- Email: [email protected]
Identification and characterization of reactive metabolites in myristicin-mediated mechanism-based inhibition of CYP1A2
Chemico-Biological Interactions 237 (2015) 133–140
Contents lists available at ScienceDirect
Chemico-Biological Interactions journal homepage: www.elsevier.com/locate/chembioint
Identification and characterization of reactive metabolites in myristicin-mediated mechanism-based inhibition of CYP1A2 Ai-Hong Yang, Xin He ⇑, Jun-Xiu Chen, Li-Na He, Chun-Huan Jin, Li-Li Wang, Fang-Liang Zhang, Li-Jun An School of Chinese Materia Medica, Tianjin University of Traditional Chinese Medicine, Tianjin 300193, PR China
a r t i c l e
i n f o
Article history: Received 21 November 2014 Received in revised form 6 June 2015 Accepted 9 June 2015 Available online 16 June 2015 Keywords: Myristicin Reactive metabolites Mechanism-based inhibition Cytochrome P450 enzymes Human liver microsomes
a b s t r a c t Myristicin belongs to the methylenedioxyphenyl or allyl-benzene family of compounds, which are found widely in plants of the Umbelliferae family, such as parsley and carrot. Myristicin is also the major active component in the essential oils of mace and nutmeg. However, this compound can cause adverse reactions, particularly when taken inappropriately or in overdoses. One important source of toxicity of natural products arises from their metabolic biotransformations into reactive metabolites. Myristicin contains a methylenedioxyphenyl substructure, and this specific structural feature may allow compounds to cause a mechanism-based inhibition of cytochrome P450 enzymes and produce reactive metabolites. Therefore, the aim of this work was to identify whether the role of myristicin in CYP enzyme inhibition is mechanism-based inhibition and to gain further information regarding the structure of the resulting reactive metabolites. CYP cocktail assays showed that myristicin most significantly inhibits CYP1A2 among five CYP enzymes (CYP1A2, CYP2D6, CYP2E1, CYP3A4 and CYP2C19) from human liver microsomes. The 3.21-fold IC50 shift value of CYP1A2 indicates that myristicin may be a mechanism-based inhibitor of CYP1A2. Next, reduced glutathione was shown to block the inhibition of CYP1A2, indicating that myristicin utilized a mechanism-based inhibition. Phase I metabolism assays identified two metabolites, 5-allyl-1-methoxy-2,3-dihydroxybenzene (M1) and 1’-hydroxymyristicin or 2’,3’-epoxy-myristicin (M2). Reduced glutathione capturing assays captured the glutathione-M1 adduct, and the reactive metabolites were identified using UPLC-MS2 as a quinone and its tautomer. Thus, it was concluded that myristicin is a mechanism-based inhibitor of CYP1A2, and the reactive metabolites are quinone tautomers. Additionally, the cleavage process of the glutathione-M1 adduct was analyzed in further detail. This study provides additional information on the metabolic mechanism of myristicin inhibition and improves risk evaluation for this compound. Ó 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Myristicin, 1-allyl-3,4-methylenedioxy-5-methoxybenzene (Fig. 1), is commonly found in parsley, carrot, black pepper, nutmeg, and many natural oils and flavoring agents [1–3]. For example, in the essential oil of mace, myristicin is the major component of the aromatic ether fraction. Myristicin is also one of main components in nutmeg volatile oils [4]. Some herbs containing myristicin, such as nutmeg, mace and dill preparations, Abbreviations: MBI, mechanism-based inhibition; RMs, reactive metabolites; CYPs, cytochrome P450 enzymes; GSH, reduced glutathione; HLMs, human liver microsomes; NADPH, b-nicotinamide adenine dinucleotide 2’-phosphate reduced tetrasodium salt; MI complex, metabolite intermediate complex. ⇑ Corresponding author at: School of Chinese Materia Medica, Tianjin University of Traditional Chinese Medicine, Tianjin 300072, PR China. Tel.: +86 22 59596231; fax: +86 22 59596153. E-mail address: [email protected] (X. He). http://dx.doi.org/10.1016/j.cbi.2015.06.018 0009-2797/Ó 2015 Elsevier Ireland Ltd. All rights reserved.
are also used in traditional medicine to treat rheumatism, cholera, psychosis, stomach cramps, nausea, diarrhea, and anxiety [5,6]. Myristicin has been shown to induce glutathione S-transferase [7] and inhibit benzo[a]pyrene-induced tumorigenesis [8]. However, traditional medicines can also cause adverse reactions, especially when taken inappropriately or in overdoses. Myristicin has weak monoamine oxidase inhibitor properties that are responsible for some cardiovascular symptoms [9]. Myristicin displayed apoptotic activity mediated by the activation of caspases in hamster ovary CHO cells [10], and it also altered mitochondrial membrane function in human leukemia K562 cells, inducing apoptosis and down-regulating DNA damage response genes [11]. Myristicin has also been reported to cause psychotropic effects [12] and even death [5] at high dosages in humans, in which myristicin is metabolized to 3-methoxy-4,5-methylenedioxyampheta mine, the hallucinogenic effect of which was more pronounced than that of mescaline [12,13]. Therefore, as the toxicity of this
134
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
obtained from Sigma-Aldrich (Milwaukee, WI, USA). Pooled HLMs, which were purchased from the Research Institute for Liver Diseases Co., Ltd. (Shanghai, China), were prepared from human liver tissues under Chinese organ donation regulations with the full consent of the patients. All other reagents were of high-performance liquid chromatography grade. Fig. 1. Structural formula of myristicin.
2.2. CYP cocktail assays—screening for significant myristicin-mediated inhibition of CYPs
compound is related to its xenobiotic metabolism, the metabolism of myristicin should be more thoroughly studied to improve risk evaluation. Tertiary aminopropiophenones were identified as urinary metabolites from rats and guinea pigs treated with myristicin [14], while rat liver converted myristicin into 3-methoxy-4, 5-methylenedioxy amphetamine [15]. In vivo and in vitro studies of rats demonstrated that myristicin could also be transformed into 5-allyl-1-methoxy-2,3-dihydroxybenzene and 1’-hydroxymyristicin [16]. Through the study of purified recombinant human liver CYPs (cytochrome P450 enzymes) and immunoinhibition assays, Yun et al. determined that the oxidation of myristicin to 5-allyl-1-methoxy-2,3-dihydroxybenzene was catalyzed by CYP3A4 and CYP1A2 in humans [17]. Although these metabolites were characterized in the above reports, the effects of myristicin on the CYPs involved in its metabolism, in particular the identification of RMs (reactive metabolites) in myristicin-mediated MBI (mechanism-based inhibition), remain unknown. The metabolism of drugs by CYPs to form RMs that bind covalently to a catalytic site or sites of the enzyme itself leads to irreversible inhibition of the enzymes. This phenomenon is referred as MBI. MBI inhibitors have very specific features that make them recognizable by in vitro testing. Fontana et al. [18] summarized the most common substructures causing the MBI of CYPs, including methylenedioxyphenyl-containing compounds such as myristicin and safrole. The drug-induced MBI of CYPs is initiated by conversion of the drug into highly RMs, and this group of drugs may be involved in metabolism-dependent drug toxicities. In this work, we demonstrated that myristicin is an MBI inhibitor of CYP1A2 and determined that its reactive metabolites are quinone tautomers using the following assays: (i) Using a CYP cocktail assay, the effects of myristicin on specific marker reactions of five CYP isoforms were measured in HLMs (human liver microsomes). Myristicin inhibited CYP1A2 more strongly than it inhibited the other CYPs. (ii) It could be estimated that myristicin may be the MBI inhibitor of CYP1A2 according to IC50 shift value. (iii) In phase I metabolic reactions, NADPH-dependent inhibition assays were used to identify the generation of two metabolites (M1 and M2). (iv) In GSH capturing assays, GSH acted as a trapping agent to identify the structure of the RMs. It was concluded that myristicin could inhibit CYP1A2 through an MBI mechanism and produce electrophilic reactive metabolites with tautomeric structures (quinone and its tautomer). Such information concerning the MBI by this compound and its RMs can be helpful for thoroughly understanding the drug metabolism of myristicin.
The same experimental conditions described above in 2.2. Cocktail assays were used in the IC50 shift assays except that six different concentrations of myristicin were used: 0.5, 1.6, 4, 8, 20 and 50 lM. The assays were performed in triplicate for all test specimens. The IC50 values were calculated using GraphPad Prism version 5.0 (GraphPad Software, San Diego, CA, USA). Based on the IC50 shift value, we predicted that GSH would be able to block the inhibition of CYP1A2 by myristicin in the presence of NADPH and designed an assay to test this prediction. The same experimental conditions described above were used except for the following two exceptions: (1) Phenacetin was chosen as a probe substrate for CYP1A2, and (2) the final incubation concentration of myristicin was 2 lM, which was close to its IC50 value of 1.796 lM. The assays were performed in triplicate. The relative activity (%) of CYP1A2 was monitored to determine whether GSH was able to block inhibition.
