Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Resveratrol (trans-3',4',5-trihydroxystilbene) is a phytochemical found in grapes and grape products such as wine, as well as other food items, and thus is a constituent of human diets (Soleas et al., 1997). Resveratrol exerts potent antioxidant and anti-inflammatory activities, which may be responsible for the beneficial effects of wine consumption in the prevention of cardiovascular disease (Constant, 1997). Recently, resveratrol was shown to inhibit the formation of preneoplastic lesions in mammary glands and block tumorigenesis in a two-stage model of skin cancer in mice exposed to the model aryl hydrocarbon (AH) dimethylbenz[a]anthracene (Jang et al., 1997). Potent carcinogens, AHs are activated to genotoxic metabolites by the cytochrome P-4501A (CYP1A) family of enzymes, which catalyze the oxidation of the hydrocarbon to a variety of diol epoxides, which are potent binders of DNA (Szeliga and Dipple, 1998; Guengerich and Shimada, 1998). The induction of CYP1A1 gene transcription by AHs begins by their binding and activating the aryl hydrocarbon receptor (AHR), a cytosolic protein that, on ligand binding, translocates to the nucleus and with its partner, the aryl hydrocarbon nuclear translocator, interacts with the promoter of the CYP1A1 gene (Rowlands and Gustafsson, 1997). This results in an up-regulation of transcription and a subsequent increase in CYP1A1 mRNA and enzyme levels. Inhibition of AHR-mediated signal transduction or CYP1A enzyme activity may be important mechanisms in the chemopreventive effect of several dietary and synthetic compounds that have been associated with a reduced risk of carcinogenesis (Wattenberg, 1996; Ciolino et al., 1998a; Singh et al., 1998).

Recently, we demonstrated that resveratrol inhibits the induction of CYP1A1 transcription by the halogenated hydrocarbon tetrachlorodibenzo-p-dioxin (TCDD) (Ciolino et al., 1998b). TCDD is the most potent known ligand of the AHR, but there are several other AHR ligands that are important carcinogens. Unlike TCDD, these carcinogens, the AHs, are activated by the CYP1A enzymes to genotoxic metabolites. The goals of the current study were to determine whether resveratrol also inhibits AHR-mediated signal transduction caused by AHs, to examine the effect of resveratrol on AH metabolism, and to determine whether resveratrol directly inhibits CYP1A enzyme activity. We therefore examined the effect of resveratrol on CYP1A enzymatic activity and CYP1A1 expression induced by the AH benzo[a]pyrene (B[a]P) in HepG2 human hepatoma cells. B[a]P is a potent environmental carcinogen to which humans are exposed in cigarette smoke, industrial byproducts, and cooked meat (Lijinsky, 1991; Petruzzelli et al., 1998). We demonstrate that resveratrol inhibits B[a]P metabolism by directly inhibiting CYP1A enzymatic activity. Resveratrol also inhibits the B[a]P-induced increase in CYP1A1 expression, thus preventing an increase in carcinogen bioactivation capacity, by blocking the binding of the activated AHR with the CYP1A1 promoter. Resveratrol was equally effective in inhibiting CYP1A1 expression and enzyme activity of another AH, the mammary carcinogen dimethylbenzanthracene (DMBA), in MCF-7 human mammary epithelial carcinoma cells. Thus, resveratrol may be a potent chemopreventive agent in vivo as a result of its inhibitory effects on both CYP1A enzymatic activity and AHR-mediated signal transduction.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. HepG2 and MCF-7 cells were obtained from the American Type Culture Collection (Rockville, MD). RPMI 1640, glutamine, FBS, trypsin/EDTA, and PBS were purchased from BioFluids (Rockville, MD). B[a]P, DMBA, resveratrol, HEPES, EDTA, dithiothreitol (DTT), glycerol, poly(deoxyinosinic/deoxycytidylic acid) [poly(dI/dC)], sodium molybdate, ethoxyresorufin (ETRF), resorufin, Tris · HCl, and protease inhibitors were purchased from Sigma Chemical Co. (St. Louis, MO). [³H]B[a]P (specific activity, 74 Ci/mmol) and [³H]resveratrol (specific activity, 7.2 Ci/mmol) were obtained from Moravek (Brea, CA). [³²P]dCTP and [³²P]dATP were obtained from DuPont-New England Nuclear (Boston, MA). Reverse transcription-polymerase chain reaction (RT-PCR) was performed with a kit from Stratagene (La Jolla, CA). Tris/borate/EDTA (TBE) gels, TBE running buffer, and high-density sample buffer were obtained from Novex (San Diego, CA). LipofectAMINE and TRIzol reagent were purchased from Life Technologies (Rockville, MD). The chloramphenicol acetyltransferase (CAT) enzyme-linked immunosorbent assay kit was obtained from Boehringer Mannheim (Indianapolis, IN). Primers for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) PCR and pCMV vector containing -galactosidase (-Gal) were obtained from Clontech (Palo Alto, CA). Aquasure was obtained from Packard (Meriden, CT). The Bradford protein assay kit was purchased from Bio-Rad (Hercules, CA). Resveratrol and B[a]P were dissolved in dimethyl sulfoxide (DMSO), aliquoted, and stored at 80°C. Polyclonal antibody to the AHR was the kind gift of Dr. Alan Poland (Centers for Disease Control/National Institute for Occupational Safety and Health, Morgantown, WV).

