Abstract
Kava kava (Piper methysticum), an herbal remedy, is widely used for the treatment of mild to moderate cases of anxiety. The therapeutic activity is presumably achieved through multiple constituents called kavalactones. Recently, kava extracts were shown to induce CYP3A4 and activate human pregnane X receptor (PXR). This study was undertaken to test the ability of purified kavalactones to induce CYP3A23 and activate PXR. Rat hepatocytes were treated with desmethoxyyangonin, dihydrokawain, dihydromethysticin, kawain, methysticin, or yangonin, and the expression of CYP3A23 was monitored. Among the kavalactones, only desmethoxyyangonin and dihydromethysticin markedly induced the expression of CYP3A23 (∼7-fold). A similar magnitude of induction was detected with combined six kavalactones at a noninductive concentration when individually used. The induced expression, however, was markedly reduced or completely abolished if dihydromethysticin, desmethoxyyangonin, or both were excluded from the mixtures. Interestingly, regardless of whether dihydromethysticin or desmethoxyyangonin was used alone or together with other kavalactones, similar amounts of total kavalactones were needed to produce comparable induction, suggesting that the inductive activity of dihydromethysticin and desmethoxyyangonin is additively/synergistically enhanced by other kavalactones. In addition, treatment with dihydromethysticin, desmethoxyyangonin, or pregnenolone 16α-carbonitrile (PCN) markedly increased the levels of CYP3A23 mRNA, and inhibition of mRNA synthesis abolished the induction. In contrast to PCN, dihydromethysticin and desmethoxyyangonin only slightly activated rat or human PXR. These findings suggest that the induction of CYP3A23 by dihydromethysticin and desmethoxyyangonin involves transcription activation, probably through a PXR-independent or PXR-involved indirect mechanism.
Kava (P. methysticum) is a shrub that grows perennially in the islands of the South Pacific. Local inhabitants have used kava for thousands of years as a beverage in social, recreational, and ceremonial events (Singh and Blumenthal, 1997; Singh and Singh, 2002; Whitton et al., 2003). Kava is known to promote relaxation and has recently gained popularity worldwide because of its role in treating a variety of subclinical disorders (Volz and Kieser, 1997; Pittler and Ernst, 2000; Martin et al., 2002; Singh and Singh, 2002). In addition to being used to treat mild or moderate anxiety, kava preparations are used for insomnia, stress, menopause, and muscle fatigue. Experiments with animal models and epidemiological data have shown that kava has a wide spectrum of therapeutic properties, such as sedative, anxiolytic, analgesic, and neuroprotective effects (Jamieson and Duffield, 1990; Martin et al., 2000; Wheatley, 2001; Singh and Singh, 2002; Cagnacci et al., 2003). It has been postulated that these activities are achieved through multiple mechanisms, such as blockade of voltage-gated sodium channels (Schirrmacher et al., 1999), enhanced interactions between ligand and corresponding receptors (e.g., aminobutyric acid type A receptor) (Davies et al., 1992), modulation on neurotransmitter release (Baum et al., 1998), inhibition of enzyme activity (e.g., cyclooxygenase-2) (Wu et al., 2002), and decrease in cytokine release (e.g., tumor necrosis factor-α; TNF-α) (Hashimoto et al., 2003).
Kavalactones are recognized as the active constituents responsible for the observed therapeutic activities (Singh and Singh, 2002). As many as 18 kavalactones have been identified, and six of them contribute as much as 95% to the overall pharmacological activity (Bilia et al., 2002; Dharmaratne et al., 2002; Singh and Singh, 2002). These kavalactones include desmethoxyyangonin (DMY), dihydrokawain, dihydromethysticin (DMS), kawain, methysticin, and yangonin. The relative abundance of the kavalactones varies markedly among different parts of the plant and also depends on where the plant grows (Singh and Blumenthal, 1997; Dinh et al., 2001). Structurally, the active kavalactones exist as the (+) isomer, and all kavalactones contain a benzene ring and a methoxy lactone connected via an ethyl or ethylene bridge (Dharmaratne et al., 2002). Because of the similarities on the basic structure, kavalactones exhibit overlapping pharmacological properties. For example, all six kavalactones cause a marked decrease on the release of TNF-α, although the relative potency varies by as much as 10-fold (Hashimoto et al., 2003). Interestingly, kavalactones have been shown to exert synergistic rather than additive activity when given together, suggesting that kavalactones have more than one molecular target (Keledjian et al., 1988; Singh and Singh, 2002).
