Chem. Res. Toxicol. 1996, 9, 729-736
729
Mechanism-Based Inactivation of Hepatic Ethoxyresorufin O-Dealkylation Activity by Naturally Occurring Coumarins Yingna Cai,† Wanda Baer-Dubowska,† Mike J. Ashwood-Smith,‡ Oluna Ceska,‡ Sanro Tachibana,§ and John DiGiovanni*,† The University of Texas M. D. Anderson Cancer Center, Science Park-Research Division, Department of Carcinogenesis, P.O. Box 389, Smithville, Texas 78957, and Department of Biology and Department of Botany, University of Victoria, Victoria, British Columbia V8W 2YZ, Canada Received December 11, 1995X
Several naturally occurring coumarins contained in the human diet have been found to be effective inhibitors and inactivators of murine hepatic ethoxyresorufin O-dealkylase (EROD) and pentoxyresorufin O-dealkylase in vitro [Cai, Y., Bennett, D., Nair, R. V., Ceska, O., Ashwood-Smith, M., and DiGiovanni, J. (1993) Chem. Res. Toxicol. 6, 872-879]. In the present study, these same coumarins decreased the content of cytochrome P450 (P450) in either 3-methylcholanthrene (MC)- or phenobarbital-induced murine hepatic microsomes but did not have a major effect on heme content. Detailed in vitro studies with [14C]coriandrin, which selectively inhibits and inactivates P450 1A1-mediated EROD activity, demonstrated that it covalently bound, in a preferential manner, to hepatic microsomal protein from MC-pretreated mice. A linear relationship was observed between covalent binding and loss of EROD activity. The inclusion of electrophile trapping agents in the incubations significantly inhibited the covalent binding of [14C]coriandrin to microsomal protein. In addition, the covalent binding of [14C]coriandrin was decreased 46% by 7,8-benzoflavone (7,8-BF), 58% by a monoclonal antibody with specificity toward MC-induced form(s) of P450, and 60% by ethoxyresorufin, implicating the bioactivation of coriandrin by P450 1A1. Analysis by sodium dodecyl sulfatepolyacrylamide gel electrophoresis of [14C]coriandrin-bound microsomal protein from MCpretreated mice showed that [14C]coriandrin bound covalently to a protein with an approximate molecular mass of 49 kDa. Again, addition of 7,8-BF or polyclonal antibody against P450 1A1 reduced the covalent binding of [14C]coriandrin to this specific protein band. Interestingly, coriandrin was also found to be a potent inhibitor and inactivator of purified human P450 1A1. These results demonstrate that certain coumarins to which humans are exposed in the diet are bioactivated by P450 1A1 to reactive intermediates that subsequently form covalent adducts with the apoprotein, effectively destroying enzyme activity. Thus, certain naturally occurring coumarins may have a significant effect on human health.
Introduction The cytochrome P450 (P450)1 superfamily consists of many enzymes that play a central role in both metabolic activation and detoxication of a variety of xenobiotics, including drugs, chemical toxicants, and carcinogens. To date, 12 P450 gene families comprising over 100 enzymes have been identified in mammalian species (1). Although there is some overlap in substrate specificity among the enzymes, individual enzymes may prefer a specific substrate (2). Many synthetic as well as naturally occurring compounds can inhibit P450-mediated reactions (reviewed in refs 3 and 4), including polycyclic * Author to whom correspondence and reprint requests should be addressed. † The University of Texas M. D. Anderson Cancer Center. ‡ Department of Biology, University of Victoria. § Department of Botany, University of Victoria. X Abstract published in Advance ACS Abstracts, May 1, 1996. 1 Abbreviations: P450, cytochrome P450; 1-EP, 1-ethynylpyrene; EROD, ethoxyresorufin O-dealkylase; MC, 3-methylcholanthrene; PROD, pentoxyresorufin O-dealkylase; PB, phenobarbital; β-NF, β-naphthoflavone; NBS, hybrid cells formed from RGNS-1 myeloma cells and spleen cells from unimmunized mice; G-6-P, glucose 6-phosphate; G-6PD, glucose-6-phosphate dehydrogenase; β-ME, β-mercaptoethanol; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; DLPC, dilauroylphosphatidylcholine; 7,8-BF, 7,8-benzoflavone; EH, epoxide hydrolase; ER, ethoxyresorufin; SCZ, sodium semicarbazide.