Cell Culture. HepG2 and MCF-7 cells were grown in RPMI 1640 supplemented with 2 mM glutamine and 10% FBS (growth medium). Cells were subcultured weekly using 0.25% trypsin/0.05% EDTA. All experiments were carried out at 37°C and 5% CO₂ on confluent cells in growth medium except where noted.

B[a]P Metabolism. HepG2 cells were treated with 1 µM B[a]P for 24 h. Microsomes were prepared as described previously (Ciolino et al., 1998a). Microsomes (400 µg) were incubated with 5 nM [³H]B[a]P and 1 mM NADPH in the presence of DMSO or resveratrol for 10 min, and the amount of [³H]B[a]P was metabolized to water-soluble compounds was determined as described by Chae et al. (1991). As a negative control, microsomes isolated from untreated HepG2 cells were tested in a similar manner: there was no [³H]B[a]P metabolized by these microsomes.

Microsomal CYP1A1/1A2 Activity. CYP1A1/1A2 activity was determined by ethoxyresorufin-O-deethylase (EROD) activity assay in the following manner: 10 µg of microsomes were brought up to 100 µl with PBS, pH 7.2. ETRF (400 nM) was added, along with DMSO or the indicated concentrations of resveratrol. The reaction was initiated by the addition of 250 µM NADPH. The reaction mixture was transferred to a 96-well plate, and EROD activity was determined in a CytoFluor II multiwell fluorescence plate reader (PerSeptive Biosystems, Framingham, MA), with an excitation wavelength of 530 nm and emission at 590 nm. For Fig. 3B, 250 µM NADPH and 100 to 1600 nM ETRF were added to 3 ml of PBS, pH 7.2. Aliquots of 410 µl were removed to which DMSO or the indicated concentration of resveratrol was added. The reaction was initiated by the addition of 45 µg of microsomal protein (final volume, 450 µl) and gently vortexed. Four 100-µl aliquots (10 µg/assay) of each were removed and placed in a 96-well plate, and EROD activity was determined. A standard curve was constructed using resorufin.

CYP1A1/1A2 Enzyme Activity in Intact Cells. Confluent cells in 24-well plates were treated with 1 ml of growth medium containing 100 nM B[a]P for HepG2 or 500 nM DMBA for MCF-7 for 9 h in the presence of DMSO or the indicated concentrations of resveratrol. At the end of the incubation, the medium was removed, and the wells were washed two times with fresh growth medium. EROD activity was determined in intact cells as described by Kennedy and Jones (1994) using 5 µM ETRF in growth medium as a substrate in the presence of 1.5 mM salicylamide to inhibit conjugating enzymes. The assay was carried out at 37°C. The fluorescence of resorufin generated from the conversion of ETRF by CYP1A1/1A2 was measured every 10 min for 60 min as described above.

