The role of chrysin and the Ah receptor in induction of the human UGT1A1 gene in vitro and in transgenic UGT1 mice

Authors

  • Jessica A. Bonzo,

    1. Laboratory of Environmental Toxicology, Departments of Pharmacology, Chemistry & Biochemistry, University of California, San Diego, La Jolla, CA
    2. Biomedical Sciences Graduate Program at UCSD, CA
    Search for more papers by this author
  • Alain Bélanger,

    1. Molecular Endocrinology and Oncology Research Center, CHUL Research Center and the Faculty of Medicine, Laval University, Quebec, Canada
    Search for more papers by this author
  • Robert H. Tukey

    Corresponding author
    1. Laboratory of Environmental Toxicology, Departments of Pharmacology, Chemistry & Biochemistry, University of California, San Diego, La Jolla, CA
    • Departments of Pharmacology, Chemistry, and Biochemistry, Leichtag Biomedical Research Building, Room 211, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA, 92093-0722
    Search for more papers by this author
    • fax: 858-822-0363


  • Potential conflict of interest: Nothing to report.

Abstract

The flavonoid chrysin is an important dietary substance and induces UGT1A1 protein expression in cell culture. As a representative of the class of dietary flavonoids, clinical investigations have been considered as a means of inducing hepatic UGT1A1 expression. We demonstrate the necessity of a xenobiotic response element (XRE) in support of chrysin induction of UGT1A1 in the human hepatoma cell line HepG2. Receptor binding assays confirm that chrysin is a ligand for the Ah receptor by competition with [3H]2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). However, key differences in Ah receptor recognition and activation of UGT1A1 by chrysin exist when compared with classical mechanisms of UGT1A1 induction by TCDD. Ah receptor degradation, an indicator of Ah receptor activation, does not occur after chrysin treatment, and chrysin cannot transactivate the Ah receptor in a TCDD-dependent fashion. Knock-down of the Ah receptor by siRNA indicates that chrysin uses the Ah receptor in conjunction with other factors through MAP kinase signaling pathways to maximally induce UGT1A1. Most importantly, oral treatment of chrysin to transgenic mice that express the human UGT1 locus is unable to induce UGT1A1 expression in either the small intestine or liver. Conclusion: Although the implications for chrysin as an atypical agonist of the Ah receptor are intriguing at the molecular level, the relevance of chrysin-induced transcription for the purpose of clinical therapies or to regulate phase 2–dependent glucuronidation is questionable given the lack of in vivo regulation of human UGT1A1 by chrysin in a transgenic animal model. (HEPATOLOGY 2007;45:349–360.)

Humans are exposed to a wide variety of natural polyphenolic plant compounds called flavonoids through food as well as the growing dietary supplement industry. The flavonoid family of compounds has gained popularity because they are believed to elicit a wide range of biochemical and pharmacological properties, with one of the most investigated being cancer preventive properties.1 Along with their antioxidant properties, the beneficial effects of flavonoids have been attributed to their ability to induce phase 2 xenobiotic metabolism,2 which is central to the removal and detoxification of drugs, xenobiotics, and carcinogens. Considered one of the most important of the phase 2 detoxifying gene families, the UDP-glucuronosyltransferases (UGTs) catalyze the formation of endobiotic and xenobiotic β-D-glucopyranosiduronic acids (glucuronides) creating polar, hydrophilic compounds that dissolve in the aqueous urine and bile for excretion.3 Because most of the dietary flavonoids are consumed through diet, their biological properties as inducers of phase 2 metabolism require adequate oral bioavailability necessary to activate the signaling and receptor pathways that lead to induction of drug metabolism. Although the use of flavonoids such as chrysin (5,7-dihydroxyflavone) has been advocated in the fight against cancer, they have also been discussed as a tool to induce UGT1A1 activity,4 with the intention that increases in UGT1A1 activity would improve the detoxification of selective toxic cancer chemotherapeutic agents.

The UGT1A family of enzymes is encoded by the UGT1 locus and along with the UGT2 proteins are responsible for a significant portion of all phase 2 drug metabolism.5 Of the nine functional UGT1A enzymes, UGT1A1 is unique in that it is the sole glucuronidation reaction leading to bilirubin elimination.3, 6, 7 This is significant because Gilbert's syndrome results from clinical diagnoses of mild non-hemolytic hyperbilirubinemia (reviewed in Tukey and Strassburg8), a finding that has drawn awareness to the importance of UGT1A1 in clinical drug metabolism. The most common genotype leading to Gilbert's syndrome is the inheritance of the UGT1A1 promoter sequence containing [A(TA)7TAA] (UGT1A1*28),9 which has been speculated to lead to an estimated 70% reduction in transcriptional activity compared with wild-type UGT1A1 expression, which is represented by a normal [A(TA)6TAA] sequence. One of the most severe of the clinical implications associated with reduced expression of UGT1A1 has been the cause of gastrointestinal tract toxicities associated with therapeutic doses of the prodrug irinotecan,10 a camptothecin derivative that is bioactivated by serum carboxylesterases to the active SN-38 topoisomerase inhibitor. SN-38 is removed from the circulation by glucuronidation, and patients that are either heterozygous or homozygous for the UGT1A1*28 allele are predisposed to SN-38–initiated GI toxicities.11 In this clinical setting, it has been postulated that a diet rich in flavonoids may counter the toxicity by promoting transcriptional activation of UGT1A1 in intestinal tissue, therefore diminishing the toxic actions of SN-38.4 Thus, it is important to understand the cellular and molecular steps that underlie induction and expression of UGT1A1 by flavonoids both in vivo and in vitro.

