Abstract
Naturally occurring isothiocyanates, such as benzyl isothiocyanate (BITC), are potent and selective inhibitors of carcinogenesis induced by a variety of chemical carcinogens. These effects appear to be mediated through favorable modification of both phase I and II enzymes involved in carcinogen metabolism. The inactivation of rat and human cytochromes P450 (P450s) in microsomes and the reconstituted system by BITC was investigated. BITC is a mechanism-based inactivator of rat P450s 1A1, 1A2, 2B1, and 2E1, as well as human P450s 2B6 and 2D6. BITC was most effective in inactivating P450s 2B1, 2B6, 1A1, and 2E1, whereas the activities of human P450 2C9 and rat P450 3A2 were not altered. The concentrations required for half-maximal inactivation (KI) of P450s 1A1, 1A2, 2B1, and 2E1 were 35, 28, 16, and 18 μM, respectively. The corresponding values forkinact were 0.26, 0.09, 0.18, and 0.05 min−1, respectively. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of P450 2B1 inactivated by [14C]BITC indicated specific and covalent modification of the P450 apoprotein by a metabolite of BITC. High-performance liquid chromatography analysis of the BITC metabolites revealed that benzylamine was the major metabolite and there were lesser amounts of benzoic acid, benzaldehyde,N,N′-di-benzylurea, andN,N′-di-benzylthiourea. Presumably, BITC was metabolized to the reactive benzyl isocyanate intermediate that covalently modified the P450 apoprotein or hydrolyzed to form benzylamine. BITC was an efficient inactivator of P450 2B1 with a partition ratio of approximately 11:1. This irreversible inactivation of P450s by BITC could contribute significantly to its chemopreventative action.
The cytochromes P450 (P450s) play an important role in the oxidative metabolism and detoxification of various drugs and carcinogens (Guengerich, 1991). P450 enzymes are able to incorporate one of the two atoms of an O2 molecule into a broad variety of substrates with concomitant reduction of the other oxygen atom by two electrons to produce H2O (Groves and Han, 1995). The resultant increases in polarity of the metabolites usually facilitate excretion or further detoxification of the products formed. Since P450s have also been shown to play key roles in the activation of a variety of carcinogens (Guengerich, 1991), the inhibition of P450-dependent carcinogen activation, especially by dietary substances, has been extensively studied. Naturally occurring isothiocyanates are released from their glucosinolate precursors through the activity of the enzyme myrosinase after chewing or maceration of cruciferous vegetables such as cabbage, cauliflower, and broccoli (Fenwick et al., 1983). Benzyl isothiocyanate (BITC) is released in significant concentrations from cabbage, radishes, Indian cress, garden cress, and mustard spinach (Fenwick et al., 1983).
Isothiocyanates have been shown to be potent and selective inhibitors of carcinogenesis induced by a variety of chemical carcinogens such as tobacco-derived nitrosamines and polycyclic aromatic hydrocarbons (Chung, 1992). These effects are partly due to the direct inhibition and/or down-regulation of the P450 responsible for carcinogen activation, resulting in decreased amounts of ultimate carcinogens formed (Zhang and Talalay, 1994). In addition, isothiocyanates have been shown to induce certain phase II enzymes responsible for the detoxification of electrophilic intermediates formed during phase I metabolism (Zhang and Talalay, 1994). The relative importance of these two mechanisms might differ among isothiocyanates, depending on their specificity to influence the specific enzymes involved, and must be determined individually. The important role of P450 enzymes in the metabolism of endogenous compounds, drug metabolism, and the detoxification of numerous xenobiotics (Guengerich, 1991) also has practical implications for this approach to chemoprevention. A third mechanism involving suppression of tumor promotion by an undefined mechanism has also been reported for BITC (Wattenberg, 1981).
Yang and coworkers (1994) have described the inhibition of several P450s, including P450s 1A1, 1A2, 2A1, 2B1, and 2E1, involved in carcinogen activation by isothiocyanates. BITC and phenethyl isothiocyanate (PEITC) have been shown to inhibitN-nitrosodimethylamine demethylation activity with IC50 values of 8 to 9 μM. The mechanism of inhibition by PEITC involves both competitive and metabolism-dependent inhibition of P450 2E1 (Ishizaki et al., 1990). BITC and PEITC have also been studied extensively for their role in the prevention of lung cancer. PEITC inhibits lung carcinogenesis induced by the tobacco-specific nitrosamine 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), whereas BITC is ineffective in this regard (Chung, 1992). In turn, BITC effectively inhibits benzo(a)pyrene (BaP)-induced lung tumors in A/J mice, whereas PEITC is ineffective (Lin et al., 1993). The synergistic combination of isothiocyanates in the prevention of human cancer such as lung cancer in individuals resistant to smoking cessation is also under investigation (Hecht, 1997).
