Proc. Natl. Acad. Sci. USAVol. 79, pp. 5513-5517, September 1982Biochemistry

Inhibition of the mutagenicity of bay-region diol epoxides ofpolycyclic aromatic hydrocarbons by naturally occurringplant phenols: Exceptional activity of ellagic acid

(ultimate carcinogen/Ames' Salmonella typhimurium/Chinese hamster V79 cells/kinetics of epoxide disappearance)

ALEXANDER W. WOOD*, MOU-TUAN HUANG*, RICHARD L. CHANG*, HAROLD L. NEWMARKt,ROLAND E. LEHR§, HARUHIKO YACIW, JANE M. SAYERS, DONALD M. JERINA¶, ANDALLAN H. CONNEY*Departments of *Biochemistry and Drug Metabolism, and tTechnical Development, Hoffmann-La Roche Inc., Nutlev, New Jersey 07110; §Department of Chemistry,Universitv of Oklahoma, Norman, OK 73019; and ¶Laboratory of Bioorganic Chemistry, National Institute of Arthritis, Metabolism, and Digestive and KidneyDiseases, National Institutes of Health, Bethesda, Maryland 20205

Communicated by James A. Miller, June 17, 1982

ABSTRACT Ferulic, caffeic, chlorogenic, and ellagic acids,four naturally occurring plant phenols, inhibit the mutagenicityand cytotoxicity of (+)-7,8,8a-dihydroxy-9a,10a-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (B[a]P 7,8-diol-9,10-epoxide-2), theonly known ultimate carcinogenic metabolite of benzo[a]pyrene.The mutagenicity of 0.05 nmol of B[a]P 7,8-diol-9,10-epoxide-2 instrain TA100 of Salmonella typhimurium is inhibited 50% by in-cubation ofthe bacteria and the diol epoxide with 150 nmol offeru-lic acid, 75 nmol of caffeic acid, 50 nmol of chlorogenic acid or,most strikingly, 1 nmol of ellagic acid in the 0.5-ml incubationmixture. A 3-nmol dose of ellagic acid inhibits mutation inductionby 90%. Ellagic acid is also a potent antagonist of B[a]P 7,8-diol-9,10-epoxide-2 in Chinese hamster V79 cells. Mutations to 8-aza-guanine resistance induced by 0.2 ,AM diol epoxide are reducedby 50% when tissue culture media also contains 2 ,uM ellagic acid.Similar to results obtained with the bacteria, ferulic, caffeic, andchlorogenic acids are approximately two orders of magnitude lessactive than ellagic acid in the mammalian cell assay. The anti-mutagenic effects of the plant phenols result from their direct in-teraction with B[a]P 7,8-diol-9,10-epoxide-2, because a concen-tration-dependent increase in the rate of diol epoxide disappearancein cell-free solutions of 1:9 dioxane/water, pH 7.0, is observedwith all four phenols. In parallel with the mutagenicity studies,ellagic acid is 80-300 times more effective than the other phenolsin accelerating the disappearance of B[a]P 7,8-diol-9,10-epoxide-2. Ellagic acid at 10 ,M increases the disappearance of B[a]P 7,8-diol-9,10-epoxide-2 by approximately 20-fold relative to the spon-taneous and hydronium ion-catalyzed hydrolysis of the diol epox-ide at pH 7.0. Ellagic acid is a highly potent inhibitor of the mu-tagenic activity of bay-region diol epoxides of benzo[a]pyrene,dibenzo[a,h]pyrene, and dibenzo[a,ijpyrene, but higher concen-trations of ellagic acid are needed to inhibit the mutagenic activityof the chemically less reactive bay-region diol epoxides ofbenz[a]anthracene, chrysene, and benzo[c]phenanthrene. Thesestudies demonstrate that ellagic acid is a potent antagonist of theadverse biological effects of the ultimate carcinogenic metabolitesof several polycyclic aromatic hydrocarbons and suggest that thisnaturally occurring plant phenol, normally ingested by humans,may inhibit the carcinogenicity of polycyclic aromatic hydrocarbons.

Polycyclic aromatic hydrocarbons are relatively inert environ-mental precarcinogens that must be metabolized by mammalianenzymes to their biologically active products (ultimate carcin-ogens). Mutagenicity, metabolism, DNA-binding, and tumor-igenicity studies during the past 8 years have shown that benzo-

ring diol epoxides, in which the epoxide moiety forms part ofthe bay region (1, 2) of the hydrocarbons, are ultimate carcin-ogens of polycyclic hydrocarbons (3-6).

