Inhibition of CYP3A-mediated Midazolam Metabolism by Kaempferia Parviflora

Abstract Kaempferia parviflora (KP) extract has recently attracted attention in Japan as a dietary supplement; however, there is little information regarding food-drug interactions (FDIs). The current study was conducted to clarify the FDI of KP extract via inhibition of cytochrome P450 3A (CYP3A), a typical drug-metabolizing enzyme. The inhibitory effects of KP extract and its main ingredients, 5,7-dimethoxyflavone (5,7-DMF) and 3,5,7,3’,4’-pentamethoxyflavone (3,5,7,3’,4’-PMF), on CYP3A-mediated midazolam 1’-hydroxylation (MDZ 1’-OH) activity were investigated in human liver microsomes. In addition, the effect of a single oral treatment with KP extract (135 mg/kg) on oral MDZ (15 mg/kg) metabolism was investigated in rats. Serum MDZ concentration was analyzed and pharmacokinetic parameters were compared with the control group. KP extract competitively inhibited MDZ 1’-OH activity with an inhibition constant value of 78.14 µg/ml, which was lower than the estimated concentration in the small intestine after ingestion. Furthermore, KP extract, 5,7-DMF, and 3,5,7,3’,4’-PMF inhibited the activity in a time-, NADPH-, and concentration-dependent manner. In vivo study showed that administration of KP extract to rats 2 h before MDZ significantly increased the area under the serum concentration-time curve and the maximum concentration of MDZ significantly by 2.3- and 1.9- fold, respectively (p < 0.05). Conversely, administration of MDZ 18 h after KP extract treatment displayed a weaker effect. These results suggest that KP extract competitively inhibits CYP3A-mediated MDZ metabolism, and that this inhibition may be time-dependent but not irreversible. This work suggests an FDI through CYP3A inhibition by KP extract.


Introduction
Dietary supplements are widely used to maintain and im-prove health and beauty. In recent years, self-medication has been promoted in Japan in response to the increase in medical expenses in an aging society with a declining birthrate, and the market for supplements is expected to expand further in the future. Among them, products using herbal extracts are classified as foods in Japan, but some products, such as St. John's wort, saw palmetto, and echinacea, are used as drugs in European countries [1][2][3] . In Japan, approval review regulations for herbal medicinal products were announced in 2007, and red vine leaf extract and chest berry extract have been approved as direct over-the-counter (OTC) medicines. Since these OTC drugs and supplements derived from herbal extracts are easily available at drug stores and online retailers, the number of people using them has been increasing.
However, it is very important to pay attention to drug interactions in order to safely use drugs. In particular, drug interactions mediated by inhibition of the typical drug-metabolizing enzyme, cytochrome P450 3A (CYP3A), is known to be an important factor in the incidence of adverse effects with concomitant drugs. Drug-drug interactions caused by CYP3A inhibition have been reported in many drugs such as simvastatin and nifedipine, and careful attention has been paid to them 4,5) .
In addition to drug-drug interactions, food-drug interactions (FDIs) have also been reported. As a typical example of such FDIs, grapefruit juice (GFJ) has been shown to inhibit small intestinal CYP3A activity, thereby increasing the blood levels of concomitant drugs metabolized by this enzyme, such as felodipine, triazolam, and simvastatin, and increasing the possibility of adverse effects [6][7][8] . Therefore, it is important to pay attention to FDIs to ensure the safe use of these combinations. However, there are limited reports of such FDIs; therefore, in many cases, it is uncertain whether their combined use is safe. In general, dietary supplements tend to be regarded as safe, and physicians and pharmacists rarely pay attention to their combined use with medicines. This can lead to unexpected adverse effects. For these reasons, providing information on FDIs and making it available to physicians, pharmacists, and patients is very important for accurate and safe self-medication.
Since KP extract may be used by both healthy people and patients taking prescription medications, it is predicted that this supplement will be ingested in combination with drugs.
There have been few reports on FDIs through CYP3A inhibition by KP extract. KP extract reportedly inhibited CYP3A activity in an in vitro experiment using mouse liver microsomes 17) . In addition, Ochiai et al reported that administration of 5,7-DMF to mice for ten days increased the area under the serum concentration-time curve (AUC) and elimination half-life (t 1/2 ) of midazolam (MDZ), which is a probe drug of CYP3A, and decreased hepatic CYP3A expression 18) . However, both reports used mice and whether the same effects will also be observed in humans is unclear. In addition, it is important to investigate the effect of KP extract treatment, which is a mixture of several polymethoxyflavones other than 5,7-DMF, in vivo.
Mechanism-based inhibition (MBI) is an important mechanism in drug interactions. In this case, the reactive metabolic intermediate produced by CYP irreversibly inactivates this enzyme, which can lead to long-lasting severe adverse effects. It should be noted that GFJ is a typical example that causes MBI, and thus even foods can cause this severe inhibition 19) . Therefore, it is important to investigate whether KP extract causes MBI of CYP3A, since this enzyme is involved in the metabolism of many clinically important drugs. Therefore, the current study was conducted to clarify the inhibitory effects of KP extract and its main ingredients, 5,7-DMF and 3,5,7,3',4'-PMF, on CYP3A-mediated MDZ 1'-hydroxylation in human liver microsomes (HLM). In addition, we examined the effect of KP extract administration on MDZ metabolism in rats in vivo. The aim of this study was to investigate FDI through the inhibition of CYP3A by KP extract and to provide information on the safety of KP extract during medication. This information is very important for promoting the proper use of dietary supplements.
A pooled sample of HLM (pooled from 10 donors) was purchased from the Non-Profit Organization Human and Animal Bridging Research Organization (Chiba, Japan). All other chemicals used were of analytical grade.