The administration of isothiocyanates to rodents may produce either increases or decreases in microsomal P450 content and activity. The effects appear to be dependent on the experimental conditions, the isothiocyanate used, the treatment regimen, the target tissue examined, and the specific monooxygenase measured (Zhang and Talalay, 1994). Acute administration of PEITC to rats decreased liver P450 2E1 activity, whereas P450 2B1 activity and content increased approximately 10-fold, without any significant effect on lung P450 2B1 and P450 1A2 activities (Guo et al., 1992). However, chronic administration of PEITC increased levels of both P450 2B1 and P450 2E1 with related increases in carcinogen toxicity (Smith et al., 1993). Epidemiological data suggest that the consumption of fruit and vegetables decreases the risk for cancer development (Block et al., 1992), and isothiocyanates might therefore contribute significantly to these effects through modulation of P450 activities.
The mechanism of P450 inhibition by isothiocyanates is thought to involve reversible competitive inhibition as well as metabolism-dependent inhibition of P450 2E1 (Ishizaki et al., 1990) and P450 1A2 (Smith et al., 1996) by PEITC. We have recently also reported the mechanism-based inactivation of P450 2B1 (Goosen et al., 2000) and P450 2E1 (Moreno et al., 1999) by BITC. The inactivation of P450s 2B1 and 2E1 involves the formation of a reactive intermediate that covalently modifies the P450 apoprotein.
In the current report, the mechanism-based inactivation of several P450 isozymes by BITC, both in microsomes and in the reconstituted system, is described. In addition, BITC was found to be metabolized to the reactive benzyl isocyanate intermediate by P450 2B1. This intermediate then covalently modifies the P450 apoprotein.
Materials and Methods
Chemicals.
Phenobarbital, pyridine, β-napthoflavone, pregnenolone-16α-carbonitrile, p-nitrophenol, 4-nitrocatechol, erythromycin, dilauroyll-α-phosphatidylcholine (DLPC), NADPH, bovine serum albumin (BSA), and catalase were purchased from Sigma Chemical Co. (St. Louis, MO). Benzyl isothiocyanate (BITC), benzylamine, benzoic acid, 7-ethoxycoumarin, dimethyl sulfoxide , and sodium dithionite were purchased from Aldrich Chemical Co. (Milwaukee, WI). Resorufin, 7-ethoxyresorufin, 7-methoxyresorufin, 7-benzyloxyresorufin, and 7-ethoxy-4-(trifluoromethyl)coumarin (7-EFC) were purchased from Molecular Probes Inc. (Eugene, OR). 7-Hydroxy-4-(trifluoromethyl)coumarin (7-HFC) was from Enzyme Systems Products (Livermore, CA). N,N′-Di-benzylurea andN,N′-di-benzylthiourea were provided by Dr. M.-S. Lee (Wayne State University, Detroit, MI). HPLC-grade acetonitrile, ethyl acetate, and methanol were purchased from Fisher (Pittsburgh, PA). Hyperfilm-MP was obtained from Amersham Pharmacia Biotech (Cleveland, OH). Topp3 Escherichia coli cells were obtained from Stratagene (La Jolla, CA). [14C]BITC labeled at the α-carbon with a specific activity of 56 mCi/mmol and chemical purity >97% by HPLC was kindly provided by Dr. F.-L. Chung (American Health Foundation, Valhalla, NY). All other materials were of reagent grade and obtained from commercial sources.
Purification of P450 2B1 and P450 Reductase.
P450 2B1 was prepared from liver microsomes of fasted male Long-Evans rats (175–190 g; Harlan Sprague-Dawley, Indianapolis, IN) according to the method ofSaito and Strobel (1981). These rats were treated with 0.1% phenobarbital in their drinking water for 12 days. The cDNA for rat NADPH-cytochrome P450 reductase (hereafter referred to as “reductase”) within the expression plasmid pOR263 was expressed in E. coli Topp3 cells. The rat liver reductase was expressed and purified as described (Hanna et al., 1998).
Preparation of Microsomes.
Microsomes were prepared from liver homogenates of male Fisher 344 rats (175–190 g; Harlan Sprague-Dawley) as described previously (Coon et al., 1978). P450 2B1 was induced by i.p. injection of 100 mg/kg phenobarbital in water for 3 days. P450 2E1 was induced by i.p. injection of 100 mg/kg pyridine in water for 3 days. P450s 1A1 and 1A2 were induced by i.p. injections of 80 mg/kg β-napthoflavone in corn oil for 3 days. P450 3A2 was induced by i.p. injection of 50 mg/kg pregnenolone-16α-carbonitrile in corn oil for 3 days. Animals were fasted for 18 h after the last dose and sacrificed.
Microsomes from human lymphoblastic cells expressing P450 2B6 were from Gentest Corp. (Woburn, MA). Human liver P450s 2C9 and 2D6 were coexpressed with reductase in E. coli (He et al., 1999). The pCW vectors containing P450s 2C9 and 2D6 and the pACYC vector-containing reductase provided by Dr. Thomas Friedberg (University of Dundee, UK) were transformed into JM109 cells provided by Dr. Kan He (Parke-Davis Pharmaceutical Research, Ann Arbor, MI). The P450 concentrations of all preparations were determined from the reduced carbon monoxide difference spectra (Omura and Sato, 1964), recorded on a DW2 UV-visible spectrophotometer (SLM Aminco, Urbana, IL) equipped with an OLIS spectroscopy operating system (On-Line Instrument Systems, Inc., Bogart, GA).
Microsomal Enzyme Activity Assays.