Chemical and kinetic studies have demonstrated that bay-region diol epoxides are subject to general acid catalysis to formchemically unreactive and biologically inactive tetraols (7-12)as well as to form covalent adducts with low molecular weightnucleophiles such as p-nitrothiophenol and 2-mercaptoethanol(7, 8, 12). Thus a chemical basis exists for a rational approachto the identification of nontoxic compounds that can block theadverse biological effects of the ultimate carcinogens of poly-cyclic aromatic hydrocarbons. We have recently shown (13) thatriboflavin 5'-phosphate (FMN) can markedly diminish the mu-tagenicity of the trans diastereomerll of the bay-region diolepoxide of benzo[a]pyrene (B[a]P 7,8-diol-9, 10-epoxide-2) bya mechanism involving complex formation followed by generalacid catalyzed hydrolysis of the diol epoxide to tetraols. Theseresults with FMN have demonstrated the feasibility of blockingthe adverse biological effects of ultimate carcinogens and haveserved as an impetus to identify additional compounds witheven better capacity to interact with and detoxify bay-regiondiol epoxides. In the present study, we find that ferulic, caf-feic, chlorogenic, and ellagic acids, four naturally occurringplant phenols that share the common structural elements of ameta- or para-hydroxylated benzoic or cinnamic acid (Fig. 1)can inhibit the mutagenicity of the ultimate carcinogenic me-tabolite of B[a]P. One of these phenols, ellagic acid, is an ex-

Abbreviations: B[a]P, benzo[a]pyrene; B[a]P 7,8-diol-9,10-epoxide-1,(±)-7f3,8a-dihydroxy-9/3,10,3epoxy-7,8,9, 10-tetrahydrobenzo[a]pyrene;B[a]P 7,8-diol-9, 10-epoxide-2, (±)-73,8a-dihydroxy-9a, 10a-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene; B[a]P H4-9,10-epoxide, 9,10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene; DB[a,h]P 1,2-diol-3,4-epoxide-2,(+)- 1,B,2a-dihydroxy-3a,4a-epoxy- 1,2,3,4-tetrahydrodibenzo[a,h]-pyrene; DB[a,i]P 3,4-diol-1,2-epoxide-2, (+)-3a,4,3-dihydroxy-la,2a-epoxy-1,2,3,4-tetrahydrodibenzo[a,i]pyrene; B[a]A 3,4-diol-1,2-epox-ide-2, (+)-3a,4f3-dihydroxy-la,2a-epoxy-1,2,3,4-tetrahydrobenz-[a]anthracene; chrysene 1,2-diol-3,4-epoxide-2, (+)-1f,32a-dihydroxy-3a,4a-epoxy-1,2,3,4-tetrahydrochrysene; B[c]Phen 3,4-diol-1,2-epox-ide-2, (+ )-3a, 4/-dihydroxy-la,2a-epoxy-1,2, 3,4-tetrahydro-benzo[c]phenanthrene; Me2SO, dimethyl sulfoxide. All compoundsare racemic mixtures where optical enantiomers are possible.t Present address: Ludwig Institute for Cancer Research, Toronto,Ontario, Canada M4Y 1M4.

11 As indicated in the abbreviations, benzo-ring diol epoxides of transdihydrodiols exist as diastereomeric pairs in which the benzylic hy-droxyl group and the epoxide oxygen are either cis (isomer-I series)or trans (isomer-2 series) to each other. Fig. 1 depicts the bay-regiondiol epoxide-2 isomer of benzo[a]pyrene.

5513

The publication costs ofthis article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertise-ment" in accordance with 18 U. S. C. § 1734 solely to indicate this fact.

Proc. Natd Acad. Sci. USA 79 (1982)

ceptional antagonist of the mutagenicity of bay-region diolepoxides of several polycyclic aromatic hydrocarbons.

MATERIALS AND METHODSMaterials. Ferulic, caffeic, chlorogenic, and ellagic acids

were obtained from Aldrich (Milwaukee, WI), and all four com-pounds were essentially pure as determined by reverse-phasehigh-pressure liquid chromatography. For the kinetic studies,caffeic, ferulic, and ellagic acids were recrystallized fromethanol/water, methanol/water, and isopropyl alcohol/water,respectively. The bay-region diol epoxides of B[a]P (7, 8), di-benzo[a,h]pyrene (DB[a,h]P) (14), dibenzo[a,i]pyrene (DB[a,i]P)(14), benz[a]anthracene B[a]A (15), chrysene (16, 17), andbenzo[c]phenanthrene (B[c]Phen) (12) were synthesized as pre-viously described, as were B[a]P H4-9, 10-epoxide (18) andB[a]P 4,5-oxide (19). All compounds were free of detectableimpurities, and the diol epoxides were free of diastereomericcontamination, as determined by mass spectral and nuclearmagnetic resonance analyses. Media for culturing bacterial andmammalian cells were obtained from Becton Dickinson (Cock-eysville, MD) and GIBCO, respectively. Fetal calf serum wasobtained from Reheis (Kankakee, IL). Dimethyl sulfoxide(Me2SO) was distilled from calcium hydride under reducedpressure, and dioxane was distilled from sodium.

