Evaluation of the inhibition potential of plumbagin against cytochrome P450 using LC-MS/MS and cocktail approach

Plumbagin (5-hydroxy-2-methyl-1,4-naphthoquinone), a natural naphthoquinone compound isolated from roots of Plumbago zeylanica L., has drawn a lot of attention for its plenty of pharmacological properties including antidiabetes and anti-cancer. The aim of this study was to investigate the effects of plumbagin on CYP1A2, CYP2B1/6, CYP2C9/11, CYP2D1/6, CYP2E1 and CYP3A2/4 activities in human and rat liver and evaluate the potential herb-drug interactions using the cocktail approach. All CYP substrates and their metabolites were analyzed using high-performance liquid chromatography–tandem mass spectrometry (LC-MS/MS). Plumbagin presented non-time-dependent inhibition of CYP activities in both human and rat liver. In humans, plumbagin was not only a mixed inhibitor of CYP2B6, CYP2C9, CYP2D6, CYP2E1 and CYP3A4, but also a non-competitive inhibitor of CYP1A2, with Ki values no more than 2.16 μM. In rats, the mixed inhibition of CYP1A2 and CYP2D1, and competitive inhibition for CYP2B1, CYP2C11 and CYP2E1 with Ki values less than 9.93 μM were observed. In general, the relatively low Ki values of plumbagin in humans would have a high potential to cause the toxicity and drug interactions involving CYP enzymes.

Plumbago zeylanica L. (family: Plumbaginaceae), also known as Chitrak, is a medicinal plant and mainly distributed around Africa and Asia including India and China 1,2 . It is originally documented in "Tang Ben Cao" in China and widely used in prescription for treating many diseases such as rheumatic joint pain, blood stasis amenorrhea and malignant sore in Traditional Chinese medicine 3 . Plumbagin (5-hydroxy-2-methyl-1,4-naphthoquinone, Fig. 1a) isolated from roots of Plumbago zeylanica L. is an active naphthoquinone and responsible for its therapeutic effects 4 . Previous studies have reported that plumbagin possesses lots of pharmacological properties including antidiabetic, anti-inflammatory, antibacterial and anti-immediate allergic reaction activities 1,5 . In particular, plumbagin presents significant anticancer activity 6 . For example, plumbagin can induce apoptosis and autophagy of human prostate cancer cells via sirtuin1-and PI3K/Akt/mTOR-mediated pathways 7 . And it also can kill human gastric cancer cells by induction of SH2-containing protein tyrosine phosphatase 1 8 . Therefore, Plumbagin is considered to be a promising antitumor candidate for drug development.
The cytochrome P450 (CYP) enzymes play an important role in metabolism of xenobiotics and endogenous substances. Almost 90% drugs are metabolized by CYP superfamily 9,10 . Among the CYP isoforms identified, the major human CYP isoforms (CYP1A2, CYP2B6, CYP2C9, CYP2D6, CYP2E1 and CYP3A4) are responsible for 80% of the CYP-mediated metabolism [11][12][13] . However, the inhibition and induction of CYP by herbs often occur, which is directly linked to herb-drug interactions (HDIs) and result in the toxicity and/or therapeutic failure 14,15 . In consequence, U.S. Food and Drug Administration (FDA) has recommended to use CYP-associated metabolic studies in vitro to predict the potential HDIs, which is the major attrition in drug development 16 . Recently, one study only found the inhibitory effects of plumbagin on CYP1A2, CYP2C19 and CYP3A4 17 . However, the mechanisms of CYP inhibition remain unclear. Further studies should be carried out to illustrate the mechanisms of plumbagin on CYP enzymes. and pooled rat liver microsomes (c). Chemical structure was produced using ChemDraw Professional 15.0. Results were mean ± SD of quadruplicate assays. The data was fit to log (plumbagin concentration) and normalized response equations using GraphPad Prism 5.0.
