Inhibition of catechol O-methyltransferase-catalyzed O-methylation of 2- and 4-hydroxyestradiol by quercetin. Possible role in estradiol-induced tumorigenesis.

Catecholestrogens have been postulated to mediate the induction of kidney tumors by estradiol in male Syrian hamsters. In this study, we examined the mechanism of inhibition by quercetin of the catechol O-methyltransferase-catalyzed O-methylation of catecholestrogens as a basis for the previously reported enhancement of estradiol-induced tumorigenesis by this flavonoid. In hamsters treated with 50 micrograms of [6,7-3H]estradiol, quercetin increased concentrations of 2- and 4-hydroxyestradiol in kidney by 80 and 59%, respectively. In animals treated with two 10-mg estradiol implants, quercetin also decreased by 63-65% the urinary excretion of 2- and 4-hydroxyestradiol monomethyl ethers. Taken together, these results demonstrate the in vivo inhibition of the O-methylation of catecholestrogens by quercetin. S-Adenosyl-L-homocysteine, produced by the methylation of catecholestrogens, noncompetitively inhibited the O-methylation of 2- and 4-hydroxyestradiol by hamster kidney cytosolic catechol O-methyltransferase (IC50 approximately 10-14 microM). Due to the rapid O-methylation of quercetin itself, quercetin decreased renal concentrations of S-adenosyl-L-methionine by approximately 25% in control or estradiol-treated hamsters and increased concentrations of S-adenosyl-L-homocysteine by 5-15 nmol/g of wet tissue, which was estimated to cause a 30-70% inhibition of the enzymatic O-methylation of catecholestrogens. Quercetin or fisetin (a structural analog) inhibited the O-methylation of 2- and 4-hydroxyestradiol by a competitive plus noncompetitive mechanism (IC50 approximately 2-5 microM). These results suggest that the in vivo O-methylation of catecholestrogens is inhibited more by S-adenosyl-L-homocysteine than by quercetin. The accumulation of 2- and 4-hydroxyestradiol during co-administration of estradiol and quercetin may enhance metabolic redox cycling of catecholestrogens and thus estradiol-induced kidney tumorigenesis.

The chronic administration of natural or synthetic estrogens such as estradiol (E 2 ) 1 or diethylstilbestrol to male Syrian hamsters induces kidney tumors with an incidence approaching 100% (1). Estrogens are complete carcinogens, i.e. tumor initiators and promoters in this animal model, which is thus useful for studying the multiple roles of estrogens in the development of hormone-associated cancers. The metabolic redox cycling of CE metabolites or of diethylstilbestrol between their hydroquinone and quinone forms has been established as a mechanism of metabolic activation (2), because this process generates potentially mutagenic free radicals in addition to chemically reactive estrogen semiquinone/quinone intermediates (3,4,5). These reactive chemical species mediate damage to DNA and other cellular components in the target organ of carcinogenesis in analogy to some classical chemical carcinogens (4 -8) and thus may participate in the initiation of tumors. This mechanistic hypothesis has been probed by the co-administration of quercetin to hamsters treated chronically with E 2 (9), because this flavonoid is a good inhibitor of the in vitro O-methylation of CE by purified porcine liver COMT (9). Quercetin itself is also an excellent substrate for COMT (10). The inhibition of the enzymatic conversion of CE to methoxyestrogens in hamster kidney by quercetin was expected to result in an accumulation of CE, the substrates for metabolic redox cycling, and thus to enhance the induction of renal tumors by E 2 . Consistent with this hypothesis, a diet supplemented with quercetin significantly enhanced the severity of E 2 -induced tumorigenesis (9), but it did not enhance (rather, slightly decreased) the induction of tumors by diethylstilbestrol, 2 which may directly undergo metabolic redox cycling without initial conversion to catechol metabolites (2). The enhancement of E 2 -induced kidney tumorigenesis by quercetin also contrasts sharply with the inhibition by this flavonoid of tumors in several other animal models. For instance, this flavonoid decreases the incidence of 7,12-dimethylbenz(a)anthracene-and N-nitrosomethylurea-induced mammary tumors in rats (11) and azoxymethanol-induced colonic neoplasms in mice (12). Taken together, these observations suggest that the selective enhancement by quercetin of E 2 (but not diethylstilbestrol)induced kidney tumorigenesis in hamsters may be due to a specific potentiating effect rather than to a nonspecific cocarcinogenesis by this flavonoid. The inhibition of renal Omethylation of 2-and 4-OH-E 2 , which may accumulate in kidneys of hamsters treated with E 2 and quercetin, may increase the concentrations of substrates for metabolic redox cycling of CE (2), increase the production of potentially mutagenic free radicals (3)(4)(5), and thereby potentiate E 2 -induced renal tumorigenesis.