2. Materials and methods
2.4. Phase I metabolism—identification of the metabolites
2.1. Materials
Mixtures of 50 lL myristicin (stock concentration: 200 lM; final concentration: 50 lM) and 50 lL HLMs (stock concentration: 8 mg protein/mL; final concentration: 2 mg protein/mL) were pre-incubated 5 min at 37 °C. Subsequently, 100 lL NADPH (stock concentration: 2 mM; final concentration: 1 mM) were added to the above reaction mixture. After incubation at 37 °C for 2 h, the reaction was stopped by adding 400 lL ice-cold methanol. After centrifugation at 10,000 rpm for 10 min, the supernatant was
Myristicin was purchased from the Chinese National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). Phenacetin, midazolam, dextromethorphan, (s)-mephenytoin, chlorzoxazone and NADPH (b-nicotinamide adenine dinucleotide 2’-phosphate reduced tetrasodium salt) were obtained from Sigma Chemical Co. (St. Louis, MO, USA). GSH was
CYP cocktail assays were employed to analyze and screen the myristicin-mediated inhibition of five CYP isoforms. Phenacetin (CYP1A2), dextromethorphan (CYP2D6), chlorzoxazone (CYP2E1), midazolam (CYP3A4), and (s)-mephenytoin (CYP2C19) were chosen as probe substrates for each corresponding CYP. The assays consisted of two steps: (i) A mixture of 20 lL myristicin (stock concentration: 200 lM or 2 mM) and 20 lL HLMs (stock concentration: 8 mg protein/mL) were pre-incubated in the presence or absence of 40 lL NADPH (denoted as +/ NADPH; stock concentration: 2 mM) for 30 min at 37 °C. For the ( )NADPH reaction, 40 lL of 0.1-M potassium phosphate buffer, pH 7.4, was used instead of NADPH. (ii) Next, 80 lL of the probe substrates and 100 lL NADPH were added to the 20 lL pre-incubation mixtures and incubated for 30 min under the same conditions. The reactions were stopped by adding a 2-fold volume (400 lL) of ice-cold methanol containing 50-ng/ml carbamazepine. Control reactions were performed without myristicin. The final concentrations of the components were as follows: myristicin (5 lM or 50 lM), HLMs (0.2 mg protein/mL), and NADPH (1 mM). The final incubation concentrations of the probe substrates were fixed around their reported Km values: phenacetin (10 lM), dextromethorphan (2.5 lM), chlorzoxazone (20 lM), midazolam (5 lM), and (s)-mephenytoin (20 lM). The assays were performed in triplicate for all test specimens. 2.3. IC50 shift assays and the blocking of CYP1A2 inhibition by GSH—an evaluation of myristicin MBI potential
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
collected and analyzed by UPLC-ESI-MS2. A control reaction was performed without myristicin. 2.5. GSH capture of the reactive metabolites The incubation method used was the same as described in 2.4. Phase I Metabolism expect for the addition of GSH (at a final concentration of 1 mM) before pre-incubation. The supernatant was subsequently collected and analyzed by UPLC-MS2. 2.6. Metabolite analysis by UPLC-MS2 CYP cocktail and IC50 shift assays: All chromatography was performed using an HPLC system from Shimadzu Corporation with a controller (model CBM 20A), a column oven (model CTO-20A), two pumps (model LC-20AC), and an auto sampler (model SIL-20AHT). A mass spectrometer with an API4000 Q trap and analyst software 1.5.2 chromatographic workstation was used. An Agilent ZORBAX XDB C18 column (3.5 lm, 2.1 mm 50 mm) was used. The mobile phase consisted of 0.1% aqueous formic acid (A) and 0.1% formic acid in acetonitrile (B) with a linear-gradient elution at a flow rate of 0.45 mL/min. The elution program was optimized and conducted as follows: Mobile phase A was maintained at 98% for 0.5 min, and then a linearly programmed gradient dropped mobile phase A from 98% to 2% over 1.5 min and maintained it at 2% for 1.5 min. Next, a linearly programmed gradient ramped mobile phase A to 98% in 0.01 min and maintained it at 98% for 1.49 min. The incubations were performed and the residual activities were analyzed by HPLC-MS as described by Bertelsen [19]. Phase I and GSH capture assays: The UPLC-MS2 method was performed using an ACQUITY Ultra Performance Liquid Chromatography System coupled to a Quattro Premier XE triple-quadrupole mass spectrometer (Waters Corp., Milford, MA, USA) equipped with a binary solvent manager, a sample manager, and a Masslynx V4.1 chromatographic work station. The residues were dissolved in 100 lL methanol, and an aliquot of 10 lL was analyzed. An ACQUITY UPLC BEH C18 column (1.7 lm, 2.1 mm 100 mm) was used. The instrument settings were as follows: ESI+; source temperature, 120 °C; desolvation temperature, 350 °C; column temperature, 35 °C; capillary voltage, 3.2 kV; dissolution gas flow, 600 L/h; and cone gas flow, 50 L/h. The mobile phase consisted of 0.1% aqueous formic acid (A) and methanol (B) with a linear-gradient elution at a flow rate of 0.2 mL/min. The HPLC gradient program used was as follows: mobile phase A was maintained at 85% for 1 min, and then a linearly programmed gradient dropped mobile phase A from 85% to 0% over 6 min. Next, mobile phase A was ramped to 85% again in 0.1 min toward the end of the analysis. 3. Results and discussion 3.1. CYP cocktail assays—screening for significant myristicin-mediated inhibition of CYPs The CYP cocktail inhibition assay used in this study was a high-throughput approach that allowed the simultaneous evaluation of the metabolic activities of five major human CYP isoforms on their corresponding probe substrates when incubated with myristicin and HLMs. The effect of myristicin on the activities of the five CYP isozymes (CYP1A2, CYP2D6, CYP2E1, CYP3A4 and CYP2C19) at microsomal protein concentrations of 0.2 mg/ml is shown in Fig. S1 in the Supporting Information. The concentrations and retention times for the metabolites are listed in Table S1 in the
135
Supporting Information. Myristicin inhibition can be observed from its effects on the relative activities (%) of each enzyme for its specific substrate. The relative activity (%) is defined as the ratio of the amount of metabolites from specific substrates in the drug group versus the control group. To determine whether the effect was concentration-dependent, myristicin was incubated at two concentrations (5 and 50 lM) with each probe substrate with or without NADPH. The results showed that myristicin exerts some inhibitory effect on human CYP2E1, CYP2C19 and CYP1A2, but it exerts the strongest inhibitory effect on CYP1A2, with the relative activity values of 8.4% and 28.8% at 50 and 5 mM, respectively, in (+)NADPH. The results of the cocktail assays showed that the inhibition of CYPs by myristicin was concentration-dependent, and CYP1A2 was identified as the CYP that could be inhibited by myristicin most significantly among the five CYPs tested. The IC50 of CYP1A2 was estimated to be lower than 10 lM because the low and high relative activity values were 8.4% and 28.8%, respectively, when the incubation concentrations of myristicin were 50 and 5 lM, respectively, in (+)NADPH. In contrast, the IC50 values of the other four CYPs were estimated to be higher than 10 lM. The IC50 value is an indicator of inhibitory effects. According to accepted conventions, IC50 < 1 lM suggests strong inhibition, 1 lM < IC50 < 10 lM suggests medium inhibition, and IC50 > 10 lM suggests weak inhibition and may have no clinical relevance. Thus, it was necessary to perform IC50 shift assays for CYP1A2 to study the inhibitory mechanism of myristicin, but it was unnecessary to assay the other four enzymes. This cocktail screening approach can improve the efficiency and reduce the cost of testing. 3.2. IC50 shift assays and the blocking of CYP1A2 inhibition by GSH—an evaluation of myristicin MBI potential Because myristicin exhibited the most significant inhibition of CYP1A2, assays were designed to measure the IC50 shift. Six concentrations (0.5, 1.6, 4, 8, 20 and 50 lM) of myristicin were chosen to calculate IC50 values for CYP1A2; the control reactions were performed without myristicin. The IC50 values (the concentration of an inhibitor necessary to cause 50% inhibition of the original enzyme activity) were determined by GraphPad Prism version 5.0 (GraphPad Software, San Diego CA) using a previously described method [20]. The IC50 values +/ NADPH were calculated to be 1.796 and 5.764 lM, respectively (Fig. 2A), resulting in a 3.21-fold IC50 shift. IC50 shift value can be used as a judgment criterion of MBI, some respondents stated that a clear criterion is not established, whereas most cited shift values of 1.2- to 3-fold as indicative of a positive finding, with one respondent citing a 10-fold shift as being clearly positive [21]. Moreover, Obach et al. also reported that if a test compound is an MBI inhibitor, the pre-incubation process will result in a leftward shift in the IC50 curve relative to a control when the cofactor NADPH is omitted in the pre-incubation period [22]. Therefore, in this work, our preliminary experiments indicated that myristicin might be an MBI inhibitor of CYP1A2, and RMs mediated by CYP1A2 are likely to be produced in the process of myristicin metabolism. Because myristicin might be an MBI inhibitor of CYP1A2, this compound therefore may produce RMs, which can be captured by exogenous nucleophiles such as GSH. We predicted that GSH would block the inhibition of CYP1A2 by myristicin in the presence of NADPH. As shown in Fig. 2B, when the concentration of myristicin (2 lM) was fixed near the IC50 value of 1.796 lM, the relative activity value was 48.6%; this result agrees with the results of the IC50 assay. The relative activity value was nearly restored (91.2%) when GSH was added. Therefore, it can be concluded that GSH blocks the inhibition of CYP1A2 by myristicin, indicating that some
136
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
Fig. 2. A, IC50 curves of myristicin on CYP1A2. The X-axis is the logarithmic value of inhibitor concentration, and the Y-axis is relative activity (%). B, GSH blocking the inhibition of CYP1A2 by myristicin (2 lM).
