smart et al., 1986b), which cannot reach the cytochrome P-450 system known to be embedded in lipid. Addition of the cytosolic fraction to the activation system, following termination of the microsomal metabolism, enhanced the mutagenicity of IQ, as we have previously described (Abu-Shakra et al., 1986). Incorporation of ellagic acid to the cytosolic fraction did not influence the mutagenic response, showing that the antimutagenic effect of the phenol does not result from the inhibition of the cytosolic activation pathway of the mutagen. As ellagic acid was reported to inhibit the mutagenicity of the ultimate carcinogen, benzo[a]pyrene 7,8-diol-9,10-epoxide, by an acid-catalysed direct interaction (Sayer et al., 1982; Wood et al., 1982), we investigated the possibility that the phenol could scavenge the microsome-generated reactive intermediates of IQ presumed to be the N-hydroxylamine. When microsomal metabolism was terminated, addition of ellagi¢ acid into the system at pH 7.4 gave rise to a decrease in the mutagenicity of IQ only at the highest dose of the phenol, thus making this mechanism of action unlikely. Ellagic acid readily complexes with divalent cations (Press and Hardcastle, 1969), and the possibility that the phenol reduces the mutagenicity of IQ by removing the magnesium ions was examined. Surprisingly, absence of magnesium ions did not alter the mutagenie response of IQ, questioning the role of Mg 2+ as cofactor in this system. Moreover, the antimutagenic effect of ellagi¢ acid was independent of the presence of magnesium ions. It has been suggested that a possible mechanism of the antimutagenic and anticarcinogenic action of ellagic acid involves direct interaction with DNA that results in its protection from electrophiles (T¢¢1, 1986). When the bacteria were exposed to various concentrations of ellagic acid, which was subFCT 30/4..--C
sequently removed by washing with fresh nutrient before incubation with IQ, a modest but concentration-dependent decrease in mutagenic response was observed, indicating that such an interaction between the phenol and DNA may make a limited contribution to its antimutagenic effect. Certainly the planarity of ellagic acid would facilitate such interaction. However, such a mechanism cannot be established unequivocally unless the formation of IQ-DNA adducts is quantitated. Similarly, when N-methyl-N-nitrosourea was used as the model mutagen it was concluded that ellagic acid acted by blocking the O6-methylation of guanine (Dixit and Gold, 1986). Moreover, ellagic acid interacted covalently with calf thymus DNA (Teel, 1986) and this author suggested that ellagic acid may act by masking the mutagen-binding sites of DNA. In rive, additional mechanisms may contribute to the reported anticarcinogenic effect of ellagic acid. Ellagic acid may stimulate detoxicating mechanisms, such as glutathione S-transferase activity, as previously reported (Das et al., 1985); furthermore, metabolites of ellagic acid, rather than the parent compound, may interfere with carcinogen activation in rive. As the phenol is very poorly absorbed it was administered ip in order to ensure that significant systemic concentrations were achieved (Smart et al., 1986b). Precipitated ellagic acid was present in the abdominal cavity, as observed in previous studies (Smart et al., 1986b). When DCNB was used as the glutathione-accepting substrate, a significant inhibition of glutathione S-transferase activity was observed, which was accompanied by a decrease in cytosolic protein content. These findings are in contrast to reports where ip or oral administration of ellagic acid enhanced hepatic glutathione S-transferase activity when DCNB and benzo[a]pyrene 4,5oxide were used as substrates (Das et al., 1985). It is possible that the effect of ellagic acid on glutathione S-transferases is isozyme specific and/or animalspecies related. However, the most marked effect of ellagic acid administration observed in this study was a nearly 50% decrease in total cytochrome P-450 content accompanied with a decrease in all mixedfunction oxidase activities studied, although statistical significance was not always achieved. These findings may represent an expression of ellagic acidinduced hepatotoxicity and it must be pointed out
294
A.D. AYRTONet al.
that a loss in body weight occurred during this short treatment with the phenol. Furthermore, the use of the same dose regimen used here, but administered to mice, resulted in toxicity (Lesca, 1983), and it is tempting to conclude that the inhibition of polycyclic aromatic hydrocarbon-induced neoplasia by ellagic acid reflects nothing more than decreased metabolic activation, being the consequence of non-selective destruction of hepatic cytochromes P-450. Lending support to this conclusion is the fact that acute and chronic ellagic acid administration to mice depressed hepatic and pulmonary cytochrome P-450 proteins and associated enzyme activities (Das et aL, 1985). Smart et al. (1986a), using a variety of experimental protocols, failed to demonstrate any decrease in the levels of benzo[a]pyrene-DNA adduct formation in the lung and skin following administration of nontoxic doses of ellagic acid, and the phenol did not afford protection against 3-methylcholanthreneinduced tumorigenesis. It is difficult to envisage how a very poorly absorbed compound such as ellagic acid (Smart et al., 1986b) can reach sufficiently high tissue levels to protect against chemically induced carcinogenesis, particularly when the phenol is administered to animals at low levels in water (3 mg/litre) (Mukhtar et aL, 1986). Virtually no ellagic acid was detected in the blood, lung and liver of mice after feeding 1% of ellagic acid in the diet for l wk (Smart et al., 1986b). Even the less polar derivative 3-O-decylellagic acid, which retains the same antimutagenic activity as the parent compound towards the ultimate carcinogen of benzo[a]pyrene (i.e. the 7,8-diol-9,10-epoxide) did not influence the binding of benzo[a]pyrene to pulmonary or epidermal DNA when administered to mice under conditions that produce high tissue concentrations (Smart et al., 1986b). It therefore seems unlikely that dietary ellagic acid would be an effective anticarcinogen in vivo. Acknowledgement--The authors acknowledge with gratitude financial support of part of this work by the Ministry of Agriculture, Fisheries and Food (Grant No. 530). REFERENCES