Kava extracts or kavalactones generally have favorable pharmacological and toxicological profiles (Singh and Singh, 2002; Whitton et al., 2003); however, cases with severe liver injury have been reported (Humberston et al., 2003; Teschke et al., 2003). The predominant liver damage suggests that kava-induced toxicity is due to the formation of metabolites rather than the parent constituents. Although the overall metabolism of kava extracts or kavalactones remains to be determined, they have been shown to profoundly alter the catalytic activities of P450 enzymes (Mathews et al., 2002; Unger et al., 2002), a superfamily of enzymes that are involved in the metabolism of almost all xenobiotics. Incubation of kavalactones (∼100 μM) decreases P450-mediated catalysis by as much as 80%, notably on CYP1A2, 2C9, 2C19, 2D6, and 3A4. The kavalactone-mediated inhibition is highly concentration-dependent with modest changes on the P450 spectrum, suggesting that kavalactones act competitively and are indeed P450 substrates (Mathews et al., 2002; Unger et al., 2002). In addition to competitive inhibition, total kava extracts are found to markedly increase the expression of CYP3A4, an enzyme that is involved in the metabolism of more than 50% of drugs and other xenobiotics (Raucy, 2003).
The aim of the present study was to test the ability of purified kavalactones to induce CYP3A23. Among six pharmacological kavalactones, only desmethoxyyangonin and dihydromethysticin markedly induce CYP3A23 accompanied by increased levels of CYP3A23 mRNA. Selective subtraction of dihydromethysticin, desmethoxyyangonin, or both from combined six kavalactones markedly reduced or completely abolished the induction. Interestingly, similar amounts of total kavalactones were needed to produce comparable induction, regardless of whether dihydromethysticin or desmethoxyyangonin was used alone or together with other kavalactones. These findings establish that dihydromethysticin and desmethoxyyangonin are two CYP3A inducers and that their inductive activity is additively/synergistically enhanced by other kavalactones.
Materials and Methods
Chemicals and Supplies. Dexamethasone, PCN, rifampicin, HBSS, Williams' E medium (WEM), and β-actin antibody were purchased from Sigma-Aldrich (St. Louis, MO). Dulbecco's modified Eagle's medium, ITS, LipofectAMINE, and Plus Reagent were obtained from Invitrogen (Carlsbad, CA). Collagenase (type I) was obtained from Worthington Biochemical Corporation (Lakewood, NJ). The goat anti-rabbit IgG conjugated with horseradish peroxidase or alkaline phosphatase and the chemiluminescent kit were obtained from Pierce (Rockford, IL). Kits for luciferase detection and the null-Renilla luciferase plasmid were obtained from Promega (Madison, WI). Delipidated and normal fetal bovine sera were obtained from HyClone (Logan, UT). The CYP3A23 antibody was purchased from Research Diagnostics (Flanders, NJ). The procedure for kavalactone purification and structural determination was described elsewhere (Dharmaratne et al., 2002). Unless otherwise specified, all other reagents were purchased from Fisher Scientific Co. (Pittsburgh, PA). Male Sprague-Dawley rats (∼220 g) were purchased from Harlan Teklad (Madison, WI). The use of animals was approved by the Institutional Animal Care and Use Committee.