Mutagenesis Assays with Bacteria. StrainTA100 of histidine-dependent Salmonella typhimurium was obtained from B.Ames (University of California, Berkeley, CA) and was culturedas described (20). Test compounds were dissolved in anhydrousMe2SO. The bacteria (2 X 108), suspended in 0.5 ml of isotonicphosphate-buffered saline (5 mM potassium phosphate/150mM sodium chloride, final pH 7.0), were preincubated for 1min at 370C with 10 A1 ofMe2SO or various-amounts ofthe plantphenols added in 10 1 of Me2SO. At the end of this time, thediol epoxide, tetrahydroepoxide, or arene oxide was added tothe incubation mixture in 15 ,ul of Me2SO. After an additional5-min incubation at 37°C, 2 ml of top agar was added, .and theentire mixture was poured onto a histidine-deficient Vogel-Bonner agar medium in a 15 X 100 mm Petri dish. Mutationsto histidine-independent growth were assessed by counting themacroscopic colonies of bacteria after a 2-day incubation of theplates at 37C. All experiments were performed in triplicate.

Mutagenesis Assays with Mammalian Ceils. Line V79-6 ofChinese hamster cells was a generous gift of E. H. Y. Chu

HO C H=CH - COOH

H3COFERULIC ACID

OH OH

HO HH=CH -COO

HO H OH

CHLOROGENIC ACID

HO jH =CH-COOH

HOCAFFEIC ACID

H

HO OH

HO

ELLAGIC ACID

HOOH

B|a|P DIOL EPOXIDE-2

FIG. 1. Structures of ferulic, caffeic, chlorogenic, and ellagic acids,and of B[a]P 7,8-diol-9,10-epoxide-2.

(University of Michigan, Ann Arbor, MI). Assays assessing cy-totoxicity and intrinsic mutagenicity were adapted from theprocedure of Chu (21) and were performed as described (20).The V79 cells were seeded in 60-mm culture dishes at concen-trations of 102 cells per dish (4 dishes) and 104 cells per dish(16 dishes) for experiments assessing cell survival and muta-genicity, respectively. After the cells were cultured for 18 hrin 5 ml of medium, the plant phenol was added in 10 ,1 of an-hydrous Me2SO, and B[a]P 7,8-diol-9,10-epoxide-2 was added20 min later in 15 ,ul of Me2SO. Control samples were treatedwith an equal amount ofMe2SO. After a 1-hr incubation at 37TC,the medium and compounds were removed from the culturedish, and the cells were washed once with 5 ml of phosphate-buffered saline (9.5 mM sodium phosphate/140 mM sodiumchloride/3 mM potassium chloride, final pH 7.2) and then cul-tured in 5 ml of fresh medium. To assess cell survival, culturedishes that contained 102 cells were incubated at 37°C for 7 days,at which time the macroscopic colonies of cells were fixed withmethanol, stained with Giemsa stain, and counted. To assessmutagenic activity, the culture dishes containing 104 cells wereincubated at 37°C for 3 days. At this time, mutant cells resistantto 8-azaguanine were selected by addition of 10 ,g of 8-aza-guanine per ml of culture medium. On day 7, the medium wasreplaced with 5 ml of fresh medium containing the same amountof 8-azaguanine. One week later, colonies resistant to 8-aza-guanine were fixed with methanol, stained with Giemsa stain,and counted.

Kinetics of Diol Epoxide Disappearance. Rates for the dis-appearance of B[a]P 7,8-diol-9, 10-epoxide-2 in the presence ofthe phenolic compounds were measured at 25°C in 1:9 (vol/vol) dioxane/water solutions, 0.1 M in NaC104. The pH wasmaintained at 7.0 with 2 mM Tris-HCI buffer. Reactions werefollowed by measuring absorbance at 346 nm with a Cary 219spectrophotometer. Reactions were initiated by addition of5-20 ,1u of a stock solution of the diol epoxide in dioxane to 3ml ofeach reaction mixture containing a buffered solution oftheappropriate phenol to give a final diol epoxide concentration of7-9 jM (with caffeic, ferulic, and chlorogenic acids) or 1.6-3.3LM (with ellagic acid). In all experiments the ratio of phenolto diol epoxide was equal to or greater than 5.3: 1, and at thelowest diol epoxide-to-phenol ratios no significant deviationsfrom pseudo-first-order kinetics were observed. Pseudo-first-order rate constants for reactions of the diol epoxide were de-termined as described (13). Apparent second-order rate con-stants, kapp, at pH 7.0, were determined from the slopes of lin-ear plots of the observed pseudo-first-order rate constants vs.the concentrations of the phenols.