Abu-Shakra A:, Ioannides C. and Walker R. (1986) Metabolic activation of 2-amino-3-methylimidazo (4,5-f) quinoline by hepatic preparations--contribution of the cytosolic fraction and its significanceto strain differences. Mutagenesis I, 367-370. Burke M. D. and Mayer R. T. (1974) Ethoxyresorufin: direct fluorimetric assay of a microsomal O-dealkylation which is preferentially inducible by 3-methylcholanthrene. Drug Metabolism and Disposition 2, 583-588. Chang R. L., Huang M.-T., Wood A. W., Wong C.-Q., Newmark H. L., Yagi H., Sayer J. M., Jerina D. M. and Conney A. H. (1985) Effect of ellagic acid and hydroxylated flavonoids on the tumorigenicity of benzo(a)pyrene and ( + )-7fl,g~t-dihydroxy-9~t,10~-epoxy-7,8,9,10-tetrahydrobenzo(a)pyrene on mouse skin and in the newborn mouse. Carcinogenesis 6, 1127-1133. Das M., Bickers D. R. and Mukhtar H. (1985) Effect of ellagic acid on hepatic and pulmonary xenobiotic metab-
olism in mice: studies on the mechanism of its antiearcinogenic action. Carcinogenesis 6, 1409-1413. Dixit R. and Gold B. (1986) Inhibition of N-methyl-Nnitrosourea-induced mutagenicity and DNA methylation by ellagic acid. Proceedings of the National Academy of Sciences of the U.S.A. 83, 8039-8043. Graham S., Dayal H., Swanson M., Mittleman A. and Wilkinson G. (1978) Diet in the epidemiology of cancer of the colon and rectum. Journal of the National Cancer Institute 61, 709-714. Guengerich F. P., Dannan G. W., Wright S. T., Martin M. V. and Kaminsky L. S. (1982) Purification and characterization of cytochrome P-450s. Xenobiotica 12, 701-716. Habig W. H., Pabst M. J. and Jakoby W. B. (1974) Glutathione-S-transferases. The first enzymatic step in mercapturic acid formation. Journal of Biological Chemistry 29, 7130-7139. Ioannides C. and Parke D. V. (1975) Mechanism of induction of hepatic microsomal drug metabolising enzymes by a series of barbiturates. Journal of Pharmacy and Pharma. cology 27, 739-746. Kato R. (1986) Metabolic activation of mutagenic heterocyclic amines from protein pyrolysates. CRC Critical Reviews in Toxicology 16, 307-348. Kato R., Kamataki T. and Yamazoe Y. (1983) N-Hydroxyiation of carcinogenic and mutagenic aromatic amines. Environmental Health Perspectives 49, 21-25. Lesca P. (1983) Protective effects of ellagic acid and other plant phenols on benzo(a)pyrene-induced neoplasia in mice. Carcinogenesis 4, 1651-1653. Lewis D. V., Ioannides C. and Parke D. V. (1986) Molecular dimensions of the substrate binding site of cytochrome P-448. Biochemical Pharmacology 35, 2179-2185. Lowry O. H., Rosebrough N. J., Farr A. L. and Randall A. J. (1951) Protein measurements with the Folin phenol reagent. Journal of Biological Chemistry 193, 265-275. Lubet R. A., Mayer R. T., Cameron J. W., Nims R. W., Burke M. D., Wolf T. and Guengerich F. P. (1985) Dealkylation of pcntoxyresorufin: a rapid and sensitive assay for measuring induction of cytochrome(s) P-450 by phenobarbital and other xenobiotics in the rat. Archives of Biochemistry and Biophysics 238, 43-48. Mandal S., Ahuja A., Shivapurkar N. M., Cheng S.-J., Groopman, J. D. and Stoner G. D. (1987) Inhibition of aflatoxin BI mutagenesis in Salmonella typhimurium and DNA damage in cultured rat and human tracheobronchial tissues by ellagic acid. Carcinogenesis $, 1651-1656. Maron D. M. and Ames B. N. (1983) Revised methods for the Salmonella mutagenicity test. Mutation Research 113, 173-215. Mukhtar H., Das M. and Bickers D. R. (1986) Inhibitionof 3-methylcholanthrene-induced skin tumorigenicity in BALB/c mice by chronic oral feeding of trace amounts of ellagic acid in drinking water. Cancer Research 46, 2262-2265. Mukhtar H., Das M., Del Tito B. J. and Bickers D. R. (1984) Protection against 3-methylcholanthrene-induced skin tumorigenesis in BALB/c mice by ellagic acid. Biochemical and Biophysical Research Communications 119, 751-757. Ohgaki H., Takayama S. and Sugimura T. (1991) Carcinogenicitiesof heterocyclicamines in cooked food. Mutation Research, 259, 399-410. Omura T. and Sato, R. (1964) The carbon monoxide pigment of liver microsomes. I. Evidence for its haemoprotein nature. Journal of Biological Chemistry 239, 2370-2378. Overvik E. and Gustafsson J.-A. (1990) Cooked-food mutagens: current knowledge of formation and biological significance. Mutagenesis 5, 437-446.
Ellagic acid inhibits the mutagenieity of IQ Phillipson C. E., Godden P. M. M., Lum P. Y., loannides C. and Parke D. V. (1984) Determination of cytochrome P-448 activity in biological tissues. Biochemical Journal 221, 81-88. Press R. E. and Hardcastle D. (1969) Some physieochemical properties of ellagic acid. Journal of Applied Chemistry 19, 247-251. Sayer J. M., Yagi H., Wood A. W., Conney A. H. and Jerina D. M. (1982) Extremely facile reaction between the ultimate carcinogen benzo(a)pyrene-7,8-diol-9,10epoxide and ellagic acid. Journal of the American Chemical Society 104, 5562-5564. Shugart L. and Kao J. (1984) Effect of ellagic and caffeic acids on covalent binding of benzo(a)pyrene to epidermal DNA of mouse skin in organ culture. International Journal of Biochemistry 16, 571-573. Smart R. C., Huang M.-T., Chang R. L., Sayer J. M., Jerina D. M., Wood A. W. and Conney A. H. (1986a) Effect of ellagic acid and 3~)-decylellagic acid on the formation of benzo(a)pyrene-derived DNA adducts in vivo and on the tumorigenicity of 3-methylcholanthrene in mice. Carcinogenesis 7, 1669-1675. Smart R. C., Huang M.-T., Chang R. L., Sayer J. M., Jerina D. M. and Conney A. H. (1986b) Disposition of the naturally occurring antimutagenic plant phenol, ellagic acid, and its synthetic derivatives, 3~)-decylellagic acid and 3,Y-di-O-methylellagic acid in mice. Carcinogenesis7, 1663-1667.
