1754 Biol. Pharm. Bull. 26(12) 1754—1760 (2003) Vol. 26, No. 12

Screening of the Inhibitory Effect of Vegetable Constituents on the Aryl Hydrocarbon Receptor-Mediated Activity Induced by 2,3,7,8-Tetrachlorodibenzo-p-dioxin

,a a a b a Yoshiaki AMAKURA,* Tomoaki TSUTSUMI, Kumiko SASAKI, Takashi YOSHIDA, and Tamio MAITANI a Division of Foods, National Institute of Health Sciences; 1–18–1 Kamiyoga, Setagaya-ku, Tokyo 158–8501, Japan: and b Faculty of Pharmaceutical Sciences, Okayama University; 1–1–1 Tsushima, Okayama 700–8530, Japan. Received June 12, 2003; accepted August 22, 2003

The aryl hydrocarbon receptor (AhR) is a ligand-activated nuclear transcription factor that mediates re- sponses to environmental contaminants such as dioxins, which have many adverse health effects. We performed a preliminary screening of the inhibitory effects of vegetable constituents on 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)-induced activation of AhR using the AhR-based bioassay for dioxins, the Ah-Immunoassay. Ninety veg- etable constituents including flavonoids, tannins, saponins, terpenes, etc., were assayed in vitro. Among them, flavones, flavonols, anthraquinones, piperine, coumestrol, brevifolincarboxylic acid, and resveratrol showed marked inhibitory effects on AhR-based bioassay activation by TCDD, and their effects were dose dependent. Curcumin, carnosol, and capsaicin also inhibited the activation of AhR in this assay, although to a lesser degree. These results suggest that several vegetable constituents might play a role in protection against dioxin toxicity. Key words vegetable constituent; aryl hydrocarbon receptor; dioxin; food; Ah-immunoassay

The toxic expression system of dioxins is generally agreed by TCDD by applying AhR-based bioassays for dioxins and on based on various studies.1—3) Briefly, potent dioxins showed that inhibitory effects were exerted by extracts of known as xenobiotic ligands, most notably 2,3,7,8-tetra- sage, green leafy vegetables, and citrus.13) In addition, the in- chlorodibenzo-p-dioxin (TCDD), enter cells and specifically teraction of vegetable constituents with AhR was determined bind to the aryl hydrocarbon receptor (AhR). AhR is located using an in vitro reporter gene assay.14) The present study in the cytoplasm as one component of a complex composed was performed to evaluate the inhibitory activity of 90 veg- of two molecules of a 90-kDa heat shock protein. After lig- etable constituents against TCDD-induced activation of AhR and binding, AhR dissociates from the complex and translo- using an AhR-based bioassay for dioxins, the Ah-immunoas- cates to the nucleus and dimerizes with the AhR nuclear say (Ah-I), as an experimental model. translocator protein (ARNT). AhR-ARNT heterodimers act as transcriptional activators by binding to specific DNA se- MATERIALS AND METHODS quences termed xenobiotic-response elements (XREs) in the enhancer regions of proteins involved in xenobiotic metabo- Samples and Reagents Dimethyl sulfoxide (DMSO) lism and function as ligand-activated transcription factors and TCDD (98%) were purchased from Wako Pure Chemical mediating the pleiotropic biological responses to dioxins. Industries, Ltd. (Osaka, Japan). The following compounds These biological responses include teratogenesis, carcinogen- were isolated from the species indicated in parentheses15—17): esis, epithelial hyperplasia, and induction of drug-metaboliz- Agrimoniin (Agrimonia japonica); brevifolincarboxylic acid ing enzymes. (Phyllanthus flexuosus); gemin A, , and tellima- AhR activation caused by the binding of dioxins is as- grandin I (Geum japonicum); geraniin (Geranium thun- sumed to be one of the first and key steps in the development bergii); procyanidins B2 and C1 (Theobroma cacao); and ca- of dioxin toxicity. Accordingly, the inhibition of AhR activa- suarinin and procyanidin B4 (Cowania mexicana). Pentagal- tion would be expected to provide protection against such loylglucose was obtained by methanolysis of tannic acid.18) toxicity. The main route of dioxin contamination in humans Naringin 4Ј-O-a- was isolated from a natural food appears to be through food,4—7) and therefore if factors that additive, enzymatically modified naringin.19) Their purities inhibit AhR activation are present in our daily diet, they were greater than 90% as determined by HPLC. The other might play a role in protection against dioxin toxicity, conse- compounds tested were purchased from Funakoshi (Tokyo, quently leading to a lower risk of dioxins to human health. Japan), Wako, Tokyo Kasei (Tokyo, Japan), Sigma-Aldrich AhR ligands such as dioxins are artificial products that (St. Louis, MO, U.S.A.), and Nakahara Science (Gifu, have recently appeared, and AhR might primarily function as Japan). Most chemicals were of the highest grade commer- a regulatory receptor for exogenous natural products such as cially available and used without further purification (mostly vegetable constituents. Recently, several investigations have Ͼ98%). The Ah-I kit was obtained from Kubota (Osaka, studied the effects of dietary components such as flavonoids, Japan). etc. on the binding of TCDD to AhR, and it was found that Evaluation Using AhR-Based Bioassay To evaluate the these components exert antagonistic activity.8—12) However, AhR-based activity, the Ah-I was applied as an experimental fundamental studies on the relationship between the AhR model. The Ah-I method is a receptor-binding assay using pathway and food constituents are still limited. cytosol containing AhR extracted from mammalian liver Based on this background information, we reported the cells and immunologically measures the dioxin level utilizing screening for the inhibitory effects on the activation of AhR an antigen-antibody reaction. This is a technique to detect