Induction of UGT1A1 catalytic activity by chrysin has been reported in primary human hepatocytes,12 the human hepatoma cell line HepG2,12–15 and the human intestinal cell line Caco-2.16 As well as inducing UGT1A1, chrysin weakly induces CYP1A1, which is almost exclusively controlled through activation of the Ah receptor.17–19 Additionally, a precedence for flavonoid interaction with the Ah receptor has been established with the agonist β-naphthoflavone as well as Ah receptor antagonists epigallocatechin gallate,20 3′-methoxy-4′-nitroflavone,21 and the indole derivative 3,3′-diindolylmethane.22 Chrysin has been shown to induce luciferase reporter gene activity regulated by xenobiotic response elements (XREs),17, 18 and one report demonstrated the necessity for a functional XRE in the UGT1A1 promoter.13 We report here that chrysin is able to induce UGT1A1 gene expression through an XRE, yet activation of the Ah receptor by chrysin is distinct from activation by the classical Ah receptor ligand 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Partial regulation of UGT1A1 gene expression by chrysin provides evidence that the Ah receptor serves to participate in a protective role by modulating phase 2 drug metabolism. However, when the application of chrysin induction of UGT1A1 was tested in a transgenic mouse model (Tg-UGT1) that expresses the UGT1 locus,23 oral chrysin administration was rapidly metabolized by glucuronidation, thus preventing adequate plasma levels needed for induction of the UGT1 locus in vivo. This finding questions the usefulness of chrysin and other flavonoids in therapies in the prevention of drug-induced toxicities or other uses targeted for inducing phase 2 drug metabolism.

Abbreviations

Ah, arylhydrocarbon; B[a]P, benzo[a]pyrene; DMSO, dimethylsulfoxide; DTT, dithiothreitol; EDTA, ethylenediaminetetra-acetic acid; GRDBD, glucocorticoid receptor DNA-binding domain; GRE, glucocorticoid-binding enhancer sequence; HEDG, 25 mM HEPES pH 7.4, 1.5 mM EDTA, 10% glycerol, 1 mM DTT; si-RNA AhR, siRNA directed against the human Ah receptor; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin; TCDF, 2,3,7,8-tetrachlorodibenzo-furan; UGT, UDP-glucuronosyltransferases; WT, wild type; XRE, xenobiotic response element.

Materials and Methods

Chemicals and Reagents.

TCDD was obtained from Wellington Laboratories (Guelph, ON, Canada). Chrysin (5,7-dihydroxyflavone), dimethylsulfoxide (DMSO), benzo[a]pyrene (B[a]P), and corn oil were all purchased from Sigma-Aldrich (St. Louis, MO). The [3H]TCDD (specific activity, 27.5 Ci/mmol) was purchased from EaglePicher Pharmaceutical (Lenexa, KS), and 2,3,7,8-tetrachlorodibenzo-furan (TCDF) was from Cambridge Isotope Laboratories (Andover, MA). UO126 (1,2-diamino-2,3-dicyano-1,4-bis[2-aminophenylthio] butadiene) and PD98059 (2′-amino-3′-methoxyflavone) were purchased from Cell Signaling (Beverly, MA). SP600125 (1,9-pyrazoloanthrone) and SB203580 (4-(4-fluorophenyl)-2-(4-methylsulfinylphenyl)-5-(4-pyridyl)1H-imidazole) were from Calbiochem (San Diego, CA). All chemicals were dissolved in DMSO, and final DMSO concentration in cell culture did not exceed 0.1%. The transfection reagent GenePorter2 was from Genlantis (San Diego, CA), and Lipofectamine 2000 was from Invitrogen (Carlsbad, CA). The anti–β-actin antibody was purchased from Sigma-Aldrich. The p-P44/42 and P44/42 antibodies were purchased from Cell Signaling. The rabbit anti-mouse Ah receptor and rabbit anti-mouse Arnt antibodies were gifts of Dr. Christopher Bradfield (University of Wisconsin, Madison, WI); the mouse anti-human UGT1A1 antibody was a gift of Dr. Joseph K. Ritter (Virginia Commonwealth University, Medical College of Virginia, Richmond, VA). The horseradish peroxidase–conjugated secondary antibodies were purchased from Cell Signaling.

Cell Culture.

The human hepatoma cell line HepG2 (American Type Culture Collection, Manassas, VA) and mouse hepa1c1c7 cells (gift of Dr. James Whitlock, Stanford University, Stanford, CA) were cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum. MH1A1L cells were derived from HepG2 cells that stably express a human UGT1A1-promoter luciferase reporter gene.24 TV101L cells were derived from HepG2 cells that stably express a human CYP1A1-luciferase reporter gene.25 MH1A1L and TV101L cells were cultured as described except G418 (Geneticin) was added to 0.8 mg/mL. Cells were incubated in a humidified incubator under 5% CO2 at 37°C.

Animals and Treatments.

A mouse line heterozygous for the human UGT1 locus was generated as previously described.23 The Tg-UGT1 strain showed strong induction of UGT1A1 by TCDD in small intestine, large intestine, and liver. Animals were housed 2 to 3 in plastic cages with hardwood chips for bedding in a 12-hour light, 12-hour dark cycle with water and food (#7912, Harlan-Teklad, Indianapolis, IN) ad libitum. For plasma chrysin analysis, nine wild-type (WT) and eight Tg-UGT1 mice were given a single oral dose of 100 μL (50 mg/kg) chrysin in vehicle (60% corn oil, 40% DMSO). Blood was collected 60 minutes after the dose and plasma separated. For analysis of UGT1A1 induction, 3 animals per group were orally gavaged with 100 μL vehicle for 7 days, 100 μL 50 mg/kg chrysin for 7 days, or 100 μL 100 mg/kg benzo[a]pyrene for 3 days. Mice were anesthetized by isoflurane inhalation, and the liver was perfused with ice-cold 1.15% KCl. Small intestine, large intestine, and stomach were dissected lengthwise and rinsed in cold 1.15% KCl. Organs were immediately frozen on dry ice after collection and stored at −80°C. All animal experiments were carried out following University of California San Diego Institutional Animal Care and Use guidelines.

Transient Transfection and Luciferase Assays.