Hepatocyte Culture and Treatment. Primary cultures of rat hepatocytes were isolated from adult male Sprague-Dawley rats (∼220 g) by a modified two-step collagenase digestion method, essentially as described previously (Zhu et al., 2000). Rat liver was perfused through the portal vein with calcium-free HBSS buffer containing 0.5 mM EGTA for ∼1 to 2 min at a flow rate of 28 ml/min until the liver was completely blanched (∼1–2 min), followed by perfusion for 5 to 8 min with WEM containing collagenase (150 U of collagenase activity per milliliter of medium). The liver was further perfused with calcium-free HBSS containing EGTA. Hepatocytes were dispersed from the digested liver in WEM without collagenase and washed three times by low-speed centrifugation (100–150g, 5min). The resulting cell pellet was then suspended in WEM containing 10% fetal bovine serum, ITS supplement, and dexamethasone (Zhu et al., 2000), and viability was determined by trypan blue exclusion. Hepatocytes were then plated onto a collagen-coated culture plate at a density of 6 × 105 hepatocytes per well (12-well plate). The cells were allowed to attach for 3 to 4 h at 37°C in a humidified chamber with 95%/5% air/CO2. Culture plates were then gently swirled, and the medium containing unattached cells was then aspirated. Fresh WEM containing ITS supplement and dexamethasone (100 nM) was added to each well, and the cultures were returned to the humidified chamber. Unless otherwise specified, the cells were maintained for 48 h before initiating treatment with chemicals. Groups of cultures (n = 3 individual rats) were then treated for 24 to 72 h with kavalactones, PCN, rifampicin, or the same volume of vehicle DMSO (final concentration, 0.1%), with daily changes of medium containing appropriate chemicals or solvent.
Western Analysis. Hepatocytes were washed once with cold phosphate-buffered saline and lysed in Tris-HCl buffer (50 mM, pH 7.4) containing SDS and Triton X-100 (0.1% each). The cell lysates were subjected to centrifugation at 12,000g for 15 min at 4°C to remove insoluble precipitates. Cell lysates (usually 6.5 μg) were resolved by 7.5% SDS-polyacrylamide gel electrophoresis in a minigel apparatus and transferred electrophoretically to nitrocellulose membranes (Bio-Rad, Hercules, CA). After nonspecific binding sites were blocked with 5% nonfat milk, the blots were incubated with an antibody against CYP3A23 or β-actin. The anti-CYP3A23 antibody recognized a single protein band with a molecular weight of ∼52 kDa in the lysates of hepatocytes. This antibody showed no cross-reactivity toward CYP3A2, a closely related enzyme of CYP3A23. The primary antibodies were subsequently localized with the goat anti-rabbit IgG conjugated with horseradish peroxidase or alkaline phosphatase (Xie et al., 2003). Horseradish peroxidase activity was detected with a chemiluminescent kit (SuperSignal West Pico; Pierce), whereas alkaline phosphatase activity was detected with the coupled substrates nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl-phosphate (Bio-Rad). The chemiluminescent signal was captured by KODAK Image Station 2000, and the relative intensities were quantified by KODAK 1D Image Analysis Software (Eastman Kodak, Rochester, NY).
Reverse Transcription-Coupled Polymerase Chain Reaction (RT-PCR). The mRNA levels of CYP3A23 were determined by RT-PCR experiments with a ThermoScript I kit (Invitrogen) as described previously (Li et al., 2003). Total RNA (2 μg) was subjected to the synthesis of the first-strand cDNA in a total volume of 30 μl with random primers and ThermoScript I reverse transcriptase. The reactions were incubated initially at 50°C for 30 min and then at 60°C for 60 min after additional reverse transcriptase was added. The cDNAs were then diluted by 100 times and subjected to PCR amplification (10 μl of the diluted cDNA). The cycling parameters were 95°C for 30s, 52°C for 30s, and 68°C for 30s for a total of 23 cycles. The primers for CYP3A23 amplification were 5′-ACAGCCCTATCCGATATGGA-3′ (forward) and 5′-CTGCCCTTGTTCTCCTTGC-3′ (reverse), and the primers for β-actin amplification were 5′-GATCTTGATCTTCATGGTGCTAGG-3′ (forward) and 5′TTGTAACCAACTGGGACGATATGG-3′ (reverse). The PCR-amplified products were resolved by agarose gel electrophoresis and detected by Typhoon 9410 (Amersham Biosciences Inc., Piscataway, NJ).