Scientific RepoRts | 6:28482 | DOI: 10.1038/srep28482 dextromethorphan (CYP2D1/6), chlorzoxazone (CYP2E1) and midazolam (CYP3A2/4) 16 , alternatively named the n-in-one assay, was adopted to monitor several CYP activities in a single experiment, which was much comprehensive, time and resource efficient 10,18 . In this approach, a liquid chromatography tandem mass spectrometry (LC-MS/MS) was usually used for the simultaneous detection of the formation of metabolites of these probe substrates to calculate the enzymatic reaction rates, which could be used to evaluate the corresponding CYP enzyme activities 19 . To our knowledge, this is the first study to report the potential effects of plumbagin on CYP2B1/6, CYP2C9/11, CYP2E1 and CYP2D1/6, and demonstrate the inhibitory mechanisms of CYPs. Therefore, the present study could be helpful to evaluate the potential HDIs involved with plumbagin comprehensively.
To further determine the types of CYP inhibition by plumbagin in HLMs, enzyme kinetics studies were carried out with various substrates concentrations in the presence or absence of plumbagin. The Lineweaver-Burk transformation of the enzyme velocities versus substrates concentrations showed that the types of inhibition by plumbagin for these selected CYPs were mixed inhibition on CYP2B6, CYP2C9, CYP2D6, CYP2E1 and CYP3A4, but non-competitive for CYP1A2. The K i values were determined using the secondary Lineweaver-Burk plot, and then calculated at 0. Since α values of CYP2B6, CYP2C9, CYP2D6, CYP2E1 and CYP3A4 were not equal to 1, the types of inhibition by plumbagin for these five CYPs were mixed inhibition. However, the type of inhibition by plumbagin for CYP1A2 was non-competitive inhibition due to α value of CYP1A2 approximately equal to 1 (Table 1, Fig. 3).
Enzyme kinetic analysis for inhibition of CYPs by plumbagin in rat liver. The time-dependent inhibition experiments were also performed and no time-dependent inhibition was observed in RLMs (Fig. 2). The Lineweaver-Burk transformation of the enzyme velocities versus substrates concentrations showed that the types  Fig. 4). Moreover, the enzyme inhibition kinetics of CYP3A2 did not conform to the classical Michaelis-Menten kinetics (Fig. 5), which would be discussed in Discussion section.  Discussion Plumbago zeylanica L. has been widely used as traditional herbal medicine among many Chinese minorities for many years. Meanwhile, the medicinal plant is the main ingredient of Qubai Babuqi Pian (Sinopharm XinJiang Pharmaceutical Co., Ltd.) which is used to treat leucoderma in Uighur medicine. Nowadays, herbal medicines, especially being consumed along with modern drugs, are increasingly popular when patients are in several diseases or conditions 11,14 . When they are co-administered over overlapping time periods, the possibility of HDIs mediated by CYPs might exist 20,21 . However, the inhibition effects of plumbagin on CYP activities were rarely concerned. Among all CYP isoforms in humans, CYP3A4, CYP2C9, CYP1A2, CYP2E1, CYP2D6 and CYP2B6 are relative abundant, which accounts about for 40%, 20%, 13%, 10%, 2% and 2% in total CYPs respectively, and responsible for metabolism of the majority of clinically prescribed drugs 14,[22][23][24] . CYP1A2 is involved in the metabolism of a variety of clinically using drugs, such as clozapine, tacrine, tizanidine, theophylline, and several important endogenous compounds 22,[24][25][26] . CYP2B6 is responsible for approximately 3-8% of widely used pharmaceuticals and also reported to have high expression in breast tumors 23,27,28 . CYP2C9 contributes to the metabolism of not only commonly prescribed drugs which accounts for 10-20%, but also fatty acids, prostanoids and steroid hormones 22 . CYP2D6 expressing in liver and extrahepatic organs mediates approximately 25% of marketed drugs, such as antidepressants and antipsychotics 29 . Meanwhile, CYP2D6 with decreased activity is implicated in Parkinson disease in Caucasian populations 29 . CYP2E1 as a natural ethanol-inducible enzyme metabolically activates various carcinogens. Increased CYP2E1 activity is linked to the carcinogenic process 30 . Furthermore, more than 50% of drugs were metabolized by CYP3A4 including antibiotics, antiviral drugs, benzodiazepines, calcium channel blockers, statins and immunosuppressive drugs 31,32 . However, to our knowledge, the inhibition effects of plumbagin on these important CYPs have not been reported except for CYP1A2 and CYP3A4 17 . In particular, the mechanisms of inhibition of all six CYPs were completely investigated for the first time. The purpose of our work was to investigate the effects of plumbagin on the major CYP activities both in human and rat liver and assess the potential HDIs involved plumbagin or Plumbago zeylanica L.