electrophilic RMs are formed. Further characterization of these RMs requires a GSH capture assay. 3.3. Phase I metabolism—identification of the metabolites The positive ion electrospray UPLC-LC/MS analysis of an incubation of myristicin with HLMs is summarized in Table 1. Compared with the control group, two metabolites of myristicin with molecular weights of 180 and 208 were detected at 6.00 and 7.04 min, respectively. The structure of each metabolite was determined on the basis of chromatographic behavior and characteristic mass spectrometric fragmentation features, which were generated by electrospray ionization MS, MS2 spectra. The peak eluting at 7.44 min by UPLC-LC/MS corresponded to unmodified myristicin. Myristicin showed a pseudo-molecule ion [M + H]+ at m/z 193 in the full-scan mass spectrum, which was consistent with the molecular formula (C11H12O3). Metabolites M1 and M2 had UPLC retention times of 6.00 and 7.04 min, respectively. Both exhibited pseudo-molecule ions [M + H]+ at m/z 181 and 209, which confirmed the deduced molecular formulas of C10H12O3 and C10H12O4, respectively. Representative selective ion current chromatograms of myristicin and its metabolites and MS/MS2 spectra of [M + H]+ of myristicin and its metabolites after incubation in HLMs and NADPH are shown in Figs. S2 and S3 in the Supporting Information. The MS2 spectrum of myristicin provided a number of characteristic fragment ions at m/z 178.18, 165.17, and 151.88, which is useful fragment information for metabolite identification. The myristicin cleavage process was analyzed in detail (Fig. 3A). Based on the MS2 spectra of the two cleavage metabolites, the fragment information and cleavage processes of these metabolites have also been inferred and are shown in Fig. 3B. It was inferred that the metabolites may be 5-allyl-1-methoxy-2,3-dihydroxyben zene (M1) and 1’-hydroxymyristicin or 2’,3’-epoxy-myristicin (M2). With respect to M1 (5-allyl-1-methoxy-2,3-dihydroxybenzene), Lee et al. [16] also identified it as the major metabolite of myristicin using GC-MS. The major metabolic pathway from myristicin to M1
Table 1 LC-MS2 data of myristicin and its metabolites after incubation with HLMs in phase I. Metabolite
Molecular formula
Molecular weight (M)
[M + H]+ m/z
Tr (min)
Major fragment ions in MS2 m/z
Myristicin
C11H12O3
192
193
7.44
M1
C10H12O3
180
181
6.00
M2
C11H12O4
208
209
7.04
193.06, 165.17, 181.25, 152.92, 209.10, 181.47,
178.18, 151.88 163.40, 149.04 194.36, 153.14
may involve the cleavage of the methylenedioxyphenyl moiety and loss of the methylene carbon as carbon dioxide [23] or carbon monoxide [24,25]. With respect to metabolite M2 (1’-hydroxymyristicin or 2’,3’-epoxy-myristicin), an earlier study showed that 1’-hydroxymyristicin was one of the metabolites obtained from the in vitro and in vivo metabolism of myristicin [16]. The metabolism of other allyl-benzene compounds, such as safrole and estragole, can also result in 1’-hydroxy derivatives [26,27]. In addition, terminal olefins are usually converted to epoxides, which are considered to be an active intermediate substructure after metabolic activation [26,28]. Therefore, the identity of metabolite M2 may be 1’-hydroxymyristicin or 2’,3’-epoxy-myristicin. 3.4. GSH capture of the reactive metabolites To determine conclusively whether myristicin is an MBI inhibitor of CYP1A2, a variety of methods can be used including the kinetic analysis of MBI and the identification of the covalent adducts to GSH. In this work, the identification of covalent adducts by GSH trapping was adopted. GSH acted as a trapping agent to identify the structure of RMs [29], and this classical approach is widely employed in the detection of RMs [30]. A comprehensive understanding of the fragmentation behavior of GSH [31] was helpful for analyzing the behavior of GSH-myristicin adducts. In this work, compared with the control group, incubating GSH with myristicin in HLMs resulted in one obvious peak of myristicin at 5.10 min in a positive-mode chromatogram (Fig. 4A). The extracted ion chromatograms showed a pseudo-molecule ion [GSH + M12H + H]+ at m/z 486.47 (Fig. 4B), which confirmed the deduced molecular formula (C20H27N3O9S), and the resulting MS2 spectrum provided a number of characteristic fragment ions at m/z 411.00, 393.17, 357.02, 339.65, 305.93, 287.85, 254.14, 237.21, 211.04, etc. (Fig. 4C), which is useful fragment information for metabolite identification and is also in agreement with previous reports [31]. The structural information of all of the representative fragment ions in this GSH-M1 adduct has been analyzed completely, and the detailed cleavage process of the GSH-M1 adduct for MS2 is shown in Fig. 5. Constant-neutral-loss scanning for 129 Da in positive-ion mode has long been considered a standard approach in GSH-conjugate screening [32]. In this work, the detection of a neutral loss with an exact mass of 129.45 Da (the m/z 357.02 Da fragment derived from GSH-M1 losing 129 Da in positive-ion mode) confirmed the conjugation of GSH and M1, and the RMs can be inferred to be a quinone and its tautomer. As shown in Fig. 6A, myristicin can be oxidized to form a quinone (4-allyl-6-methoxy-o-quinone) that can be converted into its tautomeric form, (Z)-4-allylidene-2-hyd roxy-6-methoxycyclohexa-2,5-dien-1-one, the oxidation process of which is similar to that of safrole [33]. The quinone and its tautomeric species can interconvert rapidly. The tautomers are
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
137
Fig. 3. The fragment information and cleavage processes of myristicin (A) and the metabolites (B) analyzed by UPLC-MS2.
unstable and short-lived and can be easily hydroxylated and reduced to M1 in phase I. However, when GSH was added, quinone was captured to form the GSH-M1 adduct. As shown in Fig. 6B, GSH reacts directly with the quinone ring to form a quinone-GSH conjugate, as reported in other examples [34–36]. Furthermore, Bolton et al. [37] also investigated whether allyl-benzene compounds could be oxidized to their corresponding quinones or p-quinone methides (the tautomer of quinone) and showed that GSH could trap the resultant reactive electrophiles. It has also been confirmed using UV, NMR and mass spectrometry that the major GSH adduct results from the addition of GSH to the quinone. However, when GSH was added after an initial incubation of 10 min, the GSH conjugates of the two tautomers were both observed, and GSH trapping at various times revealed increases in the GSH adducts of the tautomer at the expense of the quinone
conjugates. Thus, the formation of tautomer-GSH adducts should also be considered because it is possible to be trapped by a nucleophile (such as GSH) in certain conditions. Therefore, further investigation of allyl-benzene compound metabolism is warranted. Additionally, metabolite M2 was not captured by GSH. UPLC-MS2 analysis of the GSH adduct indicated that the oxidation forming the quinone tautomers occurred on the methylenedioxyphenyl moiety of myristicin, presumably through the initial formation of a di-hydroxyl alkenylbenzene (M1); the quinone can bind to GSH by an addition reaction. Thus, we have confirmed that myristicin is an MBI inhibitor of CYP1A2, and RMs with tautomeric quinone structures act as the highly reactive species. Fontana et al. [18] have described the different binding modes of MBI: (a) the formation of an MI (metabolite intermediate) complex; (b) the formation of heme adducts; and (c) binding to amino
138
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
Fig. 4. GSH capture in myristicin with HLMs. (A) Representative selective ion current chromatogram of the GSH-M1 adduct in HLMs, Tr 5.10 min. (B) The MS of the GSH-M1 adduct. (C) The MS2 of the GSH-M1 adduct (quinone GSH adduct).
Fig. 5. The fragment information and cleavage processes of the GSH-M1 adduct analyzed from MS2 in the GSH capture of myristicin with HLMs.
acids within the enzyme active site. Methylenedioxyphenyl (1,3-benzodioxole) compounds can form a metabolite intermediate carbene at the methylene group, and the carbene may bind to the CYP heme iron and generate an MI complex [38]. Carbene
formation has been reported to be a mechanism of inhibition of P450s by 1,3-benzodioxoles [39]. Additionally, goldensealderived 1,3-benzodioxoles form metabolic-intermediates with P450 isoforms [40]. Which binding mode does myristicin utilize
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
139
Fig. 6. Metabolic activation pathway for myristicin and the reaction process of the quinone-GSH adduct. The RMs, o-quinone (4-allyl-6-methoxy-o-quinone) and its tautomer ((Z)-4-allylidene-2-hydroxy-6-methoxycyclohexa-2,5-dien-1-one), can be interconverted and are in rapid equilibrium.
in its mechanism-based inhibition of CYP1A2? Could a carbene form in the biotransformation of myristicin? Finding out the answers to these questions will be helpful for studying the myristicin-mediated inhibition of CYPs in more depth.
Changjiang Scholars and Innovative Research Team in University (PCSIRT, No. IRT_14R41).
Appendix A. Supplementary data 4. Conclusion In summary, we have demonstrated that myristicin is an MBI inhibitor of CYP1A2 through CYP cocktail screening, IC50 shift calculations and GSH capture assays in HLMs. We also used UPLC-MS2 to identify the highly reactive metabolites (a quinone and its tautomer) through a detailed characterization of the myristicin cleavage processes and the GSH-M1 adduct. Myristicin can be oxidized to form two metabolites. Among them, metabolite M1 with the structure 5-allyl-1-methoxy-2,3-dihydroxybenzene has been captured by GSH and can be oxidized to form two chemical tautomeric forms. While the quinone and its tautomer are unstable and short-lived, it can be inferred that they are the reactive metabolites that covalently bind to the catalytic site of the CYP1A2 enzyme, leading to myristicin-mediated MBI. The identification of this MBI mechanism and the detection of RMs underscores the safety issues associated with myristicin. Furthermore, the use of UPLC-MS2 to confirm the MBI of myristicin and identify its RMs was sensitive, reliable, efficient, and cost-effective enough to allow for the routine identification and characterization of the RMs of other compounds. Conflict of interest The authors declare that there are no conflicts of interest. Transparency Document The Transparency document associated with this article can be found in the online version.
Acknowledgments This work was supported by National Natural Science Foundation of China [NSFC, Nos. 81373890 and 81430096], the General Financial Grant from the China Postdoctoral Science Foundation (Grant No. 2014M551039), and the Program for
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cbi.2015.06.018.
