Abstract
In the present study, we evaluated the metabolicO-methylation of several catechol-containing tea polyphenols by human placental catechol-O-methyltransferase (COMT). (−)-Epicatechin, (+)-epicatechin, and (−)-epigallocatechin were good substrates for metabolic O-methylation by placental cytosolic COMT (150–500 pmol/mg of protein/min), but (−)-epicatechin gallate and (−)-epigallocatechin gallate were O-methylated at much lower rates (<50 pmol/mg of protein/min). When (−)-epicatechin was used as substrate, its O-methylation by human placental COMT showed dependence on incubation time, cytosolic protein concentration, incubation pH, and concentration ofS-adenosyl-l-methionine (the methyl donor). Analysis of cytosolic COMT from six human term placentas showed that the O-methylation of increasing concentrations of (−)-epicatechin or (−)-epigallocatechin follows typical Michaelis-Menten kinetics, with Km andVmax values of 2.2 to 8.2 μM and 132 to 495 pmol/mg of protein/min for (−)-epicatechin and 3.9 to 6.7 μM and 152 to 310 pmol/mg of protein/min for (−)-epigallocatechin, respectively. Additional analysis revealed that COMT-catalyzedO-methylation of (−)-epicatechin and (−)-epigallocatechin was strongly inhibited in a concentration-dependent manner byS-adenosyl-l-homocysteine (IC50 = 3.2–5.7 μM), a demethylated product ofS-adenosyl-l-methionine. This inhibition byS-adenosyl-l-homocysteine follows a mixed (competitive plus noncompetitive) mechanism of enzyme inhibition. In summary, several catechol-containing tea polyphenols are rapidlyO-methylated by human placental cytosolic COMT. This metabolic O-methylation is subject to strong inhibitory regulation by S-adenosyl-l-homocysteine, which is formed in large quantities during theO-methylation of tea polyphenols.
Catechol-O-methyltransferase (COMT),2 an enzyme ubiquitously present in high activity in humans and rodents, catalyzes the metabolic O-methylation of endogenous catecholamines and catechol estrogens (Axelrod and Tomchick, 1958; Axelrod, 1966; Guldberg and Marsden, 1975; Thakker and Creveling, 1990; Zhu and Conney, 1998a). Metabolic O-methylation of catecholamines (epinephrine, norepinephrine, and dopamine) by COMT inactivates their neurohormonal or neurotransmitter activities. More recent studies suggested that metabolic O-methylation of endogenous catechol estrogens plays a role in protection against estrogen-induced cancers by inactivating the potentially genotoxic 4-hydroxyestrogens and by forming the antitumorigenic 2-methoxyestradiol (reviewed by Zhu and Conney, 1998a,b).
COMT also catalyzes the metabolic O-methylation of many catechol-containing xenobiotics (Axelrod and Tomchick, 1958; Axelrod, 1966; Guldberg and Marsden, 1975; Thakker and Creveling, 1990; Zhu et al., 1994). An earlier study by Zhu et al. (1994) showed that quercetin and fisetin, two catechol-containing dietary flavonoids, are rapidlyO-methylated by cytosolic COMT, with rates two to three orders of magnitude higher than for endogenous catecholamines. Like these dietary flavonoids, several of the common tea polyphenols, such as (−)-epicatechin, (+)-epicatechin, (−)-epigallocatechin, (−)-epicatechin gallate, and (−)-epigallocatechin gallate (structures shown in Fig. 1), also contain a catechol structure and share certain degrees of structural similarity to them. These catechol-containing tea polyphenols constitute a major fraction of the water-extractable polyphenols contained in tea leaves (Balentine, 1992; Zhu et al., 1998). Several recent studies showed that the plasma levels of (−)-epigallocatechin and other catechol-containing tea polyphenols can easily reach lower micromolar concentrations in human subjects after oral administration of tea products (Lee et al., 1995; Unno et al., 1996; Li et al., 2000). We report here, for the first time, that several common catechol-containing tea polyphenols are very rapidlyO-methylated by human COMT. Our findings are interesting, and also point to the need to determine whether some of the biological actions of tea polyphenols may be caused by their methylated products or may result from their potential inhibition of COMT-catalyzedO-methylation of endogenous catecholamines and catechol estrogens.
Materials and Methods
Chemicals.