To evaluate the potential HDIs, the cocktail approach, frequently used for CYP inhibition studies [33][34][35] , has been developed for monitoring CYP activities in a single run by LC-MS/MS with high sensitivity and selectivity. According to FDA guidance, classical probe substrates including phenacetin (CYP1A2), bupropion (CYP2B1/6), tolbutamide (CYP2C9/11), dextromethorphan (CYP2D1/6), chlorzoxazone (CYP2E1) and midazolam (CYP3A2/4) were selected. The concentrations of probe substrates were close to the K m values of each CYP-catalyzed reaction according to published papers 16,33 . Given that the linear initial rates of metabolites formation for all selected CYPs with minimal substrates depletion and the automatic inactivation of CYPs in 37 °C comprehensively, the incubation time was chosen to 20 min. Moreover, the microsomal protein concentration in HLMs was 0.5 mg/ml, which was the most frequently employed, and high protein concentrations ought to be avoided because of the possibility of protein binding 18 .
To investigate whether inhibition of CYPs by plumbagin was time-dependent, time-dependent inhibition assays were performed with various plumbagin concentrations close to IC 25 , which was the most sensitive point for detecting a time-dependent inhibitor 36,37 . Too low concentrations of plumbagin may miss inactivation, but too high concentrations of plumbagin may cause much reversible inhibition to make the effect of time-dependent inhibition not obvious 36 . Meanwhile, considering not only automatic inactivation of some CYPs but also weak inhibitors which could not be identified within a short inactivation time, 20 min was used for this time-dependent inhibition assays. From Fig. 2, the activity of CYP2D6 in HLMs decreased slightly in a time-dependent manner without any inhibitor, which could be explained for easily automatic inactivation of CYP2D6 in HLMs. On the other hand, the inhibition effects on several CYPs by plumbagin were reduced with pre-incubation with NADPH, which could be due to metabolic depletion of plumbagin leading to diminished inhibitory capability 36,38 . Therefore, no time-dependent inhibition was observed in both HLMs and RLMs, which showed reversible inhibition of CYP1A2, CYP2B1/6, CYP2C9/11, CYP2D1/6, CYP2E1 and CYP3A2/4 by plumbagin.
The present study showed that the types of inhibition by plumbagin were mixed inhibition on CYP2B6, CYP2C9, CYP2D6, CYP2E1 and CYP3A4, but non-competitive for CYP1A2 in HLMs and the K i values were 1.62 μ M, 2.16 μ M, 1.46 μ M, 0.65 μ M, 0.88 μ M and 0.15 μ M, respectively. Correspondingly, mixed inhibition on CYP1A2 and CYP2D1, but competitive for CYP2B1, CYP2C11 and CYP2E1 were observed in RLMs and the K i values were 2.64 μ M, 9.49 μ M, 9.93 μ M, 7.85 μ M, and 6.28 μ M, respectively. Thus, plumbagin was a potent inhibitor on the above CYPs in both humans and rats.
Based on the inhibition results of HLMs and RLMs, some differences should be noticed such as the types of CYP inhibition and inhibitory intensity. The K i values of inhibition on CYPs by plumbagin in HLMs are generally lower than those in RLMs. It may be attributed to variations in CYP levels, inducibilities and the existence of different CYP isoforms of the same protein family or subfamily in different species 39 . For example, CYP2C11 accounts for about 50% in total rat CYPs while human CYP2C9 being a homologue of the rat CYP2C11 with 77% homology only accounts for 20% of total human CYPs 40 . In addition, the enzyme inhibition kinetic of CYP3A2 disagreed with the classical Michaelis-Menten kinetics in RLMs, but the enzyme inhibition kinetic of CYP3A4 in HLMs conformed to that when the concentrations of midazolam were no more than 5 μ Μ . The differences between CYP3A2 and CYP3A4 could also be a reflection of species differences. Consequently, extrapolating results from animals to humans should be cautious.