Activation of the Pregnane X Receptor (PXR). CV-1 cells were plated in 24-well plates in Dulbecco's modified Eagle's medium supplemented with 10% delipidated fetal bovine serum at a density of 9 × 104 cells per well and cultured overnight. Transfection was conducted by lipofection with LipofectAMINE and Plus Reagent as described previously (Song et al., 2004). Transfection construct mixtures contained 200 ng of rat or human PXR, 150 ng of reporter, and 10 ng of the null-Renilla luciferase plasmid (Li et al., 2004). The expression constructs encoding full-length rat or human PXR were obtained by cDNA-trapping hybridization (Zhang et al., 1999; Song et al., 2004). The reporter CYP3A23-DR3-luc was prepared by inserting four copies of DR3 (5′-GATCAGACAGTTCATGAAGTTCATCTAGATC-3′) into the pGL3 promoter luciferase vector. Cells were transfected for 3 h, and the medium was replaced with fresh medium. After 24 h, the medium was changed with the same medium containing a kavalactone, PCN, or rifampicin. The transfected cells were incubated for an additional 24 h and washed once with PBS and collected by scraping. The collected cells were then subjected to a freeze/thaw cycle. The reporter enzyme activities were assayed with the Dual-Luciferase Reporter Assay System as described previously (Song et al., 2004). This system contained two substrates that were used to determine the activity of two luciferases sequentially. The firefly luciferase activity, which represented the reporter activity, was initiated by mixing aliquot of lysates (10 μl) with Luciferase Assay Reagent II. Then the firefly luminescence was quenched, and the Renilla luminescence was simultaneously activated by adding Stop and Glo reagent to the sample tube. The firefly luminescence was normalized based on the Renilla luminescence signal, and the ratio of treatment over control served as -fold activation (Song et al., 2004).
Other Analyses. Purification and structural identification of kavalactones were conducted as described previously (Dharmaratne et al., 2002). Protein concentrations were determined with a bicinchoninic acid assay (Pierce) with albumin as standard. Data are presented as mean ± S.D. of at least three separate experiments except where the results of blots are shown, in which case a representative experiment is depicted in the figures. Comparisons between two values were made with Student's test at p < 0.05.
Results
Desmethoxyyangonin and Dihydromethysticin Markedly Induce the Expression of CYP3A23. Kavalactones are recognized as the active constituents responsible for various pharmacological activities in kava extracts (Singh and Singh, 2002). Six kavalactones constitute 80% of total kavalactones and contribute as much as 95% to the overall pharmacological activities (Bilia et al., 2002; Dharmaratne et al., 2002; Singh and Singh, 2002). These kavalactones are desmethoxyyangonin, dihydrokawain, dihydromethysticin, kawain, methysticin, and yangonin. To determine whether the pharmacologically active kavalactones are also responsible for the induction of CYP3A enzymes, rat hepatocytes were treated with purified kavalactones for 48 h, and the levels of CYP3A23 were determined. At a low concentration (2 μM), none of the kavalactones caused changes on the levels of CYP3A23 (Fig. 1A). In contrast, with the exception of kawain, a moderate or marked induction was detected when kavalactones were used at 50 μM. Treatment with desmethoxyyangonin and dihydromethysticin increased the expression of CYP3A23 by as much as 7-fold, whereas yangonin, methysticin, and dihydrokawain were less effective (∼2-fold). All data were collected from three independent experiments (three rats).