(−)-Epicatechin, (+)-epicatechin, (−)-epicatechin gallate, (−)-epigallocatechin, (−)-epigallocatechin gallate,S-adenosyl-l-methionine, dithiothreitol, and Tris-HCl were purchased from the Sigma Chemical Co. (St. Louis, MO).S-[methyl-3H]Adenosyl-l-methionine (specific activity, 11.2–13.5 Ci/mmol) was obtained from NEN Life Science Products (Boston, MA). All organic solvents used in this study were of HPLC grade and obtained from Fisher Scientific Co. (Springfield, NJ).
Preparation of Human Placental Cytosolic Fraction.
Placentas were obtained from term pregnant women (at 36–40 weeks of gestation) undergoing normal vaginal delivery at the St. Peter's University Medical Center (New Brunswick, NJ). The procedure for procurement of human placental tissues used in this study was approved by the institutional review boards of both Rutgers University (Piscataway, NJ) and the St. Peter's Medical Center (New Brunswick, NJ). Within 30 min after the placenta was expelled, a portion (∼50–100 g) was snap-frozen in liquid nitrogen for immediate transport to the laboratory for long term storage in a −80°C freezer.
On the day of cytosol preparation, the placental samples were first thawed at room temperature and then rinsed with ice-cold normal saline (0.9%) to remove blood clots. The chorionic membrane was removed with a pair of sharp eye surgical scissors. The tissues were minced in 3 volumes of ice-cold 0.1 M Tris-HCl/KCl solution (pH 7.4) and sequentially homogenized with a Tri-R homogenizer (Model K41) for 2–3 min and then with a Polytron homogenizer for another 2–3 min. Tissue homogenates were centrifuged at 9000g for 15 min, and supernatants were pooled and filtered through two layers of cheesecloth to remove lipid clots. The filtrates were recentrifuged at 105,000g (4°C) for 90 min. The supernatants were combined, and aliquots of the cytosolic preparations were stored at −80°C until used. The protein concentration was determined by using a protein assay kit (Bio-Rad, Richmond, CA) with BSA as the standard.
Human Placental COMT-Catalyzed O-Methylation of Catechol-Containing Tea Polyphenols.
The reaction mixture was prepared at ∼4°C and consisted of a 5 to 30 μM concentration of a tea polyphenol substrate, 1.0 mg of human placental cytosolic protein per milliliter, 25 or 50 μMS-adenosyl-l-methionine (containing 0.25 μCi ofS-[methyl-3H]adenosyl-l-methionine), 1.2 mM magnesium chloride, and 1.0 mM dithiothreitol in a final volume of 1 ml of Tris-HCl buffer (10 mM, pH 7.4). The reaction was initiated by addition of human placental cytosolic protein and was carried out at 37°C for 20 min under constant mild shaking. The reaction was arrested by placing the tubes on ice followed immediately by addition of 0.5 ml of ice-cold water and extraction of the3H-containing methylated catechin with 5 ml of ethyl acetate. After centrifugation for 20 min at approximately 1000g, portions of the organic supernatants (2 ml) were accurately removed for measurement of radioactivity in a scintillation counter (model LS 5000TD; Beckman Instruments, Berkeley, CA). Blank values obtained from incubations without placental cytosolic protein were determined in each individual assay and subtracted. The blank radioactivity counts were usually <1/5 of the values obtained from incubations in the presence of placental cytosols. By using the3H-labeled monomethylated products of (−)-epicatechin and (−)-epigallocatechin (isolated from HPLC), the extraction efficiencies for the O-methylated (−)-epicatechin and (−)-epigallocatechin were found to be similar (both above 93%). The rate (velocity) for the O-methylation of various tea polyphenol substrates by human placental COMT was expressed as picomoles of methylated tea polyphenol formed per milligram of protein per minute.
Liquid Chromatography-Mass Spectrometry Analysis of Methylated Tea Polyphenols.
To determine the structures of the methylated tea polyphenols, 30 μM (−)-epicatechin or (−)-epigallocatechin was first incubated in vitro for 30 min with 1.0 mg of human placental cytosolic protein per milliliter, 50 μMS-adenosyl-l-methionine, 1.2 mM magnesium chloride, and 1.0 mM dithiothreitol in a final volume of 1 ml of Tris-HCl buffer (10 mM, pH 7.4). The reaction mixture was extracted twice with 5 ml of ethyl acetate. The organic solvent extracts were combined and dried under a stream of nitrogen, and the residues were redissolved in 100 μl of 20% acetonitrile in water by vortex mixing. The mixture was then centrifuged at ∼15,000g for 10 min, and a 50-μl aliquot of the supernatant was injected into the HPLC for analysis of methylated tea polyphenol metabolites according to the method for tea catechins described by Lee et al. (1995). The incubations described above were first done with tea catechins andmethyl-3H-labeledS-adenosyl-l-methionine to determine the retention times of the radioactive metabolites. Parallel incubations of the catechins with only nonradioactiveS-adenosyl-l-methionine were performed for collection of metabolites for mass spectrometric analysis.