As mentioned above, when investigating the enzyme inhibition kinetic of CYP3A2, we found it did not conform to the classical Michaelis-Menten kinetics in RLMs. In fact, metabolic activation was achieved by incubation of plumbagin in the presence of CYP3A2 when the concentrations of midazolam are less than 5 μ Μ in RLMs 41 . In this report, however, the unusual effects of plumbagin on CYP3A2 could be explained using an atypical kinetics model. Atypical Michaelis-Menten kinetics including substrate inhibition, partial inhibition, activation, auto-activation and biphasic metabolism has been demonstrated for some CYP3A-mediated reactions 42,43 . In this study, CYP3A2 displayed auto-activation kinetics. Multisite hypothesis was a possible mechanistic model of atypical kinetics. CYP3A enzyme except for two or more midazolam-binding sites contains an activator binding site which is closely related to the substrate binding sites 42 . Plumbagin might lead to metabolic activation by binding to the activator binding sites. These speculations still need to be further proved by experiments.
Previous pharmacokinetic study of plumbagin by oral administration (100 mg/kg) in rats showed that the peak concentration (C max ), half-life (t 1/2 ) and area under the curve (AUC) of plumbagin in the plasma was 0.35 μ g/ml (1.86 μ M), 1028 min and 271.9 min* μ g/ml, respectively 3 . Therefore, the rat plasma concentrations were close to the IC 50 and K i values determined in this study, thus suggesting plumbagin could quite probably decrease the activities of CYP1A2, CYP2B1, CYP2C11, CYP2D1, CYP2E1 and CYP3A2 in vivo. In fact, these CYP homologues in humans are responsible for 80% of the CYP-mediated metabolism including a number of drugs with narrow therapeutic ranges. Therefore the potential HDIs should be brought to attention and further studies in vivo are needed to confirm the interactions of plumbagin with CYP isoforms in humans.
In conclusion, the data in this study demonstrated that plumbagin exhibited potently broad-spectrum inhibitory effects toward CYP enzymes including CYP1A2, CYP2B1/6, CYP2C9/11, CYP2D1/6, CYP2E1 and CYP3A2/4 with lower K i values in both human and rat liver. In particular, plumbagin was not only a mixed inhibitor of CYP2B6, CYP2C9, CYP2D6, CYP2E1 and CYP3A4, but also a non-competitive inhibition of CYP1A2 in humans. These results provided useful guidance for safe and effective usage of plumbaign in clinic, especially in combination with other drugs.  Preparation of rat liver microsomes. In this study, the rats were fasted overnight and killed by cervical dislocation before removal of the liver. The liver was excised, rinsed with ice-cold saline (0.9% NaCl w/v), weighed and homogenized in a 0.05 mM Tris/KCl buffer (pH 7.4). Subsequently, the homogenate was centrifuged at 10,000 × g at 4 °C for 30 min. The supernatant was then centrifuged at 105,000 × g at 4 °C for 60 min. The pellet was reconstituted with 0.05 mM Tris/KCl buffer (pH 7.4) and stored at − 150 °C until use. The protein concentration of the liver microsomes was determined by a protein quantitative assay using bicinchoninic acid 22 .

Microsomal incubation.