Microsomal P450 1A1 activity was measured with 7-ethoxyresorufin as the substrate, and P450 1A2 activity was measured with 7-methoxyresorufin as the substrate (Burke et al., 1994). The primary reaction mixtures contained 10 μM P450 from microsomes of β-napthoflavone-treated rats in 50 mM Tris-HCl (pH 7.5); 50 mM MgCl2; 1, 10, or 100 μM BITC (in 1 μl of CH3OH/100 μl) or solvent in control samples. The primary microsomal incubation mixtures were all preincubated for 3 min at 30°C before initiation of the reactions with NADPH or water in reactions without NADPH. At 0 and 10 min after addition of NADPH (0.8 mM), the P450 1A1 or 1A2 activity was measured by transferring 20 μl (P450 1A1) or 40 μl (P450 1A2) of the primary reaction mixture into 980 μl or 960 μl, respectively, of a secondary reaction mixture containing either 5 μM 7-ethoxyresorufin (for 1A1) or 7-methoxyresorufin (for 1A2), 50 mM Tris-HCl (pH 7.5), 50 mM MgCl2, and 0.2 mM NADPH. Secondary reaction mixtures were quenched after 5 min (1A1) or 15 min (1A2) with 334 μl of ice-cold CH3CN before determining the fluorescence at room temperature on a SLM-Aminco model SPF-500 C spectrofluorometer with excitation at 522 nm and emission at 586 nm. For the determination of the kinetic parameters with BITC, the primary mixtures were incubated with increasing concentrations of BITC, and aliquots were removed at the indicated times. The residual 7-ethoxyresorufin or 7-methoxyresorufin activity remaining was measured as described above.
Microsomal P450 2B1 activity was assayed with 7-benzyl-oxyresorufin as the substrate (Nerarkur et al., 1993). The primary reaction mixtures contained 1.5 μM P450 2B1 from microsomes of phenobarbital-treated rats in 50 mM Tris-HCl buffer (pH 7.4); 1, 10, or 100 μM BITC (in 1.0 μl of CH3OH/100 μl); or solvent in control samples. At 0 and 10 min after addition of NADPH (1.2 mM), the P450 2B1 activity was assayed by transferring 25-μl aliquots to 975 μl of a secondary reaction mixture containing 5 μM benzyloxyresorufin and 0.2 mM NADPH in 50 mM Tris-HCl buffer (pH 7.4). The reaction was quenched after 3 min with 750 μl of ice-cold CH3OH. The resorufin product was measured spectrofluorometrically with excitation at 522 nm and emission at 586 nm. For the determination of the kinetic parameters with BITC, the primary mixtures were incubated with increasing concentrations of BITC. Aliquots were removed at the indicated times, and residual 7-benzyloxyresorufin activity was measured as described above.
The activities of P450s 2B6, 2C9, and 2D6 were determined using 7-EFC as a substrate (Buters et al., 1993; He et al., 1999). The primary reaction mixtures contained 160 nM P450 2B6 or 2.2 μM P450 2C9 in 50 mM potassium phosphate buffer (pH 7.4); 1, 10, or 100 μM BITC (in 1 μl of CH3OH/100 μl); or solvent in control samples. Primary reactions for P450 2D6 contained 1.1 μM P450 2D6 in 50 mM HEPES buffer (pH 7.5) containing 20% glycerol, 0.5 mM EDTA, and BITC as described before. Microsomal activity was assayed at 0 and 10 min after addition of 1.2 mM NADPH. The P450 2B6 activity was assayed by transferring 40 μl of the primary reaction mixture into 960 μl of a secondary reaction mixture containing 100 μM 7-EFC, 40 μg of BSA, 0.2 mM NADPH, and 50 mM potassium phosphate buffer (pH 7.4). The secondary reaction was allowed to proceed for 10 min at 30°C and was quenched by the addition of 334 μl of ice-cold CH3CN. The P450 2C9 activity was assayed by transferring 50 μl of the primary reaction to 250 μl of a secondary reaction mixture containing 100 μM 7-EFC, 120 μg/ml BSA, 10 mM MgCl2, and 0.2 mM NADPH in 50 mM potassium phosphate buffer (pH 7.4). After incubation for 10 min at 30°C, the reactions were stopped with 100 μl of ice-cold 30% CH3CN in 0.1 M Tris-HCl (pH 9.0) and centrifuged at 16,000g for 10 min before removal of the supernatant. The P450 2D6 activity was assayed by transferring 52 μl of the primary reaction to 948 μl of a secondary reaction mixture containing 200 μM 7-EFC, 40 μg/ml BSA, 0.2 mM NADPH in 50 mM HEPES buffer (pH 7.5) as above. After incubation for 15 min at 30°C, the reactions were stopped with 334 μl of ice-cold CH3CN. The formation of the deethylation product (7-HFC) was measured spectrofluorometrically at room temperature with the excitation at 410 nm and emission at 510 nm. The product formed was quantified based on a standard curve constructed using known amounts of 7-HFC.