To evaluate the in vivo inhibition of the O-methylation of CE metabolites by quercetin, we first determined the concentrations of unmetabolized CE in kidney and of methoxyestrogens (the O-methylated CE) in urine of hamsters treated with E 2 alone or in combination with quercetin. Because of the rapid metabolic O-methylation of quercetin itself (10), we examined the effect of this flavonoid on the pools of SAM, the cofactor and methyl donor in the O-methylation of CE, and of SAH, the demethylated product of SAM (10). Kinetic analyses of the effect of SAH, a potential noncompetitive inhibitor, on the COMT-catalyzed O-methylation of CE were carried out with a cytosolic COMT preparation from hamster kidney to evaluate the contribution of increased tissue levels of SAH to the in vivo inhibition of O-methylation of CE. These kinetic studies were also expected to assist in understanding the nature of the noncompetitive component of inhibition of the O-methylation of CE by quercetin as observed previously (9). Finally, we have also studied the in vitro inhibitory effect of quercetin and fisetin on the O-methylation of CE catalyzed by cytosolic COMT of hamster kidney and also determined the circulating and tissue levels of unmetabolized quercetin available as an inhibitor.
Our study demonstrates that administration of quercetin to hamsters inhibited the O-methylation of CE metabolites in kidney and increased renal concentrations of these reactive estrogen metabolites. The correlation of increased concentrations of CE metabolites in kidney of quercetin-treated hamsters with increased severity of E 2 -induced renal tumorigenesis observed previously (9) supports the postulated role of these reactive estrogen metabolites in the induction of hormonal cancers.
Animals-2-4-month-old male Syrian hamsters, purchased from Harlan Sprague-Dawley (Houston, TX), were housed in a facility accredited by the American Association for Accreditation of Laboratory Animal Care and had free access to rodent chow and water throughout the experiment. The animals were allowed to acclimatize for at least one week prior to any experimentation.
Preparation of Hamster Kidney Cytosolic Fractions-All procedures were carried out at 0 -4°C. Kidneys from 2-month-old male Syrian hamsters were homogenized in 1.4% potassium chloride solution containing 10 mM EDTA, pH 7.4. Tissue homogenates were centrifuged at 9,000 ϫ g for 10 min, and supernatants were pooled and filtered through two layers of cheesecloth to remove lipid clots. The filtrates were then recentrifuged at 105,000 ϫ g (4°C) for 60 min, and the supernatant cytosolic fractions were filtered (0.45-m pore size). The proteins in the filtrate were precipitated by slowly adding ice-cold ethanol to a final concentration of 80%. The protein precipitates were collected by centrifugation at 9,000 ϫ g for 10 min, and then resuspended in 10 mM Tris-HCl (pH 7.4) to a protein concentration of 2 mg/ml. Aliquots of these cytosolic preparations were stored at Ϫ80°C.
