Chem. Res. Toxicol. 2001, 14, 841-848
841
Catechin Metabolism: Glutathione Conjugate Formation Catalyzed by Tyrosinase, Peroxidase, and Cytochrome P450 Majid Y. Moridani, Hugh Scobie, Par Salehi, and Peter J. O’Brien* Faculty of Pharmacy, University of Toronto, Ontario M5S 2S2, Canada Received November 10, 2000
The metabolic pathways of dietary flavonoids are still largely unknown. In the present work, mass spectrometry and UV-vis spectroscopy studies were used to show that the naturally occurring flavonoid catechin underwent enzymatic oxidation by tyrosinase in the presence of glutathione (GSH) to form mono-, bi-, and tri-glutathione conjugates of catechin and monoand bi-glutathione conjugates of a catechin dimer. A hydroxylated catechin adduct was also detected. Using UV spectroscopy, it was shown that the catechol B-ring of catechin was oxidized by tyrosinase to form an o-quinone which could be reduced back to catechin with potassium borohydride or reacted with GSH to form glutathione conjugates. The catechin-glutathione conjugates formed had much lower distribution coefficient values than catechin itself. When peroxidase and hydrogen peroxide were used instead of tyrosinase, only mono-glutathione conjugates were formed but not bi-glutathione conjugates or hydroxylated adducts. 1H NMR evidence showed that three different mono-glutathione conjugates on ring B of catechin were formed by peroxidase and hydrogen peroxide. Rat liver microsomes and NADPH or cumene hydroperoxide also catalyzed catechin-glutathione conjugate formation which was prevented by benzylimidazole, a P450 2E1 inhibitor. Catechin cytotoxicity toward isolated hepatocytes was also markedly enhanced by hydrogen peroxide or cumene hydroperoxide and was prevented by benzylimidazole, suggesting that catechin could be metabolically activated by P450 peroxidase activity to form cytotoxic quinoid species.
Introduction Structurally, flavonoids are divided into five main classes: i.e., flavonols, flavones, catechins, flavanones, and flavanols (1). Catechins are present in tea (2), wine (3), fruits (4, 5), chocolate (6), and legumes (7). The five major catechins are (+)-catechin, (-)-epicatechin, (-)epigallocatechin, (-)-epicatechin gallate, and (-)-epigallocatechin gallate (Figure 1). Catechins are potentially beneficial to human health as they are strong antioxidants, anticarcinogens, antiinflammatory agents, and inhibitors of platelet aggregation in in vivo and in vitro studies (8-10). There is some epidemiological evidence that the consumption of tea, a rich source of catechins, could reduce the risk of coronary heart disease, mortality, and stroke (11). In vitro studies have demonstrated that catechins may lower the risk of coronary heart disease by protecting low-density lipoproteins (LDL) against lipid peroxidation and oxidative stress (12). Animal in vivo studies also show that catechin effectively prevents cardiotoxicity caused by the anticancer drug doxorubicin (13). Catechin or epigallocatechin-3-gallate (a major tea flavonoid) has recently been shown by several investigators to inhibit the growth and metastatic potential (lung tumor formation) of B16-BL6 or B16F-10 melanoma cells in mice in vivo (14, 15). The molecular mechanisms involved in the anticancer activity of catechins were not * To whom correspondence should be addressed. Phone: (416) 9782716. Fax: (416) 978-8511. E-mail:
[email protected].
discussed, but one explanation could be that melanoma cells are unique in that they contain tyrosinase which catalyzes the oxidation of some phenols/catechols to form cytotoxic o-quinones. Tyrosinase substrates are being developed for use as anti-melanoma prodrugs (16), and the following suggests that catechin could be such a substrate for use as an anti-melanoma prodrug. Previously, catechin was shown to be oxidized by mushroom tyrosinase or grape polyphenol oxidase to form different dimers involving C-C or C-O-C interflavon linkages between the B-ring of a catechin and the A-ring of a second catechin (17). However, no evidence of o-quinone formation or possible cytotoxicity was provided. In the following, we have used mass spectrometry and UV-vis spectroscopy to show that catechin can be metabolized by tyrosinase to form a cytotoxic o-quinone, which reacts with GSH to form various mono-, bi-, and tri-glutathione conjugates of catechin and a catechin dimer. A hydroxylated catechin adduct was also detected (Figure 2). Mass spectrometry and 1H NMR analyses showed that catechin formed only mono-glutathione conjugate by peroxidase/H2O2/GSH oxidizing system. We have also found that rat liver microsomes and NADPH or cumene hydroperoxide also catalyzed catechinglutathione conjugate formation from catechin, which was inhibited by benzylimidazole, a P450 inhibitor. Benzylimidazole also prevented hepatocyte cytotoxicity induced by catechin, suggesting that catechin is metabolically activated by P450 to a cytotoxic reactive intermediate.