Microsomal P450 2E1 activity was determined usingp-nitrophenol as the substrate. The primary reaction mixtures contained 5.5 μM P450 from microsomes of pyridine-treated rats in 100 mM potassium phosphate buffer (pH 6.8); 1, 10, or 100 μM BITC (in 1 μl of CH3OH/100 μl); or solvent in control samples. At 0 and 10 min after addition of NADPH (1 mM), the P450 2E1 activity was assayed by transferring 50-μl aliquots of the primary reaction mixture into 450 μl of a secondary reaction mixture containing 0.1 mM p-nitrophenol, 10 mM ascorbate (fresh), and 1 mM NADPH in 100 mM potassium phosphate buffer (pH 6.8). The reaction mixtures were incubated at 30°C. After 10 min the reactions were quenched with 100 μl of 1.5 M perchloric acid and left on ice for 20 min. The samples were centrifuged at 16000g at 4°C. The supernatant was carefully removed before addition of NaOH (1 M final) and spectrophotometric quantitation of the product at 490 nm. For the determination of the kinetic parameters with BITC, the primary mixtures were incubated with increasing concentrations of BITC, and aliquots were removed at the indicated times for the determination of residual p-nitrophenol activity remaining as described above.
Microsomal P450 3A2 activity was determined using erythromycin as a substrate. N-Demethylation of erythromycin was quantified by measuring the formation of formaldehyde (Nash, 1953). The primary reaction mixtures contained 4.1 μM P450 from microsomes of pregnenolone-16α-carbonitrile-treated rats in 50 mM potassium phosphate buffer (pH 7.4); 1, 10, or 100 μM BITC (in 1 μl of CH3OH/100 μl); or solvent in control samples. At 0 and 10 min after addition of NADPH (1.2 mM), the P450 3A2 activity was assayed by transferring 20 μl of the primary reaction mixture into 480 μl of a secondary reaction mixture containing 1 mM erythromycin and 1 mM NADPH in 50 mM potassium phosphate buffer (pH 7.4). After incubation for 10 min at 30°C, the reactions were stopped by adding 250 μl of 60% trifluoroacetic acid and the amount of formaldehyde formed was measured spectrophotometrically according toNash (1953).
SDS-PAGE Analysis for Specificity and Irreversibility of Binding.
Inactivation of purified P450 2B1 by BITC was investigated using a reconstituted mixture containing 4 μM P450 2B1, 4 μM reductase, 200 μg/ml DLPC, 55 μM [14C]BITC, 90 U of catalase, and 50 mM Tris-HCl (pH 7.4) in a total volume of 85 μl. The primary incubation mixture was incubated at 30°C for 3 min before initiation of the reaction with 1.2 mM NADPH or water in reactions without NADPH. At 0 and 12 min after addition of NADPH, aliquots of 5 μl (20 pmol of P450 2B1) were taken from the primary reaction and added to a secondary reaction similar to that used for the determination of residual P450 2B6 activity. The formation of 7-HFC was measured as described for P450 2B6.
Aliquots containing 0.1 nmol of P450 2B1 were removed at 0 and 12 min and diluted with sample loading buffer, boiled for 3 min, and loaded on a 10% polyacrylamide gel and electrophoresed with the buffer system described by Laemmli (1970). After electrophoretic separation of the proteins, the gels were either stained with 0.25% Coomassie Blue R-250 or transferred to a Immobilon-PSQ polyvinylidene difluoride microporous membrane with 25 mM Tris base containing 192 mM glycine, and 30% methanol at 100 mV for 1 h. Autoradiography was performed by exposing the membrane to Biomax MR film supplemented with a Biomax intensifying screen (Eastman Kodak Co., Rochester, NY) for 1 week at −80°C.
Metabolism of BITC by P450 2B1.
Metabolites formed during incubation with P450 2B1 were analyzed by HPLC and gas chromatography-mass spectrometry (GC-MS). Purified P450 2B1 and reductase were reconstituted with DLPC for 1 h at 4°C. The reaction mixture contained 95.8 μM [14C]BITC, 2 μM P450 2B1, 2 μM reductase, 200 μg/ml DLPC, 400 U of catalase, and 50 mM potassium phosphate buffer (pH 7.4) in a total volume of 0.5 ml. The reactions were initiated with 1.2 mM NADPH or by adding water to the reactions without NADPH. After 30 min at 30°C the reactions were stopped by addition of 1 ml of ice-cold ethyl acetate. The samples were extracted two more times with ethyl acetate, and the organic phases were pooled. The remaining reaction mixture was adjusted to pH 11.0 with NaOH and extracted twice with 1-ml portions of ethyl acetate. The reaction mixture was re-adjusted to pH 3.0 with HCl and extracted twice with 1 ml of ethyl acetate. Approximately 85% of the14C label was recovered. The extracts were combined and dried using a Speed-Vac (Savant, Farmingdale, NY). To prevent loss of the volatile metabolites, 50 μl of dimethyl sulfoxide was added and the residual ethyl acetate was evaporated completely. Metabolites were resolved by HPLC on a 5-μm reversed-phase C18 column (4.6 × 250 mm, Rainin, Ultrasphere-ODS) (Varian, Walnut Creek, CA). The HPLC system consisted of a Waters (Milford, MA) 490E programmable variable wavelength detector, Waters 501 HPLC pumps, a Waters system interface module, and a fraction collector (model 201, Gilson, Middleton, WI). The system was operated through the Maxima 820 chromatography workstation from Waters. The solvent system consisted of solvent A (5% CH3CN/1% acetic acid/H2O) and solvent B (80% CH3CN/1% acetic acid/H2O) adjusted to pH 4.2 with 5 M potassium hydroxide. Initial conditions were 5% B at a flow rate of 1 ml/min, increasing to 35% B in 5 min, then to 65% B in 30 min, and finally 90% B in 5 min. The solvent was maintained at 90% B for 5 min before returning to initial conditions. Fractions were collected every 0.6 min and monitored by liquid scintillation counting on a liquid scintillation counter (model LS-5801, Beckman, Berkeley, CA).