Luciferase constructs containing sections of the human UGT1A1 promoter were previously described.24 Briefly, segments were cloned from a BAC library of the UGT1 locus with sequences published under GenBank accession number AF297093. PCR products were cloned into the pGL3-basic or pGL3-promoter vectors (Promega, Madison, WI). Plasmids pCMV/GRBD/mAhR83-805 and p(GRE)2T105Luc were kindly provided by Dr. Lawrence Poellinger (Karolinska Institute, Stockholm, Sweden). The plasmid pCMV/GRDBD/mAhR 83-805 contains an N-terminal zinc finger DNA-binding domain of the glucocorticoid receptor linked to the C-terminal amino acids 83-805 of the mouse Ah receptor. This construct can be activated by ligand as measured by glucocorticoid-binding enhancer sequence (GRE)-driven luciferase activity when co-transfected with the p(GRE)2T105Luc plasmid.26 Plasmids pMEV-MEK1-WT (P1030a) and pMEV-MEK1-DN (P1030b) were purchased from Biomyx (San Diego, CA).

Luciferase assays in the stably transfected cells MH1A1L and TV101L were carried out as described with alterations.27 Cells were treated for 48 hours as indicated and lysed on the plates in lysis buffer [1% Triton X-100, 25 mM Tricine, pH 7.8, 15 mM MgSO4, 4 mM ethylenediaminetetraacetic acid (EDTA), 1 mM dithiothreitol (DTT)]. Cell lysates were collected by centrifugation at 10,000g. Supernatant (20 μL) was mixed with 100 μL reaction buffer (25 mM Tricine, 15 mM MgSO4, 4 mM EDTA, 15 mM KPO4, pH 7.8, 1 mM DTT, and 2 mM ATP). Reaction was started by the addition of 50 μL luciferin (0.3 mg/mL), and light output measured for 2 seconds using a LMax II384 luminometer (Molecular Devices, Sunnyvale, CA). Results were normalized by Bradford protein assay and expressed as fold induction of vehicle control. All experiments were done in triplicate.

Dual luciferase assays were conducted following the manufacturer's instructions (Promega). Briefly, HepG2 cells were transfected in 12-well plates for 24 hours with 500 ng UGT1A1-luc and 20 ng phRL-SV40 following manufacturer instructions. Cells were treated for 48 hours, lysed, and 20 μL lysate used for dual-luciferase analysis using the LMax II384 luminometer. Firefly luciferase values were normalized to Renilla luciferase and protein concentration and reported as fold increase over control treatment. All experiments were conducted in triplicate.

Protein Preparations.

Whole cell, cytosolic, and nuclear protein extracts were prepared from HepG2 and Hepa1c1c7 cells as described.27 Microsomal preparations were isolated from mouse tissue as described.23 All protein concentrations were determined using the Bradford protein assay.

Electrophoretic Mobility Shift Assay.

As described,24, 27 nuclear extracts from HepG2 cells treated for 48 hours were incubated on ice for 15 minutes with 2.2 μg poly(dI-dC) and 1 μg salmon sperm DNA in HEDG buffer (25 mM HEPES pH 7.4, 1.5 mM EDTA, 10% glycerol, 1 mM DTT). A 32P-labeled XRE oligonucleotide probe (1 × 106 cpm) was added and the reaction incubated at room temperature for 15 minutes. For competition experiments, 200-fold excess unlabeled UGT-XRE probe or UGT-mutated XRE (CACGCA mutated to ACCGCA) was added to the binding reaction before addition of labeled UGT-XRE probe. For antibody competitions, 3 μL of either the anti-Ah receptor or anti-Arnt antibody was added to the binding reaction. Loading dye was added and the proteins resolved on a 6% non-denaturing polyacrylamide gel. Radioactively bound proteins were visualized by exposure to a phosphoimager plate and scanning with a Molecular Dynamics Storm 840 scanner (Amersham Biosciences, Piscataway, NJ). Relative band intensities were determined using ImageQuant 5.2 (Molecular Dynamics).

Western Blot Analysis.

All Western blots were performed using NuPAGE gel electrophoresis units as outlined by the manufacturer (Invitrogen). Protein aliquots were heated at 70°C in loading buffer and resolved on 4% to 12% Bis-Tris gels under reducing conditions, and protein was transferred to a nitrocellulose membrane using a semi-dry transfer system (Novex, Invitrogen). The membrane was blocked with 5% nonfat dry milk in Tris-buffered saline for 1 hour at room temperature, followed by incubation with primary antibodies in Tris-buffered saline overnight. Membranes were then washed and incubated for 1 hour with horseradish peroxidase–conjugated secondary antibody at room temperature. The conjugated horseradish peroxidase was detected using ECL Plus Western blotting detection system (Amersham), and blots were exposed to X-ray film.

Ligand Binding Assay.

Analysis of ligand binding to the Ah receptor was carried out by examining the ability of chrysin to interfere with [3H]TCDD binding to the cytosolic receptor.28, 29 In this assay, 500 μL hepa1c1c7 cytosol (5 mg/mL) was incubated with 2 nM [3H]TCDD and indicated concentrations of chrysin. To monitor competitive binding, a control binding assay was conducted that included 1 μM TCDF. Cytosols and chemicals were incubated at 4°C for 2 hours. To remove free [3H]TCDD, dextran-coated charcoal (5 mg/5 mg protein) was added to each binding reaction, the mixture was incubated for 10 minutes on ice, and the dextran-charcoal was removed by centrifugation. Cytosol (300 μL) was layered onto linear 10% to 30% sucrose density gradients prepared in HEDG buffer (25 mM HEPES, pH 7.5, 1 mM EDTA, 1 mM DTT, 10% glycerol). Gradients were centrifuged at 4°C for 16 hours at 235,000g in a Beckman Coulter SW60 Ti rotor. The gradients were collected by puncturing the bottom of the tube and collecting 150 μL per fraction, and the radioactivity in each fraction was determined by liquid scintillation counting.

siRNA Knockdown of the Ah Receptor.