∗ To whom correspondence should be addressed. e-mail: [email protected] © 2003 Pharmaceutical Society of Japan December 2003 1755 the reactivity of AhR with dioxins and dioxin-like com- fluenced by glycosidation in the B-ring. pounds on an ELISA plate without using living cells and is Among the flavanones, hesperetin and naringenin inhibited useful for screening the biological toxicity of dioxins. There- the activation of AhR induced by TCDD with EC70 values of fore the Ah-I is used as a simple dioxin-monitoring 14.6 and 27.7 m M, respectively. The flavanone glycosides method.20,21) The Ah-I was employed as follows. The cytosol (naringin, hesperidin, naringin 4Ј-O-a-glucoside, taxifolin,

(200 ml) was added to the sample, which as dissolved in and fustin) did not inhibit the activity at the EC70 level. The DMSO at final concentrations of 0.5—50 m M, or to DMSO tendency of glycosides to weaken these activities was similar alone as the control, the mixture was preincubated for to those of flavones and flavonols. 20 min, and then incubated with 5 nM TCDD (final concen- The isoflavones mostly did not show an inhibitory effect tration of 0.025 nM) for 2 h at 30 °C (the final DMSO concen- on activation of AhR induced by TCDD in bioassay at the tration was 1% in culture medium). After incubation, the for- EC70 level, and they slightly inhibited the activity at high mation of the AhR-TCDD complex was determined using an concentrations of 25—50 m M (ca. 10—20% inhibition).