RESULTSThe dose-dependent inhibitory effects of ferulic, caffeic, chlo-rogenic, and ellagic acids on the mutagenicity of 0.05 nmol ofB[a]P 7,8-diol-9,10-epoxide-2 in strain TA100 of Salmonella ty-phimurium are shown in Fig. 2. Mutations induced by the diolepoxide were inhibited 50% by doses of 150, 75, and 50 nmol,respectively, of ferulic, caffeic, and chlorogenic acids. Morestrildngly, only 1 nmol of ellagic acid was needed to reduce themutagenicity of B[a]P 7,8-diol-9,10-epoxide-2 by 50%, and 3nmol of ellagic acid reduced mutations by 90%. Comparison ofthese results with those previously reported (13) for riboflavin5'-phosphate (FMN) (dotted line Fig. 2) indicated that FMNwas 6-20 times more potent than ferulic, caffeic, and chloro-genic acids but was only about 1/10th as active as ellagic acid.Thus 21M ellagic acid was sufficient to inhibit the mutagenicityof B[a]P 7,8-diol-9,10-epoxide-2 by 50% in the bacterial testsystem.

5514 Biochemistry: Wood et aL

Proc. Natl. Acad. Sci. USA 79 (1982) 5515

UV

0

20 -. \'

0

0

z 40-

Uj W~~~~~~~~~~~L

100 200 0 1.0 2.0 3.0

r 40 -CELLAGIC ACID(nmol)(

FIG..Efectf ferliccaffic, hloroenicandella icaiso

ck:~~~ ~ ~ ~ ~ ~~~~Ai0~~~~~~~~~~~~~20

ChmorogenicAcid

FthN Coffeic

[\EI~~o~Ic1AcAcid

Ellagic ~ ciIcdwsasoetihbtro h yooi Fg

100 200 300 400 500

COMPOUND (nmol)

FIG. 2. Effect of ferulic, caffeic, chlorogenic, and ellagic acids on

the mutagenic activity of B[a]P 7,8-diol-9,10-epoxide-2 in strain TA100

of S. typhimurium. The intrinsic mutagenicity of 0.05 nmol of the diolepoxide in the absence of any added phenolic compound was 1,013121 histidine-independent revertants per plate (n = 12). The back-

ground mutation rate was 121 24 (n = 6). All data points representthe mean of three replicate determinations.

Ellagic acid was also a potent inhibitor of the cytotoxic (Fig.3) and mutagenic (Fig. 4) effects ofB[a]P 7,8-diol-9,9O-epoxide-2 in Chinese hamster V79 cells. Although only 1% of the mam-malian cells survived exposure to 0.6 nmol/ml of the diol epox-

ide in the absence of ellagic acid, cell survival was increased to

50% by an 8-fold molar excess (5 nmol/ml) of ellagic acid (Fig.3). 8-Azaguanine-resistant mutants induced in the V79 cells bya 1-hr incubation with B[a]P 7,8-diol-9, 10-epoxide-2 at 0.2

nmol/ml were reduced by 50% when the cells were pretreatedwith a 10-fold molar excess (2 nmol/ml) of ellagic acid. As wasfound in the bacterial test system, considerably higher concen-trations of chlorogenic, caffeic, and ferulic acids were requiredto block the mutagenicity (Fig. 4) and toxicity (data not shown)of B[a]P 7,8-diol-9, 10-epoxide-2, but the relative protectiveeffect of these three compounds was essentially the same in themammalian and bacterial cells. No cytoxicity (>80% cell sur-vival) or mutagenicity was observed in the Chinese hamster V79cellswhen relatively high concentrations ofellagic acid (20 nmol/ml), chlorogenic acid (500 nmol/ml), caffeic acid (500 nmol/ml), or ferulic acid (500 nmol/ml) were incubated with the cellsin the absence of B[a]P 7,8-diol-9, 10-epoxide-2.The effect of ellagic acid on the mutagenicity of several bay-

-j 80

Wfi 20'l//1.0 nmol/ml

i-

5 10 15 20ELLAGIC ACID (nmol/mI )