References [1] L.W. Wulf, C.W. Nagel, A.L. Branen, Analysis of myristicin and falcarinol in carrots by high pressure liquid chromatography, J. Agric. Food Chem. 26 (1978) 1390–1393. [2] S.G. Yates, R.E. England, Isolation and analysis of carrot constituents: myristicin, falcarinol, and falcarindiol, J. Agric. Food Chem. 30 (1982) 317–320. [3] A.W. Archer, Determination of safrole and myristicin in nutmeg and mace by high-performance liquid chromatography, J. Chromatogr. 438 (1988) 117–121. [4] U. Stein, H. Greyer, H. Hentschel, Nutmeg (myristicin) poisoning-report on a fatal case and a series of cases recorded by a poison information center, Forensic Sci. Int. 118 (2001) 87–90. [5] S. Jana, G.S. Shekhawat, Anethum graveolens: an Indian traditional medicinal herb and spice, Pharmacogn. Rev. 4 (2010) 179–184. [6] Y. Ozaki, S. Soedigdo, Y.R. Wattimena, A.G. Suganda, Antiinflammatory effect of mace, aril of Myristica fragrans Houtt., and its active principles, Jpn. J. Pharmacol. 49 (1989) 155–163. [7] H. Ahmad, M.T. Tijerina, A.S. Tobolar, Preferential overexpression of a class MU glutathione S-transferase subunit in mouse liver by myristicin, Biochem. Biophys. Res. Commun. 236 (1997) 825–828. [8] G.Q. Zheng, P.M. Kenney, J. Zhang, L.K. Lam, Inhibition of benzo[a]pyreneinduced tumorigenesis by myristicin, a volatile aroma constituent of parsley leaf oil, Carcinogenesis 13 (1992) 1921–1923. [9] E.B. Truitt Jr, G. Duritz, E.M. Ebersberger, Evidence of monoamine oxidase inhibition by myristicin and nutmeg, Proc. Soc. Exp. Biol. Med. 112 (1963) 647–650. [10] C. Martins, C. Doran, A. Laires, J. Rueff, A.S. Rodrigues, Genotoxic and apoptotic activities of the food flavourings myristicin and eugenol in AA8 and XRCC1 deficient EM9 cells, Food Chem. Toxicol. 49 (2011) 385–392. [11] C. Martins, C. Doran, I.C. Silva, C. Miranda, J. Rueff, A.S. Rodrigues, Myristicin from nutmeg induces apoptosis via the mitochondrial pathway and down regulates genes of the DNA damage response pathways in human leukaemia K562 cells, Chem. Biol. Interact. 218 (2014) 1–9. [12] A.T. Shulgin, Possible implication of myristicin as a psychotropic substance, Nature 210 (1966) 380–384. [13] D.A. Kalbhen, Nutmeg as a narcotic. A contribution to the chemistry and pharmacology of nutmeg (Myristica fragrans), Angew. Chem. Int. Ed. 10 (1971) 370–374. [14] E.O. Oswald, L. Fishbein, B.J. Corbett, M.P. Walker, Urinary excretion of tertiary amino methoxy methylenedioxy propiophenones as metabolites of myristicin in the rat and guinea pig, Biochim. Biophys. Acta 244 (1971) 322–328. [15] U. Braun, D.A. Kalbhen, Evidence for the biogenic formation of amphetamine derivatives from components of nutmeg, Pharmacology 9 (1973) 312–316. [16] H.S. Lee, T.C. Jeong, J.H. Kim, In vitro and in vivo metabolism of myristicin in the rat, J. Chromatogr. B 705 (1998) 367–372.
140
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
[17] C.H. Yun, H.S. Lee, H.Y. Lee, S.K. Yim, K.H. Kim, E. Kim, S.S. Yea, F.P. Guengerich, Roles of human liver cytochrome P450 3A4 and 1A2 enzymes in the oxidation of myristicin, Toxicol. Lett. 137 (2003) 143–150. [18] E. Fontana, P.M. Dansette, S.M. Poli, Cytochrome P450 enzymes mechanism based inhibitors: common sub-structures and reactivity, Curr. Drug Metab. 6 (2005) 413–454. [19] K.M. Bertelsen, K. Venkatakrishnan, L.L. Von Moltke, R.S. Obach, D.J. Greenblatt, Apparent mechanism-based inhibition of human CYP2D6 in vitro by paroxetine: comparison with fluoxetine and quinidine, Drug Metab. Dispos. 31 (2003) 289–293. [20] C. Jin, X. He, F. Zhang, L. He, J. Chen, L. Wang, L. An, Y. Fan, Inhibitory mechanisms of celastrol on human liver cytochrome P450 1A2, 2C19, 2D6, 2E1 and 3A4, Xenobiotica (2015), http://dx.doi.org/10.3109/00498254.2014. 1003113. [21] S.W. Grimm, H.J. Einolf, S.D. Hall, K. He, H.K. Lim, K.H. Ling, C. Lu, A.A. Nomeir, E. Seibert, K.W. Skordos, G.R. Tonn, R. Van Horn, R.W. Wang, Y.N. Wong, T.J. Yang, R.S. Obach, The conduct of in vitro studies to address time-dependent inhibition of drug-metabolizing enzymes: a perspective of the pharmaceutical research and manufacturers of America, Drug Metab. Dispos. 37 (2009) 1355– 1370. [22] K. Venkatakrishnan, R.S. Obach, Drug-drug interactions via mechanism-based cytochrome P450 inactivation: points to consider for risk assessment from in vitro data and clinical pharmacologic evaluation, Curr. Drug Metab. 8 (2007) 449–462. [23] F.X. Kamienski, J.E. Casida, Importance of demethylenation in the metabolism in vivo and in vitro of methylenedioxyphenyl synergists and related compounds in mammals, Biochem. Pharmacol. 19 (1970) 91–112. [24] M.W. Anders, J.M. Sunram, C.F. Wilkinson, Mechanism of the metabolism of 1,3-benzodioxoles to carbon monoxide, Biochem. Pharmacol. 33 (1984) 577– 580. [25] L.S. Yu, C.F. Wilkinson, M.W. Anders, Generation of carbon monoxide during the microsomal metabolism of methylenedioxyphenyl compounds, Biochem. Pharmacol. 29 (1980) 1113–1122. [26] A.B. Swanson, E.C. Miller, J.A. Miller, The side-chain epoxidation and hydroxylation of the hepatocarcinogens safrole and estragole and some related compounds by rat and mouse liver microsomes, Biochim. Biophys, Acta 673 (1981) 504–516. [27] D. Thompson, D. Constantin-Teodosiu, B. Egestad, H. Mickos, P. Moldéus, Formation of glutathione conjugates during oxidation of eugenol by microsomal fractions of rat liver and lung, Biochem. Pharmacal. 39 (1990) 1587–1595.
[28] J. Schmidt, P. Kotnik, J. Trontelj, Zˇ. Kenz, L.P. Mašicˇ, Bioactivation of bisphenol A and its analogs (BPF, BPAF, BPZ and DMBPA) in human liver microsomes, Toxicol. In Vitro 27 (2013) 1267–1276. [29] S. Nakayama, H. Takakusa, A. Watanabe, Y. Miyaji, W. Suzuki, D. Sugiyama, K. Shiosakai, K. Honda, N. Okudaira, T. Izumi, O. Okazaki, Combination of GSH trapping and time-dependent inhibition assays as a predictive method of drugs generating highly reactive metabolites, Drug Metab. Dispos. 39 (2011) 1247–1254. [30] L.A. Peterson, M.E. Cummings, C.C. Vu, B.A. Matter, Glutathione trapping to measure microsomal oxidation of furan to cis-2-butene-1,4-dial, Drug Metab. Dispos. 33 (2005) 1453–1458. [31] C. Xie, D.F. Zhong, X.Y. Chen, A fragmentation-based method for the differentiation of glutathione conjugates by high-resolution mass spectrometry with electrospray ionization, Anal. Chim. Acta 788 (2013) 89–98. [32] T.A. Baillie, M.R. Davis, Mass spectrometry in the analysis of glutathione conjugates, Biol. Mass Spectrom. 22 (1993) 319–325. [33] S.F. Zhou, C.C. Xue, X.Q. Yu, G. Wang, Metabolic activation of herbal and dietary constituents and its clinical and toxicological implications: an update, Curr. Drug Metab. 8 (2007) 526–553. [34] T.J. Monks, S.S. Lau, Toxicology of quinone-thioethers, Crit. Rev. Toxicol. 22 (1992) 243–270. [35] J.L. Bolton, M.A. Trush, T.M. Penning, G. Dryhurst, T.J. Monks, Role of quinones in toxicology, Chem. Res. Toxicol. 13 (2000) 135–160. [36] Z.Z. Fang, K.W. Krausz, F. Li, J. Cheng, N. Tanaka, F.J. Gonzalez, Metabolic map and bioactivation of the anti-tumour drug noscapine, Br. J. Pharmacol. 167 (2012) 271–1286. [37] J.L. Bolton, N.M. Acay, V. Vukomanovic, Evidence that 4-allyl-o-quinones spontaneously rearrange to their more electrophilic quinone methides: potential bioactivation mechanism for the hepatocarcinogen safrole, Chem. Res. Toxicol. 7 (1994) 443–450. [38] M. Murray, Mechanisms of inhibitory and regulatory effects of methylenedioxyphenyl compounds on cytochrome P450-dependent drug oxidation, Curr. Drug Metab. 1 (2000) 67–84. [39] G. Simonneaux, P. Le Maux, Carbene complexes of heme proteins and iron porphyrin models, Top. Organomet. Chem. 17 (2006) 83–122. [40] P. Chatterjee, M.R. Franklin, Human cytochrome P450 inhibition and metabolic-intermediate complex formation by Goldenseal extract and its methylenedioxyphenyl components, Drug Metab. Dispos. 31 (2003) 1391– 1397.