295
Teel R. W. (1986) Ellagic acid binding to DNA as a possible mechanism for its antimutagenic and anticarcinogenic action. Cancer Letters 30, 329-336. Ted R. W., Babcock M. S., Dixit R. and Stoner G. D. (1986) Ellagic acid toxicity and interaction with benzo(a)pyrene 7,8-dihydrodiol in human bronchial epithelial cells. Cell Biology and Toxicology 2, 53-62. Ullrich V. and Weber P. (1972) The O-dealkylation of 7-ethoxycoumarin by liver microsomes. Hoppe-Seyler's Z. Physiologic Chemie 353, 1171-1177. Williams C. H., Jr. and Kamin H. (1962) Microsomal triphosphopyridine nucleotide cytochrome c reductase of liver. Journal of Biological Chemistry 237, 587-595. Wood A. W., Huang M.-T., Chang R. L., Newmark H. L., Lehr R. E., Yagi H., Sayer J. M., Jerina D. M. and Conney A. H. (1982) Inhibition of the mutagenicity of bay-region dioi epoxides of polycyclic aromatic hydrocarbons by naturally occurring plant phenols: exceptional activity of ellagic acid. Proceedings of the National Academy of Sciences of the U.S.A. 79, 5513-5517. Yamazoe Y., Shimada M., Maeda K., Kamataki T. and Kato R. (1984) Specificity of four forms of cytochrome P-450 in the metabolic activation of several aromatic amines and benzo(a)pyrene. Xenobiotica 14, 549-552.
0278-6915/92 $5.00 + 0.00 Copyright © 1992Pergamon Press Ltd
A N T I M U T A G E N I C I T Y OF ELLAGIC ACID TOWARDS THE FOOD M U T A G E N IQ: INVESTIGATION INTO POSSIBLE MECHANISMS OF ACTION A. D. AYRTON*, D. F. V. LEWIS, R. WALKERand C. IOANNIDES'[" Molecular Toxicology Research Group, Division of Toxicology, School of Biological Sciences, University of Surrey, Guildford, Surrey GU2 5XH, UK (Accepted 18 December 1991)
Abstract--The ability of the plant phenol ellagic acid to inhibit the mutagenicity of the food mutagen IQ was evaluated using Salmonella typhimurium strain TA98 in the Ames mutagenicity test. Ellagic acid caused a concentration-dependent decrease in the S-9- and microsome-mediated mutagenicity of IQ. The plant phenol did not interact directly with the IQ-derived mutagenic species and did not modify the cytosol-mediated activation of the promutagen. At the concentrations used in the mutagenicity studies, ellagic acid failed to inhibit microsomal mixed-function oxidase activity, including that mediated by the P450I family responsible for the bioactivation of IQ, despite being an essentially planar molecule as indicated by computer-graphic analysis. The inhibitory effect of ellagic acid was independent of its ability to chelate Mg2+ . However, pre-incubation of ellagic acid with the bacteria, followed by removal of the plant phenol, did not completely prevent the inhibitory effect of the phenol on the mutagenicity of IQ. Intraperitoneal administration of ellagic acid to rats caused a decrease in total cytochrome P-450 levels and related activities as well as in cytosolic glutathione S-transferase activity. Finally, the possibility that the reported anticarcinogenic action of ellagic acid reflects nothing more than non-selective destruction of hepatic cytochromes P-450, and thus reduced chemical activation, is considered.
INTRODUCTION Naturally occurring plant polyphenols, such as ellagic acid (2,3,7,8-tetrahydroxy [1] benzopyrano[5,4,3-cde] [1] benzopyran-5,10-dione), have been claimed to possess antimutagenic and anticarcinogenic activities towards major groups of environmental carcinogens, including polycyclic aromatic hydrocarbons, mycotoxins and nitrosamines (Dixit and Gold, 1986; Mandal et al., 1987; Mukhtar et al., 1986). These studies were provoked by initial reports that ellagic acid was a potent inhibitor of the mutagenicity of the bay-region diol epoxides, the ultimate carcinogenic intermediates of polycyclic aromatic hydrocarbons, in the Ames test (Sayer et aL, 1982; Wood et al., 1982) and inhibited the binding of 3-methylcholanthrene to D N A (Mukhtar et al., 1984). In human bronchial epithelial cells, however, ellagic acid increased the toxicity of the 7,8-diol of benzo[a]pyrene, the precursor of the ultimate carcinogen (Teel et al., 1986). Other workers failed to detect any effect of ellagic acid on the metabolism of benzo[a]pyrene by mouse skin in organ culture (Shugart and Kao, 1984). Finally, studies conducted *Present address: Rh6ne-Poulenc Agrochimie, Division Protection des Cultures, 14-20 rue Pierre Baizet, 69263 Lyon, Cedex 09, France. tTo whom all correspondence should be addressed. Abbreviations: DCNB ffi 1,2-dichloronitrobenzene; Glu-P-I = 2-amino-6 methyldipyrido[l,2-a:Y,2'-d]imidazole; IQ ffi 2-amino-3-methylimidazole[4,5-f]quinoline.