For GC-MS analysis of metabolites, purified P450 2B1 and reductase were reconstituted as before. The reaction mixtures contained 50 μM BITC, 3 μM P450 2B1, 3 μM reductase, 200 μg/ml DLPC, 400 U of catalase, and 50 mM potassium phosphate buffer (pH 7.4) in a total volume of 1.0 ml. The reactions were initiated by addition of 1.4 mM NADPH and incubated for 30 min at 30°C. The reactions were stopped by the addition of 1.5 ml of methylene chloride on ice. This was followed by two extractions with 1.5 ml of methylene chloride, the pH of the mixtures was adjusted to 3.0 and then 11.0, and the mixtures were re-extracted twice with 1.5 ml of methylene chloride at each pH. The extracts were combined and dried using a Speed-Vac (Savant) to approximately 500 μl. The extracts from five separate reactions were combined and dried over anhydrous sodium sulfate before final evaporation in the Speed-Vac to approximately 5 μl.
Analysis by GC-MS was carried out on a Finnigan MAT 4500 mass spectrometer (ThermoQuest, San Jose, CA) coupled to a HP 5890 gas chromatograph via a heated interface. This GC-MS system employed a Galaxy data system, manufactured by LGC Inc. (Los Gatos, CA). Gas chromatographic separation employed a DB-5 capillary column (30 m × 0.32 mm i.d. × 1.0-μm film thickness) purchased from Altech Associates, Inc. (Deerfield, IL). Helium gas flow was maintained at approximately 10-psi head pressure, and the column was installed in a splitless configuration. The gas chromatograph temperature program was initiated at 50°C and raised at 10°C/min to 275°C. Mass spectrophotometric conditions were as follows: electron impact ionization, with 70-eV electron energy. The ion source temperature was maintained at 150°C.
Statistics.
Data were analyzed where appropriate using the Student's t test. A p value of <0.05 was considered to be a statistically significant difference.
Results
Effect of BITC on the Activities of P450s.
The effects of BITC on the activities of P450s 1A1, 1A2, 2B1, 2B6, 2C9, 2D6, 2E1, and 3A2 were examined. Microsomes were incubated with different concentrations of BITC or solvent added to control samples as described underMaterials and Methods. Kinetic parameters (Table1) were calculated from the rates of inactivation of the P450s incubated with increasing concentrations of BITC (Fig. 1). Both P450 1A1 and P450 1A2 were inactivated in a time- and concentration-dependent manner. A concentration of 100 μM BITC resulted in 62% inactivation of the 7-ethoxyresorufin (P450 1A1) or 52% inactivation of the 7-methoxyresorufin (P450 1A2) activity, following 10-min incubations (Fig. 2). Kinetic studies revealed that, although the concentrations of BITC required for half-maximal inactivation (KI) of P450 1A1 and P450 1A2 were similar, the rate of inactivation of P450 1A1 was much faster than that of P450 1A2 (Table 1). This difference was also seen in Fig.2 where 10 μM BITC resulted in a statistically significant loss of P450 1A1 activity with no significant loss in P450 1A2 activity. With P450 1A1 and especially P450 1A2 a loss in activity at time 0 was observed suggesting inhibition due to carryover into the secondary reaction with higher concentrations of BITC (Fig. 1). These data indicate that BITC is more effective as an inactivator of P450 1A1 activity when compared with inactivation of P450 1A2.
P450s 2B1 and 2B6 were both inactivated by BITC to similar extents in time- and concentration-dependent manners. A concentration of 100 μM BITC resulted in an 88% and 70% inactivation of the 7-benzyloxyresorufin (P450 2B1) and the 7-EFC (P450 2B6) activity, respectively (Fig. 2). There was no statistically significant difference in the amount of P450 2B1 or 2B6 activity remaining following incubation with 1 or 10 μM BITC. A concentration as low as 10 μM already resulted in approximately 50% loss in activity of both these enzymes following a 10-min incubation. P450 2E1 activity was inactivated by 71% when incubated with 100 μM BITC and by 27% when incubated with 10 μM BITC (Fig. 2). As shown in Table 1, theKI value for the inactivation of P450 2E1 was similar to that required for the inactivation of P450 2B1. However, the rate of inactivation (kinact) was much slower for P450 2E1 than the rate of inactivation of P450 2B1.