RT-PCR. Confluent cells were treated with DMSO (control), 100 nM B[a]P (HepG2), or 500 nM DMBA (MCF-7) in the presence of DMSO or resveratrol for 6 h. The cells were washed twice with PBS, and total RNA was isolated using TRIzol reagent. cDNA was synthesized from 10 µg of total RNA using a Stratagene RT-PCR kit as instructed. Semiquantitative PCR for CYP1A1, CYP1A2, and GAPDH was performed in the presence of 1.5 µCi of [³²P]dATP. Hot start was performed by premixing AmpliTaq (Perkin-Elmer, Foster City, CA) with anti-Taq antibody (Clontech). Primer sequences and PCR conditions for CYP1A1 were as described by Dohr et al. (1995), and those for CYP1A2 were as described by Chung and Bresnick (1994). Primer sequences for GAPDH were from Clontech, and PCR was carried out as directed. The optimum cycle number that fell within the exponential range of response for CYP1A1 (22 cycles), CYP1A2 (25 cycles), or GAPDH (17 cycles) was used. After PCR, 5 µl of high-density sample buffer was added to the samples, and they were subjected to electrophoresis on a 10% TBE gel in 1× TBE running buffer. The gel was dried, and the results were visualized and quantified on a Bio-Rad GS-363 Molecular Imaging System (Hercules, CA). Graphs of the resulting data were generated by normalizing CYP1A1 and CYP1A2 to GAPDH.

CAT/-Gal Assays. Cells were plated at 60,000 cells/well in 24-well plates. After 24 h, the cells were transiently transfected with 12.0 µg of a CAT reporter vector containing the full-length rat CYP1A1 promoter (Sogawa et al., 1986) using LipofectAMINE as directed. To control for transfection efficiency, the cells were cotransfected with 1.0 µg of pCMV vector containing -Gal. After an additional 24 h, the cells were treated with DMSO (control), 250 nM B[a]P (HepG2), or 500 nM DMBA (MCF-7) in the presence of DMSO (control) or resveratrol for 6 h. The amount of CAT transcription was determined using an enzyme-linked immunosorbent assay as directed. Transcription of -Gal was determined by measuring enzyme activity according to the method of Rosenthal (1987). The amount of CAT transcription was normalized to -Gal transcription.

Electrophoretic Mobility Shift Assay (EMSA) for AHR. HepG2 cells were treated with DMSO, 1 µM B[a]P, or 1 µM B[a]P in the presence of 5 or 10 µM resveratrol in growth media for 2 h. Nuclear protein was isolated, and EMSA was performed according to the method of Denison et al. (1988). Synthetic oligonucleotides containing the AHR-binding site of the xenobiotic response element (XRE) of the CYP1A1 promoter (Chen and Tukey, 1996) were labeled with [³²P]dCTP. The binding reactions were carried out for 30 min and contained 5 µg of nuclear protein, 1 µg of poly(dI/dC), and ~50,000 cpm of labeled probe in a final volume of 20 µl of binding buffer (25 mM Tris, pH 7.9, 50 mM KCl, 1 mM MgCl₂, 1.5 mM EDTA, 0.5 mM DTT, and 5% glycerol). To determine specificity of binding to the oligonucleotide, a 200-fold excess of unlabeled XRE or 0.9 µg of a polyclonal antibody to human AHR was added to extract from B[a]P-treated cells. DNA-protein complexes were separated under nondenaturing conditions on a 4% polyacrylamide gel using 0.5× TBE (45 mM Tris, pH 7.5, 45 mM boric acid, 2 mM EDTA) as a running buffer. The gels were dried, and the DNA-protein complexes were visualized on a Bio-Rad GS-363 Molecular Imaging System.

AHR Ligand-Binding Assay. HepG2 cells were grown to confluence in 175-cm² flasks. The cells were washed once in PBS, harvested by trypsination, and pelleted by centrifugation at 800g for 10 min at 4°C. The pellet was washed once in cold PBS, repelleted as above, and resuspended in cold buffer (25 mM HEPES, 1 mM EDTA, 1 mM DTT, 20 mM sodium molybdate, and 10% glycerol, pH 7.4), with protease inhibitors as described above. The cells were homogenized by 30 strokes with a Dounce glass homogenizer on ice, and the homogenate was centrifuged at 100,000g for 60 min at 4°C. The supernatant (cytosol) was removed, and protein was assayed according to the method of Bradford (1976). The cytosol was aliquoted and stored at 80°C. Specific binding to the AHR was measured by sucrose density gradient centrifugation as described by Raha et al. (1990). One mg of cytosolic protein was incubated with 5 nM [³H]B[a]P in the presence of DMSO (control), 5 µM B[a]P (positive control), or 0.5 or 5 µM resveratrol in a total volume of 500 µl of the above buffer for 2 h at 4°C. Samples were applied to 5 to 30% (w/v) linear sucrose density gradients in 12 ml of Beckman Quick-Seal rotor tubes (Palo Alto, CA). The gradients were centrifuged at 4°C for 2 h at 63,000 rpm (372,000g) in a Beckman VTI-65-1 rotor. Twenty-five fractions of 7 drops each (~500 µl) were collected from the bottom of the tubes and assayed for radioactivity using Aquasure scintillation fluid. The gradients were calibrated with catalase (11S; fraction 10) and BSA (4S; fraction 20).