Inhibition of COMT-catalyzed O-Methylation of CE by Quercetin and SAH-The COMT-catalyzed O-methylation of CE was carried out as described previously (10,13,14). The reaction mixture consisted of 250 -500 g of cytosolic protein from hamster kidney, 1.2 mM MgCl 2 , 200 M SAM iodide (containing 0.5 Ci of [methyl-3 H]SAM), 1 mM dithiothreitol, and varying concentrations of CE in a final volume of 1.0 ml of Tris-HCl buffer (50 mM, pH 7.4). The reaction was started by addition of cytosolic protein of hamster kidney, and carried out at 37°C for 30 min. The reaction was arrested by rapidly cooling to ice temperatures. The reaction mixture was then immediately extracted with 7 ml of ice-cold n-heptane. After centrifugation at 1000 ϫ g for 10 min, 3-ml aliquots of the organic extracts were analyzed for radioactivity content by liquid scintillation counting (Beckman Instruments, model LS 5000TD).
Urinary Concentrations of Methoxyestrogen Metabolites-2-4-monthold male Syrian hamsters received for 1 week either a control diet or a diet supplemented with 3% quercetin (by weight). The animals were then treated with two subcutaneous implants of 10 mg of E 2 . Portions (100 l) of the first 48-h urine samples following E 2 implantation were collected in metabolic cages (purchased from Lab Products Inc., Maywood, NJ) and were extracted 3 times with 7 ml of diethyl ether for the determination of concentrations of unconjugated methoxyestrogens. For the determination of total methoxyestrogen pools (unconjugated and conjugated metabolites), 100-l aliquots of urine were first hydrolyzed at 37°C for 24 h by ␤-glucuronidase (10,000 units) and sulfatase (2000 units) in a final volume of 0.5 ml of sodium acetate buffer (0.5 M, pH 6.0). The enzymatic hydrolyses were arrested by the addition of 1 ml of 1 M citric acid and 0.1 ml of 5 N HCl. The mixtures were then extracted 3 times with 7 ml of diethyl ether. The combined ether extracts were evaporated to dryness under a stream of nitrogen gas and then assayed by gas chromatography as described previously (15,16).
Tissue Concentrations of 2-OH-E 2 and 4-OH-E 2 -To avoid variations caused by ingestions of differing amounts of quercetin by individual hamsters, 30 mg of quercetin (in 2 ml of syrup) was administered to each hamster (8 weeks old) by intragastric intubation. Control animals received 2 ml of syrup (vehicle). 15 min after quercetin or vehicle administration, each hamster received an intraperitoneal injection of 50 g of E 2 (in 50 l of corn oil, containing 100 Ci of [6,7-3 H]E 2 ). The animals were decapitated 1 h after E 2 injection, and the kidneys and livers were immediately removed and washed 2-3 times in ice-cold normal saline solution. Two kidneys or 0.4 -0.6 g of liver tissues were weighed and homogenized for 1-2 min in 3 ml of 0.2 N HCl containing 5 mM ascorbic acid. The homogenates were extracted 3 times with 10 ml of ethyl acetate saturated with 0.2 N HCl. The pooled extracts were treated with 1 g of anhydrous sodium sulfate for 30 min to remove any water component, transferred to another container, and dried under a stream of nitrogen gas. After the dried extracts were redissolved in 200 l of methanol, 800 l of 100 mM Tris base buffer (pH ϳ8.3) and 25 mg of neutral alumina were added to adsorb CE metabolites. The adsorbed CE were eluted with 0.3 N HCl, extracted with ethyl acetate (saturated with 0.2 N HCl), and separated on thin-layer chromatographic plates as described previously (17,18,19). The overall extraction efficiency (approximately 40%) was determined based on the extraction efficiency of known amounts of 14 C-labeled 2-OH-E 2 and 4-OH-E 2 added to the tissues from untreated animals.