Contents lists available at ScienceDirect
Chemico-Biological Interactions journal homepage: www.elsevier.com/locate/chembioint
Identification and characterization of reactive metabolites in myristicin-mediated mechanism-based inhibition of CYP1A2 Ai-Hong Yang, Xin He ⇑, Jun-Xiu Chen, Li-Na He, Chun-Huan Jin, Li-Li Wang, Fang-Liang Zhang, Li-Jun An School of Chinese Materia Medica, Tianjin University of Traditional Chinese Medicine, Tianjin 300193, PR China
a r t i c l e
i n f o
Article history: Received 21 November 2014 Received in revised form 6 June 2015 Accepted 9 June 2015 Available online 16 June 2015 Keywords: Myristicin Reactive metabolites Mechanism-based inhibition Cytochrome P450 enzymes Human liver microsomes
a b s t r a c t Myristicin belongs to the methylenedioxyphenyl or allyl-benzene family of compounds, which are found widely in plants of the Umbelliferae family, such as parsley and carrot. Myristicin is also the major active component in the essential oils of mace and nutmeg. However, this compound can cause adverse reactions, particularly when taken inappropriately or in overdoses. One important source of toxicity of natural products arises from their metabolic biotransformations into reactive metabolites. Myristicin contains a methylenedioxyphenyl substructure, and this specific structural feature may allow compounds to cause a mechanism-based inhibition of cytochrome P450 enzymes and produce reactive metabolites. Therefore, the aim of this work was to identify whether the role of myristicin in CYP enzyme inhibition is mechanism-based inhibition and to gain further information regarding the structure of the resulting reactive metabolites. CYP cocktail assays showed that myristicin most significantly inhibits CYP1A2 among five CYP enzymes (CYP1A2, CYP2D6, CYP2E1, CYP3A4 and CYP2C19) from human liver microsomes. The 3.21-fold IC50 shift value of CYP1A2 indicates that myristicin may be a mechanism-based inhibitor of CYP1A2. Next, reduced glutathione was shown to block the inhibition of CYP1A2, indicating that myristicin utilized a mechanism-based inhibition. Phase I metabolism assays identified two metabolites, 5-allyl-1-methoxy-2,3-dihydroxybenzene (M1) and 1’-hydroxymyristicin or 2’,3’-epoxy-myristicin (M2). Reduced glutathione capturing assays captured the glutathione-M1 adduct, and the reactive metabolites were identified using UPLC-MS2 as a quinone and its tautomer. Thus, it was concluded that myristicin is a mechanism-based inhibitor of CYP1A2, and the reactive metabolites are quinone tautomers. Additionally, the cleavage process of the glutathione-M1 adduct was analyzed in further detail. This study provides additional information on the metabolic mechanism of myristicin inhibition and improves risk evaluation for this compound. Ó 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Myristicin, 1-allyl-3,4-methylenedioxy-5-methoxybenzene (Fig. 1), is commonly found in parsley, carrot, black pepper, nutmeg, and many natural oils and flavoring agents [1–3]. For example, in the essential oil of mace, myristicin is the major component of the aromatic ether fraction. Myristicin is also one of main components in nutmeg volatile oils [4]. Some herbs containing myristicin, such as nutmeg, mace and dill preparations, Abbreviations: MBI, mechanism-based inhibition; RMs, reactive metabolites; CYPs, cytochrome P450 enzymes; GSH, reduced glutathione; HLMs, human liver microsomes; NADPH, b-nicotinamide adenine dinucleotide 2’-phosphate reduced tetrasodium salt; MI complex, metabolite intermediate complex. ⇑ Corresponding author at: School of Chinese Materia Medica, Tianjin University of Traditional Chinese Medicine, Tianjin 300072, PR China. Tel.: +86 22 59596231; fax: +86 22 59596153. E-mail address: [email protected] (X. He). http://dx.doi.org/10.1016/j.cbi.2015.06.018 0009-2797/Ó 2015 Elsevier Ireland Ltd. All rights reserved.
are also used in traditional medicine to treat rheumatism, cholera, psychosis, stomach cramps, nausea, diarrhea, and anxiety [5,6]. Myristicin has been shown to induce glutathione S-transferase [7] and inhibit benzo[a]pyrene-induced tumorigenesis [8]. However, traditional medicines can also cause adverse reactions, especially when taken inappropriately or in overdoses. Myristicin has weak monoamine oxidase inhibitor properties that are responsible for some cardiovascular symptoms [9]. Myristicin displayed apoptotic activity mediated by the activation of caspases in hamster ovary CHO cells [10], and it also altered mitochondrial membrane function in human leukemia K562 cells, inducing apoptosis and down-regulating DNA damage response genes [11]. Myristicin has also been reported to cause psychotropic effects [12] and even death [5] at high dosages in humans, in which myristicin is metabolized to 3-methoxy-4,5-methylenedioxyampheta mine, the hallucinogenic effect of which was more pronounced than that of mescaline [12,13]. Therefore, as the toxicity of this
134
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
obtained from Sigma-Aldrich (Milwaukee, WI, USA). Pooled HLMs, which were purchased from the Research Institute for Liver Diseases Co., Ltd. (Shanghai, China), were prepared from human liver tissues under Chinese organ donation regulations with the full consent of the patients. All other reagents were of high-performance liquid chromatography grade. Fig. 1. Structural formula of myristicin.
2.2. CYP cocktail assays—screening for significant myristicin-mediated inhibition of CYPs
compound is related to its xenobiotic metabolism, the metabolism of myristicin should be more thoroughly studied to improve risk evaluation. Tertiary aminopropiophenones were identified as urinary metabolites from rats and guinea pigs treated with myristicin [14], while rat liver converted myristicin into 3-methoxy-4, 5-methylenedioxy amphetamine [15]. In vivo and in vitro studies of rats demonstrated that myristicin could also be transformed into 5-allyl-1-methoxy-2,3-dihydroxybenzene and 1’-hydroxymyristicin [16]. Through the study of purified recombinant human liver CYPs (cytochrome P450 enzymes) and immunoinhibition assays, Yun et al. determined that the oxidation of myristicin to 5-allyl-1-methoxy-2,3-dihydroxybenzene was catalyzed by CYP3A4 and CYP1A2 in humans [17]. Although these metabolites were characterized in the above reports, the effects of myristicin on the CYPs involved in its metabolism, in particular the identification of RMs (reactive metabolites) in myristicin-mediated MBI (mechanism-based inhibition), remain unknown. The metabolism of drugs by CYPs to form RMs that bind covalently to a catalytic site or sites of the enzyme itself leads to irreversible inhibition of the enzymes. This phenomenon is referred as MBI. MBI inhibitors have very specific features that make them recognizable by in vitro testing. Fontana et al. [18] summarized the most common substructures causing the MBI of CYPs, including methylenedioxyphenyl-containing compounds such as myristicin and safrole. The drug-induced MBI of CYPs is initiated by conversion of the drug into highly RMs, and this group of drugs may be involved in metabolism-dependent drug toxicities. In this work, we demonstrated that myristicin is an MBI inhibitor of CYP1A2 and determined that its reactive metabolites are quinone tautomers using the following assays: (i) Using a CYP cocktail assay, the effects of myristicin on specific marker reactions of five CYP isoforms were measured in HLMs (human liver microsomes). Myristicin inhibited CYP1A2 more strongly than it inhibited the other CYPs. (ii) It could be estimated that myristicin may be the MBI inhibitor of CYP1A2 according to IC50 shift value. (iii) In phase I metabolic reactions, NADPH-dependent inhibition assays were used to identify the generation of two metabolites (M1 and M2). (iv) In GSH capturing assays, GSH acted as a trapping agent to identify the structure of the RMs. It was concluded that myristicin could inhibit CYP1A2 through an MBI mechanism and produce electrophilic reactive metabolites with tautomeric structures (quinone and its tautomer). Such information concerning the MBI by this compound and its RMs can be helpful for thoroughly understanding the drug metabolism of myristicin.
The same experimental conditions described above in 2.2. Cocktail assays were used in the IC50 shift assays except that six different concentrations of myristicin were used: 0.5, 1.6, 4, 8, 20 and 50 lM. The assays were performed in triplicate for all test specimens. The IC50 values were calculated using GraphPad Prism version 5.0 (GraphPad Software, San Diego, CA, USA). Based on the IC50 shift value, we predicted that GSH would be able to block the inhibition of CYP1A2 by myristicin in the presence of NADPH and designed an assay to test this prediction. The same experimental conditions described above were used except for the following two exceptions: (1) Phenacetin was chosen as a probe substrate for CYP1A2, and (2) the final incubation concentration of myristicin was 2 lM, which was close to its IC50 value of 1.796 lM. The assays were performed in triplicate. The relative activity (%) of CYP1A2 was monitored to determine whether GSH was able to block inhibition.