in vivo demonstrated that ellagic acid, topically applied to the skin of mice, afforded protection against the induction of skin tumours by 3-methylcholanthrene and 7,12-dimethylbenzanthracene (Lesca, 1983; Mukhtar et al., 1984). In the same animal species, the plant phenol was reported to inhibit the benzo[a]pyrene-induced pulmonary tumorigenesis (Lesca, 1983). It has been proposed that ellagic acid could exert its antimutagenic and anticarcinogenic actions through a combination of one or more of the following distinct mechanisms: (1) inhibition of the enzymes responsible for bioactivation of the carcinogen, such as the cytochrome P-450-dependent mixed-function oxidases; (2) stimulation of the enzymes involved in the detoxication of reactive intermediates, such as the glutathione S-transferases; and (3) through direct interaction of the phenol with the reactive intermediates to form inactive adducts, as previously described (Das et al., 1985; Wood et al., 1982). A fourth possible mechanism is that ellagic acid may interact with D N A in such a way as to decrease the number of binding sites available for interaction with the ultimate mutagens/carcinogens (Teel, 1986). It is now widely recognized that a major group of genotoxic carcinogens are the imidazo-quinolines and -quinoxalines and carbolines generated during the cooking of protein-rich foods (Overvik and Gustaffson, 1990). These chemicals are potent bacterial mutagens in the presence of appropriate activation systems (Kato et al., 1983) and are carcinogenic in
289
290
A.D. AYRTONet aL
mice and rats, giving rise to tumours in multiple sites including the liver (Ohgaki et aL, 1991). Since these compounds are likely to be ingested simultaneously with plant polyphenols, they provide an appropriate model for investigating the possible mechanisms of action of ellagic acid following dietary exposure, and this is the purpose of the present study. MATERIALSAND METHODS IQ (2-amino-3-methylimidazo [4,5-f]quinoline), GIu-P-I (2-amino-6-methyldipyrido [1,2-a:3,2'-d]imidazole) (Wako Fine Chemicals, 4040 Neuss 1, Germany), ethoxyresorufin, pentoxyresorufin, resorufin (Molecular Probes, Eugene, OR, USA), Aroclor 1254 (Robens Institute of Health and Safety, Guildford, Surrey, UK), 1,2-dichloronitrobenzene (Alldrich Chemicals Co. Ltd, Gillingham, Dorset, UK), abscissic acid (Fluka AG, Buchs, Switzerland), ellagic acid, menadione, chlorogenic acid, ethoxycoumarin, cytochrome c, bovine serum albumin and all cofactors (Sigma Co., Poole, Dorset, UK) were purchased. The Salmonella typhimurium strain TA98 was a kind gift from Professor B. N. Ames, University of California, Berkeley, CA, USA. Male Wistar albino rats (150-200 g) were purchased from the Experimental Biology Unit, University of Surrey. Induction of the hepatic mixedfunction oxidase system was achieved by a single ip injection of Aroclor 1254 (500 mg/kg), dissolved in corn oil (200 mg/ml), the animals being killed on the fifth day following administration. The effect of ellagic acid on the hepatic mixed-function oxidase system was investigated following single daily ip administrations of the plant phenol (100 mg/kg) for 5 days, whereas controls received the corresponding volume of vehicle, 0.9% (w/v) NaCI in 0.1 mM-phosphate buffer pH 7.4, all animals being killed 24 hr after the last dose. Hepatic post-mitochondrial supernatant (9000g supernatant, S-9 fraction), microsomal fraction (105,000g pellet resuspended in homogenizing medium) and cytosolic fraction (105,000g supernatant) were prepared as previously described (loannides and Parke, 1975). The following assays were carried out on the microsomal fraction: the O-dealkylations of ethoxyresorufin (Burke and Mayer, 1974), pentoxyresorufin (Lubet et al., 1985) and ethoxycoumarin (Ullrich and Weber, 1972), the NADPH-dependent reduction of cytochrome c Table 1. Plant phenol
(Williams and Kamin, 1962) and total c y t o c h r o m e P-450 levels (Omura and Sato, 1964); on the cytosolic fraction glutathione S-transferase activity was assayed using 1,2-dichloronitrobenzene (DCNB) as the acceptor substrate (Habig et al., 1974). Protein was determined on both hepatic fractions (Lowry et al., 1951) using bovine serum albumin as standard. The mutagenicities of IQ and GIu-P-I were determined in the Ames test (Maron and Ames, 1983), using S. typhimurium strain TA98. In experiments where microsomes only were used for the activation, the activation system was supplemented with glucose6-phosphate dehydrogenase (1 unit/plate). In studies where the role of the cytosol was evaluated, t h e bacteria, IQ and microsomal activation system were pre-incubated for 20 min, at 37°C in a shaking water-bath, before termination of the microsomal reaction by the addition (100/d/plate) of menadione (900/zM). This was followed by the addition of either buffer or the cytosolic fraction. IQ, GIu-P-1 and the plant antimutagens were always dissolved in dimethylsulphoxide so that the total amount of t h e solvent in the activation system was constant and never exceeded 100/zl. At the concentrations used, none of the phenols was toxic to the bacteria (results not shown). The molecular geometry of ellagic acid was determined using the COSMIC package (written by J. G. Vinter, A. Davies and M. R. Saunders of SKF Ltd). The molecular plot, which is viewed in, and perpendicular to, the molecular plane, was generated using the PLUTO crystallographic package. The following van der Waals radii were utilized: C, 1.6 A; O, 1.4 A and H, 1.2 A. Statistical evaluation was carried out using Student's t-test RESULTS
Of the plant compounds studied, only ellagie acid produced a concentration-dependent decrease in t h e Aroclor 1254-induced mutagenicity of IQ (Table 1) and was therefore investigated further. A more detailed study showed that, when post-mitochondrial fractions from Aroclor 1254-induced rats were used as activation system, the inhibition of the IQ-mediated mutagenic response by ellagic acid was related to the concentration both of the plant phenol and of t h e mutagen (Fig. IA). When isolated microsomes (i.e. in the absence of cytosol) were used as the activation system, the mutagenic response of IQ was considerably lower than that following activation by t h e
Amestest* Histidine revertantsper platet at a plant phenol concentration (M) of:
Effect of various phenols on the mutagcnicity of IQ in the 0
Chlorogenic acid 6879 + 1467 Ellagic acid 10352 ± 860 Abscissicacid 4537± 69
10 -7
10 -6
10 -5
10 -4
8410 + 1105 9087 ± 292 6578 ± 330
7681 ± 1460 8520 ± 370
7452 ± 440 7656 ± 495
6579 ± 674 1424 ± 201 5649± 490
4990± 1072
5538± 588
*The test was carried out using Salmonella typhimurium strain TA98, IQ (I00 ng/plate) and Aroclor 1254-induced hepatic post-mitochonddal activation systems (I0%, v/v). tResults are presented as mean + SD for triplicates. The spontaneous reversion rate of 25 ± 6 has already been subtracted.