A SMARTpool siRNA directed against the human Ah receptor (si-AhR, cat #M-004990) was purchased from Dharmacon RNA Technologies (Boulder, CO). To measure knockdown of Ah receptor protein, HepG2 cells were plated in 6-well plates and transfected for 48 hours with 50 nM negative control siRNA (Dharmacon, cat #D-001210-01) or 50 nM si-AhR following the Lipofectamine (Invitrogen) protocol for siRNA transfection. Whole-cell extracts were collected as described and 20 μg protein used for Western blot. Ah receptor band intensities were quantified using a GS-800 calibrated densitometer (Bio-Rad) and normalized to actin loading. Results shown are representative of 6 independent transfections.

For luciferase assays, the −1612CYP1A1-promoter or the UGT1A1 PR3.7 luciferase construct was co-transfected with 50 nM negative control siRNA or 50 nM si-AhR as well as phRL-SV40 for 36 hours following the Lipofectamine protocol. Cells were subsequently treated with DMSO, 1 nM TCDD, or 20 μM chrysin for 24 hours. Firefly luciferase values were normalized to Renilla luciferase and protein concentration and reported as fold increase over control treatment. All experiments were conducted in triplicate.

Quantification of Plasma Chrysin Levels.

Chrysin was purchased from Sigma. Chrysin glucuronide was obtained from enzymatic assays using liver microsomes. Briefly, media from enzymatic assays were diluted with 0.1% formic acid and loaded on Strata X cartridges (60 mg; Phenomenex, Torrance, CA) preconditioned with methanol followed by 0.1% formic acid. After loading the sample, the cartridge was washed with ultrapure water and ethyl acetate to remove the unreacted chrysin. The chrysin glucuronide was eluted with a mixture of acetonitrile:water (90:10) and evaporated under nitrogen, diluted in methanol, and the purity of the compound was confirmed by HPLC-MS. An aliquot was treated with β-glucuronidase and the residue was quantified with a calibration curve of chrysin. The concentration of chrysin obtained for the aliquot digested by β-glucuronidase was then converted in a concentration of chrysin glucuronide.

Plasma samples (10-50 μL) were diluted with 1 mL water and applied to Strata X 60-mg solid-phase extraction cartridges that had been preconditioned with methanol and water. The loaded cartridges were washed sequentially with water and 10% methanol. The washed cartridges were placed under full vacuum. The analytes were eluted with acetonitrile:water (90:10), and the solvent was evaporated under a stream of nitrogen (25°C). The residue was dissolved in mobile phase and then transferred into conical vial for injection into the mass spectrometer. Determinations of chrysin and chrysin glucuronide were performed using a standard curve containing chrysin (1-50 ng/mL) and chrysin glucuronide (1-50 ng/mL) extracted in the same conditions as sample.

The HPLC-45 system consisted of a mass spectrometer (model API 3000, Perkin Elmer Sciex, Thornhill, Canada). It was operated in multiple reactions monitoring mode and equipped with an electrospray ionization interface in positive ion mode and a HPLC pump plus autosampler Model 2690 (Waters, Milford, MA). Chromatographic separation was achieved with a C6 phenyl column 3 μm packing material, 100 × 4.6 mm (Phenomenex). Isocratic condition with 50% methanol:40% acetonitrile:10% water:0.1% acetic acid with a flow rate of 0.9 mL/min were used to elute chrysin and chrysin glucuronide.

Results

Chrysin Regulates the UGT1A1 Gene Through an XRE.

The flavonoid chrysin induces the protein expression of UGT1A1 in the HepG2 and Caco-2 cell lines.12–16 To confirm whether this increase in protein expression is attributable to regulation of the promoter, a HepG2 cell line that stably expresses the −3712UGT1A1 promoter driving luciferase (MH1A1L) was treated with increasing concentrations of chrysin. The UGT1A1 promoter responded in a dose-dependent manner (Fig. 1A). The UGT1A1 promoter was previously identified as being TCDD responsive,24 and the XRE core sequence (CACGCA) was located between −3309 and −3304 bp. Chrysin induced the 3.7 kb UGT1A1 luciferase construct 6-fold to 9-fold compared with 2-fold to 4-fold by TCDD treatment (Fig. 1B). Conversely, treatment of the HepG2 cell line–derived TV101L cells, which express a stable −1612CYP1A1 promoter driving luciferase,25 resulted in minimal but significant induction by chrysin (Fig. 1C). Activation of the human CYP1A1 and UGT1A1 promoter implicates activation of the Ah receptor by chrysin.

Figure 1.

Chrysin induction of UGT1A1 and CYP1A1. Induction of UGT1A1-luc and CYP1A1-luc were measured in MH1A1L24 and TV101L25 cells, respectively. (A) MH1A1L cells were treated with increasing concentrations of chrysin for 48 hours. (B) MH1A1L cells were treated with either DMSO, TCDD (10 nM), or chrysin (20 μM) for 48 hours. (C) TV101L cells were treated with either DMSO, TCDD (10 nM), or chrysin (20 μM) for 48 hours. In all experiments, luciferase readings were normalized to protein concentration and displayed as fold increase over DMSO treatment. All experiments were done in triplicate, and significant increases from DMSO treatment are indicated (*P ≤ 0.05; **P ≤ 0.005; ***P ≤ 0.0005).

To determine whether the Ah receptor is directly linked to induction of the UGT1A1 promoter by chrysin, further deletions in the UGT1A1 promoter24 were examined in transient transfection experiments (Fig. 2). Within a 380-bp enhancer fragment are located the binding sites for the pregnane X-receptor (DR3), the constitutive androstane receptor (CAR) DR4, the Ah receptor (XRE),24, 30 and a putative peroxisome proliferator-activated receptor alpha binding site (DR1). A region between −3377 and −3289 that contains the DR4 and XRE was shown to be essential for chrysin induction (Fig. 2A). Mutation of the XRE core sequence in the 380-bp enhancer region resulted in complete loss of chrysin induction (Fig. 2B). These data support previous findings13 that the XRE element originally identified in the enhancer region of the UGT1A1 gene24 is responsible for induction by chrysin.

Figure 2.