FIG. 3. Effect of ellagic acid on the toxicity of several concentra-tions of B[a]P 7,8-diol-9,10-epoxide-2 in Chinese hamster V79 cells.Each data point represents the mean of four replicate determinations.

region diol epoxides was determined in strain TA100 of S. ty-phimurium (Table 1). Ellagic acid had equivalent antimutagenicactivity toward B[a]P 7,8-diol-9, 10-epoxide-1, B[a]P 7,8-diol-9,10-epoxide-2, and B[a]P H4-9, LO-epoxide, a highly mutagenicepoxide in which the hydroxyl groups of the angular benzo-ringhave been replaced by hydrogen. Ellagic acid was much lesseffective in blocking the mutagenicity of the B[a]P 4,5-oxide.(K-region oxide), a primary oxidative metabolite of B[a]P thathas little or no carcinogenic activity (22). Ellagic acid was alsoa potent inhibitor of the mutagenic activity of the bay-regiondiol epoxides of DB[a,h]P and DB[a,i]P but was less effectivein inhibiting mutations induced by the bay-region diol epoxidesof B[a]A, chrysene, and B[c]Phen.

Ferulic, caffeic, chlorogenic, and ellagic acids accelerate the

cnU)

40 10z5

0

z20 20 1. nmollogic Ceruic\m

00 AcAcd

a.

c25 50 7 0 0' 50 500@ ~ ~ ~ LAI ACIOUD (nmol /ml )

FIG. . Effect of elhorogenic and aconcid en

themuagniitof B[a]P 7,8-diol-9,10-epoxide-2in0.2es nmolmstrV9cl)n

Chinese um 20ableV7 .Ellagicthe adsencioanyaddedupheno c

poun tytoadB[a]P 7,8-diol-9,10-epoxide-2,induce an veag od12outnt

perf10esuriveingblclls. Theaboutaenplatin efficienc w~asP 70%.x

ZKrgo xd) Acidayoiaiemtblt fBaPta

FG.4Efetoferulic,caffeic, chlorogenic, and ellagic acidsoneeathtemtgncit fBaP78dil91-pxd-2(. mlm)iChns ase 7 el.I h asneo n de hnlccmpon _] ,-il910eoie2idcda aeaeo 0 uatpe0 uvvn el.Teaslt ltn fiinyws7%

Biochemistry: Wood et al.

Proc. Natl. Acad. Sci. USA 79 (1982)

Table 1. Effect of ellagic acid on the mutagenic activity of bay-region diol epoxides and other oxides of several polycyclicaromatic hydrocarbons

Histidine Inhibition by ellagic acid, %Amount, revertants 0.625 1.25 2.50 12.5 25 50 75 100

Compound nmol* per platet nmol nmol nmol nmol nmol nmol nmol nmolB[a]P diol epoxide-2 0.05 1,190 50 77B[a]P diol epoxide-1 0.05 1,060 23 42 58 -

B[a]P H4-9,10-epoxide 0.05 1,240 37 55 70 - -

B[a]P 4,5-oxide 0.50 890 - - - 18 22 35 60DB[a,h]P diol epoxide-2 0.15 950 35 55 77 -

DB[a,i]P diol epoxide-2 0.15 1,360 36 54 73B[a]A diol epoxide-2 0.15 830 - - 3 28 58 73 85B[c]Phen diol epoxide-2 0.15 440 - 5 10 18 16 36Chrysene diol epoxide-2 0.50 360 - 15 21 34* Amounts of the compounds were selected from linear portions of their dose-response curves with strain TA100 of S. typhimurium.t Average number (n = 3) of histidine-independent revertants induced per plate in strain TA100 of S. typhimurium by the indicated amounts ofeach epoxide, after subtraction of the spontaneous mutation frequency of 140 revertants per plate.

disappearance of B[a]P 7,8-diol-9,10-epoxide-2 from aqueoussolution (1:9 dioxane/water, pH 7/0.1 M NaC1O4) as shownby the second-order rate constants for diol epoxide disappear-ance in Table 2. This observation indicates that the antimuta-genic activity of the phenolic compounds results from their in-teraction with the diol epoxide, rather than with the target cells.If the most active compound, ellagic acid, is assigned an arbi-trary rate of 100 for its reaction with the diol epoxide, the rel-ative rates for chlorogenic, caffeic, and ferulic acids are 1.1, 0.6,and 0.3, respectively. This order of reactivity, as well as themagnitude of the differences in rate among the phenolic com-pounds, parallels closely their effectiveness in inhibiting mu-tagenesis. The effect of ellagic acid on diol epoxide disappear-ance is particularly striking. Under the experimental conditionsdescribed above, the half-life of the diol epoxide was decreasedfrom 40 min in the absence of ellagic acid to 2 min in the pres-ence of 10 AM ellagic acid. The data in Table 2 also show thatthe rate constant for the disappearance of the diol epoxide inthe presence of ellagic acid is about 12 times greater than theanalogous constant calculated for the disappearance of the diolepoxide in the presence of FMN, a finding that is consistentwith the approximate 10-fold difference in antimutagenic po-tency between FMN and ellagic acid (Fig. 2).