The peaks corresponding to methylated (−)-epicatechin or methylated (−)-epigallocatechin were collected from the HPLC column and were dried under a stream of nitrogen. They were then analyzed by liquid chromatography-mass spectrometry (Finnegan LCQ; Thermoquest Corp., San Jose, CA) using electrospray/mass spectrometry with an ion trap operated in the negative ion mode.
Results
Measurement of Metabolic Formation of Methylated Tea Catechins by HPLC.
Incubation of (−)-epigallocatechin with the cytosol of a human placenta (NP5C) in the presence ofS-[methyl-3H]adenosyl-l-methionine resulted in formation of a less polar radioactive metabolite peak, with a retention time of 22.6 min (Fig. 2, A and B). Similarly, incubation of (−)-epicatechin with this human placental cytosol in the presence ofS-[methyl-3H]adenosyl-l-methionine (0.25 μCi) resulted in formation of two less polar radioactive metabolite peaks, with retention times of 27.6 and 30.3 min, respectively (Fig. 2, A and C). The ratio of the two radioactive peaks was ∼5:1.
Confirmation by Mass Spectrometry of the Metabolic Formation of Monomethylated Tea Catechins.
Multiple incubations of (−)-epigallocatechin or (−)-epicatechin with nonradioactive S-adenosyl-l-methionine and a placental cytosol (NP5C) followed by extraction with ethyl acetate and isolation by HPLC of the 22.6-min metabolite peak from (−)-epigallocatechin and the 27.6-min metabolite peak from (−)-epicatechin resulted in sufficient material for mass spectrometric analysis. Analysis by liquid chromatography-mass spectrometry of the metabolite derived from (−)-epigallocatechin (Mr = 306) revealed a base peak (deprotonated ion) with a mass of 319 Da, confirming the identity of this metabolite as a monomethylated metabolite of (−)-epigallocatechin. Similarly, analysis of the major metabolite derived from (−)-epicatechin (Mr = 290) revealed a base peak (deprotonated ion) with a mass of 303 Da, confirming the identity of this metabolite as a monomethylated metabolite of (−)-epicatechin. The minor metabolite formed from (−)-epicatechin was not further investigated.
Kinetic Studies on the O-Methylation of Catechol-Containing Tea Polyphenols by Human Placentas.
When (−)-epicatechin (30 μM) was used as substrate for metabolicO-methylation by human placental cytosol (NP5C), formation of methylated products was dependent on incubation time (linear up to 45 min; Fig. 3, top panel), cytosolic protein concentration (linear up to 1.0 mg/ml; Fig. 3, middle panel), and the concentration of the supporting cofactorS-adenosyl-l-methionine (Fig. 3, bottom panel). The dependence of O-methylation onS-adenosyl-l-methionine concentration follows typical Michaelis-Menten kinetics, with aKm value of ∼24 μM as calculated from a double reciprocal plot (not shown).
We also determined the pH dependence (from pH 4.5 to 13) for metabolicO-methylation of (−)-epicatechin by the cytosol from two human term placentas (NP5C and NP10C) (Fig.4). Rates of (−)-epicatechinO-methylation by placental cytosol were highest at pH 7.5, suggesting that the normal intracellular pH would be optimal for the metabolic O-methylation of this tea polyphenol in vivo.
Under optimized conditions for in vitro O-methylation, (−)-epicatechin, (+)-epicatechin, and (−)-epigallocatechin at a 30 μM concentration were all found to be rapidly O-methylated by the cytosol of a representative human placenta (NP5C), with metabolic rates of 417, 312, and 409 pmol/mg of protein/min, respectively (Fig. 5). In contrast, (−)-epicatechin gallate and (−)-epigallocatechin gallate wereO-methylated at much lower rates (<50 pmol/mg of protein/min) under the same reaction conditions (Fig. 5).