10.1021/tx000235o CCC: $20.00 © 2001 American Chemical Society Published on Web 05/26/2001
842
Chem. Res. Toxicol., Vol. 14, No. 7, 2001
Moridani et al.
Figure 1. Chemical structures of catechins.
Figure 2. Proposed scheme for the enzymatic oxidation of catechin by tyrosinase.
Experimental Procedures Chemicals. Catechin, 4-tert-butylcatechol, potassium borohydride, tyrosinase (mushroom), horseradish peroxidase, hydrogen peroxide, glutathione (GSH),1 oxidized glutathione (GSSG), tris(hydroxymethyl)aminomethane (Tris), sodium phosphate monobasic, sodium phosphate dibasic, diethylenetriaminopentaacetic acid (DETAPAC), 5,5′-dithio-bis(2-nitrobenzoic acid) (DTNB), 3-(N-morpholino)propanesulfonic acid (MOPS), di1 Abbreviations: GSH, glutathione; P450, cytochrome P450; NQO, NAD(P)H:quinone oxidoreductase; DNFB, 2,4-dinitro-fluorobenzene; MOPS, 3-(N-morpholino)propanesulfonic acid; DETAPAC, diethylenetriaminopenta-acetic acid; DMSO, dimethyl sulfoxide.
methyl sulfoxide (DMSO), trichloroacetic acid, and 1-octanol were obtained from Sigma/Aldrich Chemical Co. (Canada). The stock solutions of polyphenols were prepared in DMSO. Other chemicals were dissolved in Millipore filtered water or buffer. The stock solution of DTNB was prepared in Tris/HCl buffer 0.1 M (pH 8.94). UV-Vis Spectroscopy. The spectra of a solution containing catechin (100 µM) and tyrosinase (20 units/mL) were recorded in the absence and presence of GSH (400 µM) using a Pharmacia Ultraspec 1000 spectrophotometer with the proprietary Swift 1000 software to record the data. The spectra of the mixture were obtained when GSH was added to the solution either before or after the addition of tyrosinase. The control spectrum was
Glutathione Conjugation during Catechin Oxidation that of a catechin solution (100 µM) in a Tris/HCl buffer [0.1 M (pH 7.4) containing DETAPAC (1 mM)]. A similar experiment was carried out in the presence of peroxidase (0.1 µM)/H2O2 (200 µM) instead of tyrosinase (20 units/mL). The UV-vis spectra of solutions containing catechin (100 µM), GSH (400 µM), and tyrosinase (10 units/mL) were also obtained when tyrosinase was added to a mixture of catechin and GSH in a total volume of 5 mL of phosphate buffer 0.1 M (pH 4.0). After 10 min of incubation at room temperature, the solution was extracted three times with 5 mL of ethyl acetate. The aqueous phase was separated and centrifuged after which its UV-vis spectra was recorded. GSH Depletion Assay. Tyrosinase (20 µL; 2500 units/mL) was added to a mixture of catechin (50 µM) and GSH (200 µM) in a final volume of 1 mL Tris/HCl buffer (0.1 M, pH 7.4). The mixture was preincubated for 30 min at 37 °C. A 250 µL aliquot was added to trichloroacetic acid (25 µL; 30% w/v), vortexed and left for 5 min. A 100 µL aliquot of the supernatant was then added to a mixture of DTNB (25 µL; 2 mg/mL) and Tris/HCl buffer (875 µL; 0.1 M, pH 8.94) and vortexed. The absorbance of the solution was monitored at 412 nm (18). The standard curve