Human P450 2D6 was also inactivated by BITC in a mechanism-based manner, but the rate of inactivation appeared to be very slow since approximately 75% of the activity remained following a 10-min incubation with 100 μM BITC. As shown in Fig. 2, BITC did not cause a statistically significant metabolism-dependent inhibition of human P450 2C9 or rat P450 3A2 activities when incubated with up to 100 μM BITC.
Specificity of BITC Binding.
The specificity of the radioactive labeling of the proteins in the reconstituted incubation mixture by BITC was investigated by separation of the proteins using SDS-PAGE followed by autoradiography as described under Materials and Methods. Radiolabeled BITC was bound to all proteins in the reconstituted system (Fig. 3A) in the absence of NADPH. This nonspecific labeling of proteins did not result in any loss of catalytic activity. Samples incubated with BITC and NADPH under conditions where more than 80% of the 7-EFCO-deethylation activity was lost showed a marked increase in radiolabel only on the P450 band (Fig. 3A, lane 4). That this increase was NADPH- and BITC-specific and not due to uneven loading of samples can be seen from the Coomassie Blue stain shown in Fig. 3B. These data indicate that the labeling of P450 2B1 by a metabolite of BITC was covalent and specific for P450 2B1.
Analysis of Metabolites of BITC.
The metabolites formed by incubating radiolabeled BITC with P450 2B1 in the reconstituted system were analyzed by reversed-phase HPLC with UV detection at 254 and 286 nm or by liquid scintillation counting as described underMaterials and Methods. Seven metabolites were separated from the ethyl acetate extracts of samples incubated with NADPH as detected by liquid scintillation counting of fractions (Fig.4). The peak retention time of the major metabolite, which accounted for almost 50% of the total product formed, corresponded to benzylamine (Table2). The other metabolites identified by coelution with authentic standards were identified as benzoic acid, benzaldehyde, N,N′-di-benzylurea, andN,N′-di-benzylthiourea. In control samples without NADPH, two products corresponding to benzylamine andN,N′-di-benzylthiourea were also detected (Fig.4). However, the amount of benzylamine formed in the control reaction was less than 30% of the amount formed in the experimental reactions (Table 2). There was no significant difference between the amount ofN,N′-di-benzylthiourea formed in reactions with or without NADPH. Two additional products eluting at 4.8 and 12.7 min were detected by liquid scintillation counting and accounted for less than 16% of the total products formed in the samples incubated with NADPH.
The amounts of product formed, as shown in Table 2, were quantified by integration of the liquid scintillation data since only benzaldehyde and N,N′-di-benzylthiourea were formed in sufficient amounts to be quantified by UV detection. The average yield of product specifically formed in the NADPH-dependent reaction was 10.6 nmol of product/nmol of P450 2B1. The enzyme lost more than 90% of its original activity under the conditions used to determine product formation. Therefore, the partition ratio, defined as the number of metabolite molecules produced per enzyme molecule inactivated (Silverman, 1996), is approximately 11:1.
The identities of benzylamine, benzoic acid, benzaldehyde, andN,N′-di-benzylurea were confirmed by GC-MS analysis as described under Materials and Methods. The product ion spectra and GC retention times were compared with the authentic standards. Benzylamine eluted at 7.6 min and gave characteristic ion fragments at m/z (relative %) 107 (M+, 61.1%), 106 (100%), 91 (6.2%), and 77 (41.3%) by loss of H, NH, and CH2, respectively. Benzoic acid eluted at 9.4 min and gave characteristic ion fragments atm/z (relative %) 122 (M+, 45.5%), 105 (54.6%), and 77 (47.9%) by loss of OH and CO, respectively. Benzaldehyde eluted at 7.1 min and gave characteristic ion fragments at m/z (relative %) 106 (M+, 69.2%), 105 (54.6%), and 77 (100%) by loss of H and CO, respectively.N,N′-Di-benzylurea eluted at 8.3 min and gave no parent ion fragment with characteristic ion fragments atm/z (relative %, fragment lost) 163 (4.5%, M+-C6H5), 134 (100%, M+-PhCH2NH), 106 (50.5%, M+-PhCH2NHCO), 91 (35.6%, M+-PhCH2(NH)2CO), and 77 (13.4%, M+-Ph(CH2)2(NH)2CO). BITC eluted at 11.2 min under these conditions.
Discussion
The results reported here demonstrate that BITC is a mechanism-based inactivator of rat P450s 1A1, 1A2, 2B1, and 2E1, and human P450s 2B6 and 2D6. Human P450 2C9 and rat P450 3A2 were not inactivated by BITC. P450s 1A1 and 1A2 were inactivated in a time- and NADPH-dependent manner. BITC was shown to be an efficient inactivator of P450 1A1 with KI andkinact values of 35 μM and 0.26 min−1, respectively. TheKI for P450 1A2 (28 μM) was similar to the KI value obtained for P450 1A1, but the rate of inactivation was much slower than that of P450 1A1. These results are in good agreement with the IC50values reported for the inhibition of ethoxyresorufin activity (54 μM) by BITC in microsomes from 3-methylcholanthrene-induced rats (Conaway et al., 1996). It is possible that the reactive intermediate responsible for the inactivation of P450 1A2 is released more readily from the active site of the enzyme, thereby increasing the partition ratio for inactivation and effectively decreasing the rate of inactivation (Silverman, 1996). It is also possible that BITC is preferentially oxidized to a different product, possibly benzaldehyde, instead of being desulfurated to benzyl isocyanate as shown in Fig.5. One component of the inhibition of P450 1A1 and P450 1A2 by BITC was also seen to be metabolism-independent as evidenced by the inhibition of catalytic activity when assayed at time 0 (Fig. 1). Both of these enzymes have been investigated extensively for their roles in carcinogen activation and metabolism. Many reports implicate a role for increased P450 1A1 levels in lung cancer (McLemore et al., 1990), and the inhibition of P450 1A1 may be involved in the inhibition of BaP-induced lung carcinogenesis by BITC (Lin et al., 1993). P450 1A2 is involved in the metabolic activation of tobacco-derived NNK (Smith et al., 1996), aflatoxin B1, and other carcinogenic aryl amines and heterocyclic amines (Guengerich, 1995).