Statistical Analysis. Statistical analyses were performed using StatView Statistical Analysis software (SAS Institute, San Francisco, CA). Differences between group mean values were determined by a one-factor ANOVA, followed by Fisher's protected least-significant difference post hoc analysis for pairwise comparison of mean values.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Effects of Resveratrol on Metabolism of B[a]P. Incubation of [³H]B[a]P with microsomes isolated from B[a]P-induced HepG2 cells and NADPH for 10 min resulted in the almost complete conversion of B[a]P to water-soluble metabolites (data not shown). This conversion results solely from CYP1A activity because the assay mixtures lacked cofactors for phase 2 (conjugating) enzymes (Chae et al., 1991). The addition of resveratrol caused a concentration-dependent decrease in the amount of B[a]P converted to water-soluble metabolites (Fig. 1).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Effect of resveratrol on B[a]P metabolism. Microsomes from B[a]P-induced HepG2 cells were incubated for 10 min with [3H]B[a]P, and the indicated concentrations of resveratrol and the amount of water-soluble [3H]B[a]P metabolites generated was determined. n = 4 ± S.E. There was a significant decrease in [3H]B[a]P metabolism in the presence of all concentrations of resveratrol (p < .05).

Effect of Resveratrol on Activity of CYP1A. We analyzed the effect of resveratrol on the activity of CYP1A using the EROD assay, which is specific for the bioactivation capacity of the CYP1A enzyme family. To test whether resveratrol directly inhibits EROD activity, we examined the effect of resveratrol on EROD activity in microsomes isolated from induced cells. As shown in Fig. 2A, resveratrol inhibited microsomal EROD activity, with an IC₅₀ value of approximately 1 µM. Analysis of the kinetics of inhibition by double-reciprocal (Lineweaver-Burk) plot demonstrated that the K_m value of the enzyme toward the substrate was increased in the presence of resveratrol (Fig. 2B). The K_i value was calculated as 0.42 µM.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Effect of resveratrol on microsomal EROD activity. A, EROD activity in 10 µg of microsomes isolated from induced HepG2 cells was determined in the presence of the indicated concentrations of resveratrol. n = 4 ± S.E. Error bars are less than symbols. There was a significant decrease EROD activity in the presence of all concentrations of resveratrol (p < .05). B, EROD activity in 10 µg of microsomes was determined in the presence of the indicated concentrations of ETRF and 0, 0.5, or 1 µM resveratrol. n = 4.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Effect of resveratrol on cellular EROD activity induced by B[a]P. HepG2 cells were incubated with 100 nM B[a]P for 9 h in the presence of the indicated concentrations of resveratrol and cellular EROD activity was determined. n = 4 ± S.E. There was a significant decrease EROD activity in the presence of all concentrations of resveratrol (p < .05).

Effect of Resveratrol on Expression of CYP1A1. The treatment of HepG2 cells with B[a]P resulted in an approximately 6-fold increase in the mRNA levels of CYP1A1 and CYP1A2, the two major carcinogen activating enzymes in these cells (Fig. 4). The increase in both CYP1A1 and CYP1A2 mRNA was inhibited by resveratrol in a concentration-dependent manner. The transcription of an AH-sensitive CAT reporter vector that contains the CYP1A1 promoter was examined. The treatment of cells transiently transfected with this vector with B[a]P caused a 5-fold increase in CAT transcription, which was inhibited by resveratrol in a concentration-dependent manner (Fig. 5).