S0893-228x(95)00208-6 CCC: $12.00
aromatic acetylenes (5) and flavonoids (6, 7). These inhibitors may be classified as either reversible or irreversible inhibitors. Mechanism-based inactivators of P450 are a category of irreversible inhibitors that require metabolic activation by P450 to form reactive intermediates, leading to the formation of covalent adducts (8). Without exception, this covalent binding ultimately leads to the loss of the catalytic function of the P450 molecule (8). Furthermore, some inactivators appear to selectively inactivate specific P450 enzymes (reviewed in ref 9). For example, as reported by Hopkins et al., 1-ethynylpyrene (1-EP) effectively inactivated P450 1A1 but not P450 2B1 (5). Mechanism-based inactivators of P450 provide valuable probes to study the catalytic site of the enzyme molecule. Interestingly, inactivators such as 1-EP have also displayed anticarcinogenic effects in experimental carcinogenesis systems (10). A number of naturally occurring coumarins have been found to both inhibit and inactivate P450-mediated enzyme activity. Earlier studies (11, 12) indicated that xanthotoxin (7, Chart 1) and several linear furanocoumarins were irreversible inhibitors of P450 enzymes both in vitro and in vivo. Recently, we reported that certain naturally occurring coumarins, including the linear furanocoumarins imperatorin (6, Chart 1) and isopimpinellin © 1996 American Chemical Society
730 Chem. Res. Toxicol., Vol. 9, No. 4, 1996 Chart 1. Chemical Structures of Naturally Occurring Coumarins Used in the Present Study
Cai et al.
action of their reactive intermediates with the apoprotein.
Experimental Procedures
(5, Chart 1), the linear furoisocoumarin coriandrin (3, Chart 1) found in cilantro leaves (13), and the simple coumarin ostruthin (2, Chart 1), inactivated either P450 1A1-mediated ethoxyresorufin O-dealkylase (EROD) activity in hepatic microsomes from 3-methylcholanthrene (MC)-pretreated mice or P450 2B1-mediated pentoxyresorufin O-dealkylase (PROD) activity in phenobarbital (PB)-induced liver microsomes (14). In each case, metabolic activation of the coumarins was necessary to achieve inhibition, suggesting that the coumarins were mechanism-based inactivators. Mechanism-based inactivators of P450 appear to work by one of three potential mechanisms: (i) alkylation of the P450-prosthetic heme group (e.g., olefins and phenylhydrazine) (15, 16); (ii) degradation of the prosthetic heme group (e.g., halocarbons) (17); and (iii) covalent binding to the apoprotein (e.g., 2-ethynylnaphthalene) (18, 19). Limited studies by Mays et al. (20) indicated that xanthotoxin (7), a furanocoumarin, could be metabolically activated by liver microsomes from both β-napthoflavone (β-NF) and PB-induced rats to an electrophilic intermediate which then covalently bound to the P450 apoprotein. However, the specific mechanism of inactivation of P450-mediated EROD activity has not been determined for most of these coumarins, including coriandrin (3), a common component in the human diet. The present investigation is a continuation of our initial studies, which were aimed at determining the effects of naturally occurring coumarins on xenobiotic metabolizing enzymes, including P450s. The primary goal of the present investigation was to demonstrate whether these coumarins could be bioactivated by P450 1A1 to reactive intermediates that would covalently bind to the apoprotein, resulting in a subsequent loss of the enzyme activity. The following parameters were investigated: (i) the effect of selected coumarins on the level of P450 and heme content in MC-induced liver microsomes; (ii) the effects of inducers on covalent binding of [14C]coriandrin to liver microsomes; (iii) the ability of [14C]coriandrin to covalently bind to P450; (iv) the relationship between covalent binding of [14C]coriandrin and loss of EROD enzyme activity; (v) the role of P450 1A1 in bioactivation of [14C]coriandrin in liver microsomes; and (vi) the ability of coriandrin to inactivate human P450 1A1. The results indicate that certain naturally occurring coumarins to which humans are routinely exposed can inactivate P450 by covalent inter-