The magnitude of the induction was studied as a function of the concentrations (0–100 μM). Three kavalactones were used: kawain, desmethoxyyangonin, and dihydromethysticin. The selection of these kavalactones was based on their relative potency on the induction. Kawain caused no induction activity, whereas the other two elicited the highest induction (Fig. 1A). As shown in Fig. 1B, both desmethoxyyangonin and dihydromethysticin induced CYP3A23 in a concentration-dependent manner, and the overall induction was comparable between these two kavalactones, particularly when higher concentrations (≥ 50 μM) were assayed. The maximum induction was ∼8-fold and occurred at either 50 μM (dihydromethysticin) or 100 μM (desmethoxyyangonin). In contrast, no induction was detected with kawain, even at 100 μM (Fig. 1B). Next we performed a time course study for the induction. Hepatocytes were treated with desmethoxyyangonin, dihydromethysticin, or PCN for 24 to 72 h. PCN was used because this steroid has been shown to efficaciously induce CYP3A23 (Wrighton et al., 1985). As shown in Fig. 1C, a marked induction was detected as early as 24 h with all three chemicals; however, the overall magnitude of the induction varied markedly. In the 24-h cultures, PCN and dihydromethysticin comparably induced CYP3A23 (slightly higher than 4-fold), whereas desmethoxyyangonin was less effective (∼3-fold). Similar induction patterns were detected in the 48-h cultures among these chemicals. The induction magnitudes, however, were markedly higher than those with 24-h cultures (Fig. 1C). For example, desmethoxyyangonin caused a 7-fold induction in the 48 h-cultures but only a 3-fold induction in the 24-h cultures. Further prolonged treatment (72 h) exhibited differential effects depending on the chemical. Higher induction was detected with PCN (10- versus 8-fold), lower induction was detected with dihy-dromethysticin (7- versus 8-fold), and no changes were detected with desmethoxyyangonin (7-fold) compared with those in the 24-h cultures.
Combined Kavalactones Induce CYP3A23 in an Additive/Synergistic Manner. The study with individual kavalactones clearly demonstrated that these chemicals, although structurally similar, differ markedly in the induction of CYP3A23. It has been shown that various kavalactones have additive/synergistic effects when given together (Keledjian et al., 1988; Singh and Singh, 2002). We next examined whether other kavalactones modulate the inductive activity of dihydromethysticin and desmethoxyyangonin. Three types of experiments were performed: 1) all six kavalactones were mixed at an equal molar concentration (15 μM); 2) mixtures were made to exclude dihydromethysticin, desmethoxyyangonin, or both; and 3) dihydromethysticin and desmethoxyyangonin were assayed at a concentration equal to the concentration (15 μM) of individual kavalactones in the mixture or a concentration equal to that of total kavalactones (e.g., 75 or 90 μM). When dihydromethysticin and desmethoxyyangonin were mixed together, they were used with an equal amount but at half of the concentration used individually.
The results from the interaction study among kavalactones are summarized in Fig. 2. Treatment with mixed kavalactones markedly increased the expression of CYP3A23 (lane 1, DMSO versus lane 2, mixture). The increase was reduced or abolished depending on the chemical(s) excluded from the mixtures. The exclusion of DMY reduced the induction from 7- to 4-fold (Fig. 2A), whereas a slightly smaller reduction was observed when DMS was excluded (from 7- to 5-fold) (Fig. 2B). More importantly, no induction was detected when both dihydromethysticin and desmethoxyyangonin were excluded (Fig. 2C), suggesting that these two kavalactones are the major contributors to the inductive activity in the total mixture.
We next examined the inductive activity of dihydromethysticin or desmethoxyyangonin at a concentration equal to that of each or total kavalactones in the mixtures (15, 75, and 90 μM). At 15 μM, neither dihydromethysticin nor desmethoxyyangonin induced CYP3A23 (Fig. 2, A and B, lane 4). At 75 or 90 μM, a marked induction (∼7-fold) was detected with both kavalactones (Fig. 2, A and B, lanes 5 and 6). Interestingly, the magnitude of the induction was similar to that elicited by the total mixture (six kavalactones) (Fig. 2, A and B, lane 2). The total mixture had a concentration of 15 μM for each kavalactone and a total kavalactone concentration of 90 μM (Fig. 2, A and B, lane 2). Therefore, the relative induction potency, based on the concentration of total kavalactones, was very similar, regardless of whether dihydromethysticin or desmethoxyyangonin was used alone or together with other kavalactones. In contrast, combined dihydromethysticin and desmethoxyyangonin, at a total concentration of 75 or 90 μM (half from each), caused a slightly higher induction (8–9-fold) than that by the total mixture, although no induction was detected when they were used at a combined concentration of 15 μM (7.5 μM each) (Fig. 2C).