The inactivation of P450s 2B1 and 2B6 displayed characteristics of mechanism-based inactivation, including a time- and concentration-dependent loss in catalytic activity. The dependence on NADPH for inactivation indicated that BITC had to be metabolized to a reactive intermediate responsible for the inactivation process. BITC was shown to be the most selective in inactivating P450 2B1 and 2B6 compared with the other P450s examined in this study. TheKI for the inactivation of P450 2B1 in microsomal preparations was 16 μM, andkinact was 0.18 min−1. These results were comparable to theKI andkinact of 5.8 μM and 0.66 min−1 determined using purified P450 2B1 (Goosen et al., 2000). The inactivation of human P450 2B6 by BITC has biological importance, because this enzyme has been shown to activate several carcinogens, including BaP, NNK, and aflatoxin B1 (Code et al., 1997). Although P450 2B6 is expressed in low levels in human liver (Guengerich, 1995), it has been found to be expressed in lung and uterine endometrium and is induced in patients with breast cancer (Hellmold et al., 1998). One would therefore expect tissue-specific activation of carcinogens.
The KI andkinact for P450 2E1 inactivation in microsomal preparations were 18 μM and 0.05 min−1, respectively, and this is in accordance with values obtained using purified P450 2E1 (Moreno et al., 1999). It appears that BITC is a more efficient inactivator of P450 2B1 than P450 2E1 as reflected in the slower inactivation rate and larger partition ratio for inactivation of P450 2E1. The partition ratio for inactivation of P450 2B1 determined from the amount of product released per mole of enzyme inactivated is approximately 11 compared with 27 for P450 2E1. This could explain the in vivo effects on these isozymes described earlier. P450 2D6 is involved in the metabolism of a significant number of drugs, and different phenotypes may be associated with diseases such as Parkinsonism and various cancers (Guengerich, 1995). BITC also inactivated P450 2D6 in a mechanism-based manner. However, the rate of inactivation was slow since 100 μM BITC resulted in only approximately 20% loss of activity in 10 min. Human P450 2C9 and rat P450 3A2 were not inactivated by BITC.
The mechanism of inactivation of hepatic microsomal P450s is thought to proceed through one of three characterized pathways (Osawa and Pohl, 1989): formation of a reactive intermediate that covalently modifies the heme moiety, destruction of the heme with cross-linking to the apoprotein, or covalent modification of the P450 apoprotein. The separation of P450 2B1 by SDS-PAGE followed by autoradiography, indicated a specific increase in radiolabeled metabolite associated with the P450 apoprotein. Modification of the heme is not implicated, because the heme would be dissociated from the protein under the conditions of SDS-PAGE and previous studies indicated no apparent heme modification (Goosen et al., 2000). The radiolabel remained bound to the protein under denaturing conditions, and this indicates a covalent modification of the apoprotein by a metabolite of BITC. Stoichiometric calculations revealed that approximately 1 mol of radiolabeled metabolite was associated per mole of enzyme inactivated (Goosen et al., 2000). Taken together, these data suggest that a critical amino acid in the active site of the enzyme is modified during metabolism, which then prevents further binding or catalysis of substrate.