Ellagic acid inhibits the mutagenicity of IQ post-mitochondrial (S-9) fraction (Fig. 1B). The mutagenic activity of IQ was once again inhibited by ellagic acid in a concentration-dependent fashion. Finally, neither ellagic acid nor the other plant cornpounds, at the concentrations used, displayed any mutagenic effect (results not shown). When the microsomal activation of IQ was terminated by the addition of menadione, addition of the cytosol, as expected, enhanced the mutagenic potency of the carcinogen. When incorporated into the cytosol, ellagic acid did not influence the mutagenicity of IQ (Fig. 2). When ellagic acid was incorporated into the buffer, instead of the cytosol, the mutagenicity of IQ was not significantly altered, except for a decrease at the highest concentration of the phenol, 5 x 10 -s M (Fig. 2). At the concentrations used in the mutagenicity studies, ellagic acid had no effect on the O-dealkylations of ethoxyresorufin or pentoxyresorufin, or the NADPH-dependent reduction of cytochrome c. B
10,000
(A) 8000
c
4000
2000
o
', 0
20
40
60
80
I
I
1 O0
120
IQ concentration (ng/plate) 1200
_=
-
(B)
ca- ~ 1000
--
4=# C
800
-
>
600
C
400
.=
I
:6 .m
-I-
--
500
--
400
--
300
--
200
--
100
-
~...........---t
x
O. .~
~, ~>
',~
= -l-
0
10-7
,
, ,=~,ld 10-6
= , =l,,,I
i
, ~,,=,,I
10-5
10-4
Ellagic acid c o n c e n t r a t i o n (M)
Fig. 2. Effect of ellagic acid on the cytosolic potentiation of the microsome-mediated mutagenicity of IQ. IQ (10 ng) was preincubated with Salmonella typhimurium strain TA98 and an Aroclor 1254-induced hepatic microsomal activation system (10%] for 20 rain at 37°C in a shaking water-bath. The reaction was terminated by addition of 100/tl menadione (900#M). Cytosol (&) or buffer ( 1 ) containing various concentrations of ellagic acid were then added and a further 20-min pre-incubation was carried out at 37°C. Results are presented as mean + SD for triplicates. The spontaneous reversion rate of 25 + 7 has already been subtracted.
6000
c
t~
600
291
200
0 o
20
40
60
80
100
120
IQ concentration (ng/plate)
Fig l. Inhibition of the mutagenicity of IQ by ellagic acid. IQ, Salmonella typhimurium strain TA98 and hepatic S-9 (A) or microsomal (B) activation system (10%) derived from Aroclor 1254-induced rats were incubated in the absence (&) and presence of ellagic acid at concentrations of 5 X 10-TM ( I ) , 10-6 M (rq), 5 x 10-6M (O), I0-SM (A) and 5 x 10-~M (O). Each point represents the mean of triplicates. The spontaneous reversion rate of 22 + 5 has already been subtracted.
When Glu-P- 1 was used as the premutagen (20-80 ng! plate) in the Ames test and microsomes from Aroclor 1254-rats as the activation system, ellagic acid (5 x 10-7-5 x 10 -5 M) had no significant influence on the mutagenic response (results not shown). Presence of magnesium ions did not modify the mutagenic effect of IQ; similarly, the inhibitory effect of ellagic acid was independent of the presence of magnesium ions (Table 2). Pre-incubation of the bacteria with ellagic acid, followed by the removal of the plant phenol by successive washings with nutrient broth, did not completely prevent the inhibitory effect of the phenol on the mutagenicity of IQ (Table 3). When administered to rats by the ip route, precipitated ellagic acid was present in the abdominal cavity and body-weight loss was observed in the animals treated with the phenol. The same treatment gave rise to a decrease in the microsomal levels of total cytochrome P-450, N A D P H cytochrome c reductase activity and microsomal and cytosolic protein levels, but not all effects, were statistically significant (Table 4). Similarly, the O-deethylation ofethoxycoumarin and cytosolic glutathione S-transferase activity were significantly inhibited; the O-dealkylations of ethoxyresorufin and pentoxyresorufin were also reduced by the ellagic acid treatment but the effect was not statistically significant (Table 5). Microsomes from ellagic acid-treated rats did not differ significantly from preparations from control animals in their ability to convert GIu-P-I to mutagens in the Ames test (Table 5). The computer-optimized dimensions of ellagic acid were: length, 12.1 A.; width 10.2A and depth 3.2A,
A . D . AYRTON et al.
292
Table 2. Effect of Mg 2+ ions and ellagic acid on the mutagenicity of IQ* Histidine revertants per platet Supplementation of activation system
+ MgCI 2
- MgCI2
-Ellagic acid
2711 + 474 779 _+ 140
2638 ± 284 736 __ 160
*The study was carried out using Salmonella typhimurium strain TA98. The bacteria, IQ (20 ng) and microsomes (10%, v/v) from Aroclor 1254-treated rats, in the presence and absence of MgCI 2 (8 raM) and/or ellagic acid (5 x l0 -5 M) were pre-incubated for 20 minutes at 37°C in a shaking water-bath. ?Results are presented as mean + SD for three plates. The spontaneous reversion rate of 25 _+ 5 has already been subtracted.
giving rise to an area/depth 2 of 12.0, showing that it is essentially a planar molecule (Fig. 3). DISCUSSION
Consumption of diets with a high vegetable content has been repeatedly correlated with reduced human cancer incidence at a number of sites (Graham et al., 1978). This observation has led to the evaluation of various plant compounds as possible anticarcinogens that may contribute to reduced cancer risk. The most extensively studied is ellagic acid, which was shown to inhibit the mutagenicity of a number of promutagens and ultimate mutagens in the Ames test (Dixit and Gold, 1986; Mandal et al., 1987; Wood et al., 1982) and to diminish the carcinogenicity of chemical carcinogens such as polycyclic aromatic hydrocarbons (Mukhtar et al., 1984 and 1986; Lesca, 1983); however, other workers failed to demonstrate an anticarcinogenic effect of ellagic acid towards benzo[a]pyrene and 3-methylcholanthrene (Chang et al., 1985; Smart et al., 1986a). In the present study we investigated the possible antimutagenic effect of ellagic acid towards IQ because both compounds can be present in food. Ellagic acid decreased the S-9-mediated mutagenicity of IQ, which was virtually abolished at a phenol concentration of 10 -s M. One or more of the following mechanisms may be responsible for the antimutagenic effect of ellagic acid: (1) ellagic acid or one of its metabolites may inhibit the microsomal Table 3. Mutagenicity of IQ to bacteria treated with ellaglc acid* Ellagic acid concentration (M)
Histidine revertants per plate
0 5 x 10 -7 5 x 10 -6 5 x 1 0 -s 5 x 10 -4
792 + 60 781+38 734+47 728+50 659 ± 78
*Salmonella typhimurium strain TA98 and ellaglc acid were pre-incubated for I hr at 37°C in a shaking water-bath. The bacteria were then washed three-times with nutrient broth and were then mixed with IQ (10ng) and Aroclor 1254-induced microseines (10%, v/v). Results are presented as mean + SD of tripficates. The spontaneous reversion rate of 29 + 9 has already been subtracted.