for GSH measurement gave a regression coefficient of greater than 0.99 over the range of 5-200 µM GSH concentrations (data not shown). HPLC Analysis of GSH and GSSG. A modified method reported by Reed et al. (19) was used for the HPLC analysis of GSSG. An aliquot (800 µL) of the reaction mixture prepared for GSH depletion assay (preincubated at 37 °C for 30 min) was added to metaphosphoric acid (200 µL; 25% w/v) in a glass tube, vortexed, left for 30 min at room temperature, and centrifuged. The supernatant (500 µL) and freshly prepared iodoacetic acid (50 µL; 15 mg/mL water) were co-transferred to a glass tube containing sodium bicarbonate (∼100-150 mg). This was then vortexed, left up to 1 h or overnight at room temperature in a dark room to which 2,4-dinitro-fluorobenzene (DNFB) (500 µL; 1.5% w/v prepared in ethanol) was added, vortexed, left to stand at room temperature in a dark room for a period of 4-6 h for immediate HPLC processes. An auto-injector (WISP 710B, Waters Scientific Ltd) was used to inject a sample (50 µL) of the reaction mixture into an HPLC column (µBondapak NH2 (aminopropylsilyl bonded amorphous silica) 125°A, 10 µm, 3.9 × 300 mm, Waters Scientific Ltd.). A gradient mobile phase comprises two solvent systems: solvent A, methanol/water 80:20, and solvent B, methanol/acetate buffer 80:20, used to elute the sample. Acetate buffer was prepared by the addition of sodium acetate trihydrate (270 g) and acetic acid glacial (378 mL) to Millipore water (128 mL). The HPLC pump (model 501, Waters) was programmed at the flow rate of 1 mL/min for the mobile phase of A/B (min) with a ratio of 90/ 10 (0), 10/90 (25), 90/10 (27), and 90/10 (30). The LC Spectrophotometer (Lambada-Max, model 481, Waters Scientific Ltd) was set at 365 nm to detect GSH and GSSG, with retention times of 17.5 and 20.5 min, respectively. A software (Maxima 800) and an IBM-compatible computer were employed for analysis of the data and integration. Distribution Coefficient Value Measurement for Catechin. A modified version of an automated method was used to determine the distribution coefficient value for catechin and its glutathione conjugates (20). To 25 mL of MOPS buffer [(50 mM, pH 7.4, and 1 mM DETAPAC) presaturated by 1-octanol overnight] containing 100 µM catechin, an aliquot of 1-octanol (5-10 mL) (presaturated with MOPS buffer) was added at time intervals of 30 min during which the solution was stirred over a magnetic stirrer. The absorbance of the aqueous phase was monitored at 280 nm before and after each 1-octanol addition. All solutions were stored and manipulated at room temperature. The distribution coefficient value of catechin between the two phases was calculated from Dpart ) [(A1 - A2)/A2 × Vw/Vo], where Dpart is the distribution coefficient value at pH 7.4, A1 and A2 are the absorbance of aqueous phase before and after the addition of 1-octanol, respectively. Vw and Vo are the volume of aqueous and 1-octanol phases, respectively.