View larger version (33K): [in this window] [in a new window]   Fig. 4.   Effect of resveratrol on CYP1A1 and CYP1A2 mRNA induced by B[a]P. HepG2 cells were treated for 6 h with DMSO (Control) or 100 nM B[a]P in the presence of DMSO or resveratrol at the indicated concentrations. Total RNA was isolated, cDNA was synthesized, and the amount of mRNA was determined by PCR. The results were visualized and quantified by phosphorimaging. For the graph, the levels of CYP1A1 and CYP1A2 were normalized to GAPDH level. n = 3 ± S.E. There was a significant decrease in CYP1A1 and CYP1A2 mRNA in cells treated with resveratrol (p < .05).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effect of resveratrol on B[a]P-induced CYP1A1 promoter-driven transcription. HepG2 cells were transiently transfected with a CAT reporter vector containing the CYP1A1 promoter, as well as a vector containing -Gal, and then treated for 6 h with DMSO or 250 nM B[a]P in the presence of DMSO or resveratrol at the indicated concentrations. The levels of CAT and -Gal were determined as described in Experimental Procedures, and the amount of CAT transcription was normalized to the amount of -Gal. n = 4 ± S.E. There was a significant decrease in CAT transcription in the presence of resveratrol (p < .05).

Mechanism of Inhibition of CYP1A1 Expression. We examined the effect of resveratrol on the interaction of the B[a]P-activated AHR with the XRE of the CYP1A1 promoter. As shown in the EMSA in Fig. 6, B[a]P caused an increase in the amount of activated nuclear AHR binding to a ³²P-labeled oligonucleotide representing the XRE of CYP1A1 promoter. The addition of unlabeled XRE to nuclear extracts of B[a]P-treated cells abolished XRE binding, whereas treatment with a polyclonal antibody to human AHR significantly reduced XRE binding, demonstrating the specificity of the gel shift. The increase in XRE binding caused by B[a]P was inhibited by resveratrol.

View larger version (54K): [in this window] [in a new window]   Fig. 6.   Effect of resveratrol on B[a]P-induced binding of the AHR to the XRE of CYP1A1. HepG2 cells were treated for 2 h with DMSO (control) or 1 µM B[a]P in the presence of DMSO or the indicated concentrations of resveratrol. Nuclear protein was isolated, and the amount of activated AHR in 5 µg of protein was measured by EMSA. For competition studies, nuclear protein from B[a]P-treated cells was preincubated with a 200-fold excess of unlabeled XRE or with 0.9 µg of a polyclonal antibody to the AHR. This experiment was repeated two times with similar result.

View larger version (29K): [in this window] [in a new window]   Fig. 7.   Effect of resveratrol on the binding of [3H]B[a]P to HepG2 cytosol. HepG2 cytosol was incubated with 5 nM [3H]B[a]P in the presence of DMSO (control), 5 µM unlabeled B[a]P, 0.5 µM (100-fold) resveratrol, or 5 µM (1000-fold) resveratrol or with 5 nM [3H]resveratrol. The specific binding of [3H]B[a]P or [3H]resveratrol was analyzed by sucrose density gradient centrifugation. This experiment was repeated three times with different cytosolic preparations with similar results. Res, resveratrol.

Effect of Resveratrol on DMBA-Induced CYP1A1 Expression in MCF-7 Cells. The expression of CYP1A1 induced by another AH, the mammary carcinogen DMBA, was examined in MCF-7 human mammary epithelial carcinoma cells. The treatment of MCF-7 cells with DMBA resulted in an increase in EROD activity from undetectable levels to 1.20 ± 0.14 pmol/min/well. Resveratrol inhibited DMBA-induced EROD activity in intact MCF-7 cells, with an IC₅₀ value of approximately 500 nM (Fig. 8A). Resveratrol also inhibited the induction of CYP1A1 mRNA by DMBA (Fig. 8B) and inhibited CYP1A1 promoter-controlled transcription (Fig. 8C).

View larger version (19K): [in this window] [in a new window]   Fig. 8.   Effect of resveratrol on DMBA-induced CYP1A1 expression in MCF-7 cells. A, confluent MCF-7 cells were incubated for 9 h with 500 nM DMBA in the presence of DMSO or the indicated concentrations of resveratrol and the EROD activity was determined. n = 4 ± S.E. B, MCF-7 cells transfected with the CYP1A1 promoter containing CAT reporter vector were treated for 6 h with 500 nM DMBA and the indicated concentrations of resveratrol. C, MCF-7 cells were incubated with 500 nM DMBA and the indicated concentrations of resveratrol for 6 h, and the amount of CYP1A1 mRNA and GAPDH mRNA was determined by RT-PCR. n = 3 ± S.E. There was a significant decrease in all parameters in the presence of resveratrol (p < .05).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