Ah-I kit, and the absorbance was determined at 405 nm using Genistein had a slight inhibitory effect at the EC70 level of a microplate reader. The AhR activity was calculated as: 40.8 m M and equol, which is regarded as the final metabolite {1Ϫ[(AϪB)Ϫ(CϪD)]/(AϪB)}ϫ100, where A is the ab- of isoflavones, also exhibited inhibition of activity at sorbance of the control with TCDD added, B is the ab- 41.2 m M. Although most isoflavone themselves have low in- sorbance of the control with DMSO added, C is the ab- hibitory effects less than the EC70 level, it is interesting that sorbance of the sample solution with TCDD added, and D is their metabolite has an inhibitory effect at the EC70 level. the absorbance of the sample solution with DMSO added. Based on these results in flavonoids, the inhibitory effect on All experiments were carried out in duplicate or triplicate. TCDD-induced activation of AhR in flavonoids is generally The values obtained with DMSO alone were considered as in the following order: flavonesϭflavonolsϾϾflavanonesϾ 100% of the control value. To define the potency of each of isoflavonesϭflavonoid glycosides. the compounds tested, AhR activity was expressed as con- Among the remaining compounds tested, anthraquinones centration–response curves. The concentrations producing showed remarkable inhibition of AhR activation comparable AhR activity equal to 70% of the maximal response to to that of flavones and flavonols. Aloe-emodin and emodin

TCDD in controls were calculated and expressed as the EC70 had the strongest potencies among the tested samples at the value. EC70 levels of 0.5 and 0.6 m M, respectively. In addition, alizarin showed a marked inhibitory effect at 3.2 m M. RESULTS AND DISCUSSION Piperine, resveratrol, brevifolincarboxylic acid, coume-

strol, and shikonin inhibited activation in this assay (EC70 The chemical structures of the tested compounds are values were 2.7, 3.9, 3.9, 5.6, and 5.6 m M, respectively). Cap- shown in Fig. 1. We used the Ah-I as a model in our prelimi- saicin, carnosol, and curcumin had EC70 values of 28.1, 30.7, nary experiments and evaluated the influence of individual and 35.4 m M, respectively. Among them, resveratrol and cur- vegetable constituents on the AhR pathway induced by cumin have been reported to show antagonist activity on TCDD. Figure 2 shows the dose–response curves plotted on a AhR.10,12) As they also had inhibitory effects in this present log scale for some individual samples. Most showed an in- assay system, it is suggested that they are candidate prophy- hibitory effect on AhR activation at high concentrations lactic agents for the prevention of dioxin toxicity. around the 50 m M level. Some showed marked inhibitory ef- Figure 3 depicts the molecular models obtained as the fects at low concentrations of 0.5—2.5 m M. minimum energy conformation of TCDD and the five com-

Table 1 shows the inhibitory potencies (EC70) of the tested pounds that inhibited the TCDD-induced AhR activation compounds on AhR-based bioassay activation. Most samples (apigenin, resveratrol, emodin, brevifolincarboxylic acid, and had weak inhibitory effects less than the EC70 level even at piperine). These compounds have molecular sizes and planar high concentrations (inhibitory effect of ca. 20%). Of the structures similar to those of TCDD. Therefore it is sug- tested compound groups, flavonoids had the highest in- gested that they could have antagonist-like activity for AhR. hibitory activity. Among the flavonoids, the flavones and the On the other hand, hydrolyzable tannins as shown in Table flavonols had strong inhibitory potencies on AhR activation 1 and Fig. 2 inhibited the activation of this AhR-based bioas- induced by TCDD. Among flavones, the EC70 values for say by TCDD. As shown in Fig. 1, the structures of hydrolyz- chrysin, luteolin, apigenin, and baicalein were 0.7, 1.8, 1.9, able tannins (monomers: pentagalloylglucose, pedunculagin, and 5.1 m M, respectively. The flavone glycosides showed , , and geraniin; dimers: gemin A lower inhibitory effects than aglycone (EC70 values of and agrimoniin) have large molecular size and many hy- apigetrin and baicalin were 5.2 and 36.1, respectively). droxyl groups in bulky structures. Therefore their chemical Among the flavonols, kaempferol, quercetin, myricetin, characteristics differ from those of other inhibitory com- and morin strongly inhibited AhR-mediated activation by pounds such as flavonoids and anthraquinones. They might