Tissue Concentrations of SAM and SAH-Tissue concentrations of SAM and SAH were determined according to procedures described by Chun et al. (20). Briefly, kidneys were immediately removed from decapitated animals, rinsed in cold saline and blotted on filter paper. About 1 g of freshly excised kidney tissue was homogenized in 1.5 M HCl solution (1:4 (w:v)). After centrifugation at 3000 ϫ g for 15 min, the deproteinized supernatant was chromatographed using a Dowex 50 (H ϩ ) (resin bed, 4 ϫ 0.5 cm) column pre-equilibrated with 0.1 M HCl to adsorb SAM and SAH. SAM and SAH were then recovered by elution of the column with 10 ml of 6 M HCl. Thiodiglycol (20 l) was added to the 6 M HCl eluate prior to evaporation under reduced pressure. Samples were then redissolved in 1 ml of water and analyzed by HPLC as described previously (20).
Tissue Concentrations of Quercetin-4-week-old male hamsters received a diet supplemented with 3% quercetin for 2 weeks or 6.5 months. Blood samples (1-3 ml), obtained by cardiac puncture of animals anesthetized with CO 2 , were centrifuged at approximately 3000 ϫ g for 10 min. Aliquots (300 l) of supernatant plasma were transferred to tubes containing 1 ml of 0.2 M Tris base buffer (pH 8.2), 5 mM ascorbic acid, and 100 mg of neutral alumina. After the cardiac puncture, the animals were immediately decapitated and kidneys were removed, weighed, and homogenized in 4 volumes (v:w) of 30% aqueous methanol. Tris base buffer (0.5 ml, pH 8.2) containing 5 mM ascorbic acid and 100 mg of neutral alumina were added. The neutral alumina (which adsorbs quercetin) was precipitated by a brief centrifugation, and the supernatant was removed with a Pasteur pipette. The neutral alumina precipitates were washed 3 times with 5 ml of 20 mM Tris base solution containing 0.2% EDTA. Quercetin was eluted from the neutral alumina with 300 l of 0.25 M HCl-50% methanol solution and analyzed by HPLC using a reversed phase C 18 column (150 ϫ 4.6 mm, particle size 5 M, Rainin Instrument Co., Torrance, CA) with UV detection at 340 nm. The HPLC system consisted of a Waters model 510 pump, a model 501 solvent delivery system, a model 490 multi-wavelength detector, and a model 740 data module. The mobile phase was 50% aqueous methanol containing 10 mM KH 2 PO 4 , adjusted to pH 2.40 with H 3 PO 4 .

Effect of Quercetin on Estrogen Metabolite Concentrations in Tissue and Urine
Tissue Concentrations of 2-OH-E 2 and 4-OH-E 2 -1 h after an intraperitoneal injection of 50 g of [6,7-3 H]E 2 to male Syrian hamsters, concentrations of 2-OH-E 2 and 4-OH-E 2 in kidney were 2.7 Ϯ 0.4 and 1.3 Ϯ 0.3 ng/g of wet tissue, respectively, and corresponding concentrations in liver were 2.1 Ϯ 0.3 and 0.6 Ϯ 0.2 ng/g of wet tissue, respectively (Fig. 1). When hamsters were treated with 30 mg of quercetin by intragastric intubation 15 min prior to the intraperitoneal injection of [6, H]E 2 , concentrations of 2-OH-E 2 and 4-OH-E 2 in kidney were increased by 80% (p Ͻ 0.01) and 59% (p Ͻ 0.01), respectively, and corresponding concentrations in liver were increased by 48% (p Ͻ 0.05) and 59%, respectively (Fig. 1).
In summary, concentrations of CE metabolites in kidney of male hamsters injected with 50 g of [6, H]E 2 were comparable to those in liver. Treatment of hamsters with quercetin significantly increased CE concentrations in kidney and concomitantly decreased urinary concentrations of methoxyestrogens. Taken together, these results demonstrate an in vivo inhibition by quercetin of the O-methylation of CE metabolites during co-treatment of hamsters with E 2 .  (Table I). Treatment of hamsters with a diet supplemented with 3% quercetin for 2 weeks lowered renal SAM concentrations by approximately 25% (p Ͻ 0.05), and concomitantly increased SAH levels by 37% (p Ͻ 0.05) and 89% (p Ͻ 0.01) in control or estrogen-treated hamsters, respectively, resulting in 1.9 -2.6-fold increases in the average SAH/SAM ratios ( Table I). Treatment of hamsters with E 2 alone did not significantly influence renal SAM and SAH concentrations (Table I).