Preparation of Kaempferia Parviflora extract and Analysis of Polymethoxyflavones
The dried rhizomes of KP (Maechu Co., Ltd., Nara, Japan) were extracted twice with 80% ethanol. After filtration, the solution was evaporated under vacuum to yield dry powder.
The decrease in the natural logarithm of the % residual MDZ 1'-OH activity against pre-incubation time was plotted for each inhibitor concentration, and the apparent inactivation rate constant (K obs ) values were described as the negative slopes of the lines. The maximal inactivation rate constant (k inact ) and the inhibitor concentration causing half-maximal inactivation (K I ) were estimated by double-reciprocal plot of the K obs against the inhibitor concentration. The efficiency of inactivation was calculated as the ratio of k inact to K I .

Animal Treatment and Sampling
Male Sprague-Dawley rats (Japan Laboratory Animals, Inc., Tokyo, Japan), weighing 200-230 g, were used in this study. The rats were housed in a temperature-controlled room (23°C) with a 12 h light (6:00 AM-6:00 PM) and dark (6:00 PM-6:00 AM) cycle and fed standard laboratory chow ad libitum. The rats were fasted overnight (after 6:00 PM) before sampling. All procedures were reviewed and approved by the Institutional Animal Care and Use Committee of Showa University (Permit Number: 07002, approval date: April 1, 2017) and complied with the guidelines established by the National Institute of Health (NIH) in the Guide for the Care and Use of Laboratory Animals (8th edition, ILAR-NRC).

Determination of Serum MDZ Concentration and Pharmacokinetic Analysis
The serum concentrations of MDZ were determined using previously described HPLC methods with slight modifications [22][23][24] . Briefly, 100 µL of each serum sample was diluted with 500 µL of 0.1 N NaOH, and 5 ng of diazepam was added as an internal standard. The mixture was extracted with 8 ml of dichloromethane-pentane (1:1) for 1 min and centrifuged at 1,600 × g for 10 min. The upper organic solvent phase was transferred to a clean glass tube and evaporated to dryness under a nitrogen stream. The dried residue was dissolved in mobile phase 10 mM acetate buffer (pH 3.8):CH 3 CN:MeOH (50:38:12) at 30°C and injected into HPLC equipment with a CAPCELLPAK C18 MG II column (2.0 × 250 mm, Shiseido). As the mobile phase 50:50 mixture (v/v) of 10 mM phosphate buffer (pH 5.0) and CH 3 CN was used at a flow rate of 0.2 ml/ min. The UV detector was set at 220 nm.
The pharmacokinetic parameters of MDZ were assessed by non-compartmental analysis using MOMENT (Excel ® ) based on the moment analytic method 25,26) . The area under the serum concentration-time curve from time 0 to infinity (AUC 0-∞ ) was calculated according to the trapezoidal rule. The maximum serum concentration (C max ) was obtained from the actual data. The t 1/2 was calculated by dividing ln 2 by λ, which is the terminal elimination rate constant calculated using a linear regression analysis of at least four points from the terminal portion.

Statistical Analysis
In vitro data are expressed as the means of duplicate examinations. In vivo data are presented as mean ± standard error (SE). Differences between the groups were analyzed using Dunnett's test using JMP Pro 15 (SAS Institute Inc., Cary, NC, USA). Differences were considered statistically significant at p < 0.05.

Direct Inhibition Study
KP extract inhibited CYP3A-mediated MDZ 1'-OH activity competitively with a K i value of 78.14 µg/ml in HLM ( Fig. 2A).