Caution: MC is a carcinogen and should be considered hazardous and handled with extreme care. In addition, several of the linear furanocoumarins are photoactive and may be carcinogenic in the presence of UV light. These compounds should be considered hazardous and handled with extreme care. Materials. Imperatorin (6) and isopimpinellin (5) were purchased from Indofine Chemical Co. (Belle Mead, NJ). Coriandrin (3) was obtained as previously described (13). Ostruthin (2) and bergamottin (4, Chart 1) were obtained from Dr. Warren Steck (National Research Council of Canada, Prairie Regional Laboratory, Saskatoon, Saskatchewan). Glucose 6-phosphate (G-6-P), glucose-6-phosphate dehydrogenase (G-6-PD), NADP, β-mercaptoethanol (β-ME), sodium semicarbazide (SCZ), sodium phenobarbital (PB), and sodium dithionite were obtained from Sigma Chemical Co. (St. Louis, MO). The MC was purchased from Eastman Kodak Co. (Rochester, NY). Anti-MCP450 1-7-1p8 and NBS 1-48-5p24 were provided by Dr. Harry V. Gelboin (Laboratory of Molecular Carcinogenesis, National Cancer Institute, National Institutes of Health, Bethesda, MD) (21). Anti-rat P450 1A1 was purchased from GENETEST Co. (Woburn, MA). Purified human recombinant P450 1A1 and 1A2 and rabbit NADPH-P450 reductase were generous gifts from Dr. F. Peter Guengerich (Department of Biochemistry and Center of Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, TN) (22-24). The purity of all the coumarins was checked by melting point and HPLC analysis. All chemicals used in the current study were more than 96% pure as judged by HPLC. Other reagents were obtained commercially and were of the highest purity deemed necessary. Synthesis of [14C]Coriandrin. 14C-Labeled coriandrin was synthesized as shown in Scheme 1 (25, 26). Coriandrin (3) (79.1 µM) was demethoxylated in nitrobenzene containing AlCl3 (133.4 µM) at 100 °C for 15 min. The reaction mixture was distilled after adding concentrated HCl (0.05 mL). The residue was dissolved in 10 mL of water and extracted three times with an equal volume of ethyl acetate. The ethyl acetate solution was washed with brine and dried over anhydrous sodium sulfate. Demethoxycoriandrin was then crystallized from the concentrated solution in chloroform and methanol. This compound (66.7 µM) was methylated with [14C]iodomethane (250 µCi, 10.7mCi/mmoL) in the presence of anhydrous potassium carbonate and dry acetone at 60 °C for 24 h. The reaction mixture was filtered, and the residue was washed with dry acetone. The filtrate and washes were combined and then distilled. The residue was dissolved in chloroform and subjected to preparative TLC (developing solvent: n-hexane/acetone, 3:1 v/v) to give the reaction product (Rf: 0.18). The TLC procedure was repeated two additional times. The product was further purified by recrystallization from methanol to give [14C]coriandrin. The resulting compound was verified by melting point (142-143 °C) (13) and TLC analysis, which gave the same values as authentic coriandrin (3). The specific radioactivity of [14C]coriandrin was measured (1.9 mCi/mmol). Microsomal Preparation. Female Sencar mice (purchased from NCI, Frederick, MD) at 7-9 weeks of age were treated with MC dissolved in olive oil (80 mg/kg of body weight, single dose daily, ip) for 2 consecutive days or with PB dissolved in 0.01 M sodium-potassium phosphate buffer (pH 7.4) containing 0.15 M sodium chloride (80 mg/kg of body weight, single dose daily, ip) for 3 consecutive days. The mice were killed 24 h after the last treatment, and their livers were removed, washed twice with the buffer, and homogenized in 0.01 M sodium-potassium phosphate buffer (pH 7.4) containing 0.15 M sodium chloride using a Brinkmann homogenizer PT 10/35 (Switzerland) (setting 6, 45 s). Microsomal fractions were obtained by centrifugation of the whole homogenate at 9000g for 30 min, followed by centrifugation of the supernatant fraction at 105000g for 60 min. Microsomal pellets were resuspended in 0.01 M sodiumpotassium phosphate buffer (pH 7.4) to an approximate con-
Mechanism-Based EROD Inactivation by Coumarins
Chem. Res. Toxicol., Vol. 9, No. 4, 1996 731
Scheme 1. Synthesis of [14C]Coriandrin
Table 1. Effects of Selected Coumarins on Levels of P450 and Heme in Mouse Liver Microsomesa MC-induced
PB-induced
compounds
P450b (% of control)
hemec (% of control)
P450 (% of control)
heme (% of control)
coriandrin imperatorin isopimpinellin ostruthin bergamottin 0.5 mM phenylhydrazine
70 79 70 74 73 -
100 98 97 94 95 -
-d 76 100 12
93 93 64
a The compounds (100 µM) were incubated with microsomal preparations in either the presence or absence of an NADPH-generating system at 37 °C for 30 min. Loss of P450 and heme content was assessed by measuring UV spectra from carbon monoxide-P450 complexes and pyridine-heme complexes, respectively. The value from 0 min incubation containing coumarins, microsomes, and NADPH was taken as the control value (1.48 ( 0.05 nmol of P450 and 3.01 ( 0.17 nmol of heme/mg of protein or 2.94 ( 0.28 nmol of P450 and 4.99 ( 0.34 nmol of heme/mg of protein for MC- or PB-induced microsomes, respectively). b The results are the means of three experiments. c The results are the means of five experiments. d Not determined.