DISCUSSIONThe results of the present study indicate that the naturally oc-curring plant phenols ferulic, caffeic, chlorogenic, and ellagicacids can block the mutagenic activity of B[a]P 7,8-diol-9,10-epoxide-2, the only known ultimate carcinogen of B[a]P. Ofthefour compounds tested, ellagic acid (a phenolic lactone) wasshown to be an exceptional antimutagen and was effective atconcentrations 1/10th ofthose described (13) for FMN. Ferulic,caffeic, and chlorogenic acids all induced concentration-depen-dent decreases in the mutagenicity of B[a]P 7,8-diol-9, 10-epox-

Table 2. Apparent second-order rate constants for thedisappearance of B[a]P.7,8-diol-9,10pepoxide-2 at 250Cin 1:9 dioxane/water, ionic strength 0.1 M, pH 7.0

Compound Conc. range, ,uM kapp, M-1 A-

Ferulic acid 250-980 1.78Caffeic acid 250-980 3.13Chlorogenic acid 63-300 6.3Ellagic acid 8.8-35 560FMN 45*

* Calculated from Kekct for this compound at pH 7.0 (13).

ide-2, but these three plant phenols were considerably less ac-tive than ellagic acid and FMN.The kinetic data (Table 2) for the disappearance of B[a]P 7,8-

diol-9, 10-epoxide-2 from aqueous solution indicate that the an-timutagenic activity of ellagic acid and the other three com-pounds results from a direct interaction between the phenol andthe diol epoxide as opposed to an effect on the cells. Previousstudies have shown that phenol itself is an acid catalyst ofB[a]P7,8-diol-9, 10-epoxide-2 hydrolysis and that significant amountsof hydrocarbon-phenol adduct(s) can be formed (10). Similarly,recent studies (unpublished observations) have shown that el-lagic acid forms an adduct with B[a]P 7,8-diol-9,10-epoxide-2via a phenolic hydroxyl group, but the polyphenol is at least3,000 times more reactive than phenol. With FMN, generalacid catalysis by the phosphate monoanion is observed in theabsence of any adduct formation, and all of the B[a]P 7,8-diol-9, 10-epoxide-2 is hydrolyzed to tetraols (13). Studies on themechanism of the antimutagenic effect of ellagic acid may pro-vide a basis for the observation that ellagic acid inhibits themutagenicity of bay-region diol epoxides of B[a]P, DB[a,h]P,and DB[a,i]P more readily than the less reactive bay-region diolepoxides of B[a]A, chrysene, and B[c]Phen (Table 1). Both the-oretical quantum mechanical calculations (2) based on the 'r-electron energy changes and kinetic data for solvolysis of diolepoxides (9, 10, 12, 16, 23) have indicated that the diol epoxidesof B[a]P, DB[a,h]P, and DB[a,i]P are the most reactive of thebay-region diol epoxides that have been studied to date.

None of the four plant phenols had detectable mutagenicactivity in strain TA100 of S. typhimurium or in Chinese ham-ster V79 cells. The bacterial studies were done in both the ab-sence and presence of a hepatic 9,000 X g supernatant fractionfrom Aroclor 1254-induced rats, which serves as a metabolicactivation system for premutagens that must be metabolized totheir active products. Recent studies have suggested that caffeicand chlorogenic acids are mutagenic and clastogenic whentested in the presence of a transition metal (24, 25). However,previous reports (26-29) have indicated that transition metalscatalyze the oxidation of catechols with the generation of hy-drogen peroxide. It is likely that the mutagenicity of caffeic andchlorogenic acids in the in vitro test systems results from theformation of unphysiologically high concentrations of hydrogenperoxide or. related active oxygen species such as superoxide(26). No toxicity was observed when rats were fed a diet con-taining 0.5% -caffeic acid for 6 months. Indeed, these rats hada lower incidence of spontaneous tumors than rats fed a controldiet (30).

All four of the phenolic compounds described in this report

5516 Biochemistry: Wood et aL

Proc. Natl. Acad. Sci. USA 79 (1982) 5517

are widely distributed among fruits and vegetables that are reg-ularly consumed by humans (31-33). Caffeic and ferulic acidsare ubiquitous; high concentrations of ellagic acid are found ingrapes, certain nuts, and strawberry preserves; and chlorogenicacid is found in relatively high concentrations in coffee, apples,and potatoes.