Identification of the chrysin-responsive region within the UGT1A1 promoter. HepG2 cells were transiently transfected with the UGT1A1 promoter constructs, PR-3712, PR-3377, PR-3289, and PR-2584 (A) or ER and ER-XRE mut (B) along with a Renilla luciferase plasmid. Cells were treated for 24 hours after transfection. Firefly luciferase values were normalized by Renilla luciferase and protein concentration. Results shown are fold increase over DMSO treatment for each construct. Significant increases over DMSO treatment are indicated (*P ≤ 0.05; **P ≤ 0.005; ***P ≤ 0.0005), and significant decreases in XRE mutation constructs from the wild type are indicated (‡P ≤ 0.005).

Effects of Chrysin on TCDD Binding to the Ah Receptor.

Because chrysin treatment leads to transcriptional activation of the UGT1A1 gene through a functional XRE, this finding would suggest that chrysin is a ligand for the Ah receptor. To examine this possibility, a series of radioligand binding assays were conducted to establish the ability of chrysin to compete with [3H]TCDD binding to the Ah receptor. Using cytosols from hepa1c1c7 cells,29 2 nM [3H]-TCDD was allowed to equilibrate in solution and the nonspecifically bound label removed by extraction with dextran charcoal. Specifically bound [3H]TCDD was identified as a single peak by sedimentation through a linear sucrose gradient (Fig. 3). The specificity of [3H]TCDD binding was confirmed by complete competition for binding when 1 μM TCDF, a competitive antagonist with higher Ah receptor binding affinity than TCDD, was included in the cytosol incubation. Chrysin inhibited [3H]TCDD binding in a concentration-dependent manner, demonstrating that chrysin was capable of displacing TCDD from the active site of the receptor. These results indicate that the affinity of chrysin for the receptor is markedly lower than that for TCDD, but is capable of interfering with the binding of TCDD and may serve as a ligand for the Ah receptor.

Figure 3.

Chrysin alters TCDD-binding affinity to the Ah receptor. Hepa1c1c7 cytosols (2.5 mg) were incubated for 2 hours with 2 nM [3H]TCDD in the presence or absence of varying concentrations of chrysin (Ch). As a control experiment to document specific binding, the competitive antagonist TCDF (1 μM) was incubated with one of the reactions. Following equilibration, free [3H]TCDD was removed by dextran-coated charcoal. The cytosol was layered onto 10% to 30% discontinuous sucrose gradients and the protein separated by centrifugation for 16 hours. Fractions were collected, and the total radioactivity in each was determined by liquid scintillation counting.

The Role of the Ah Receptor in Chrysin Induction of UGT1A1.

The findings that an XRE element was required for maximal induction of UGT1A1 reporter gene activity coupled with the observation that chrysin effectively displaced TCDD from the receptor led us to analyze the potential of chrysin to activate the Ah receptor. For these studies, electromobility shift analyses were performed using the XRE sequence from both the UGT1A1 and CYP1A1 promoters (Fig. 4A). TCDD-induced activation of the Ah receptor in HepG2 cells leads to binding of the receptor to the UGT1A1 XRE sequence.24 When HepG2 cells were treated with chrysin, only weak binding of the nuclear Ah receptor occurred to both the UGT1A1 and CYP1A1 XREs (Fig. 4A). In contrast, significantly greater binding of the Ah receptor to the UGT1A1 and CYP1A1 XRE occurred after TCDD treatment. This pattern of XRE binding was not consistent with expression of UGT1A1 promoter activity (Fig. 1B), whereas chrysin exhibited far greater transcriptional activation than TCDD, suggesting the banding pattern may not be due solely to Ah receptor recruitment. Supershift analysis confirmed only partial recruitment of the Ah receptor to the UGT1A1 XRE as shown by the 60% reduction in banding intensity when the binding reaction included an anti-Ah receptor antibody (Fig. 4B). Furthermore, unlabeled mutated XRE probe did compete with the chrysin-induced band, whereas it did not compete with TCDD-induced Ah receptor recruitment to the UGT1A1 XRE, indicating that other factors may be associated at or near the UGT1A1 XRE.

Figure 4.

The effects of chrysin on Ah receptor activation. (A) Nuclear protein was collected from HepG2 cells treated for 48 hours with DMSO, 10 nM TCDD, or 20 μM chrysin. Binding reactions were carried out in the presence of 10 μg nuclear protein with either the UGT1A1 or CYP1A132P-XRE oligonucleotide. Competition reactions (+ Comp) were performed in the presence of 200 fold-excess of unlabeled probe. (B) Competition reactions were performed in the presence of TCDD- or chrysin-treated nuclear protein and 200-fold excess of either the wild-type XRE (+ XRE) or the mutated XRE (+ mXRE) unlabeled probe. Antibody supershift experiments were performed in the presence of 100 ng antibody directed against the AhR (+ AhR), Arnt (+ Arnt), or nonspecific rabbit IgG (+ IgG). PhosphorImager quantification (arbitrary units) of the AhR/Arnt/XRE binding complex is given below each lane. Results shown are representative of three independent experiments. (C) HepG2 cells were treated for 48 hours with DMSO (D), 10 nM TCDD (T), 20 μM chrysin (Ch), and cytosolic, nuclear, and whole cell extracts were collected. Twenty micrograms protein was used for Western blot analysis of Ah receptor protein using a rabbit anti-mouse Ah receptor antibody that is cross-reactive to the human Ah receptor.

The relatively weak binding of nuclear Ah receptor to the XRE sequences after chrysin treatment was consistent with the concentration of nuclear Ah receptor protein identified by Western blot analysis after chrysin treatment of HepG2 cells (Fig. 4C). With TCDD treatment, cytosolic levels of the Ah receptor decreased dramatically while accumulating in the nucleus. When HepG2 cells were treated with chrysin, a reduction in the cytosolic concentration of the receptor was not apparent, yet a detectable but low accumulation could be seen in the nucleus. It can also be noticed from Western blot analysis of whole-cell extracts that TCDD treatment leads to a loss of cellular Ah receptor content. Agonists of the Ah receptor initiate ubiquitin-mediated degradation of the Ah receptor,29, 31 However, chrysin treatment of HepG2 cells had no impact on the steady-state levels of Ah receptor content, as evident from no change in concentration of the receptor. These results indicate that chrysin is unable to activate the Ah receptor in a TCDD-dependent manner.