TCDD (EC70 values were 2.1, 2.7, 4.3, and 5.3 m M, respec- function as partial antagonists to AhR since they mostly pos- tively). Spiraeoside and tiliroside, which are flavonol glyco- sess the biphenyl structural moieties referred to as the hexa- sides, exhibited inhibition of activity at the EC70 level hydroxydiphenoyl (HHDP) group. However, the structure of (2.1 and 34.4, respectively), although the other glycosides the chiral HHDP moiety is nonplanar,22) and therefore it is quercitrin, rutin, and isoquercitrin did not inhibit activity at difficult to accept the possibility of an antagonist of this type the EC70 level. Spiraeoside showed a strong inhibitory effect of compound with the HHDP group. Among the important comparable to aglycones such as quercetin, and therefore it is properties of tannins is the formation of complexes with pro- suggested that the inhibitory effect of flavonols are little in- teins.23) Accordingly, it is considered that this characteristic 1756 Vol. 26, No. 12

Fig. 1. Chemical Structures of the Vegetable Constituents Assayed property of tannins yielded their effects in this assay system. classified into so-called phytoestrogens and they also had Further study will be required to elucidate them. structural characteristics similar to those of TCDD. Flavones In a previous report, we reported on the AhR-mediated ac- such as chrysin showed activation in the AhR-based assay at tivity of vegetable constituents at high concentrations using high concentrations. On the other hand, in this study, they an in vitro reporter gene assay.14) The active compounds were had strong inhibitory effects on AhR-mediated activity in- December 2003 1757

Fig. 1. (Continued)

duced by TCDD at the EC70 level. This suggests that most Additionally, isoflavones at high concentrations also flavones, flavonols, etc. act as both agonists and antagonists showed AhR activation, although equol exhibited only a on AhR. There are a few reports on the amounts of ingestion slight induction of activity.14) In this study, most isoflavones and absorption of flavonoids. For example, it was estimated had slight inhibitory effects less than the EC70 level, and that the daily consumption of flavonols plus flavones is 3— genistein and equol showed inhibition at the EC70 level. 80 mg.24,25) Moreover, the papers that the ingestion of 64 mg Watanabe et al. reported that the genistein and equol ap- quercetin resulted 3 h later in a concentration of 650 nM peared in the plasma at micromole levels of a few micro- quercetin in hydrolyzed plasma was reported.25,26) Another grams after the ingestion of baked soybean powder by hu- study performed in 10 healthy volunteers showed that in- mans.27) Therefore it is suggested that they might have not gested quercetin 87 mg resulted in a quercetin concentration functioned as agonists to AhR in usual intake. 25) in hydrolyzed plasma ranging from 129—843 nM after 3 h. Some anthraquinones also showed slight activation of AhR Taking into account the average intake of flavonoids in gen- at high concentrations.14) In the present study, most an- eral and their absorption amount, it is suggested that they thraquinone samples tested had potent inhibitory effects on might act as antagonists of AhR in usual intake. AhR activity at the EC70 level. On the other hand, coume- 1758 Vol. 26, No. 12

Fig. 2. Dose-Dependent Inhibitory Effect of Vegetable Constituents on the AhR-Activation Induced by TCDD in the Ah-Immunoassay Each point represents the mean of two or three replicated analyses. December 2003 1759

Table 1. Inhibitory Effect of Vegetable-Related Constituents on TCDD-Induced Activation of AhR Estimated Using the AhR-Based Bioassay

a) EC70 EC70 (m M)(m M)