2. Materials and methods
2.4. Phase I metabolism—identification of the metabolites
2.1. Materials
Mixtures of 50 lL myristicin (stock concentration: 200 lM; final concentration: 50 lM) and 50 lL HLMs (stock concentration: 8 mg protein/mL; final concentration: 2 mg protein/mL) were pre-incubated 5 min at 37 °C. Subsequently, 100 lL NADPH (stock concentration: 2 mM; final concentration: 1 mM) were added to the above reaction mixture. After incubation at 37 °C for 2 h, the reaction was stopped by adding 400 lL ice-cold methanol. After centrifugation at 10,000 rpm for 10 min, the supernatant was
Myristicin was purchased from the Chinese National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). Phenacetin, midazolam, dextromethorphan, (s)-mephenytoin, chlorzoxazone and NADPH (b-nicotinamide adenine dinucleotide 2’-phosphate reduced tetrasodium salt) were obtained from Sigma Chemical Co. (St. Louis, MO, USA). GSH was
CYP cocktail assays were employed to analyze and screen the myristicin-mediated inhibition of five CYP isoforms. Phenacetin (CYP1A2), dextromethorphan (CYP2D6), chlorzoxazone (CYP2E1), midazolam (CYP3A4), and (s)-mephenytoin (CYP2C19) were chosen as probe substrates for each corresponding CYP. The assays consisted of two steps: (i) A mixture of 20 lL myristicin (stock concentration: 200 lM or 2 mM) and 20 lL HLMs (stock concentration: 8 mg protein/mL) were pre-incubated in the presence or absence of 40 lL NADPH (denoted as +/ NADPH; stock concentration: 2 mM) for 30 min at 37 °C. For the ( )NADPH reaction, 40 lL of 0.1-M potassium phosphate buffer, pH 7.4, was used instead of NADPH. (ii) Next, 80 lL of the probe substrates and 100 lL NADPH were added to the 20 lL pre-incubation mixtures and incubated for 30 min under the same conditions. The reactions were stopped by adding a 2-fold volume (400 lL) of ice-cold methanol containing 50-ng/ml carbamazepine. Control reactions were performed without myristicin. The final concentrations of the components were as follows: myristicin (5 lM or 50 lM), HLMs (0.2 mg protein/mL), and NADPH (1 mM). The final incubation concentrations of the probe substrates were fixed around their reported Km values: phenacetin (10 lM), dextromethorphan (2.5 lM), chlorzoxazone (20 lM), midazolam (5 lM), and (s)-mephenytoin (20 lM). The assays were performed in triplicate for all test specimens. 2.3. IC50 shift assays and the blocking of CYP1A2 inhibition by GSH—an evaluation of myristicin MBI potential
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
collected and analyzed by UPLC-ESI-MS2. A control reaction was performed without myristicin. 2.5. GSH capture of the reactive metabolites The incubation method used was the same as described in 2.4. Phase I Metabolism expect for the addition of GSH (at a final concentration of 1 mM) before pre-incubation. The supernatant was subsequently collected and analyzed by UPLC-MS2. 2.6. Metabolite analysis by UPLC-MS2 CYP cocktail and IC50 shift assays: All chromatography was performed using an HPLC system from Shimadzu Corporation with a controller (model CBM 20A), a column oven (model CTO-20A), two pumps (model LC-20AC), and an auto sampler (model SIL-20AHT). A mass spectrometer with an API4000 Q trap and analyst software 1.5.2 chromatographic workstation was used. An Agilent ZORBAX XDB C18 column (3.5 lm, 2.1 mm 50 mm) was used. The mobile phase consisted of 0.1% aqueous formic acid (A) and 0.1% formic acid in acetonitrile (B) with a linear-gradient elution at a flow rate of 0.45 mL/min. The elution program was optimized and conducted as follows: Mobile phase A was maintained at 98% for 0.5 min, and then a linearly programmed gradient dropped mobile phase A from 98% to 2% over 1.5 min and maintained it at 2% for 1.5 min. Next, a linearly programmed gradient ramped mobile phase A to 98% in 0.01 min and maintained it at 98% for 1.49 min. The incubations were performed and the residual activities were analyzed by HPLC-MS as described by Bertelsen [19]. Phase I and GSH capture assays: The UPLC-MS2 method was performed using an ACQUITY Ultra Performance Liquid Chromatography System coupled to a Quattro Premier XE triple-quadrupole mass spectrometer (Waters Corp., Milford, MA, USA) equipped with a binary solvent manager, a sample manager, and a Masslynx V4.1 chromatographic work station. The residues were dissolved in 100 lL methanol, and an aliquot of 10 lL was analyzed. An ACQUITY UPLC BEH C18 column (1.7 lm, 2.1 mm 100 mm) was used. The instrument settings were as follows: ESI+; source temperature, 120 °C; desolvation temperature, 350 °C; column temperature, 35 °C; capillary voltage, 3.2 kV; dissolution gas flow, 600 L/h; and cone gas flow, 50 L/h. The mobile phase consisted of 0.1% aqueous formic acid (A) and methanol (B) with a linear-gradient elution at a flow rate of 0.2 mL/min. The HPLC gradient program used was as follows: mobile phase A was maintained at 85% for 1 min, and then a linearly programmed gradient dropped mobile phase A from 85% to 0% over 6 min. Next, mobile phase A was ramped to 85% again in 0.1 min toward the end of the analysis. 3. Results and discussion 3.1. CYP cocktail assays—screening for significant myristicin-mediated inhibition of CYPs The CYP cocktail inhibition assay used in this study was a high-throughput approach that allowed the simultaneous evaluation of the metabolic activities of five major human CYP isoforms on their corresponding probe substrates when incubated with myristicin and HLMs. The effect of myristicin on the activities of the five CYP isozymes (CYP1A2, CYP2D6, CYP2E1, CYP3A4 and CYP2C19) at microsomal protein concentrations of 0.2 mg/ml is shown in Fig. S1 in the Supporting Information. The concentrations and retention times for the metabolites are listed in Table S1 in the
135
Supporting Information. Myristicin inhibition can be observed from its effects on the relative activities (%) of each enzyme for its specific substrate. The relative activity (%) is defined as the ratio of the amount of metabolites from specific substrates in the drug group versus the control group. To determine whether the effect was concentration-dependent, myristicin was incubated at two concentrations (5 and 50 lM) with each probe substrate with or without NADPH. The results showed that myristicin exerts some inhibitory effect on human CYP2E1, CYP2C19 and CYP1A2, but it exerts the strongest inhibitory effect on CYP1A2, with the relative activity values of 8.4% and 28.8% at 50 and 5 mM, respectively, in (+)NADPH. The results of the cocktail assays showed that the inhibition of CYPs by myristicin was concentration-dependent, and CYP1A2 was identified as the CYP that could be inhibited by myristicin most significantly among the five CYPs tested. The IC50 of CYP1A2 was estimated to be lower than 10 lM because the low and high relative activity values were 8.4% and 28.8%, respectively, when the incubation concentrations of myristicin were 50 and 5 lM, respectively, in (+)NADPH. In contrast, the IC50 values of the other four CYPs were estimated to be higher than 10 lM. The IC50 value is an indicator of inhibitory effects. According to accepted conventions, IC50 < 1 lM suggests strong inhibition, 1 lM < IC50 < 10 lM suggests medium inhibition, and IC50 > 10 lM suggests weak inhibition and may have no clinical relevance. Thus, it was necessary to perform IC50 shift assays for CYP1A2 to study the inhibitory mechanism of myristicin, but it was unnecessary to assay the other four enzymes. This cocktail screening approach can improve the efficiency and reduce the cost of testing. 3.2. IC50 shift assays and the blocking of CYP1A2 inhibition by GSH—an evaluation of myristicin MBI potential Because myristicin exhibited the most significant inhibition of CYP1A2, assays were designed to measure the IC50 shift. Six concentrations (0.5, 1.6, 4, 8, 20 and 50 lM) of myristicin were chosen to calculate IC50 values for CYP1A2; the control reactions were performed without myristicin. The IC50 values (the concentration of an inhibitor necessary to cause 50% inhibition of the original enzyme activity) were determined by GraphPad Prism version 5.0 (GraphPad Software, San Diego CA) using a previously described method [20]. The IC50 values +/ NADPH were calculated to be 1.796 and 5.764 lM, respectively (Fig. 2A), resulting in a 3.21-fold IC50 shift. IC50 shift value can be used as a judgment criterion of MBI, some respondents stated that a clear criterion is not established, whereas most cited shift values of 1.2- to 3-fold as indicative of a positive finding, with one respondent citing a 10-fold shift as being clearly positive [21]. Moreover, Obach et al. also reported that if a test compound is an MBI inhibitor, the pre-incubation process will result in a leftward shift in the IC50 curve relative to a control when the cofactor NADPH is omitted in the pre-incubation period [22]. Therefore, in this work, our preliminary experiments indicated that myristicin might be an MBI inhibitor of CYP1A2, and RMs mediated by CYP1A2 are likely to be produced in the process of myristicin metabolism. Because myristicin might be an MBI inhibitor of CYP1A2, this compound therefore may produce RMs, which can be captured by exogenous nucleophiles such as GSH. We predicted that GSH would block the inhibition of CYP1A2 by myristicin in the presence of NADPH. As shown in Fig. 2B, when the concentration of myristicin (2 lM) was fixed near the IC50 value of 1.796 lM, the relative activity value was 48.6%; this result agrees with the results of the IC50 assay. The relative activity value was nearly restored (91.2%) when GSH was added. Therefore, it can be concluded that GSH blocks the inhibition of CYP1A2 by myristicin, indicating that some
136
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
Fig. 2. A, IC50 curves of myristicin on CYP1A2. The X-axis is the logarithmic value of inhibitor concentration, and the Y-axis is relative activity (%). B, GSH blocking the inhibition of CYP1A2 by myristicin (2 lM).