In summary, the addition of SAH strongly inhibited the in vitro O-methylation of 2-OH-E 2 and 4-OH-E 2 catalyzed by hamster kidney cytosolic COMT with IC 50 values of approximately 10 -14 M. Enzyme kinetic analyses revealed the inhibition of CE O-methylation by SAH to be purely noncompetitive with respect to CE substrate, but purely competitive with respect to the methyl donor SAM. Treatment of hamsters with E 2 and a 3% quercetin supplement in the diet decreased the renal pool of SAM by 25% and almost doubled SAH concentrations in the kidney. The marked increase in renal SAH levels induced by the quercetin co-treatment is estimated to inhibit the COMT-catalyzed O-methylation of CE metabolites by approximately 30 -70% based on the in vitro inhibiting activity of SAH.

Inhibition of the in Vitro O-Methylation of CE by Quercetin-
The in vitro O-methylation of 10 M 2-OH-E 2 or 4-OH-E 2 by hamster kidney cytosolic COMT was inhibited by the addition of quercetin or its structural analog, fisetin, in a concentrationdependent manner (Fig. 6)  concentrations of quercetin or fisetin, the V max values of this enzymatic O-methylation of 2-OH-E 2 and 4-OH-E 2 were markedly inhibited in a concentration-dependent manner ( Fig. 7 and Table II). The marked decreases in V max values in the presence of flavonoids indicated a substantial contribution by a noncompetitive mechanism of enzyme inhibition. In addition to the observed decreases in V max values, the K m values were simultaneously increased in the presence of quercetin or fisetin ( Fig.  7 and Table II), thus indicating a mixed (competitive plus noncompetitive) mechanism of enzyme inhibition as reported previously with a purified porcine liver COMT preparation (4).
Tissue Levels of Quercetin-The concentrations of unmetabolized quercetin in plasma, kidney, and liver of hamsters treated with a dietary supplement of 3% quercetin for 2 weeks were 0.22 Ϯ 0.19 M, 0.43 Ϯ 0.37, and 0.37 Ϯ 0.22 nmol/g of wet tissue, respectively (Table III). Similar concentrations of unmetabolized quercetin were obtained in plasma, liver, and kidney of hamsters treated with this 3% quercetin diet for 6.5 months (Table III). In contrast, concentrations in tissues from animals on a control diet were below the detection limit (Ͻ0.05 M quercetin in plasma or Ͻ0.05 nmol of quercetin/g of wet tissue; Table III).
In summary, quercetin inhibited the COMT-catalyzed O-  The CE concentrations in the kidney of male hamsters injected with 50 g of [6, H]E 2 were comparable with those in liver. In contrast, the enzyme activities in kidney catalyzing the 2-and 4-hydroxylation of E 2 are at least 1 order of magnitude lower than those in liver (21). These relatively high concentrations of CE metabolites in kidney may be explained in part by the lower detoxifying enzyme activities in this organ compared with those in liver (22). The larger increase in renal CE concentrations compared with hepatic concentrations in male hamsters co-treated with quercetin and E 2 also suggests that detoxification of CE in liver may remain intact, whereas that in kidney may be compromised. Second, the concentrations of endogenous catecholamines (substrates and competitive inhibitors of COMT-catalyzed O-methylation) in hamster kidney are more than 40-fold higher than those in liver (14), which inhibits the O-methylation of CE (14) and thereby may contribute to the high levels of CE in kidney. Finally, in addition to direct aromatic hydroxylation of parent estrogen, CE may be formed by metabolic deconjugation of estrogen conjugates such as estrogen glucuronides and methyl ethers (16). This metabolic deconjugation has been shown to be an important source of CE production in hamster kidney but is less important in liver compared with hepatic CE production by direct hydroxylation of parent estrogens (16).