Transcriptional Involvement in CYP3A23 Induction by Dihydromethysticin and Desmethoxyyangonin. In most cases, the induction of P450 enzymes is due to transcriptional activation. We next examined this possibility by performing RT-PCR and transcriptional inhibition assays. Rat hepatocytes were treated with dihydromethysticin and desmethoxyyangonin at various concentrations (0–100 μM), and total RNA was prepared. Kawain and PCN were included as controls in this study. Treatment with dihydromethysticin or desmethoxyyangonin resulted in a marked increase on the levels of CYP3A23 mRNA, and the increase correlated well with the concentrations of either kavalactone (Fig. 3A). It seemed that dihydromethysticin was slightly more potent than desmethoxyyangonin, particularly at lower concentrations (10–50 μM). As expected, no increase on the levels of CYP3A23 mRNA was detected in kawain-treated hepatocytes, whereas a marked increase was detected with PCN at all concentration points (Fig. 3A). Apparently, PCN was more potent than either dihydromethysticin or desmethoxyyangonin, based on the fact that only PCN but not the kavalactones significantly increased CYP3A23 mRNA at 2 μM, the lowest concentration used in this study (Fig. 3A). Comparable levels of β-actin mRNA were detected in all samples (Fig. 3A).
The study with RT-PCR demonstrated that the induction of CYP3A23 by dihydromethysticin or desmethoxyyangonin was achieved by increasing CYP3A23 mRNA. To determine whether the increased mRNA levels were responsible for the induction, we performed an RNA synthesis inhibition experiment with DRB. Rat hepatocytes were treated with dihydromethysticin, desmethoxyyangonin, or PCN alone or together with DRB (Clement and Wilkinson, 2000). Cell lysates were prepared and analyzed for the expression of CYP3A23 by Western blots. As expected, the levels of CYP3A23 were markedly elevated in the hepatocytes treated with dihydromethysticin, desmethoxyyangonin, or PCN (Fig. 3B); however, the increase was abolished almost completely in the hepatocytes cotreated with DRB. It should be noted that similar experiments were performed with actinomycin D, another inhibitor of RNA synthesis; however, this inhibitor seemed to stabilize CYP3A23 protein (data not shown).
Desmethoxyyangonin and Dihydromethysticin Moderately Activate PXR. The RT-PCR analysis and DRB inhibition assay suggested that the induction of CYP3A23 by dihydromethysticin or desmethoxyyangonin is achieved by increasing transcription. Several nuclear receptors, notably PXR, have been shown to interact with a vast array of structurally dissimilar chemicals and transactivate genes encoding drug-metabolizing enzymes or transporters (Lehmann et al., 1998; Francis et al., 2003). We next tested whether kavalactones activate rat or human PXR. CV-1 cells were transfected with rat or human PXR along with a luciferase reporter (CYP3A23-DR3-luc), and the luciferase activities were determined and normalized based on the Renilla luciferase activity (internal control). All kavalactones were assayed at 50 μM, whereas PCN and rifampicin were assayed at 10 μM. The concentrations were selected based on the observation that 10 μM PCN caused an induction comparable to 50 μM desmethoxyyangonin or dihydromethysticin (Fig. 1C). As expected, PCN efficaciously activated rat PXR (∼25-fold), and rifampicin activated human PXR (∼5-fold). In contrast, only weak or moderate activation (∼2-fold) on either rat or human PXR was detected with desmethoxyyangonin and dihydromethysticin (Fig. 4). The rest of the kavalactones showed no activation activity toward either PXR.
Discussion
Kava is a medicinal herb that is used as an alternative to treat many types of subclinical disorders, particularly mild and moderate anxiety (Singh and Singh, 2002). Like many other herbal products, kava has gained popularity worldwide in the past decade. Six kavalactones are recognized as the major active constituents responsible for various pharmacological activities in kava extracts (Singh and Singh, 2002). In this study, we have tested the ability of purified kavalactones to induce CYP3A23 and activate rat and human PXR. Among six kavalactones tested, only desmethoxyyangonin and dihydromethysticin efficaciously induce the expression of CYP3A23. The combination of six kavalactones causes a similar magnitude of induction at a noninductive concentration when individually assayed. The induced expression, however, is markedly reduced or completely abolished if dihydromethysticin, desmethoxyyangonin, or both are selectively excluded from the mixtures. These findings establish that dihydromethysticin and desmethoxyyangonin are two pharmacological kavalactones that contribute significantly to the overall induction of kavalactones on the expression of CYP3A23.