Microsomal incubation was carried out at 37 °C in a mixture of 200 μ l in 0.05 M Tris/HCl buffer (pH 7.4) consisting of pooled HLMs (0.5 mg/ml) or RLMs (1 mg/ml) and an NADPH-regenerating system including MgCl 2 (5 mM), G6P (10 mM), G6PDH (0.4 U/ml) and NADP (1 mM). Substrates were added into the incubation mixture and obtained the final concentrations (10 μ M phenacetin for CYP1A2, 20 μ M bupropion for CYP2B1/6, 20 μ M tolbutamide for CYP2C9/11, 5 μ M dextromethorphan for CYP2D1/6, 20 μ M chlorzoxazone for CYP2E1 and 10 μ M midazolam for CYP3A2/4), which were close to K m respectively. The proportion of organic solvent was not higher than 1% (v/v) in the incubation mixture 18,44 . After pre-incubation for 5 min, the reaction was initiated by addition of NADP. Incubations were performed for 20 min, stopped by adding 200 μ l ice-cold acetonitrile. And then 20 μ l internal standard solution was added (3-acetamidophenol 2 μ g/ml, chlorpropamide 2 μ g/ml and mebendazole 2 μ g/ml). The solution was mixed thoroughly for 3 min and centrifuged for 10 min at 16,000 × g at 4 °C. The supernatant (60 μ l) was transferred to autosampler vials, and 2 μ l was injected into LC-MS/MS system.

Determination of IC 50 values for inhibition.
The IC 50 (concentration of inhibitor to cause 50% inhibition of original enzyme activity) values were determined with the same incubation condition above. The specific concentrations of plumbagin were ranged from 0.1 μ M to 50 μ M. The IC 50 values were calculated by plotting relative activities over the logarithm of plumbagin concentrations using GraphPad Prism 5.0 (GraphPad software Inc., CA, USA).

Mechanism of inhibition of CYPs by plumbagin.
To explore the time-dependent inhibition of plumbagin on CYPs, inactivation incubation containing plumbagin whose final concentration close to IC 25 (concentration of inhibitor to cause 25% inhibition of original enzyme activity) 37  and metabolites in a short time, and to calculate the enzymatic reaction rate, which is used to evaluate the corresponding CYP enzyme activities. In this study, an Agilent 1290 LC system consisting of a degasser, a binary pump, an autosampler and a thermostatic column compartment was coupled with a 6460 triple-quadrupole mass   Chromatography separation was performed on a Phenomenex Kinetex XB-C18 column (100 × 3.00 mm, 2.6 μ M) protected by a Phenomenex C18 guard column (Torrance, CA, USA). The mobile phase consisted of solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile) using gradient elution at a flow rate of 0.3 mL/min. The optimum condition for elution was as follows: 0-2.2 min, 10-11% B; 2.2-8.5 min, 11-90% B; 8.5-9 min, 90-92% B; 9-9.3 min, 92% B; 9.3-9.6 min, 92-10% B; 9.6-11.5 min, 10% B. The temperature of column oven was maintained at 30 °C, and the injection volume was 2 μ L. The mass spectrometer was operated in the positive/negative ion-switching ESI mode, and the detection of the ions was performed in the multiple reaction monitoring (MRM) mode. Each MRM transition and the compound dependent parameters are presented in Table 2. MS/MS conditions were optimized as follows: gas temperature, 350 °C; gas flow, 10 L/min; nebulizer, 35 psi; sheath gas temperature, 350 °C; sheath gas flow, 11 L/min; collision gas (N 2 ), 1.6 MPa. A divert valve was used to divert the eluent to waste from 0 to 2 min, 10 to 11.5 min, and to MS from 2 to 10 min. Representative multiple reaction monitoring chromatograms of analytes were shown in Fig. 6.
Data analysis. All data were presented as mean ± SD. The IC 50 values were fitted by non-linear regression analysis using GraphPad Prism 5.0. The kinetic parameters were obtained by fitting velocity data to kinetic models (Michaelis-Menten) using GraphPad Prism 5.0. The Lineweaver-Burk Plot (the reciprocal of reaction velocities versus the reciprocal of substrates concentrations) was used to determine the quality of fit to a specific inhibition model. The inhibition constant (K i ) was obtained by a secondary plot using the slopes of the primary Lineweaver-Burk Plot (K m /V max versus inhibitor concentrations). α K i was obtained from a secondary plot using the y-intercepts of the Lineweaver-Burk plot (1/V max versus inhibitor concentrations). All results were analyzed in quadruplicate. One-way analysis of variance was used to estimate the significance of differences. There was statistical significance between control and test groups if p < 0.05.