Table 4. Effect of treatment with ellagicacid on some rat hepatic parameterst Parameter Body-weight gain (g) Liver/body wt x 100 NADPH-cytochrome c reductase (nmol/min/mg protein) Total cytochrome P-450 (nmol/mg protein) Cytosolic protein (mg/g liver) Microsomal protein (mg/g liver)
Control
Ellagic acid
6.6 ± 5.5 4.7 -t- 0.1 18.5 ± 1.6
--8.0 + 3.0* 5.3 ± 0.2* 11.9 ±0.3
0.47 ± 0.01
0.25 ± 0.01"
88.9 +_4.6 17.8 -k 1.8
66.5 ± 2.2* 13.8 ± 0.4
*P < 0.05. tRats were treated with single daily ip injections of ellaglc acid (100 mg/kg) for 5 days whereas controls received the corresponding dose of the vehicle, phosphate buffered saline; all animals were killed 24 hr after last administration. Results are presented as mean ± SEM for five animals.
N-hydroxylation of IQ and/or its further metabolism to the ultimate mutagen by cytosolic enzymes; (2) the phenol may interact directly and non-enzymically with the proximate and/or ultimate mutagens, rendering them inactive; (3) ellagic acid, by virtue of its affinity for divalent ions, may complex with magnesium ions, thus depriving the activation system of an important component; and (4) the plant phenol may interact with DNA in such a way as to protect it from the ultimate mutagen. All these mechanisms were investigated. The reduction in the mutagenicity of IQ by ellagic acid was maintained even when isolated microsomes, rather than the post-mitochondrial supernatant (S-9), were used as the activation system, indicating that the phenol impairs the microsomal activation of IQ. The activation of this carcinogen is selectively catalysed by the P-450I family, and especially the A2 isoenzyme (Kate, 1986) and it is conceivable that ellagic acid is a potent inhibitor of this isoenzyme. Initially, a computer-graphic analysis was carried out to establish whether ellagic acid has the necessary molecular shape to interact with this cytochrome P-450 isoenzyme (Lewis et al., 1986). With an area/depth 2 ratio of 12 it is apparent that ellagic acid is an essentially planar molecule with a high potential for interaction with the P-450I family. However, in contrast to the predicted interaction, ellagic acid at Table 5. Rat hepatic mixed-function oxidase and glutathione Stransferase activities following treatment with ellagic acid Parameter Ethoxycoumarin O-deethylase (pmol/min/mg protein) (pmol/min/nmol P-450) Ethoxyresorufin O -deethylase (pmol/min/mg protein) (pmol/min/nmol P-450) Pentoxyresorufin O-depentylase (pmol/min/mg protein) (pmol/min/nmol P-450) Glutathione S-transferase (nmol/min/mg protein) Microsomal activationof Glu-P-I (2/*g) (histidine revertants/nmol P-450)
Control
Ellagic acid
41.8 + 7.4 89.0 + 12.8
20.3 -J- 3.4* 88.9 + 23.9
92 -6 1 200 + 10
68 -6 10 270 + 40
15 4- 2 33 + 6
II + I 46 -t- 4
78 ± 5
54 + 4*
1580 -I-210
2290 + 330
*P < 0.05. ?Treatment of animals and presentation of data are as in the footnotes to Table 4.
Ellagic acid inhibits the mutagenicity of IQ
293
Fig 3. Space-filled model of ellagic acid. Model was drawn using the PLUTO computer program and utilizing the following van der Waals radii to generate the computer-graphicplot of the molecular geometry: carbon, 1.6 A; oxygen, 1.4 A and hydrogen, 1.2 A. the concentrations studied failed to inhibit the Odeethylation of ethoxyresorufin and the bioactivation of GIu-P-I to mutagens, two reactions catalysed exclusively by the P-450I proteins, A1 and A2 respectively (Guengerich et al., 1982; Phillipson et al., 1984; Yamazoe et al., 1984). This lack of inhibition may be attributed to the relatively poor lipophilicity of the polyhydroxy phenol (also compatible with its poor absorption when orally administered; Smart et al., 1986b), which cannot reach the cytochrome P-450 system known to be embedded in lipid. Addition of the cytosolic fraction to the activation system, following termination of the microsomal metabolism, enhanced the mutagenicity of IQ, as we have previously described (Abu-Shakra et al., 1986). Incorporation of ellagic acid to the cytosolic fraction did not influence the mutagenic response, showing that the antimutagenic effect of the phenol does not result from the inhibition of the cytosolic activation pathway of the mutagen. As ellagic acid was reported to inhibit the mutagenicity of the ultimate carcinogen, benzo[a]pyrene 7,8-diol-9,10-epoxide, by an acid-catalysed direct interaction (Sayer et al., 1982; Wood et al., 1982), we investigated the possibility that the phenol could scavenge the microsome-generated reactive intermediates of IQ presumed to be the N-hydroxylamine. When microsomal metabolism was terminated, addition of ellagi¢ acid into the system at pH 7.4 gave rise to a decrease in the mutagenicity of IQ only at the highest dose of the phenol, thus making this mechanism of action unlikely. Ellagic acid readily complexes with divalent cations (Press and Hardcastle, 1969), and the possibility that the phenol reduces the mutagenicity of IQ by removing the magnesium ions was examined. Surprisingly, absence of magnesium ions did not alter the mutagenie response of IQ, questioning the role of Mg 2+ as cofactor in this system. Moreover, the antimutagenic effect of ellagi¢ acid was independent of the presence of magnesium ions. It has been suggested that a possible mechanism of the antimutagenic and anticarcinogenic action of ellagic acid involves direct interaction with DNA that results in its protection from electrophiles (T¢¢1, 1986). When the bacteria were exposed to various concentrations of ellagic acid, which was subFCT 30/4..--C