centration of 2 mg of protein/mL. Protein concentrations were determined by the Lowry method (27). In Vitro Destruction of P450 and Heme. Reaction mixtures (4 mL total volume) contained MC-induced or PBinduced microsomes (1 mg of protein/mL), coumarins (100 µM), EDTA (1 mM), an NADPH-generating system (0.5 mM NADP, 0.5 mM G-6-P, and 0.3 U of G-6-PD/mL), and 0.15 M sodiumpotassium phosphate buffer (pH 7.4). Incubations were carried out for 30 min at 37 °C under air. The reactions were terminated by placing the reaction mixture on ice. P450 was then determined as described by Omura and Sato (28). As controls, some reaction mixtures were kept on ice, and some were incubated without either the coumarin or the NADPHgenerating system. For heme determination, 1 mL of the incubation mixture described above with or without NADPH was diluted with 2 mL of a 20% pyridine solution in 0.1 N sodium hydroxide. Heme content was determined as described by Ortiz de Montellano et al. (29). Metabolism of [14C]Coriandrin in Vitro. [14C]Coriandrin (1.90 mCi/mM, 3.9 µM) was added to 1 mL of incubation mixture containing 0.225 mg of microsomal protein/mL, an NADPHgenerating system (0.4 mM NADP, 0.6 mM MgCl2, 1 mM G-6P, 0.8 U of G-6-PD/mL), and 0.1 M phosphate buffer (pH 7.4). The reaction was carried out at 37 °C in air for 2, 5, 10, or 20 min. The reaction was terminated by adding 2 mL of ice-cold ethyl acetate/acetone (2:1) and then was extracted repeatedly with the same solvent (2 mL, four times; 1 mL, twice). The combined organic extracts were then evaporated under nitrogen and used for HPLC analysis. In HPLC analysis, [14C]coriandrin metabolites were separated on an Altex Ultrasphere octadecylsilane (ODS) column (4.6 mm × 25 cm) containing a DuPont series 8820 gradient. The mobile phase was a linear gradient of 0-60% acetonitrile in water over 60 min at 1 mL/min. Fractions of 0.5 mL were collected for determination of radioactivity by scintillation counting. Covalent Binding. The protein pellets remaining from in vitro metabolism experiments at 2, 5, or 10 min time point were extracted three times with 5 mL of methanol until no radioactivity was detected. Protein pellets were then dissolved in 0.5 mL of 1 N sodium hydroxide and diluted to 1 mL with water. Aliquots were taken for scintillation counting and for protein determination using the method of Lowry (27). Inhibition and Inactivation of Purified Human P450 1A1 and P450 1A2. Purified human P450 1A1 or 1A2 was reconstituted in a mixture containing 5 pmol of purified P450 1A1 or 1A2, 10 pmol of NADPH-P450 reductase, 30 µM dilauroylphosphatidylcholine (DLPC), an NADPH-generating system (0.336 mM NADP+, 0.5 mM G-6-P, 0.15 mM MgCl2, and
1 U of G-6-PD), ethoxyresorufin (ER) (5 µM), 0.1 M Tris buffer (pH 7.6), and various concentrations of coumarins in a total volume of 1 mL. Reaction mixtures were incubated for 10 min at 37 °C, and EROD activity was determined as previously described (14). For the analysis of enzyme inactivation, a two-stage incubation method was used as previously described (14). The firststage incubation mixture contained 66 pmol of purified P450 1A1, 132 pmol of NADPH-P450 reductase, 50 µM DLPC, an NADPH-generating system, 0.18 mM coriandrin (3), and 0.1 M Tris buffer (pH 7.6) in a total volume of 0.5 mL. At various times during the incubation (37 °C), an aliquot of the first-stage incubation mixture was diluted 20-fold into a second-stage incubation mixture containing 10 µM ER, 6.6 pmol of P450 1A1, and 13.2 pmol of reductase in a 1 mL total volume to determine EROD activity (14). Sodium Dodecyl Sulfate (SDS)-Polyacrylamide Gel Electrophoresis (PAGE). The samples, containing 100-200 µg of protein, were subjected to electrophoresis using a 12% acrylamide slab gel according to the Laemmli method (30). The gel was stained with Coomassie Brilliant Blue, and each lane was cut into 3 mm sections. Each gel section was then solublized in 0.5 mL of hydrogen peroxide for 24 h at 60 °C, and the radioactivity was determined by scintillation counting. In some cases, gels were subjected to a fluorographic procedure. Briefly, the gel was immersed in Me2SO for 1 h with one change of the solvent and then in 13% 2,5-diphenyloxazole in Me2SO for 3 h, followed by two washes (15 min) each of distilled water and 2% glycerol in water. The gel was dried and exposed to XAR-5 film (Kodak) for 3 weeks or 2 months at -70°C.
Results Effects of Naturally Occurring Coumarins on the Levels of P450 and Heme in Mouse Hepatic Microsomes in Vitro. In this series of experiments, various coumarins were incubated with an NADPHgenerating system and hepatic microsomes from either MC-treated or PB-treated mice, and then P450 content was determined. As shown in Table 1, all of the coumarins tested produced approximately 20-30% loss of P450 in MC-induced microsomes (a statistically significant reduction with all compounds tested, p