The greater inductive activity of dihydromethysticin and desmethoxyyangonin compared with other kavalactones points to several important structural features required for CYP3A23 induction. As shown in Fig. 5, all six kavalactones share a basic structure consisting of a benzene ring and a methoxy lactone connected by an ethyl or ethylene bridge (Dharmaratne et al., 2002). The type of connectors seems to be important for CYP3A23 induction. For example, dihydromethysticin and methysticin are structurally identical except for the connectors. Dihydromethysticin has the ethyl bridge, whereas methysticin has the ethylene connector, and yet only dihydromethysticin markedly induces the expression of CYP3A23 (Fig. 1A). The substitution on the benzene ring also contributes to the inductive activity. Yangonin and desmethoxyyangonin differ only in regard to the substitution of the benzene ring: yangonin but not desmethoxyyangonin has a methoxy substitute. However, only desmethoxyyangonin markedly up-regulates the expression of CYP3A23, suggesting that the methoxy substitution on the benzene ring negatively affects the inductive activity. The importance of a substitute on the benzene ring is further supported by the structure-related activity between dihydrokawain and dihydromethysticin. These two kavalactones have the same structure, except that dihydromethysticin has a 3′,4′-methylene dioxy substitute on the benzene ring (Fig. 5), and yet dihydromethysticin effectively induces CYP3A23, whereas dihydrokawain is much less effective (8- versus 2-fold, Fig. 1A). It seems that the 3′,4′-methylene dioxy substitute increases the inductive activity, whereas the methoxy substitute has an opposing effect.
The profound effect on the induction by small structural differences suggests that dihydromethysticin and desmethoxyyangonin induce CYP3A23 through a receptor-based mechanism(s). In this study, the RNA synthesis inhibitor DBR completely abolishes the induction by dihydromethysticin, desmethoxyyangonin, and PCN (Fig. 3B). Therefore, it is likely that these kavalactones, like PCN, induce CYP3A23 through PXR-mediated transcription activation (Francis et al., 2003). However, cotransfection experiments have demonstrated that dihydromethysticin and desmethoxyyangonin have only ∼10% of the activity of PCN in terms of activating rat PXR at the concentrations that elicit comparable induction of CYP3A23 (Figs. 1C and 4), suggesting that PXR plays an insignificant role in CYP3A23 induction in response to dihydromethysticin and desmethoxyyangonin. Alternatively, these kavalactones undergo enzymatic conversion, and their metabolites are PXR activators; however, the metabolites are effectively formed only in hepatocytes but not CV-1 cells. A third possibility is that dihydromethysticin and desmethoxyyangonin increase the levels of an endogenous PXR activator(s) in hepatocytes, which ultimately induces the expression of CYP3A23. Another receptor that likely supports the induction is the constitutive androstane receptor (CAR). Based on molecular modeling, CAR, like PXR, has a large and expandable ligand binding pocket (Jacobs et al., 2003). However, CAR shows narrower ligand specificity and is likely more discriminative than PXR in terms of interacting with structurally similar molecules like various kavalactones. Finally, several other nuclear receptors (e.g., vitamin D receptor) are reportedly involved in the induction of P450 enzymes and transporters (Ekins et al., 2002; Kast et al., 2002; Jacobs et al., 2003), although it remains to be determined whether these receptors interact with dihydromethysticin and desmethoxyyangonin to any significant extents.
It is interesting to note that the concentration of total kavalactones rather than dihydromethysticin or desmethoxyyangonin plays a more determinant role in the magnitude of the induction. Six kavalactones, mixed at a noninductive concentration (15 μM for each), cause induction similar to 90 μM dihydromethysticin or desmethoxyyangonin (Fig. 2, A and B). Selective exclusion of both dihydromethysticin and desmethoxyyangonin, however, completely abolish the inductive activity of the mixtures (Fig. 2C). These findings provide direct evidence that dihydromethysticin and desmethoxyyangonin are essential to the inductive activity of the mixture, and their inductive effect is additively/synergistically enhanced by other kavalactones. The precise mechanism(s) on the interkavalactone interactions in terms of increasing CYP3A23 expression remains to be determined. It is likely that various kavalactones share same molecular targets and that a greater presence of other kavalactones replaces dihydromethysticin and desmethoxyyangonin for binding these targets, thus relatively increasing the availability of these two kavalactones toward CYP3A23 induction. In support of this possibility, many kavalactones share similar pharmacological activities, such as inhibition of TNF-α release (Hashimoto et al., 2003).