In Vivo study
The serum concentration-time profiles and the pharmacokinetic parameters of MDZ after 2 or 18 h of single treatment with KP extract are shown in Fig. 5 and Table 1. In the KP extract treated 2 h before MDZ administration, the mean AUC 0-∞ and C max of MDZ significantly increased from 564 ± 166 to 1312 ± 55 ng/ml•h, and from 473 ± 161 to 902 ± 39 ng/ml, respectively (p < 0.05), without a change in t 1/2 . KP extract treatment 18 h before MDZ administration showed a significant increase in the C max from 473 ± 161 to 763 ± 148 ng/ml (p < 0.05). However, marginal changes were observed in the AUC 0-∞ and t 1/2 between the control and treatment groups in this study.

Discussion
In this study, we investigated whether KP extract and its main ingredients, 5,7-DMF and 3,5,7,3',4'-PMF, cause FDIs mediated by CYP3A inhibition. As a result of the in vitro study, KP extract showed competitive inhibition of CYP3Amediated MDZ 1'-OH activity ( Fig. 2A). In addition to the inhibitory effect of KP extract on CYP3A-mediated erythromycin N-demethylation activity reported in mouse liver microsomes 17) , the current study revealed that this inhibition also occurs in HLM.  It is important to predict whether this inhibition can occur in vivo. It is unclear how much of the ingredients in the extract that inhibit CYP3A activity are transferred into the blood and liver after oral administration of KP extract to humans. Thus, it is difficult to predict FDI in the liver based on in vitro results. CYP3A4 is also highly expressed in the small intestine; therefore, it is important to consider drug interactions in the small intestine 27) . Dietary supplements containing KP extract sold in Japan are labeled with a recommended daily intake of polymethoxyflavones, including 5,7-DMF and 3,5,7,3',4'-PMF, at 12 mg. Since the KP extract used in this study contained 27.5% polymethoxyflavones, the recommended daily intake was calculated to be 44 mg. The FDA recommends using a volume of 250 mL to calculate the theoretical small intestinal drug concentration after oral administration. According to this, the estimated concentration of KP extract in the small intestine was calculated to be 176 µg/ml by dividing 44 mg by 250 ml. This value is higher than 78.14 µg/ml, which corresponds to the K i value of CYP3A activity obtained in this study, suggesting that KP extract may cause an FDI mediated by CYP3A inhibition, at least in the small intestine.
In contrast, 5,7-DMF and 3,5,7,3',4'-PMF, the primary polymethoxyflavones of KP extract, did not inhibit MDZ 1'-OH activity (Fig. 2B, 2C). The KP extract used in this study contained 7.45% of 5,7-DMF and 7.2% of 3,5,7,3',4'-PMF, thus their concentrations in 78.14 µg/ml of KP extract were calculated to be 20.59 µM and 15.08 µM, respectively. None of the compounds inhibited CYP3A activity at this concentration, suggesting that these polymethoxyflavones are not involved in the inhibition of CYP3A activity by KP extract. Therefore, other constituents may be involved in the competitive inhibition of CYP3A activity by KP extract.
Among the mechanisms of CYP inhibition, MBI requires particular clinical attention because the inhibitor irreversibly binds and inactivates CYP, resulting in long-lasting severe inhibition. Therefore, we investigated whether such MBI occurs with KP extract and its main ingredients. As a result of pre-incubation of the reaction mixture containing KP extract, 5,7-DMF, or 3,5,7,3',4'-PMF, CYP3A4-mediated MDZ 1'-OH activity was inhibited in a time-dependent manner (Fig. 3A, 3B, and 3C). Furthermore, this inhibition was NADPH-dependent, suggesting MBI, in which metabolic intermediates are produced by CYP.
As described above, although 5,7-DMF and 3,5,7,3',4'-PMF did not inhibit CYP3A activity when added to the reaction mixture at the same time as MDZ, these compounds showed inhibitory effects in this experiment, including the pre-incubation step. From this result, even though the affinity of these compounds for CYP3A is weaker than that of MDZ, the metabolism of co-administered drugs may be inhibited when they are ingested before the administration of drugs that are metabolized by CYP3A, or when they are used in combination with drugs with a lower affinity for CYP3A. 5,7-DMF is reportedly not metabolized by CYP3A4, but by CYP1A1 and CYP1A2, in HLM and recombinant microsomes 28) . Therefore, the metabolic intermediates produced by these CYP enzymes may inactivate CYP3A. To evaluate the risk of time-dependent inhibition, the inactivation efficiency (k inact /K I ) was used 29) . The values of 5,7-DMF and 3,5,7,3',4'-PMF, obtained from the plots shown in Fig. 4B and 4C, were calculated to be 0.0024 and 0.000063 ml/min•µmol, respectively. The derived k inact /K I of 5,7-DMF for CYP3A was comparable to that of ingredients such as bergamottin and 6,7-dihydroxybergamottin in grapefruit juice, which is known to be a representative inactivator of CYP3A 30) . The above results suggest that KP extract causes MBI of CYP3A, and 5,7-DMF and 3,5,7,3',4'-PMF are involved in this inactivation.