An alternative explanation for the high efficiency of UGT1A1-driven luciferase activity by chrysin (Fig. 1B) may be the result of its ability to promote Ah receptor–dependent transactivation of gene transcription. To examine this possibility, an Ah receptor construct lacking the N-terminal basic helix-loop-helix domain fused to a functional glucocorticoid receptor DNA-binding domain (GRDBD) was used (Fig. 5). When co-transfected with a luciferase reporter construct under control of the glucocorticoid-binding enhancer sequence (GRE), expressed GRDBD will associate with the GRE, and the Ah receptor transactivation domain will promote transcriptional activation. Co-transfection of HepG2 cells with the GRDBD/mAhR83-805 and p(GRE)2T105Luc plasmids followed by treatment with TCDD led to a 25-fold induction of luciferase activity, demonstrating that TCDD promotes transactivation of the Ah receptor. However, when transfected cells were treated with a concentration of chrysin that promoted UGT1A1-luciferase activity (20 μM), only a 1.6-fold induction in Ah receptor–dependent transactivation was measured. In addition, co-treatment of transfected cells with TCDD and chrysin led to a nearly 50% reduction in the ability of TCDD to promote transactivation. This result indicates that although chrysin is capable of inducing CYP1A1 and UGT1A1 transcription in an XRE-dependent fashion, its ability to activate the Ah receptor and promote transactivation is occurring in a manner that is independent from the processes underlying TCDD initiated activation of the receptor.

Figure 5.

Transactivation of the Ah receptor by chrysin and TCDD. HepG2 cells were transiently transfected for 24 hours with p(GRE)2T105Luc and phRL-SV40 in the absence (gray bars) or presence (black bars) of pCMV/GRDBD/mAhR83-805 followed by 48-hour treatment with DMSO, 10 nM TCDD, or 20 μM chrysin. Firefly luciferase readings were normalized to Renilla luciferase and protein concentration and displayed as fold increase over the respective DMSO treatment. Significant increase of chrysin treatment over DMSO is indicated (*P ≤ 0.05)

To conclusively determine the role of the Ah receptor in chrysin induction of UGT1A1, siRNA directed against the Ah receptor (si-AhR) was employed. Transfection of HepG2 cells with 50 nM si-AhR results in an average 50% knockdown of Ah receptor protein (Fig. 6A). The effectiveness of Ah receptor protein knock-down was demonstrated by the complete inhibition of TCDD induction of the -1612CYP1A1 promoter luciferase (Fig. 6B). Chrysin induction of CYP1A1-promoter luciferase was also inhibited, indicating that the Ah receptor is the primary pathway of chrysin induction of CYP1A1. In contrast, chrysin induction of UGT1A1 PR3.7 luciferase was only reduced by 30% after si-AhR transfection. Therefore, we conclude that the mechanisms of induction of CYP1A1 and UGT1A1 by chrysin are divergent. Chrysin induces CYP1A1 through weak activation of the Ah receptor, whereas induction of the UGT1A1 gene is dependent not only on the Ah receptor but also additional cellular factors such as coactivators or corepressors.

Figure 6.

siRNA knockdown of Ah receptor. (A) HepG2 cells were transfected for 48 hours with either 50 nM negative control siRNA (Mock) or 50 nM si-AhR. Whole-cell extract (20 μg) was used for Western blot analysis with the anti-Ah receptor antibody or actin. Ah receptor band intensities were quantified, normalized to actin, and represented as percent of mock transfection (100%). (B) HepG2 cells were transiently transfected with the 1.6 kb CYP1A1-promoter luciferase or UGT1A1 PR3.7 luciferase, phRL-SV40, and 50 nM negative control siRNA (mock; black bars) or 50 nM si-AhR (gray bars). Thirty-six hours after transfection, cells were treated with DMSO (D), 1 nM TCDD (T), or 20 μM chrysin (C) for 24 hours. Firefly luciferase readings were normalized to Renilla luciferase and protein concentration and displayed as fold increase over the respective DMSO treatment. Significant decreases from mock transfection are indicated (*P ≤ 0.05; **P < 0.005; ***P < 0.0005).

Chrysin Activation of the UGT1A1 Promoter Occurs Through the MAPK Pathway.

ERK1/2 has been shown to be involved in Ah receptor stabilization and transactivation potential.29 Inhibition of ERK1/2 by U0126 leads to the inhibition of TCDD-initiated activation of Cyp1a1. To investigate whether similar MAP kinase pathways are involved in chrysin induction of UGT1A1 through the Ah receptor, MH1A1 cells were treated with a panel of MAP kinase inhibitors (Fig. 7A). ERK1/2 was inhibited using the MEK1 inhibitors U0126 and PD98059, JNK1/2/3 was inhibited using SP600125, and p38 was inhibited by SB203580. All MAP kinase inhibitors suppressed chrysin induction of UGT1A1 luciferase. PD98059 and SP600125, as well as being MAP kinase inhibitors, are also competitive ligands for the TCDD binding site on the Ah receptor.32 The slight increase in UGT1A1 luciferase activity with PD98059 or SP600125 treatment alone may be the result of this Ah receptor activation. Thus, these MAP kinase inhibitors may be blocking access of chrysin to the receptor, resulting in only partial reduction of luciferase activity similar to the partial reduction observed with si-AhR. However, U0126 does not compete with TCDD for Ah receptor binding and has been used to demonstrate a direct link between ERK1/2 phosphorylation and Ah receptor activation.29 Additionally, the link between MEK1 activity and chrysin induction of UGT1A1 is shown by overexpression of a dominant negative MEK1 (MEK1-DN). When compared with wild-type MEK1 (MEK1-WT) expression, the MEK1-DN inhibits basal as well as chrysin-induced UGT1A1 luciferase activity in a manner similar to treatment with UO126 (Fig. 7B). ERK1/2 involvement in chrysin induction of UGT1A1 is further implicated by the observation that chrysin treatment of HepG2 cells leads to active, phosphorylated ERK1/2 at levels similar to TCDD (Fig. 7C). We conclude that, in conjunction with the Ah receptor, chrysin activates the MAP kinase signaling pathways to achieve maximal UGT1A1 induction.