sequently removed by washing with fresh nutrient before incubation with IQ, a modest but concentration-dependent decrease in mutagenic response was observed, indicating that such an interaction between the phenol and DNA may make a limited contribution to its antimutagenic effect. Certainly the planarity of ellagic acid would facilitate such interaction. However, such a mechanism cannot be established unequivocally unless the formation of IQ-DNA adducts is quantitated. Similarly, when N-methyl-N-nitrosourea was used as the model mutagen it was concluded that ellagic acid acted by blocking the O6-methylation of guanine (Dixit and Gold, 1986). Moreover, ellagic acid interacted covalently with calf thymus DNA (Teel, 1986) and this author suggested that ellagic acid may act by masking the mutagen-binding sites of DNA. In rive, additional mechanisms may contribute to the reported anticarcinogenic effect of ellagic acid. Ellagic acid may stimulate detoxicating mechanisms, such as glutathione S-transferase activity, as previously reported (Das et al., 1985); furthermore, metabolites of ellagic acid, rather than the parent compound, may interfere with carcinogen activation in rive. As the phenol is very poorly absorbed it was administered ip in order to ensure that significant systemic concentrations were achieved (Smart et al., 1986b). Precipitated ellagic acid was present in the abdominal cavity, as observed in previous studies (Smart et al., 1986b). When DCNB was used as the glutathione-accepting substrate, a significant inhibition of glutathione S-transferase activity was observed, which was accompanied by a decrease in cytosolic protein content. These findings are in contrast to reports where ip or oral administration of ellagic acid enhanced hepatic glutathione S-transferase activity when DCNB and benzo[a]pyrene 4,5oxide were used as substrates (Das et al., 1985). It is possible that the effect of ellagic acid on glutathione S-transferases is isozyme specific and/or animalspecies related. However, the most marked effect of ellagic acid administration observed in this study was a nearly 50% decrease in total cytochrome P-450 content accompanied with a decrease in all mixedfunction oxidase activities studied, although statistical significance was not always achieved. These findings may represent an expression of ellagic acidinduced hepatotoxicity and it must be pointed out
294
A.D. AYRTONet al.
that a loss in body weight occurred during this short treatment with the phenol. Furthermore, the use of the same dose regimen used here, but administered to mice, resulted in toxicity (Lesca, 1983), and it is tempting to conclude that the inhibition of polycyclic aromatic hydrocarbon-induced neoplasia by ellagic acid reflects nothing more than decreased metabolic activation, being the consequence of non-selective destruction of hepatic cytochromes P-450. Lending support to this conclusion is the fact that acute and chronic ellagic acid administration to mice depressed hepatic and pulmonary cytochrome P-450 proteins and associated enzyme activities (Das et aL, 1985). Smart et al. (1986a), using a variety of experimental protocols, failed to demonstrate any decrease in the levels of benzo[a]pyrene-DNA adduct formation in the lung and skin following administration of nontoxic doses of ellagic acid, and the phenol did not afford protection against 3-methylcholanthreneinduced tumorigenesis. It is difficult to envisage how a very poorly absorbed compound such as ellagic acid (Smart et al., 1986b) can reach sufficiently high tissue levels to protect against chemically induced carcinogenesis, particularly when the phenol is administered to animals at low levels in water (3 mg/litre) (Mukhtar et aL, 1986). Virtually no ellagic acid was detected in the blood, lung and liver of mice after feeding 1% of ellagic acid in the diet for l wk (Smart et al., 1986b). Even the less polar derivative 3-O-decylellagic acid, which retains the same antimutagenic activity as the parent compound towards the ultimate carcinogen of benzo[a]pyrene (i.e. the 7,8-diol-9,10-epoxide) did not influence the binding of benzo[a]pyrene to pulmonary or epidermal DNA when administered to mice under conditions that produce high tissue concentrations (Smart et al., 1986b). It therefore seems unlikely that dietary ellagic acid would be an effective anticarcinogen in vivo. Acknowledgement--The authors acknowledge with gratitude financial support of part of this work by the Ministry of Agriculture, Fisheries and Food (Grant No. 530). REFERENCES
Abu-Shakra A:, Ioannides C. and Walker R. (1986) Metabolic activation of 2-amino-3-methylimidazo (4,5-f) quinoline by hepatic preparations--contribution of the cytosolic fraction and its significanceto strain differences. Mutagenesis I, 367-370. Burke M. D. and Mayer R. T. (1974) Ethoxyresorufin: direct fluorimetric assay of a microsomal O-dealkylation which is preferentially inducible by 3-methylcholanthrene. Drug Metabolism and Disposition 2, 583-588. Chang R. L., Huang M.-T., Wood A. W., Wong C.-Q., Newmark H. L., Yagi H., Sayer J. M., Jerina D. M. and Conney A. H. (1985) Effect of ellagic acid and hydroxylated flavonoids on the tumorigenicity of benzo(a)pyrene and ( + )-7fl,g~t-dihydroxy-9~t,10~-epoxy-7,8,9,10-tetrahydrobenzo(a)pyrene on mouse skin and in the newborn mouse. Carcinogenesis 6, 1127-1133. Das M., Bickers D. R. and Mukhtar H. (1985) Effect of ellagic acid on hepatic and pulmonary xenobiotic metab-