electrophilic RMs are formed. Further characterization of these RMs requires a GSH capture assay. 3.3. Phase I metabolism—identification of the metabolites The positive ion electrospray UPLC-LC/MS analysis of an incubation of myristicin with HLMs is summarized in Table 1. Compared with the control group, two metabolites of myristicin with molecular weights of 180 and 208 were detected at 6.00 and 7.04 min, respectively. The structure of each metabolite was determined on the basis of chromatographic behavior and characteristic mass spectrometric fragmentation features, which were generated by electrospray ionization MS, MS2 spectra. The peak eluting at 7.44 min by UPLC-LC/MS corresponded to unmodified myristicin. Myristicin showed a pseudo-molecule ion [M + H]+ at m/z 193 in the full-scan mass spectrum, which was consistent with the molecular formula (C11H12O3). Metabolites M1 and M2 had UPLC retention times of 6.00 and 7.04 min, respectively. Both exhibited pseudo-molecule ions [M + H]+ at m/z 181 and 209, which confirmed the deduced molecular formulas of C10H12O3 and C10H12O4, respectively. Representative selective ion current chromatograms of myristicin and its metabolites and MS/MS2 spectra of [M + H]+ of myristicin and its metabolites after incubation in HLMs and NADPH are shown in Figs. S2 and S3 in the Supporting Information. The MS2 spectrum of myristicin provided a number of characteristic fragment ions at m/z 178.18, 165.17, and 151.88, which is useful fragment information for metabolite identification. The myristicin cleavage process was analyzed in detail (Fig. 3A). Based on the MS2 spectra of the two cleavage metabolites, the fragment information and cleavage processes of these metabolites have also been inferred and are shown in Fig. 3B. It was inferred that the metabolites may be 5-allyl-1-methoxy-2,3-dihydroxyben zene (M1) and 1’-hydroxymyristicin or 2’,3’-epoxy-myristicin (M2). With respect to M1 (5-allyl-1-methoxy-2,3-dihydroxybenzene), Lee et al. [16] also identified it as the major metabolite of myristicin using GC-MS. The major metabolic pathway from myristicin to M1
Table 1 LC-MS2 data of myristicin and its metabolites after incubation with HLMs in phase I. Metabolite
Molecular formula
Molecular weight (M)
[M + H]+ m/z
Tr (min)
Major fragment ions in MS2 m/z
Myristicin
C11H12O3
192
193
7.44
M1
C10H12O3
180
181
6.00
M2
C11H12O4
208
209
7.04
193.06, 165.17, 181.25, 152.92, 209.10, 181.47,
178.18, 151.88 163.40, 149.04 194.36, 153.14
may involve the cleavage of the methylenedioxyphenyl moiety and loss of the methylene carbon as carbon dioxide [23] or carbon monoxide [24,25]. With respect to metabolite M2 (1’-hydroxymyristicin or 2’,3’-epoxy-myristicin), an earlier study showed that 1’-hydroxymyristicin was one of the metabolites obtained from the in vitro and in vivo metabolism of myristicin [16]. The metabolism of other allyl-benzene compounds, such as safrole and estragole, can also result in 1’-hydroxy derivatives [26,27]. In addition, terminal olefins are usually converted to epoxides, which are considered to be an active intermediate substructure after metabolic activation [26,28]. Therefore, the identity of metabolite M2 may be 1’-hydroxymyristicin or 2’,3’-epoxy-myristicin. 3.4. GSH capture of the reactive metabolites To determine conclusively whether myristicin is an MBI inhibitor of CYP1A2, a variety of methods can be used including the kinetic analysis of MBI and the identification of the covalent adducts to GSH. In this work, the identification of covalent adducts by GSH trapping was adopted. GSH acted as a trapping agent to identify the structure of RMs [29], and this classical approach is widely employed in the detection of RMs [30]. A comprehensive understanding of the fragmentation behavior of GSH [31] was helpful for analyzing the behavior of GSH-myristicin adducts. In this work, compared with the control group, incubating GSH with myristicin in HLMs resulted in one obvious peak of myristicin at 5.10 min in a positive-mode chromatogram (Fig. 4A). The extracted ion chromatograms showed a pseudo-molecule ion [GSH + M12H + H]+ at m/z 486.47 (Fig. 4B), which confirmed the deduced molecular formula (C20H27N3O9S), and the resulting MS2 spectrum provided a number of characteristic fragment ions at m/z 411.00, 393.17, 357.02, 339.65, 305.93, 287.85, 254.14, 237.21, 211.04, etc. (Fig. 4C), which is useful fragment information for metabolite identification and is also in agreement with previous reports [31]. The structural information of all of the representative fragment ions in this GSH-M1 adduct has been analyzed completely, and the detailed cleavage process of the GSH-M1 adduct for MS2 is shown in Fig. 5. Constant-neutral-loss scanning for 129 Da in positive-ion mode has long been considered a standard approach in GSH-conjugate screening [32]. In this work, the detection of a neutral loss with an exact mass of 129.45 Da (the m/z 357.02 Da fragment derived from GSH-M1 losing 129 Da in positive-ion mode) confirmed the conjugation of GSH and M1, and the RMs can be inferred to be a quinone and its tautomer. As shown in Fig. 6A, myristicin can be oxidized to form a quinone (4-allyl-6-methoxy-o-quinone) that can be converted into its tautomeric form, (Z)-4-allylidene-2-hyd roxy-6-methoxycyclohexa-2,5-dien-1-one, the oxidation process of which is similar to that of safrole [33]. The quinone and its tautomeric species can interconvert rapidly. The tautomers are
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
137
Fig. 3. The fragment information and cleavage processes of myristicin (A) and the metabolites (B) analyzed by UPLC-MS2.
unstable and short-lived and can be easily hydroxylated and reduced to M1 in phase I. However, when GSH was added, quinone was captured to form the GSH-M1 adduct. As shown in Fig. 6B, GSH reacts directly with the quinone ring to form a quinone-GSH conjugate, as reported in other examples [34–36]. Furthermore, Bolton et al. [37] also investigated whether allyl-benzene compounds could be oxidized to their corresponding quinones or p-quinone methides (the tautomer of quinone) and showed that GSH could trap the resultant reactive electrophiles. It has also been confirmed using UV, NMR and mass spectrometry that the major GSH adduct results from the addition of GSH to the quinone. However, when GSH was added after an initial incubation of 10 min, the GSH conjugates of the two tautomers were both observed, and GSH trapping at various times revealed increases in the GSH adducts of the tautomer at the expense of the quinone
conjugates. Thus, the formation of tautomer-GSH adducts should also be considered because it is possible to be trapped by a nucleophile (such as GSH) in certain conditions. Therefore, further investigation of allyl-benzene compound metabolism is warranted. Additionally, metabolite M2 was not captured by GSH. UPLC-MS2 analysis of the GSH adduct indicated that the oxidation forming the quinone tautomers occurred on the methylenedioxyphenyl moiety of myristicin, presumably through the initial formation of a di-hydroxyl alkenylbenzene (M1); the quinone can bind to GSH by an addition reaction. Thus, we have confirmed that myristicin is an MBI inhibitor of CYP1A2, and RMs with tautomeric quinone structures act as the highly reactive species. Fontana et al. [18] have described the different binding modes of MBI: (a) the formation of an MI (metabolite intermediate) complex; (b) the formation of heme adducts; and (c) binding to amino
138
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
Fig. 4. GSH capture in myristicin with HLMs. (A) Representative selective ion current chromatogram of the GSH-M1 adduct in HLMs, Tr 5.10 min. (B) The MS of the GSH-M1 adduct. (C) The MS2 of the GSH-M1 adduct (quinone GSH adduct).
Fig. 5. The fragment information and cleavage processes of the GSH-M1 adduct analyzed from MS2 in the GSH capture of myristicin with HLMs.
acids within the enzyme active site. Methylenedioxyphenyl (1,3-benzodioxole) compounds can form a metabolite intermediate carbene at the methylene group, and the carbene may bind to the CYP heme iron and generate an MI complex [38]. Carbene
formation has been reported to be a mechanism of inhibition of P450s by 1,3-benzodioxoles [39]. Additionally, goldensealderived 1,3-benzodioxoles form metabolic-intermediates with P450 isoforms [40]. Which binding mode does myristicin utilize
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
139
Fig. 6. Metabolic activation pathway for myristicin and the reaction process of the quinone-GSH adduct. The RMs, o-quinone (4-allyl-6-methoxy-o-quinone) and its tautomer ((Z)-4-allylidene-2-hydroxy-6-methoxycyclohexa-2,5-dien-1-one), can be interconverted and are in rapid equilibrium.
in its mechanism-based inhibition of CYP1A2? Could a carbene form in the biotransformation of myristicin? Finding out the answers to these questions will be helpful for studying the myristicin-mediated inhibition of CYPs in more depth.
Changjiang Scholars and Innovative Research Team in University (PCSIRT, No. IRT_14R41).
Appendix A. Supplementary data 4. Conclusion In summary, we have demonstrated that myristicin is an MBI inhibitor of CYP1A2 through CYP cocktail screening, IC50 shift calculations and GSH capture assays in HLMs. We also used UPLC-MS2 to identify the highly reactive metabolites (a quinone and its tautomer) through a detailed characterization of the myristicin cleavage processes and the GSH-M1 adduct. Myristicin can be oxidized to form two metabolites. Among them, metabolite M1 with the structure 5-allyl-1-methoxy-2,3-dihydroxybenzene has been captured by GSH and can be oxidized to form two chemical tautomeric forms. While the quinone and its tautomer are unstable and short-lived, it can be inferred that they are the reactive metabolites that covalently bind to the catalytic site of the CYP1A2 enzyme, leading to myristicin-mediated MBI. The identification of this MBI mechanism and the detection of RMs underscores the safety issues associated with myristicin. Furthermore, the use of UPLC-MS2 to confirm the MBI of myristicin and identify its RMs was sensitive, reliable, efficient, and cost-effective enough to allow for the routine identification and characterization of the RMs of other compounds. Conflict of interest The authors declare that there are no conflicts of interest. Transparency Document The Transparency document associated with this article can be found in the online version.
Acknowledgments This work was supported by National Natural Science Foundation of China [NSFC, Nos. 81373890 and 81430096], the General Financial Grant from the China Postdoctoral Science Foundation (Grant No. 2014M551039), and the Program for
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cbi.2015.06.018.