Flavones Phloroglucinolcarboxylic acid nc Apigenin 1.9 Protocatechuic acid nc Luteolin 1.8 Vanillin nc Baicalein 5.1 Eugenol nc Chrysin 0.7 Chlorogenic acid nc Apigetrin 5.2 nc Vitexin ncb) Brevifolincarboxylic acid 3.9 Baicalin 36.1 Curcumin 35.4 Flavonols Rosmarinic acid nc Quercetin 2.7 Resveratrol 3.9 Kaempferol 2.1 Anthraquinones and naphtoquinones Myricetin 4.3 Emodin 0.6 Morin 5.3 Aloe-emodin 0.5 Quercitrin nc Alizarin 3.2 Rutin nc Sennoside A nc Isoquercitrin nc Shikonin 5.6 Tiliroside 34.4 Condensed tannins Spiraeoside 2.1 (ϩ)-Catechin nc Flavanones (Ϫ)-Epicatechin nc Naringenin 27.7 (Ϫ)-Epigallocatechin nc Hesperetin 14.6 (Ϫ)-Epicatechin gallate nc Naringin nc (Ϫ)-Epigallocatechin gallate nc Hesperidin nc Procyanidin B1 nc (ϩ)-Taxifolin nc Procyanidin B2 nc (ϩ)-Fustin nc Procyanidin B4 nc Naringin 4Ј-O-glucoside nc Procyanidin C1 nc Isoflavones Theaflavin nc Daidzein nc Hydrolyzable tannins Glycitein nc Pentagalloylglucose 29.6 Genistein 40.8 Pedunculagin 42.0 Daidzin nc Tellimagrandin I 12.4 Glycitin nc Geraniin 27.3 Genistin nc Casuarinin 29.3 6Љ-Acetyldaidzin nc Agrimoniin 6.4 6Љ-Malonyldaidzin nc Gemin A 31.5 6Љ-Malonylgenistin nc Others 6Љ-Malonylglycitin nc Coumestrol 5.6 Equol 41.2 Carnosol 30.7 Chalcones Glycyrrhetinic acid nc Butein nc Glycyrrhizic acid nc Phloretin nc Ginsenoside Rb1 nc Phloridzin nc Caffeine nc Phenolcarboxylic acids and related compounds Theophylline nc p-Hydroxybenzoic acid nc Capsaicin 28.1 Caffeic acid nc Piperine 2.7 trans-Ferulic acid nc b-Sitosterol nc trans-Cinnamic acid nc Ginkgolide A nc p-Coumaric acid nc Lycopene nc nc Limonin nc Methyl gallate nc b-Carotene nc

Each value is the mean of two or three analyses. a) Concentrations producing AhR activity equal to 70% of the maximal response to TCDD. Calculated from the slope of the linear portion of each dose–response curve near the origin. b) Not calculated; no induction at the EC70 level observed. strol, piperine, carnosol, capsaicin, etc., showed only antago- ating dioxin toxicity. Although the present results should be nistic effects. considered to reflect general trends and further investigation Thus vegetable constituents that inhibited TCDD-induced is required in the future, it is suggested that several vegetable activation of AhR mostly had the properties of both agonists constituents might prevent dioxin toxicity associated with and antagonists of AhR in in vitro bioassays, and it was sug- AhR and they might lead to lowering the risk of dioxins to gested that they may function as antagonists in usual intake. human health. It is believed that a well-balanced diet is important in allevi- 1760 Vol. 26, No. 12