An AH found in a variety of environmental sources, B[a]P is a potent carcinogen in animal models (Phillips, 1983). It is regarded as a human carcinogen because B[a]P-DNA adducts have been detected in humans (Petruzzelli et al., 1998) and B[a]P metabolites have been shown to bind to specific DNA residues in the p53 gene, known to be "hot spots" of mutation in human carcinomas (Puisieux et al., 1991). B[a]P is metabolized to genotoxic derivatives by the action of CYP1A enzymes, which are induced by the activation of the DNA-binding capacity of the AHR for the CYP1A1 gene promoter by B[a]P. It may be argued that the up-regulation of CYP1A1 by the AHR is a protective mechanism, in that the AH metabolites generated by CYP1A activity are substrates of many detoxifying enzymes, such as UDP-glucuronosyltransferase (Grove et al., 1997) and glutathione-S-transferase (Xia et al., 1998). However, recent experiments using AHR knockout mice have indicated that the lack of the AHR, and thus an absence of CYP1A1 induction, confers protection from the deleterious effects of B[a]P (Dertinger et al., 1998). This supports previous findings that demonstrated a positive correlation between high levels of AH hydroxylase activity and increased cancer risk (Kellermann et al., 1973; Kouri et al., 1982). Induction of CYP1A1 by carcinogens via the AHR, therefore, appears to be detrimental to the organism or cell, and consequently inhibition of CYP1A1 expression and CYP1A1 activity has much promise as a mechanism of chemoprevention.

Epidemiological studies have shown that diets rich in fruits and vegetables are associated with a reduced risk of cancer (Steinmetz and Potter, 1996). Much attention has accordingly been focused on discerning the mechanisms by which phytochemicals inhibit carcinogenesis. One such phytochemical, resveratrol, has been shown to inhibit AH-induced carcinogenesis in mice (Jang et al., 1997). In the present study, we therefore examined the effect of resveratrol on carcinogen activation and the carcinogen activation pathway in HepG2 cells. These cells are derived from human liver and have been extensively used in studies of B[a]P metabolism (Christou et al., 1992; Polzer et al., 1995) and of the AHR (Kress and Greenlee, 1997; Long et al., 1998).

Resveratrol inhibited the conversion of B[a]P to water-soluble metabolites by microsomes isolated from B[a]P-induced HepG2 cells (Fig. 1). This demonstrates that resveratrol must directly inhibit enzyme activity. We therefore examined the effect of resveratrol on microsomal CYP1A activity by using the EROD assay. Resveratrol inhibited microsomal EROD activity in a concentration-dependent manner (Fig. 2A). An analysis of the kinetics of enzyme inhibition by double-reciprocal (Lineweaver-Burk) plot showed that this inhibition was of a competitive type because the K_m value for the substrate was increased in the presence of resveratrol, but the V_max value remained unchanged (Fig. 2B). To test whether resveratrol would also inhibit cellular CYP1A activity, we measured the EROD activity in intact HepG2 cells. The treatment of cells with B[a]P caused an increase in EROD activity that was completely abolished by resveratrol (Fig. 3). This demonstrates that resveratrol is capable of entering the cells and interacting with the enzyme in situ.

Although the decrease in B[a]P-induced CYP1A1 enzymatic activity in intact cells (Fig. 3) may be due to the direct inhibitory effect of resveratrol toward CYP1A enzymatic activity, it may also result from inhibition of AHR-mediated transcriptional activation. We recently demonstrated that resveratrol inhibits the increase in CYP1A1 expression caused by the halogenated hydrocarbon TCDD, a ligand of the AHR (Ciolino et al., 1998b). We therefore examined the effect of resveratrol on CYP1A1 expression induced by the AH B[a]P. The increase in CYP1A1 mRNA in HepG2 cells caused by treatment with B[a]P was inhibited by resveratrol (Fig. 4). To determine whether this inhibition occurred at the transcriptional level, we transfected cells with an AH-responsive CAT reporter vector containing the CYP1A1 promoter. Treatment of transfected cells with B[a]P resulted in an increase in CAT transcription that was inhibited by resveratrol (Fig. 5). The decrease in CYP1A1 mRNA and CYP1A1-promoter driven transcription clearly demonstrates that resveratrol inhibits CYP1A1 expression induced by B[a]P. Resveratrol also inhibited the increase in CYP1A2 mRNA caused by B[a]P (Fig. 5). Thus, the inhibition of cellular EROD activity shown in Fig. 3 may result not only from a direct inhibitory effect but also from inhibition of CYP1A1 transcription.