Ellagic acid is pharmacologically active and has been foundto control hemorrhage in animals (34, 35) and humans (36, 37),presumably by activating the intrinsic blood coagulation system(38). Hypotensive and sedative effects of ellagic acid have alsobeen reported in rodents (39). Rats fed ellagic acid at a dose of50 mg/kg per day for up to 45 days did not show any evidenceof toxicity (40), and doses of ellagic acid of 0.2 mg/kg, intra-venously, were well tolerated in humans (37).

Caffeic and ferulic acids have been shown to diminish theincidence of forestomach tumors in mice treated with B[a]P(41). The ability of both caffeic and ferulic acids to block themutagenicity of B[a]P 7,8-diol-9,10-epoxide-2 may be relatedto their anticarcinogenic activity. To our knowledge, ellagic andchlorogenic acids have not been tested for anticarcinogenic ac-tivity, although chlorogenic acid has recently been shown toblock the formation of mutagenic compounds resulting from thepyrolysis of protein (42). The results of the present study in-dicate that ellagic acid is an exceptional antagonist of the mu-tagenicity of bay-region diol-epoxides derived from severalpolycyclic aromatic hydrocarbons. Additional research is neededto determine whether or not ellagic acid can block the carcin-ogenicity of polycyclic hydrocarbons in animals.

The authors thank Mrs. Ann Marie Williams for her excellent assis-tance in the preparation of this manuscript.

1. Jerina, D. M. & Daly, J. W. (1977) in Drug Metabolism fromMicrobe to Man, eds. Parke, D. V. & Smith, R. L. (Taylor &Francis, London), pp. 13-22.

2. Jerina, D. M., Lehr, R. E., Yagi, H., Hernandez, O., Dansette,P. M., Wislocki, P. G., Wood, A. W., Chang, R. L., Levin, W.& Conney, A. H. (1976) in Metabolic Activation in MutagenesisTesting, eds. de Serres, F. J., Fouts, J. R., Bend, J. R. & Philpot,R. M. (Elsevier/North Holland Biomedical Press, Amsterdam),pp. 159-177.

3. Conney, A. H., Levin, W., Wood, A. W., Yagi, H., Lehr, R. E.& Jerina, D. M. (1978) Adv. Pharmacol. Ther. 9, 41-52.

4. Jerina, D. M., Sayer, J. M., Thakker, D. R., Yagi, H., Levin,W., Wood, A. W. & Conney, A. H. (1980) in Carcinogenesis:Fundamental Mechanisms and Environmental Effects, eds. Pull-man, B., Ts'o, P. 0. P. & Gelboin, H. V. (D. Reidel, Dordrecht,Holland), pp. 1-12.

5. Nordqvist, M., Thakker, D. R., Yagi, H., Lehr, R. E., Wood, A.W., Levin, W., Conney, A. H. & Jerina, D. M. (1979) in Molec-ular Basis of Environmental Toxicity, ed. Bhatnager, R. S. (AnnArbor Science Publishers, Ann Arbor, MI), pp. 329-357.

6. Wood, A. W., Levin, W., Chang, R. L., Yagi, H., Thakker, D.R., Lehr, R. E., Jerina, D. M. & Conney, A. H. (1979) in Poly-nuclear Aromatic Hydrocarbons, Third International Sympos-ium on Chemistry and Biology-Carcinogenesis and Mutagenesis,eds. Jones, P. W. & Leber, P. (Ann Arbor Science Publishers,Ann Arbor, MI), pp. 531-551.

7. Yagi, H., Hernandez, 0. & Jerina, D. M. (1975) J. Am. Chem.Soc. 97, 6881-6883.

8. Yagi, H., Thakker, D. R., Hernandez, O., Koreeda, M. & Jerina,D. M. (1977) J. Am. Chem. Soc. 99, 1604-1611.

9. Whalen, D. L., Montemarano, J. A., Thakker, D. R., Yagi, H.& Jerina, D. M. (1977) J. Am. Chem. Soc. 99, 5222-5224.

10. Whalen, D. L., Ross, A. M., Montemarano, J. A., Thakker, D.R., Yagi, H. & Jerina, D. M. (1979) J. Am. Chem. Soc. 101,5086-5088.