7) Djien Liem A. K., Furst P., Rappe C., Food Addit. Contam., 17, 241— 260 (2000). 8) Ashida H., Fukuda I., Yamashita T., Kanazawa K., FEBS Lett., 476, 213—217 (2000). 9) Ciolino H. P., Daschner P. J., Yeh G. C., Biochem. J., 340, 715—722 (1999). 10) Ciolino H. P., Daschner P. J., Wang T. T. Y., Yeh G. C., Biochem. Phar- macol., 56, 197—206 (1998). 11) Cilono H. P., Daschner P. J., Yeh G. C., Cancer Res., 58, 5707—5712 (1998). 12) Casper R. F., Quesne M., Rogers I. M., Shirota T., Jolivet A., Milgrom E., Savouret J.-F., Mol. Pharmacol., 56, 784—790 (1999). 13) Amakura Y., Tsutsumi T., Nakamura M., Kitagawa H., Fujino J., Sasaki K., Yoshida T., Toyoda M., Biol. Pharm. Bull., 25, 272—274 (2002). 14) Amakura Y., Tsutsumi T., Nakamura M., Kitagawa H., Fujino J., Sasaki K., Toyoda M., Yoshida T., Maitani T., Biol. Pharm. Bull., 26, 532—539 (2003). 15) Okuda T., Yoshida T., Hatano T., Prog. Chem. Org. Nat. Prod., 66, 1— 102 (1995). 16) Hatano T., Miyatake H., Natsume M., Osakabe N., Takizawa T., Ito H., Yoshida T., Phytochemistry, 59, 749—758 (2002). 17) Ito H., Miyake M., Nishitani E., Mori K., Hatano T., Okuda T., Fig. 3. Molecular Models of Constituents with Inhibitory Effect on AhR- Konoshima T., Takasaki M., Kozuka M., Mukainaka T., Tokuda H., Activation Induced by TCDD Obtained as MM2-Minimized Structure Nishino H., Yoshida T., Cancer Lett., 143, 5—13 (1999). 18) Taniguchi S., Takeda S., Yabu-uchi R., Yoshida T., Yazaki K., Phyto- (a) TCDD; (b) apigenin; (c) resveratrol; (d) emodin; (e) brevifolincarboxylic acid; (f) piperine. Energy-minimized structures were calculated with molecular mechanics using chemistry, 46, 279—282 (1997). MM2 in CS Chem 3D. 19) Akiyama T., Yamada M., Yamada T., Maitani T., Biosci. Biotechnol. Biochem., 64, 2246—2249 (2000). 20) Wheelock G. D., Hurst K. R., Babish J. G., Toxicol. Methods, 6, 41— Acknowledgments We would like to thank Dr. Takumi 50 (1996). 21) Kobayashi Y., Lundquist A., Uechi T., Ashieda K., Sasaki K., Hughes Akiyama of our institute for donating some of the standard B., Kaise T., Organohalogen Comp., 58, 337—340 (2002). chemicals. The present work was supported in part by a grant 22) Okuda T., Yoshida T., Hatano T., J. Chem. Soc. Perkin Trans. 1, 1982, from the Ministry of Health, Labour and Welfare of Japan. 9—14 (1982). 23) Haslam E., “Plant Polyphenols,” Cambridge University Press, Cam- bridge, 1989. REFERENCES 24) Rimm E. B., Katan M. B., Ascherio A., Stamper M. J., Willett W. C., Ann. Intern. Med., 125, 384—389 (1996). 1) Landers J. P., Biochem. J., 276, 273—287 (1991). 25) Manach C., Morand C., Crespy V., Demigné C., Texier O., Régérat F., 2) Poland A., Knutson J. C., Ann. Rev. Pharmacol. Toxicol., 22, 517— Rémésy C., FEBS Lett., 426, 331—336 (1998). 554 (1982). 26) Hollman P. C. H., Gaag M. V. D., Mengelers M. J. B., Van Trijp J. M. 3) Fujii-Kuriyama Y., Farumashia (Japan), 34, 431—435 (1998). P., De Vries J. H. M., Katan M. B., Free Radical Biol. Med., 21, 703— 4) Toyoda M., Uchibe H., Yanagi T., Kono Y., Hore T., Iida T., J. Food 707 (1996). Hyg. Soc. Jpn., 40, 98—110 (1998). 27) Watanabe S., Yamaguchi M., Sobue T., Takahashi T., Miura T., Arai 5) Fries G. F., J. Anim. Sci., 73, 1639—1650 (1990). Y., Mazur W., Wähälä K., Adlercreutz H., J. Nutr., 128, 1710—1715 6) Travis C. C., Hattermer-Frey H. A., Sci. Total Environ., 104, 97—127 (1998). (1991).