olism in mice: studies on the mechanism of its antiearcinogenic action. Carcinogenesis 6, 1409-1413. Dixit R. and Gold B. (1986) Inhibition of N-methyl-Nnitrosourea-induced mutagenicity and DNA methylation by ellagic acid. Proceedings of the National Academy of Sciences of the U.S.A. 83, 8039-8043. Graham S., Dayal H., Swanson M., Mittleman A. and Wilkinson G. (1978) Diet in the epidemiology of cancer of the colon and rectum. Journal of the National Cancer Institute 61, 709-714. Guengerich F. P., Dannan G. W., Wright S. T., Martin M. V. and Kaminsky L. S. (1982) Purification and characterization of cytochrome P-450s. Xenobiotica 12, 701-716. Habig W. H., Pabst M. J. and Jakoby W. B. (1974) Glutathione-S-transferases. The first enzymatic step in mercapturic acid formation. Journal of Biological Chemistry 29, 7130-7139. Ioannides C. and Parke D. V. (1975) Mechanism of induction of hepatic microsomal drug metabolising enzymes by a series of barbiturates. Journal of Pharmacy and Pharma. cology 27, 739-746. Kato R. (1986) Metabolic activation of mutagenic heterocyclic amines from protein pyrolysates. CRC Critical Reviews in Toxicology 16, 307-348. Kato R., Kamataki T. and Yamazoe Y. (1983) N-Hydroxyiation of carcinogenic and mutagenic aromatic amines. Environmental Health Perspectives 49, 21-25. Lesca P. (1983) Protective effects of ellagic acid and other plant phenols on benzo(a)pyrene-induced neoplasia in mice. Carcinogenesis 4, 1651-1653. Lewis D. V., Ioannides C. and Parke D. V. (1986) Molecular dimensions of the substrate binding site of cytochrome P-448. Biochemical Pharmacology 35, 2179-2185. Lowry O. H., Rosebrough N. J., Farr A. L. and Randall A. J. (1951) Protein measurements with the Folin phenol reagent. Journal of Biological Chemistry 193, 265-275. Lubet R. A., Mayer R. T., Cameron J. W., Nims R. W., Burke M. D., Wolf T. and Guengerich F. P. (1985) Dealkylation of pcntoxyresorufin: a rapid and sensitive assay for measuring induction of cytochrome(s) P-450 by phenobarbital and other xenobiotics in the rat. Archives of Biochemistry and Biophysics 238, 43-48. Mandal S., Ahuja A., Shivapurkar N. M., Cheng S.-J., Groopman, J. D. and Stoner G. D. (1987) Inhibition of aflatoxin BI mutagenesis in Salmonella typhimurium and DNA damage in cultured rat and human tracheobronchial tissues by ellagic acid. Carcinogenesis $, 1651-1656. Maron D. M. and Ames B. N. (1983) Revised methods for the Salmonella mutagenicity test. Mutation Research 113, 173-215. Mukhtar H., Das M. and Bickers D. R. (1986) Inhibitionof 3-methylcholanthrene-induced skin tumorigenicity in BALB/c mice by chronic oral feeding of trace amounts of ellagic acid in drinking water. Cancer Research 46, 2262-2265. Mukhtar H., Das M., Del Tito B. J. and Bickers D. R. (1984) Protection against 3-methylcholanthrene-induced skin tumorigenesis in BALB/c mice by ellagic acid. Biochemical and Biophysical Research Communications 119, 751-757. Ohgaki H., Takayama S. and Sugimura T. (1991) Carcinogenicitiesof heterocyclicamines in cooked food. Mutation Research, 259, 399-410. Omura T. and Sato, R. (1964) The carbon monoxide pigment of liver microsomes. I. Evidence for its haemoprotein nature. Journal of Biological Chemistry 239, 2370-2378. Overvik E. and Gustafsson J.-A. (1990) Cooked-food mutagens: current knowledge of formation and biological significance. Mutagenesis 5, 437-446.
Ellagic acid inhibits the mutagenieity of IQ Phillipson C. E., Godden P. M. M., Lum P. Y., loannides C. and Parke D. V. (1984) Determination of cytochrome P-448 activity in biological tissues. Biochemical Journal 221, 81-88. Press R. E. and Hardcastle D. (1969) Some physieochemical properties of ellagic acid. Journal of Applied Chemistry 19, 247-251. Sayer J. M., Yagi H., Wood A. W., Conney A. H. and Jerina D. M. (1982) Extremely facile reaction between the ultimate carcinogen benzo(a)pyrene-7,8-diol-9,10epoxide and ellagic acid. Journal of the American Chemical Society 104, 5562-5564. Shugart L. and Kao J. (1984) Effect of ellagic and caffeic acids on covalent binding of benzo(a)pyrene to epidermal DNA of mouse skin in organ culture. International Journal of Biochemistry 16, 571-573. Smart R. C., Huang M.-T., Chang R. L., Sayer J. M., Jerina D. M., Wood A. W. and Conney A. H. (1986a) Effect of ellagic acid and 3~)-decylellagic acid on the formation of benzo(a)pyrene-derived DNA adducts in vivo and on the tumorigenicity of 3-methylcholanthrene in mice. Carcinogenesis 7, 1669-1675. Smart R. C., Huang M.-T., Chang R. L., Sayer J. M., Jerina D. M. and Conney A. H. (1986b) Disposition of the naturally occurring antimutagenic plant phenol, ellagic acid, and its synthetic derivatives, 3~)-decylellagic acid and 3,Y-di-O-methylellagic acid in mice. Carcinogenesis7, 1663-1667.
295
Teel R. W. (1986) Ellagic acid binding to DNA as a possible mechanism for its antimutagenic and anticarcinogenic action. Cancer Letters 30, 329-336. Ted R. W., Babcock M. S., Dixit R. and Stoner G. D. (1986) Ellagic acid toxicity and interaction with benzo(a)pyrene 7,8-dihydrodiol in human bronchial epithelial cells. Cell Biology and Toxicology 2, 53-62. Ullrich V. and Weber P. (1972) The O-dealkylation of 7-ethoxycoumarin by liver microsomes. Hoppe-Seyler's Z. Physiologic Chemie 353, 1171-1177. Williams C. H., Jr. and Kamin H. (1962) Microsomal triphosphopyridine nucleotide cytochrome c reductase of liver. Journal of Biological Chemistry 237, 587-595. Wood A. W., Huang M.-T., Chang R. L., Newmark H. L., Lehr R. E., Yagi H., Sayer J. M., Jerina D. M. and Conney A. H. (1982) Inhibition of the mutagenicity of bay-region dioi epoxides of polycyclic aromatic hydrocarbons by naturally occurring plant phenols: exceptional activity of ellagic acid. Proceedings of the National Academy of Sciences of the U.S.A. 79, 5513-5517. Yamazoe Y., Shimada M., Maeda K., Kamataki T. and Kato R. (1984) Specificity of four forms of cytochrome P-450 in the metabolic activation of several aromatic amines and benzo(a)pyrene. Xenobiotica 14, 549-552.