Chem. Res. Toxicol., Vol. 14, No. 7, 2001 843 Apparent Distribution Coefficient Value Measurement for Catechin-Glutathione Conjugate. The MOPS buffer (25 mL) contained a mixture of catechin (100 µM) and GSH (400 µM) to which tyrosinase was added so as to give a final activity of 10 units/mL. The reaction was left for 30 min after which trichloroacetic acid (2.5 mL; 30% w/v) was added. The absorbance of the aqueous phase was monitored at the wavelength 280 nm following the addition of 1-octanol (presaturated with MOPS buffer). The distribution coefficient value was determined as described above. Mass Spectrometry Analysis. Tyrosinase (20 µL; 2500 units/mL) was added to a reaction mixture containing catechin (1 mM) and GSH (4 mM) in a total volume of 1 mL Millipore filtered water. The reaction mixture was preincubated for 5 min at room temperature prior to direct injection into a mass spectrometer (PE Sciex III, Biomolecular Mass). A similar experiment was carried out in the presence of peroxidase (3 µM)/ H2O2 (2 mM) instead of tyrosinase (50 units/mL). 1H NMR Spectroscopy and HPLC Analysis of Glutathione-Catechin Conjugates. GSH (2 mL; 100 mM) and 50 µL HRP (100 µM) were added to methanol (2.5 mL) containing catechin (60 mM) to which hydrogen peroxide (500 µL; 300 mM) was added over a period of 2 h. The reaction mixture was then transferred to a round-bottom flask and dried using a rotary evaporator under vacuum. The content was reconstituted in methanol (2 mL) for preparative thin-layer chromatography using four TLC plates (general purpose, silica gel on glass, size 10 × 20 cm, with indicator, Aldrich) and mobile phase ethyl acetate/heptane/methanol (24:5:6). A mixture of water and methanol was used to extract the glutathione-catechin conjugate from silica gel. A Sep-Pak (Chromosep, C18, 100 mg, 1 mL, Chromatographic Specialties Inc.) was employed to separate glutathione-catechin conjugate from any GSH contamination. A total of 330 µL of the extract (which contained 1 mg of the crude product) was introduced to the Sep-Pak, washed with 400 µL and 2 mL of water (pH 3.5), separately. The latter fraction which contained the pure glutathione-catechin conjugate was evaporated and dried under vacuum. This fraction contained no detectable free unreacted GSH. A sample of 10 µL from the filtrate was introduced to an HPLC column (Ultracarb 100 × 2 mm 5 µm, Phenomenex, Canada) using 10% acetonitrile/water or 5-10% acetonitrile/ phosphate buffer (10 mM, pH 3) at a flow rate of 0.2 mL/min as the mobile phase to examine its purity. A Shimadzu (SPD-6AV) UV-vis spectrophotometeric detector, a Shimadzu (LC-6A) liquid chromatograph pump, and a Shimadzu (C-R3A) Chromatopac integrator were used for the HPLC analysis of the sample. 1H NMR was carried out in DMSO-d6 in the presence and absence of 1-2 drops deuterium oxide (D2O) for proton exchange experiment by using 500 MHz 1H NMR spectroscopy (Varian UNITY plus 500 Spectrometer). Catechin-Glutathione Conjugate Formation by Rat Hepatocyte Microsomes. Adult male Sprague-Dawley rats, 250-300 g, were obtained from Charles River Canada laboratories (Montreal, Quebec), fed ad libitum, and allowed to acclimatize for 1 week on clay chip bedding. The animals were anesthetised by sodium pentobarbital (60 mg/kg of body wt). Livers were removed under sterile potassium buffer:KCl solution (1.18%, w/v, 4 °C) as previously described (21). Hepatic microsomes were prepared as previously described (22). Glucose-6-phosphate (7.5 mM) was added to a mixture of catechin (1 mM), NADP+ (0.5 mM), Mg2+ (5 mM), microsomes (1 mg/mL), and glucose-6-phosphate dehydrogenase (2.5 units/ mL), to give a total volume of 1 mL Tris/HCl buffer (0.1 M, pH 7.4 containing DETAPAC 1 mM). Alternatively, cumene hydroperoxide (1 mM) was used instead of NADPH generating system. GSH concentrations were 0.5 and 1 mM for the NADPH and cumene hydroperoxide supported rat liver microsomes, respectively. Samples were taken at 5, 10, and 20 min from the cumene hydroperoxide supported rat liver microsomes and at 30 and 60 min from the NADPH-supported rat liver microsomes. The mixture was preincubated at 37 °C from which 250 µL was
844
Chem. Res. Toxicol., Vol. 14, No. 7, 2001
Moridani et al.
Table 1. Catechin-Glutathione Conjugate Formationa GSH equivalent depletion catechin + catechin benzylimidazole tyrosinaseb (30 min) 2.2 ( 0.2 P450/NADPHc (60 min) 0.5 ( 0.1 P450/cumene hydroperoxided (5 min) 0.8 ( 0.1