References [1] L.W. Wulf, C.W. Nagel, A.L. Branen, Analysis of myristicin and falcarinol in carrots by high pressure liquid chromatography, J. Agric. Food Chem. 26 (1978) 1390–1393. [2] S.G. Yates, R.E. England, Isolation and analysis of carrot constituents: myristicin, falcarinol, and falcarindiol, J. Agric. Food Chem. 30 (1982) 317–320. [3] A.W. Archer, Determination of safrole and myristicin in nutmeg and mace by high-performance liquid chromatography, J. Chromatogr. 438 (1988) 117–121. [4] U. Stein, H. Greyer, H. Hentschel, Nutmeg (myristicin) poisoning-report on a fatal case and a series of cases recorded by a poison information center, Forensic Sci. Int. 118 (2001) 87–90. [5] S. Jana, G.S. Shekhawat, Anethum graveolens: an Indian traditional medicinal herb and spice, Pharmacogn. Rev. 4 (2010) 179–184. [6] Y. Ozaki, S. Soedigdo, Y.R. Wattimena, A.G. Suganda, Antiinflammatory effect of mace, aril of Myristica fragrans Houtt., and its active principles, Jpn. J. Pharmacol. 49 (1989) 155–163. [7] H. Ahmad, M.T. Tijerina, A.S. Tobolar, Preferential overexpression of a class MU glutathione S-transferase subunit in mouse liver by myristicin, Biochem. Biophys. Res. Commun. 236 (1997) 825–828. [8] G.Q. Zheng, P.M. Kenney, J. Zhang, L.K. Lam, Inhibition of benzo[a]pyreneinduced tumorigenesis by myristicin, a volatile aroma constituent of parsley leaf oil, Carcinogenesis 13 (1992) 1921–1923. [9] E.B. Truitt Jr, G. Duritz, E.M. Ebersberger, Evidence of monoamine oxidase inhibition by myristicin and nutmeg, Proc. Soc. Exp. Biol. Med. 112 (1963) 647–650. [10] C. Martins, C. Doran, A. Laires, J. Rueff, A.S. Rodrigues, Genotoxic and apoptotic activities of the food flavourings myristicin and eugenol in AA8 and XRCC1 deficient EM9 cells, Food Chem. Toxicol. 49 (2011) 385–392. [11] C. Martins, C. Doran, I.C. Silva, C. Miranda, J. Rueff, A.S. Rodrigues, Myristicin from nutmeg induces apoptosis via the mitochondrial pathway and down regulates genes of the DNA damage response pathways in human leukaemia K562 cells, Chem. Biol. Interact. 218 (2014) 1–9. [12] A.T. Shulgin, Possible implication of myristicin as a psychotropic substance, Nature 210 (1966) 380–384. [13] D.A. Kalbhen, Nutmeg as a narcotic. A contribution to the chemistry and pharmacology of nutmeg (Myristica fragrans), Angew. Chem. Int. Ed. 10 (1971) 370–374. [14] E.O. Oswald, L. Fishbein, B.J. Corbett, M.P. Walker, Urinary excretion of tertiary amino methoxy methylenedioxy propiophenones as metabolites of myristicin in the rat and guinea pig, Biochim. Biophys. Acta 244 (1971) 322–328. [15] U. Braun, D.A. Kalbhen, Evidence for the biogenic formation of amphetamine derivatives from components of nutmeg, Pharmacology 9 (1973) 312–316. [16] H.S. Lee, T.C. Jeong, J.H. Kim, In vitro and in vivo metabolism of myristicin in the rat, J. Chromatogr. B 705 (1998) 367–372.
140
A.-H. Yang et al. / Chemico-Biological Interactions 237 (2015) 133–140
[17] C.H. Yun, H.S. Lee, H.Y. Lee, S.K. Yim, K.H. Kim, E. Kim, S.S. Yea, F.P. Guengerich, Roles of human liver cytochrome P450 3A4 and 1A2 enzymes in the oxidation of myristicin, Toxicol. Lett. 137 (2003) 143–150. [18] E. Fontana, P.M. Dansette, S.M. Poli, Cytochrome P450 enzymes mechanism based inhibitors: common sub-structures and reactivity, Curr. Drug Metab. 6 (2005) 413–454. [19] K.M. Bertelsen, K. Venkatakrishnan, L.L. Von Moltke, R.S. Obach, D.J. Greenblatt, Apparent mechanism-based inhibition of human CYP2D6 in vitro by paroxetine: comparison with fluoxetine and quinidine, Drug Metab. Dispos. 31 (2003) 289–293. [20] C. Jin, X. He, F. Zhang, L. He, J. Chen, L. Wang, L. An, Y. Fan, Inhibitory mechanisms of celastrol on human liver cytochrome P450 1A2, 2C19, 2D6, 2E1 and 3A4, Xenobiotica (2015), http://dx.doi.org/10.3109/00498254.2014. 1003113. [21] S.W. Grimm, H.J. Einolf, S.D. Hall, K. He, H.K. Lim, K.H. Ling, C. Lu, A.A. Nomeir, E. Seibert, K.W. Skordos, G.R. Tonn, R. Van Horn, R.W. Wang, Y.N. Wong, T.J. Yang, R.S. Obach, The conduct of in vitro studies to address time-dependent inhibition of drug-metabolizing enzymes: a perspective of the pharmaceutical research and manufacturers of America, Drug Metab. Dispos. 37 (2009) 1355– 1370. [22] K. Venkatakrishnan, R.S. Obach, Drug-drug interactions via mechanism-based cytochrome P450 inactivation: points to consider for risk assessment from in vitro data and clinical pharmacologic evaluation, Curr. Drug Metab. 8 (2007) 449–462. [23] F.X. Kamienski, J.E. Casida, Importance of demethylenation in the metabolism in vivo and in vitro of methylenedioxyphenyl synergists and related compounds in mammals, Biochem. Pharmacol. 19 (1970) 91–112. [24] M.W. Anders, J.M. Sunram, C.F. Wilkinson, Mechanism of the metabolism of 1,3-benzodioxoles to carbon monoxide, Biochem. Pharmacol. 33 (1984) 577– 580. [25] L.S. Yu, C.F. Wilkinson, M.W. Anders, Generation of carbon monoxide during the microsomal metabolism of methylenedioxyphenyl compounds, Biochem. Pharmacol. 29 (1980) 1113–1122. [26] A.B. Swanson, E.C. Miller, J.A. Miller, The side-chain epoxidation and hydroxylation of the hepatocarcinogens safrole and estragole and some related compounds by rat and mouse liver microsomes, Biochim. Biophys, Acta 673 (1981) 504–516. [27] D. Thompson, D. Constantin-Teodosiu, B. Egestad, H. Mickos, P. Moldéus, Formation of glutathione conjugates during oxidation of eugenol by microsomal fractions of rat liver and lung, Biochem. Pharmacal. 39 (1990) 1587–1595.
[28] J. Schmidt, P. Kotnik, J. Trontelj, Zˇ. Kenz, L.P. Mašicˇ, Bioactivation of bisphenol A and its analogs (BPF, BPAF, BPZ and DMBPA) in human liver microsomes, Toxicol. In Vitro 27 (2013) 1267–1276. [29] S. Nakayama, H. Takakusa, A. Watanabe, Y. Miyaji, W. Suzuki, D. Sugiyama, K. Shiosakai, K. Honda, N. Okudaira, T. Izumi, O. Okazaki, Combination of GSH trapping and time-dependent inhibition assays as a predictive method of drugs generating highly reactive metabolites, Drug Metab. Dispos. 39 (2011) 1247–1254. [30] L.A. Peterson, M.E. Cummings, C.C. Vu, B.A. Matter, Glutathione trapping to measure microsomal oxidation of furan to cis-2-butene-1,4-dial, Drug Metab. Dispos. 33 (2005) 1453–1458. [31] C. Xie, D.F. Zhong, X.Y. Chen, A fragmentation-based method for the differentiation of glutathione conjugates by high-resolution mass spectrometry with electrospray ionization, Anal. Chim. Acta 788 (2013) 89–98. [32] T.A. Baillie, M.R. Davis, Mass spectrometry in the analysis of glutathione conjugates, Biol. Mass Spectrom. 22 (1993) 319–325. [33] S.F. Zhou, C.C. Xue, X.Q. Yu, G. Wang, Metabolic activation of herbal and dietary constituents and its clinical and toxicological implications: an update, Curr. Drug Metab. 8 (2007) 526–553. [34] T.J. Monks, S.S. Lau, Toxicology of quinone-thioethers, Crit. Rev. Toxicol. 22 (1992) 243–270. [35] J.L. Bolton, M.A. Trush, T.M. Penning, G. Dryhurst, T.J. Monks, Role of quinones in toxicology, Chem. Res. Toxicol. 13 (2000) 135–160. [36] Z.Z. Fang, K.W. Krausz, F. Li, J. Cheng, N. Tanaka, F.J. Gonzalez, Metabolic map and bioactivation of the anti-tumour drug noscapine, Br. J. Pharmacol. 167 (2012) 271–1286. [37] J.L. Bolton, N.M. Acay, V. Vukomanovic, Evidence that 4-allyl-o-quinones spontaneously rearrange to their more electrophilic quinone methides: potential bioactivation mechanism for the hepatocarcinogen safrole, Chem. Res. Toxicol. 7 (1994) 443–450. [38] M. Murray, Mechanisms of inhibitory and regulatory effects of methylenedioxyphenyl compounds on cytochrome P450-dependent drug oxidation, Curr. Drug Metab. 1 (2000) 67–84. [39] G. Simonneaux, P. Le Maux, Carbene complexes of heme proteins and iron porphyrin models, Top. Organomet. Chem. 17 (2006) 83–122. [40] P. Chatterjee, M.R. Franklin, Human cytochrome P450 inhibition and metabolic-intermediate complex formation by Goldenseal extract and its methylenedioxyphenyl components, Drug Metab. Dispos. 31 (2003) 1391– 1397.