We also determined the kinetic parameters (Km and Vmaxvalues) for the O-methylation of (−)-epicatechin and (−)-epigallocatechin by several representative human placentas. The rate of metabolic O-methylation of increasing concentrations of (−)-epicatechin by the cytosol of six human placentas all showed typical Michaelis-Menten curve patterns (data not shown). Double reciprocal plotting of this data showed theKm values ranging from 2.2 to 9.1 μM and the Vmax values from 189 to 495 pmol/mg of protein/min for the six placentas tested (Table1). Similarly, the rate of metabolicO-methylation of (−)-epigallocatechin by three human placentas also followed typical Michaelis-Menten kinetics (data not shown). Their Km values for theO-methylation of (−)-epigallocatechin varied from 3.9 to 6.7 μM, and their Vmax values varied from 193 to 310 pmol/mg of protein/min (Table 1).
Inhibitory Regulation of Human PlacentalO-Methylation of Tea Polyphenols byS-Adenosyl-l-homocysteine.
When 5 μM (−)-epicatechin or (−)-epigallocatechin was used as substrate, their O-methylation by cytosol from three representative human term placentas (NP12C, NP13C, and NP7C) was inhibited by S-adenosyl-l-homocysteine (the demethylated product ofS-adenosyl-l-methionine) in a concentration-dependent manner (Fig. 6). The IC50 for the inhibition byS-adenosyl-l-homocysteine were quite low (ranging from 3.2 to 5.7 μM), and the inhibition was almost complete at ≥32 μMS-adenosyl-l-homocysteine (Fig. 6).
Kinetic analysis with two human placental samples (NP12C and NP13C) showed that S-adenosyl-l-homocysteine inhibited human COMT-catalyzed O-methylation of (−)-epicatechin with a mixed (competitive plus noncompetitive) mechanism (Fig. 7). When (−)-epigallocatechin was used as a substrate, similar kinetic mechanism of enzyme inhibition byS-adenosyl-l-homocysteine was observed (data not shown).
Discussion
The present study showed that (−)-epicatechin, (+)-epicatechin, and (−)-epigallocatechin are very rapidly O-methylated by the cytosolic COMT of human term placentas. Liquid chromatography-mass spectrometry analysis confirmed that the methylated metabolites of (−)-epicatechin and (−)-epigallocatechin were the monomethylated derivatives. When compared at the same substrate concentrations (20 and 50 μM), the rates for the O-methylation of (−)-epicatechin, (+)-epicatechin, and (−)-epigallocatechin by the cytosols of two representative human placentas were 2.5 to 6 times faster than the O-methylation of 2- and 4-hydroxyestradiol, and 10 to 50 times faster than the O-methylation of dopamine (data not shown). In comparison, (−)-epicatechin gallate and (−)-epigallocatechin gallate, when assayed at the same substrate concentration (30 μM), were O-methylated by COMT at a much lower rate.
Using (−)-epicatechin as a representative substrate, itsO-methylation by human placental COMT showed characteristic dependence on the incubation pH, with optimum close to the normal intracellular pH. Our results also showed that theKm value forS-adenosyl-l-methionine (the methyl donor) for the in vitro O-methylation of (−)-epicatechin is ∼24 μM (Fig. 2), which is within the range of tissue concentrations of S-adenosyl-l-methionine as previously reported for animals and humans (Ueland, 1982; Wagner et al., 1984; Zhu and Liehr, 1996). Taken together, these data suggest that the O-methylation of tea catechins is a kinetically feasible reaction in vivo. The results of the present study are the first demonstration of the metabolism of tea polyphenols to theirO-methylated products.
It is known that human COMT is a polymorphic enzyme, with ∼25% of the Caucasians having a homozygous allele coding for a low activity form of the enzyme (Cohn et al., 1970; Weinshilboum et al., 1974;Scanlon et al., 1979; Boudikova et al., 1990). Earlier studies estimated that the enzymatic activity of the low activity form of the COMT was ∼1/4 of that of the high activity form when catecholamines were assayed as substrates. Our data also showed that considerable differences exist in the Km values (from 2.2 to 9.1 μM) and Vmax values (from 189 to 495 pmol/mg of protein/min) for the O-methylation of (−)-epicatechin by six representative human placentas (Table 1). The differences between theVmax/Km ratios are even more pronounced (from 20.8 to 225.0; Table 1). These observed differences in enzymatic activity were not due to assay variations, because the intra-assay and interassay variations calculated for two representative human placental COMT preparations were <3% and <10%, respectively.