Figure 7.

Inhibition of UGT1A1 luciferase activity by MAP kinase inhibitors. (A) MH1A1 cells were treated for 48 hours with a series of MAP kinase inhibitors alone or in combination with 20 μM chrysin: UO126 (UO, 20 μM), PD98059 (PD, 20 μM), SB203580 (SB, 20 μM), or SP600125 (SP, 20 μM). Luciferase values were normalized to protein concentration and displayed as fold increase over DMSO treatment. Significant increases over DMSO are indicated (*P ≤ 0.05; **P ≤ 0.005; ***P ≤ 0.0005). Significant decreases are indicated (†P ≤ 0.005; ‡P ≤ 0.0005). (B) HepG2 cells were transiently transfected with a wild-type MEK1 (WT) or dominant negative MEK1 (DN) construct with the UGT1A1 PR3.7 and phRL-SV40 luciferase constructs for 24 hours followed by 24 hours of treatment with DMSO (D) or 20 μM chrysin (C). Firefly luciferase readings were normalized to Renilla luciferase and protein concentration and displayed as fold increase over MEK1-WT DMSO treatment. Significant decreases from WT-MEK1 transfection are indicated (*P ≤ 0.05). (C) Phosphorylated Erk1/2 (p-P44/42) and total Erk1/2 (P44/42) were measured in whole-cell extracts from HepG2 cells treated for 48 hours with 10 nM TCDD (T), 20 μM chrysin (C), and 10 μM UO126 (UO).

Actions of Oral Chrysin Treatment in Transgenic UGT1 (Tg-UGT1) Mice.

One of the goals in defining the mechanisms underlying induction of the UGTs by chrysin and other flavonoids is to apply this knowledge to in vivo models. To examine the contribution of chrysin on human UGT1A1 expression in vivo, we have taken advantage of a recently developed transgenic mouse model (Tg-UGT1) that expresses the entire human UGT1 locus.23 The UGT1 locus in Tg-UGT1 mice has been shown to be regulated in a tissue-specific fashion that is comparable to gene expression patterns observed in humans,8 in addition to the locus being regulated by induction after activation of the Ah receptor and pregnane X receptor,23 as well as the liver X-receptor.

To confirm that chrysin is adequately absorbed in wild-type (WT) and Tg-UGT1 mice, plasma samples were analyzed for chrysin and chrysin-glucuronide 60 minutes after a single oral dose of 50 mg/kg chrysin. The average concentration of chrysin in plasma from WT and Tg-UGT1 mice is 6.9 ng/mL and 2.7 ng/mL, respectively (Fig. 8A). Metabolism of chrysin to the glucuronide was observed in WT mice (average concentration, 581 ng/mL); however, the conversion to the glucuronide was nearly doubled in Tg-UGT1 mice (average concentration, 1,124 ng/mL). The metabolic ratio of drug to metabolite in WT and Tg-UGT1 mice after a single dose of chrysin was 0.012 and 0.002, respectively. The very low metabolic ratio both in WT and Tg-UGT1 mice indicates that chrysin is rapidly metabolized after oral administration and possesses very poor bioavailability.

Figure 8.

The actions of oral chrysin treatment. (A) Wild-type (WT) and Tg-UGT1 (Tg) mice were given a single oral dose of chrysin (50 mg/kg). Plasma was collected 60 minutes after the dose and analyzed by HPLC for the presence of chrysin (circles) and the chrysin-glucuronide (squares). (B) Tg-UGT1 mice were orally gavaged for 8 days with vehicle (V, 40% DMSO/60% corn oil) or 50 mg/kg chrysin (C) in vehicle. Microsomes were prepared from liver and small intestine for Western blot analysis. Supersomes expressing UGT1A1 were used as a standard (S). Blots were incubated with the human-specific anti-UGT1A1 antibody. (C) Tg-UGT1 mice were gavaged for 3 days with vehicle or 100 mg/kg B(a)P and microsomes prepared from liver and small intestine for Western blot analysis of UGT1A1.

To determine whether oral chrysin administration can induce human UGT1A1 expression in the gastrointestinal tract and liver, Tg-UGT1 mice were administered chrysin (50 mg/kg) every day for 7 days and the expression of human UGT1A1 was evaluated in small intestine and liver microsomes by Western blot analysis. As previously described, Tg-UGT1 mice express low basal levels of UGT1A1 in liver combined with significant expression in the small intestine (Fig. 8B). Oral chrysin treatment had no effect on facilitating induction of UGT1A1 in these tissues. However, when Tg-UGT1 mice were treated orally with B[a]P (100 mg/kg) for 3 days, UGT1A1 was induced in both small intestine and liver, demonstrating that agents capable of activating the Ah receptor in vivo will ultimately stimulate UGT1A1 induction. Overall, if the molecular mechanisms defined in this report are extrapolated to humans, the findings would indicate that the high levels of glucuronidation capacity in the gastrointestinal tract serve to limit the bioavailability of consumed flavonoids and restrict the ability of these agents to regulate UGT1A1 through activation of the Ah receptor or other as yet unidentified pathways.