Our study clearly demonstrates that the COMT-catalyzed O-methylation of CE is inhibited by quercetin via two different mechanisms, i.e. the direct inhibition of COMT by quercetin and the indirect inhibition by elevated tissue concentrations of SAH. Quercetin itself is a substrate for COMT (10) and thus competitively inhibits the O-methylation of CE substrates by competing for the methylating enzyme. Although SAH inhibited COMT in a noncompetitive fashion with respect to CE substrates, the kinetic analysis revealed that SAH competitively inhibited the association of the methyl donor SAM with the methylating enzyme. Thus, SAH may decrease concentrations of the COMT⅐SAM complex and increase those of the COMT⅐SAH complex. A decrease in the concentration of the COMT⅐SAM complex is consistent with a decrease in the V max value and unchanged K m value (a noncompetitive inhibition). This noncompetitive inhibition by SAH also explains the noncompetitive component of enzyme inhibition by quercetin or fisetin in vitro, because the O-methylation of either flavonoid will increase the concentrations of SAH. In addition to these two mechanisms, an approximately 25% decrease in renal pools of SAM (the methyl donor) during quercetin administration may also be a contributing factor for the decreased metabolism of CE by O-methylation in vivo.
In hamster kidney, the inhibition of CE O-methylation by SAH likely dominates over the direct inhibition by quercetin for the following reasons. (i) Despite the chronic administration of a high dose of quercetin to animals (3% in the diet), plasma or tissue concentrations of unmetabolized quercetin do not exceed 0.5 nmol/ml or g of wet tissue, respectively. Quercetin has previously been shown to undergo rapid O-methylation and/or other conjugation reactions (10). The low concentrations of unmetabolized quercetin in blood and in tissues observed in this study are in close agreement with previous studies (23). (ii) The marked increase in renal concentrations of SAH and in SAH/SAM ratios during treatment with quercetin makes it likely that inhibition by SAH is the dominant form of inhibition in quercetin-treated animals. The increase in tissue levels of

TABLE III
Plasma and tissue levels of quercetin in male Syrian hamsters given a diet supplemented with 3% quercetin for 2 weeks or 6.5 months The quercetin content of this diet was confirmed by an HPLC analysis to be 97 Ϯ 2% of the indicated concentration, whereas the content in normal rodent chow was below the detection limit of the assay. Blood samples, obtained by cardiac puncture, and freshly exercised liver and kidney samples were analyzed for quercetin concentrations by an HPLC method described under "Materials and Methods."  (10). Based on our enzyme kinetic studies, the magnitude of increase in renal SAH concentrations (approximately 10 M) is estimated to significantly inhibit the metabolic O-methylation of CE in vivo. Thus, it is suggested that a markedly increased demand on the circulating one-carbon pool due to the O-methylation of quercetin or other catechols may result in an increase in tissue pools of SAH and a concomitant inhibition of the O-methylation in vivo of CE metabolites. This metabolic change may be the basis of the previously observed increase in the severity of kidney tumorigenesis in hamsters treated with E 2 and quercetin (9) and supports the postulated role of CE, in particular 4-OH-E 2 , in the induction of estrogen-associated tumors (4,5). CE have previously been shown to undergo metabolic redox cycling, a process to generate potentially mutagenic free radicals in addition to other chemically reactive species such as estrogen semiquinones and quinones (3, 6 -8).
Details of this mechanism of DNA damage induced by redox cycling of CE are discussed in a recent review (5).
In summary, the administration of quercetin to male Syrian hamsters treated with E 2 inhibits the O-methylation of CE metabolites and thereby increases their concentrations in tissues and decreases urinary excretion of methoxyestrogen conjugates. The increase in CE concentrations together with the previously reported increase in the severity of E 2 -induced kidney tumorigenesis in hamsters is taken as evidence in support of a critical role of redox cycling of CE metabolites and free radical generation in the induction of hormone-associated cancers.