In an attempt to identify the reactive intermediate involved in enzyme inactivation, the metabolites formed by this reaction were analyzed by HPLC and GC-MS. As shown in Table 2, the major metabolite identified was benzylamine. The formation of isocyanates from isothiocyanates was first reported by Lee (1992), who described the conversion of 2-naphthyl isothiocyanate to 2-naphthyl isocyanate, and subsequently the formation of benzyl isocyanate from BITC (Lee, 1996). Presumably, BITC is oxidatively desulfurated to the putative product benzyl isocyanate (Fig. 5). The greater electronegativity of oxygen compared with sulfur implies increased reactivity of the central carbon atom on the isocyanate functionality. Accordingly, hydrolysis of benzyl isocyanate to benzylamine is rapid and complete. Alternatively, benzyl isocyanate could also react with benzylamine to form the corresponding carbamate, N,N′-di-benzylurea. BITC in turn can also react with benzylamine to yield the thiocarbamate,N,N′-di-benzylthiourea. The formation ofN,N′-di-benzylurea is dependent on the formation of the benzyl isocyanate intermediate since it has been shown that diarylthioureas do not desulfurate readily in vivo, whereas monoarylthioureas desulfurate considerably (Lee, 1996). It is further believed that benzyl isocyanate partitions between hydrolysis or alternatively reacts with nucleophilic amino acid residues in the active site of the enzyme. This is evident from the covalent attachment of a 14C-labeled metabolite of BITC to the apoprotein of P450 2B1. These observations are similar to those made byEl-Hawari and Plaa (1977), who showed that protein binding of 1-[3H]naphthyl isothiocyanate (labeled at the 4-position of the ring) or 1-[14C]naphthyl isothiocyanate (labeled in the isothiocyanate moiety) to rat liver microsomes was NADPH-dependent. However, the reactive binding species was not identified, presumably 1-naphthyl isocyanate by analogy to the present findings. In control reactions without NADPH, the formation of benzylamine from spontaneous hydrolysis of BITC and the formation ofN,N′-di-benzylthiourea was also observed. The amount of N,N′-di-benzylthiourea was similar to that formed in experimental reactions, whereas there was an almost 4-fold increase in the formation of benzylamine in these reactions, as would be expected by the formation of the benzyl isocyanate intermediate.
A second pathway for the metabolism of BITC by cytochrome P450 2B1 was also identified in this study. BITC could be oxidized at the α-carbon to yield benzaldehyde and the thiocyanate anion. Alternatively, benzylamine could be deaminated to give benzaldehyde and ammonia. The mechanism for the formation of benzaldehyde was not investigated here, but the possible role for the formation of the thiocyanate anion in the tumor suppression by BITC (Wattenberg, 1981) favors this pathway. The subsequent oxidation of benzaldehyde would yield benzoic acid. The formation of benzoic acid was also observed in dogs, where administration of BITC resulted in the excretion of hippuric acid, the glycine conjugate of benzoic acid (Brüsewitz et al., 1977). In humans and rats, BITC is metabolized through conjugation with GSH and finally excreted as mercapturic acid (Brüsewitz et al., 1977). This second pathway might also be important in the metabolism of BITC by P450s which are not inactivated by BITC.
The extensive conjugation of BITC with GSH (Brüsewitz et al., 1977) might also be important for the in vivo effects of BITC. Conjugation is usually considered to be a detoxification process but might also act as a transport mechanism for BITC with subsequent cleavage at peripheral organs (Meyer et al., 1995). The half-life of the isocyanate is extremely short and would therefore reduce any local tissue reactions. However, conjugation of the isocyanate product with GSH could contribute to mutagenicity, because some isocyanates are known to be mutagenic and toxic (Raulf-Heimsoth and Baur, 1998). Therefore, local release of the conjugated product might also contribute to the toxicity of BITC (Hirose et al., 1998) and should be evaluated when these compounds are considered for dietary supplementation to prevent cancer.
In summary, it was clearly demonstrated that the naturally occurring isothiocyanate, BITC, acts as a mechanism-based inactivator of rat P450s 1A1, 1A2, 2B1, and 2E1 and human P450s 2B6 and 2D6. The inactivation of purified P450 2B1 probably proceeded through metabolism of BITC to the reactive benzyl isocyanate intermediate, which covalently modified the P450 apoprotein. It is believed that this inactivation of several P450s involved in carcinogen activation might contribute significantly to its chemopreventative effect.
Acknowledgments
We thank Dr. Ute M. Kent and Hsia-Lien Lin for preparation of rat liver microsomes and purification of P450 2B1 and reductase. We are also grateful to Dr. Fung-Lung Chung, who provided us with the [14C]BITC, and Dr. Mei-Sie Lee, who provided some of the authentic standards. We thank Dr. Alfin D. N. Vaz for help with the GC-MS analyses and for helpful discussions and suggestions.
Footnotes
-
Send reprint requests to: Dr. Paul F. Hollenberg, Dept. of Pharmacology, Medical Science Research Bldg. III, 1150 West Medical Center Dr., Ann Arbor, MI 48109-0632. E-mail:phollen{at}umich.edu
-
This publication was supported in part by grants from the Foundation for Pharmaceutical Education (SA Druggist) (to T.C.G.), the Potchefstroom University for Christian Higher Education (to T.C.G.), and National Institutes of Health Grants CA 16954 (to P.F.H.) and CA 46535 (to F.-L.C.) from the National Cancer Institute.
- Abbreviations:
- P450s
- cytochromes P450
- BaP
- benzo(a)pyrene
- BITC
- benzyl isothiocyanate
- DLPC
- l-α-phosphatidylcholine dilauroyl
- 7-EFC
- 7-ethoxy-4-(trifluoromethyl)coumarin
- GC-MS
- gas chromatography-mass spectrometry
- GSH
- glutathione
- 7-HFC
- 7-hydroxy-4-(trifluoromethyl)- coumarin
- reductase
- NADPH-cytochrome P450 reductase
- NNK
- 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone
- PEITC
- 2-phenethyl isothiocyanate
- PAGE
- polyacrylamide gel electrophoresis
- Received June 26, 2000.
- Accepted September 25, 2000.
- The American Society for Pharmacology and Experimental Therapeutics