Discussion

In this study, we describe the mechanism of human UGT1A1 gene induction by the dietary supplement chrysin. As demonstrated by electrophoretic mobility shift assay and expression of UGT1A1 promoter and enhancer reporter gene constructs, UGT1A1 gene induction by chrysin occurs after activation of the Ah receptor and binding to an XRE in the proximal region of the gene. Deletion of the XRE binding sequence eliminates chrysin induction. Our findings indicate that chrysin is capable of competing for TCDD binding to the Ah receptor and thus serves as a ligand for the receptor. Like TCDD activation of the Ah receptor and induction of CYP1A1, chrysin activation and induction of UGT1A1 is dependent on activation of MAP kinase pathways, indicating that TCDD and chrysin may promote induction of UGT1A1 through conserved signaling pathways. In addition, treatment of HepG2 cells with either chrysin or TCDD promotes the nuclear translocation of the Ah receptor in a manner that promotes binding of the receptor to DNA. Combined, this evidence suggests that chrysin induction of UGT1A1 is being elicited through activation of the Ah receptor.

However, in HepG2 cells, dramatic differences also have been observed in the molecular actions between TCDD and chrysin and their ability to promote induction of UGT1A1. For example, chrysin is a more effective inducer of UGT1A1 than TCDD in tissue culture. This finding is a surprise because TCDD activation of the Ah receptor predominates when compared with the ability of chrysin to induce CYP1A1 reporter activity. One might even speculate that a mechanism is in place on the UGT1A1 promoter to suppress responsiveness to the Ah receptor. Combined with the minimal impact of Ah receptor knockdown on chrysin induction of the UGT1A1 promoter, it is clear that although the Ah receptor may be recruited to the promoter, the Ah receptor is not the primary mechanism of chrysin induction of UGT1A1. Conversely, an intact XRE sequence (CACGCA), typically only associated with the Ah receptor, is critical for chrysin induction of UGT1A1, indicating other coactivators are recruited to this site in a MEK1-dependent manner.

The differential characteristics of chrysin induction of CYP1A1 and UGT1A1 are interesting. Further evidence that TCDD and chrysin differentially regulate induction in an Ah receptor–dependent fashion was demonstrated by examining the functional properties of the receptor. After treatment of HepG2 cells with TCDD, ligand binding to the Ah receptor stimulated both transactivation and proteolysis. These 2 activation processes are thought to be characteristic properties of ligands that activate the Ah receptor. In contrast, chrysin treatment of HepG2 cells does not promote transactivation of the Ah receptor in a TCDD-dependent fashion and has no effect on initiating proteolysis of the receptor. The near complete block in chrysin induction of CYP1A1-luc by si-AhR is strong evidence that chrysin is able to activate transcription through the Ah receptor only on the CYP1A1 promoter. This activation may be the result of the very weak transactivation potential of chrysin on the Ah receptor or dependent on the recruitment of coactivators.

Flavonoids have been described as health-promoting, disease-preventing dietary supplements and have been widely proclaimed to possess biological activity as cancer preventive agents.1 Part of the rationale for predicting the effectiveness of flavonoids in disease prevention stems from reports that agents such as chrysin are capable of inducing those enzymes that are involved in drug-elicited and dietary-elicited detoxification, such as the UGTs.2 Yet, limited reports are available on the effectiveness of flavonoids in regulation of the UGT genes in vivo. To directly address the potential for chrysin to induce human UGT1A1 expression, transgenic mice that express the human UGT1 locus23 were treated orally with chrysin for 7 days followed by measurement of UGT1A1 expression in liver and small intestinal microsomes. Although UGT1A1 is expressed endogenously in small intestine, oral treatment of chrysin had no effect on protein induction. A similar result was noted when Tg-UGT1 liver microsomes were analyzed. Because the oral treatment of B[a]P led to induction of UGT1A1 in both liver and small intestine, these findings indicated that oral bioavailability of chrysin was dramatically limited. This appears to be the case, as demonstrated by the rapid reduction of plasma chrysin and the accumulation of chrysin glucuronide. Clearly, chrysin's ability to induce detoxifying enzymes such as UGT1A1 is tempered by the low bioavailability of the compound, which has been suggested to occur in humans.33 Although chrysin has been convincingly demonstrated to induce UGT1A1 expression in tissue culture models,16, 34 these data indicate that oral chrysin consumption has little effect on regulating UGT1A1 expression in vivo.

Although other flavonoids have not been investigated, the utility of consuming these compounds for dietary or clinical use may be questionable. An example of the usefulness of flavonoid therapy is presented in individuals who are taking the prodrug irinotecan for solid tumor chemotherapy. Irinotecan is metabolized to the active metabolite SN-38, which is ultimately eliminated through glucuronidation.4, 35, 36 However, a heightened sensitivity to irinotecan-induced bone marrow toxicity as well as ileocolitus toxicity37 is observed in patients with reduced UGT1A1 activity because adequate elimination of SN-38 is impaired.38 With available information that flavonoids can induce UGT1A1 in both human hepatoma and intestinal cells, a diet rich in flavonoids may help to counter the toxicities elicited by SN-38 treatment.4 Yet, the experiments that we conducted in Tg-UGT1 mice would indicate that oral flavonoid therapy may have limited potential as an inducer of human UGT1A1 in vivo. Our findings are consistent with a previous clinical study demonstrating that colorectal cancer patients on irinotecan therapy in combination with chrysin treatment did not significantly improve SN-38–induced gastrointestinal tract toxicity.39

In conclusion, this study demonstrates that chrysin controls induction of UGT1A1 in a MEK-1–dependent manner involving the Ah receptor, but the ability of chrysin to modulate UGT1A1 gene expression in vivo after oral exposure is severely limited. The current study reinforces the need for additional studies to be conducted in transgenic and humanized animal models as an endpoint to verify the importance of regulatory events that have been identified in other model systems such tissue culture studies.

Acknowledgements

The authors thank Patrick Caron for technical assistance with plasma chrysin analysis. We thank Dr. Joseph K. Ritter (Medical College of Virginia, Virginia Commonwealth University, Richmond, Virginia) for making available the human-specific UGT1A1 antibody and Dr. Christopher Bradfield (University of Wisconsin, Madison, WI) for the anti-Ah receptor antibody. The authors also thank Dr. Lawrence Poellinger (Karolinska Institute, Stockholm, Sweden) for providing the pCMV/GRDBD/mAhR83-805 and p(GRE)2T105Luc plasmids.

Ancillary