To determine the mechanism of this transcriptional repression, we carried out gel shift assays to measure the amount of AHR that had been transformed to its nuclear, DNA-binding form. AHR, when activated by a ligand such as B[a]P, interacts with the XRE of the CYP1A1 promoter, inducing transcription. As shown in Fig. 6, resveratrol inhibits the binding of B[a]P-activated nuclear AHR to ³²P-labeled XRE. Thus, inhibition of CYP1A1 transcription by resveratrol is due to an inhibition of AHR-XRE interaction. This is in agreement with our previous results using TCDD to activate the AHR (Ciolino et al., 1998b). This demonstrates that resveratrol inhibits AHR-mediated signal transduction by preventing the binding of the activated receptor to the gene promoter.

Several inhibitors of CYP1A1 transcription are known to function by binding to the ligand binding site of the AHR, blocking the binding of other ligands and thereby preventing AHR activation (Santostefano et al., 1993; Ciolino et al., 1998a). To determine whether the inhibition of AHR-XRE binding by resveratrol shown in Fig. 6 results from blocking binding of B[a]P to the AHR, HepG2 cytosol was incubated with [³H]B[a]P in the presence of excess resveratrol and the specific binding was measured using sucrose density gradient centrifugation. As shown in Fig. 7, resveratrol caused a modest decrease in [³H]B[a]P binding to two peaks in the gradient. This was reproducible in several different cytosolic preparations (data not shown). However, even greater amounts of resveratrol (>1000-fold excess) did not result in a greater inhibition of binding, and [³H]B[a]P binding was not reduced to the level present in an excess of unlabeled B[a]P. Furthermore, incubation of cytosol with [³H]resveratrol demonstrated that there was no binding of resveratrol to any cytosolic component, indicating that resveratrol does not itself bind to the receptor (Fig. 7). The complete inhibition of B[a]P-induced CYP1A1 transcription and AHR-XRE interaction by resveratrol demonstrated in Figs. 4, 5, and 6 occurred at concentrations of resveratrol, relative to B[a]P, much less than that required to cause partial inhibition of B[a]P binding. Thus, the inhibitory effect of resveratrol on AHR activation, and hence on CYP1A1 expression, although it may partially result from an inhibition of B[a]P-AHR binding, would appear to be primarily through an indirect mechanism. This is consistent with our previous study (Ciolino et al., 1998b), which demonstrated that resveratrol does not inhibit the binding of TCDD to cytosolic AHR and that resveratrol inhibits the basal level of CYP1A1 transcription in the absence of exogenous ligand. The nature of this indirect mechanism of AHR inhibition is currently under study.

Our previous study (Ciolino et al., 1998b) and current studies were carried out in HepG2 cells, a cell line derived from human liver. To test whether the inhibitory effect of resveratrol on CYP1A1 transcription would occur in another cell line, we carried out several experiments in MCF-7 cells. The AHR pathway has also been well characterized in these cells (Moore et al., 1994; Dohr et al., 1995) and is similar to that found in normal human mammary cells in vitro (Larsen et al., 1998). Because these cells are derived from human mammary epithelial cells, we used the mammary carcinogen DMBA as a CYP1A1 inducer, allowing us to examine not only another cell line but also another AHR ligand. As shown in Fig. 8A-C, compared with its effect on B[a]P-induced CYP1A1 expression in HepG2 cells, resveratrol is even more effective an inhibitor of DMBA-induced CYP1A1 enzyme activity and CYP1A1 mRNA and equally effective at inhibiting DMBA-induced CYP1A1 transcription in MCF-7 cells. Thus, resveratrol is a potent inhibitor of CYP1A1 expression induced by different AHs in different cells types. Such an inhibitory effect may be responsible for the chemopreventive activity of resveratrol toward DMBA-induced mammary neoplastic changes as reported by Jang et al. (1997).

These experiments demonstrate that resveratrol affects the carcinogen activation pathway in vitro at two levels: it directly inhibits the activity of carcinogen activation enzymes, and it inhibits the increase in CYP1A1 expression caused by AHs. In concert, these two activities may be responsible for the chemopreventive effect of resveratrol toward AH-induced carcinogenesis.