To clarify whether the inhibition of CYP3A-mediated drug metabolism by the KP extract in in vitro experiments could occur in vivo, the experiments were performed in rats. To evaluate the direct inhibitory effect of KP extract, a single dose of KP extract was administered 2 h before MDZ administration. The AUC 0-∞ and C max of MDZ increased 2.3 and 1.9 times, respectively, after treatment with KP extract (Fig. 5, Table. 1). This suggests that KP extract induced drug interactions mediated by CYP3A inhibition in rats in vivo. Ochiai et al reported that administration of 5,7-DMF to mice for 10 days resulted in decreased expression of hepatic CYP3A, increased AUC 0-∞ and C max of MDZ, and KP extract (135 mg/kg) or water (control) was orally administered to rats, followed by 2 or 18 h oral administration of MDZ (20 mg/ kg). Each point represents the mean ± S.E. of 3-5 rats.
prolonged half-life 18) . In contrast, in our experiments, the effect of a single KP extract treatment was examined as a condition that does not affect expression of CYP3A, and the results obtained are considered to be mediated by CYP3A inhibition. From the results of the direct inhibition study in vitro, the low contribution of 5,7-DMF and 3,5,7,3',4'-PMF in CYP3A inhibition was considered. In addition to these polymethoxyflavones, KP extract contains many other ingredients. It is difficult to investigate the inhibitory effect of all these ingredients on CYP3A, but even if the inhibitory ability of each ingredient on CYP3A is weak, it is possible they exhibit inhibitory effects when combined.
Administration of MDZ 18 h after KP extract treatment considering MBI showed that the increase in AUC 0-∞ and C max of MDZ was less than that after 2 h of MDZ administration. The results suggestive of MBI shown in in vitro experiments using HLM and that of this in vivo experiment are inconsistent. It has been shown that polymethoxyflavones, such as 5,7-DMF and 3,5,7,3',4'-PMF contained in KP extract, are distributed mainly in the organs at much higher levels than in blood after oral administration 31) . Therefore, these polymethoxyflavones reached the liver and small intestine and may be involved in the inactivation of CYP3A. Nevertheless, weak inhibition of MDZ metabolism was observed after 18 h. One possible reason for this is that the inhibition mechanisms are time-dependent, but not irreversible. In addition, the species differences in CYP3A need to be considered. Regarding CYP3A isoforms, CYP3A4 is expressed in the human liver and small intestine, whereas in rats, CYP3A2 is expressed in the liver, and other isoforms such as CYP3A9, CYP3A18 and CYP3A62 in the small intestine 32) . These isoforms participate in the metabolism of MDZ, and MDZ is used as an indicator of CYP3A-mediated drug metabolism 33,34) . However, because of species differences, the degree and the mechanism of CYP3A inhibition by KP extract, 5,7-DMF, and PMF might vary. From the results of in vivo experiments examined under the above two conditions, it was considered that in rats in vivo, the inhibition of CYP3A by KP extract is caused by a reversible competitive mechanism, and the constituents in KP extract may decrease to less than the concentration that inhibits CYP3A at 18 h after administration.
It is unclear whether KP extract inhibits CYP3A in humans through a competitive or time-dependent mechanism. KP extract is used as a dietary supplement in Japan and may be used in combination with statins when ingested by patients with dyslipidemia to reduce abdominal fat and prevent obesity. Among the statins, simvastatin and atorvastatin are known to be metabolized by CYP3A; therefore, inhibition of this enzyme may increase their levels in the blood, resulting in an increased risk of rhabdomyolysis. KP extract is also used by the elderly population to support muscle endurance. In such cases, it may be used in combination with various drugs, and careful caution is required because CYP3A is involved in the metabolism of many drugs. It is necessary to conduct an optimal study to clarify whether KP extract causes FDIs mediated by CYP3A inhibition in humans.
In conclusion, KP extract showed competitive inhibition of CYP3A-mediated MDZ metabolism in HLM and in rats in vivo. In addition, the time-dependent inhibition of MDZ 1'-OH activity was demonstrated by KP extract, 5,7-DMF, and 3,5,7,3',4'-PMF. These results suggest that KP extract may cause FDIs through CYP3A inhibition. Therefore, this study shows that it is better for patients taking drugs metabolized by CYP3A to avoid using supplements containing KP extract to prevent possible adverse events caused by FDIs.