Competitive metabolism provides an alternative mechanism that is likely responsible for the enhanced induction of other kavalactones on dihydromethysticin and desmethoxyyangonin. Although the overall metabolism of kavalactones remains to be determined, they have been shown to profoundly alter the catalytic activities of P450 enzymes (Mathews et al., 2002; Unger et al., 2002). Total kavalactones have been found to competitively inhibit the activities of almost all P450 enzymes tested, with the exception of CYP2C8 (Mathews et al., 2002). More importantly, kavalactones with similar structures exhibit similar inhibition profiles on P450 activity. For example, dihydromethysticin and methysticin differ only on the ethyl and ethylene connector (Fig. 5) and inhibit CYP2C9, 2C19, 2D6, and 3A4 to a relatively similar extent (Mathews et al., 2002). Another possible mechanism for the observed enhancement is an involvement of kavalactones in lipid remodeling, leading to modulations on receptor-ligand interactions. Remodeling of endogenous lipids by kavalactones has been suggested to be responsible for the enhanced activity on the γ-aminobutyric acid receptor, a major molecular target that mediates the anxiolytic activity of kavalactones (Davies et al., 1992; Singh and Singh, 2002).
There are two distinct mechanisms through which kavalactones likely cause pharmacokinetic interactions among herbs and drugs: competitive inhibition and induction of CYP3A (Raucy, 2003; this study). The metabolism of kavalactones and induction of CYP3A likely increase the clearance of kava, thus decreasing kava-related toxicity (assuming that the parent compounds are the sources for toxicity). As a matter of fact, habitual kava drinkers in the South Pacific islands usually take much larger amounts compared with kava users in the West but show similar incidences of liver toxicity to the islanders who do not take kava, suggesting that kava extracts are well tolerated (Singh and Singh, 2002); however, marked induction of CYP3A likely increases toxicity of chemicals that undergo bioactivation. It has been increasingly reported in recent years that the use of kava products is associated with severe hepatotoxicity among Westerners (Teschke et al., 2003). Interestingly, many reported cases are linked to the use of medications with potential liver toxicity that had been administered concurrently (Singh and Singh, 2002), suggesting that induced expression of CYP3A indeed has toxicological significance.
In summary, we report that pharmacologically active kavalactones differ markedly in terms of their ability to induce CYP3A23, with dihydromethysticin and desmethoxyyangonin being the most potent. Their inductive activity, however, is additively/synergistically enhanced by other kavalactones that have no or less significant inductive activity. These findings provide molecular bases for preparing therapeutically active but induction-reduced kava products, thus minimizing herb-herb and herb-drug interactions resulting from the induction of drug-metabolizing enzymes.
Footnotes
-
This work was supported by National Institutes of Health Grants R01GM61988, R01ES07965, and P20 RR16457.
-
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
-
doi:10.1124/dmd.104.000786.
-
ABBREVIATIONS: DMY, desmethoxyyangonin; DMS, dihydromethysticin; TNF, tumor necrosis factor; PCN, pregnenolone 16α-carbonitrile; HBSS, Hanks' balanced salt solution; WEM, Williams' E medium; ITS, insulin-transferrin-selenium; DMSO, dimethyl sulfoxide; RT-PCR, reverse transcription-coupled polymerase chain reaction; PCR, polymerase chain reaction; PXR, pregnane X receptor; DRB, 5,6-dichlororibosidylbenzimidazole; CAR, constitutive androstane receptor.
-
↵1 These authors contributed equally to this manuscript.
- Received May 28, 2004.
- Accepted July 26, 2004.
- The American Society for Pharmacology and Experimental Therapeutics