11. Kapitulnik, J., Wislocki, P. G., Levin, W., Yagi, H., Jerina, D.M. & Conney, A. H. (1978) Cancer Res. 38, 354-358.

12. Sayer, J. M., Yagi, H., Croisy-Delcey, M. & Jerina, D. M. (1981)J. Am. Chem. Soc. 103, 4970-4972.

13. Wood, A. W., Sayer, J. M., Newmark, H. L., Yagi, H., Mi-chaud, D. P., Jerina, D. M. & Conney, A. H. (1982) Proc. Natl.Acad. Sci. USA 79, 5122-5126.

14. Lehr, R. E., Kumar, S., Cohenoor, P. J. & Jerina, D. M. (1979)Tetrahedron Lett. 3819-3822.

15. Lehr, R., Schaeffer-Ridder, M. & Jerina, D. M. (1977) Tetrahe-dron Lett. 539-542.

16. Whalen, D. L., Ross, A. M., Yagi, H., Karle, J. M. & Jerina, D.M. (1978)J. Am. Chem. Soc. 100, 5218-5221.

17. Yagi, H., Vyas, K. P., Tada, M., Thakker, D. R. & Jerina, D. M.(1982)J. Org. Chem. 47, 1110-1117.

18. Yagi, H., Holder, J. M., Dansette; P., Hernandez, O., Yeh, H.J. C., LeMahieu, R. A. & Jerina, D. M. (1976)J. Org. Chem. 41,977-985.

19. Dansette, P. M. & Jerina, D. M. (1974) J. Am. Chem. Soc. 96,1224-1225.

20. Wood, A. W., Chang, R. L., Levin, W., Lehr, R. E., Schaeffer-Ridder, M., Karle, J. M., Jerina, D. M. & Conney, A. H. (1977)Proc. Natl. Acad. Sci. USA 74, 2746-2750.

21. Chu, E. H. Y. (1971) in Chemical Mutagens: Principles andMethods for Their Detection, ed. Hollaender, A. (Plenum, NewYork), pp. 411-444.

22. Wislocki, P. G., Kapitulnik, J., Levin, W., Conney, A. H., Yagi,H. & Jerina, D. M. (1978) Cancer Lett. 5, 191-197.

23. Wood, A. W., Chang, R. L., Levin, W., Ryan, D. E., Thomas,P. E., Lehr, R. E., Kumar, S., Sardella, D. J., Boger, E., Yagi,H., Sayer, J. M., Jerina, D. M. & Conney, A. H. (1981) CancerRes. 41, 2589-2597.

24. Stich, H. F., Rosin, M., Wu, C. W. & Powrie, W. D. (1981)Mutat. Res. 90, 201-212.

25. Stich, H. F., Rosin, M., Wu, C. H. & Powrie, W. D. (1981) Can-cer Lett. 14, 251-260.

26. Aver'yanov, A. A. (1981) Biokhimiya 46, 256-261.27. Grinstead, R. R. (1964) Biochemistry 3, 1308-1314.28. Tyson, C. A. & Martell, A. E. (1972) J. Am. Chem. Soc. 94,

939-945.29. Martell, A. E. (1980) in Autooxidation in Food and Biological

Systems, eds. Simic, M. G. & Karel, M. (Plenum, New York),pp. 89-118.

30. Hirono, I., Hosaka, S., Haga, M., Sakata, M., Mori, H., Tanaka,T. & Hikino, H. (1980)J. Food Safety 4, 205-211.

31. Fairbairn, J. W. (1959) The Pharmacology ofPlant Phenolics (Ac-ademic, New York).

32. Bate-Smith, E. C. (1972) Phytochemistry 11, 1153-1156.33. Sondheimer, E. (1964) Bot. Rev. 30, 677-712.34. Botti, R. E. & Ratnoff, 0. D. (1964) J. Lab. Clin. Med. 64,

385-398.35. Girolami, A., Agostino, D. & Cliffton, E. E. (1966) Blood 27,

93-102.36. Cliffton, E. E. (1967) Am. J. Med. Sci. 254, 117-123.37. Girolami, A. & Cliffton, E. E. (1967) Thromb. Diath. Haemorrh.

17, 165-175.38. Bock, P. E., Srinivasan, K. R. & Shore, J. D. (1981) Biochemistry

20, 7258-7266.39. Bhargava, U. C., Westfall, B. A. & Siehr, D. J. (1968)J. Pharm.

Sci. 57, 1728-1732.40. Blumenberg, F. W., Claus, C., Eunecker, C. & Kassler, F. J.

(1960) Arzneim.-Forsch. 10, 223-226.41. Wattenberg, L. W., Coccia, J. B. & Lam, L. K. T. (1980) Cancer

Res. 40, 2820-2823.42. Fukuhara, Y., Yoshida, D. & Goto, F. (1981) Agric. Biol. Chem.

45, 1061-1066.

Biochemistry: Wood et aL