Earlier studies with norepinephrine and quercetin (a catechol-containing flavonoid) showed that COMT at pH 7.4 predominantly methylates the meta-hydroxyl group of these substrates (Creveling et al., 1970, 1972; Zhu et al., 1994). Our HPLC analysis of the methylated (−)-epicatechin products (Fig. 2) showed two less polar peaks with a ratio of ∼5:1. It is suspected that the major monomethylated metabolite of (−)-epicatechin is the 3′-O-methyl isomer, with the 4′-O-methyl isomer a minor metabolite (refer to Fig. 1 for the 3′- and 4′-positions). Our HPLC analysis of the methylated (−)-epigallocatechin products only showed one metabolite peak (Fig. 2). It is possible that the two identical meta-hydroxyl groups of (−)-epigallocatechin were preferentially O-methylated at pH 7.4, resulting in formation of one detectable product.
It is known that the endogenousS-adenosyl-l-homocysteine is an important feedback inhibitor for COMT-catalyzedO-methylation of a number of catechol substrates. Our results also showed that the rapid metabolic O-methylation of tea polyphenols is subject to potent inhibitory regulation by endogenous S-adenosyl-l-homocysteine, a demethylated product ofS-adenosyl-l-methionine that is formed in equimolar amounts with the O-methylated tea polyphenol substrates. Enzyme kinetic analysis showed that the inhibition byS-adenosyl-l-homocysteine of human placental COMT follows a mixed (noncompetitive plus competitive) mechanism of enzyme inhibition. This mechanism of COMT inhibition is similar to earlier observations with human liver and hamster kidney COMT, which also showed a mixed mechanism of enzyme inhibition byS-adenosyl-l-homocysteine (Ball et al., 1972; Zhu et al., 1994).
It should be noted that the potent inhibition byS-adenosyl-l-homocysteine of the metabolic O-methylation of tea polyphenols prevents the endogenous S-adenosyl-l-methionine pool from being readily depleted.S-Adenosyl-l-methionine is a universal methyl donor for the methylation of a variety of substrates in animals and humans, including the catechol-containing xenobiotics, DNA, endogenous catecholamines, and catecholestrogens. This regulatory mechanism, therefore, would be important for maintaining the homeostasis of the universal methyl donorS-adenosyl-l-methionine during metabolic O-methylation of endogenous and foreign catechols in vivo.
Several recent studies showed that many common tea polyphenols such as (−)-epigallocatechin and epicatechins can be readily absorbed in human subjects after drinking tea, and some of them can reach micromolar concentrations in the plasma (Lee et al., 1995; Unno et al., 1996; Li et al., 2000). A recent study using 3H-labeled (−)-epigallocatechin gallate also showed that this tea catechin is widely distributed in mouse tissues (Suganuma et al., 1998). It is expected that a significant fraction of the injested catechol-containing tea polyphenols will be rapidlyO-methylated in the body. Therefore, it will be of interest to evaluate the pharmacological activities of the methylated tea polyphenols. Sano et al. (1999) recently reported that two quantitatively minor methylated catechins isolated from oolong tea showed a stronger antiallergic activity than did an unmethylated catechin. Also, it will be of considerable interest to determine whether some of tea's effects are related to inhibition of theO-methylation of endogenous catecholamines or catechol estrogens.
In summary, several catechol-containing tea polyphenols are very rapidly O-methylated by human placental COMT. This metabolicO-methylation of tea polyphenols is subject to potent inhibitory regulation byS-adenosyl-l-homocysteine, which is formed in large quantities during the enzymaticO-methylation of some catechol-containing tea polyphenols. Further research is warranted to determine the pharmacological activities of the methylated tea polyphenols and the potential inhibitory effect of the catechol-containing tea polyphenols on COMT-catalyzed O-methylation of endogenous catecholamines and catechol estrogens.
Acknowledgments
The authors thank Mao-Jung Lee for advice on the HPLC separation of tea catechins; Chuan Li, Bozena Winnik, and Brian Buckley for help with the liquid chromatography-mass spectrometry analysis of theO-methylated catechins; and Chung S. Yang for helpful comments on the manuscript.
Footnotes
-
Send reprint requests to: Dr. Bao Ting Zhu, Department of Basic Pharmaceutical Sciences, College of Pharmacy, University of South Carolina, 700 Sumter Street, Columbia, SC 29208. E-mail:btzhu{at}pharm.sc.edu
-
↵1 William and Myrle M. Garbe Professor of Cancer and Leukemia Research.
-
This study was supported by National Institutes of Health Grants CA 74787, CA 49756, and ES 05022.
- Abbreviation used is::
- COMT
- catechol-O-methyltransferase
- Received January 11, 2000.
- Accepted May 31, 2000.
- The American Society for Pharmacology and Experimental Therapeutics