-Mangostin Inhibits Inhibitor- B Kinase Activity and Decreases Lipopolysaccharide-Induced Cyclooxygenase-2 Gene Expression in C6 Rat Glioma Cells

We investigated the effect of -mangostin purified from the fruit hull of the medicinal plant Garcinia mangostana on spontaneous prostaglandin E2 (PGE2) release and inducible cyclooxygenase 2 (COX-2) gene expression in C6 rat glioma cells. An 18-h treatment with -mangostin potently inhibited spontaneous PGE2 release in a concentration-dependent manner with the IC50 value of approximately 2 M, without affecting the cell viability even at 30 M. By immunoblotting and reverse-transcription polymerase chain reaction, we showed that -mangostin concentration-dependently inhibited lipopolysaccharide (LPS)-induced expression of COX-2 protein and its mRNA, but not those of constitutive COX-1 cyclooxygenase. Because LPS is known to stimulate inhibitor B (I B) kinase (IKK)-mediated phosphorylation of I B followed by its degradation, which in turn induces nuclear factor (NF)B nuclear translocation leading to transcriptional activation of COX-2 gene, the effect of -mangostin on the IKK/I B cascade controlling the NFB activation was examined. An in vitro IKK assay using IKK protein immunoprecipitated from C6 cell extract showed that this compound inhibited IKK activity in a concentration-dependent manner, with the IC50 value of approximately 10 M. Consistently -mangostin was also observed to decrease the LPSinduced I B degradation and phosphorylation in a concentration-dependent manner, as assayed by immunoblotting. Furthermore, luciferase reporter assays showed that -mangostin reduced the LPS-inducible activation of NFB–and human COX-2 gene promoter region-dependent transcription. -Mangostin also inhibited rat carrageenan-induced paw edema. These results suggest that -mangostin directly inhibits IKK activity and thereby prevents COX-2 gene transcription, an NFB target gene, probably to decrease the inflammatory agent-stimulated PGE2 production in vivo, and is a new useful lead compound for anti-inflammatory drug development. Prostaglandins (PGs), arachidonic acid (AA) metabolites of the cyclooxygenase (COX) pathway, are major mediators in the regulation of inflammation and immune function (Smith et al., 2000). In the brain, the prostaglandin E2 (PGE2) is the most abundant PG. PGE2 levels are very low or undetectable in normal conditions but can rise in response to inflammatory processes, multiple sclerosis, and AIDS-associated dementia (Fretland, 1992; Griffin et al., 1994). High levels of PGE2 can modulate the activities of multiple cell types, including neurons, glial, and endothelial cells, as well as microglia/macrophage and lymphocyte functions during inflammatory and immune processes (Weissmann, 1993). Astrocytes are a known important source of PGE2 in the CNS (Katsuura et al., 1989). Their ability to produce PGE2 upon stimulation with interleukin (IL)-1 , tumor necrosis factor(TNF), or bacterial wall protein lipopolysaccharide (LPS) has been extensively documented (Fontana et al., 1982; Mollace et al., 1998; Molina-Holgado et al., 2000). Cyclooxygenase (COX) is well known to be responsible for PG production and the rate-limiting enzymes. This enzyme This work was partly supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sport and Culture of Japan. 1 Current address: Pfizer Global Research and Development, Nagoya Laboratories, Aichi, Japan. Article, publication date, and citation information can be found at http://molpharm.aspetjournals.org. doi:10.1124/mol.104.002626. ABBREVIATIONS: PG, prostaglandin; AA, arachidonic acid; COX, cyclooxygenase; COX-1, constitutive cyclooxygenase; COX-2, inducible cyclooxygenase; PGE2, prostaglandin E2; LPS, lipopolysaccharide; I B, inhibitor B; IKK, inhibitor B kinase; CNS, central nervous system; IL, interleukin; TNF, tumor necrosis factor; NFB, nuclear factorB; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium; RT, reverse transcription; PCR, polymerase chain reaction; PAGE, polyacrylamide gel electrophoresis; ODS, octadecylsilyl; HPLC, high-performance liquid chromatography; DMSO, dimethyl sulfoxide; NS398, N-[2-(cyclohexyloxyl)-4-nitrophenyl]-methane sulfonamide. 0026-895X/04/6603-667–674$20.00 MOLECULAR PHARMACOLOGY Vol. 66, No. 3 Copyright © 2004 The American Society for Pharmacology and Experimental Therapeutics 2626/1172525 Mol Pharmacol 66:667–674, 2004 Printed in U.S.A. 667 at A PE T Jornals on O cber 4, 2017 m oharm .aspeurnals.org D ow nladed from exists as two isoforms, namely constitutive COX (COX-1) and inducible COX (COX-2). COX-2 gene is an early-inducible gene in response to many inflammatory cytokines, including IL-1, TNF, and LPS. COX-2 gene expression is controlled at the transcriptional and/or the post-transcriptional levels (Dixon et al., 2000). The promoter regions of COX-2 gene of rat, human, and mouse have been extensively analyzed (Sirois et al., 1993; Appleby et al., 1994; Kosaka et al., 1994; Inoue et al., 1995) and have been shown to contain potential binding sites for various transcriptional factors, some of which have been demonstrated to be indeed functional (Goppelt-Struebe, 1995). For example, activation of nuclear factorB (NFB) has been reported recently to actually participate in the transcriptional activation of COX-2 gene induced by IL-1 (Newton et al., 1997), TNF(Yamamoto et al., 1995), and LPS (Inoue and Tanabe, 1998). Furthermore, the LPSinduced activation of COX-2 gene evidently has been shown to be mediated by inhibitor B (I B) kinase (IKK), which specifically catalyzes I B phosphorylation followed by its degradation and the subsequent NFB nuclear translocation, leading to a stimulation of the cis-acting B elementmediated transcription (Griseavage et al., 1996). Because the fruit hull of mangosteen, Garcinia mangostana, has been used for many years as traditional medicine for the treatment of skin infection, wounds, and diarrhea in Southeast Asia—that is, the fruit hull exhibits an antiinflammatory action—we pharmacologically studied the anti-inflammatory components of the fruit hull of mangosteen. In an earlier study, we examined the effect of -mangostin, a tetraoxygenated diprenylated xanthone from the fruit hull of this plant (Fig. 1A), on AA cascade in C6 rat glioma cells, which is well known to be a useful model system for the study of PG production in the astrocytes, and demonstrated that this natural product reduces PGE2 release from C6 glioma cells with an IC50 of approximately 5 M and potently inhibits the activities of both COX-1 and COX-2 enzymes with the IC50 values of approximately 0.8 and 2 M, respectively, in vitro (Nakatani et al., 2002). Herein, we describe the first evidence that -mangostin directly inhibits IKK activity, which specifically phosphorylates I B, and thereby prevents its degradation and, as a result, induces a decrease in the expression of COX-2 protein and its mRNA by a suppression of NFB–dependent transcription. This study also demonstrated that -mangostin inhibited rat carrageenan-induced hind paw edema, an in vivo acute model of inflammation. Materials and Methods Materials. PGE2 was a generous gift from Ono Pharmaceuticals (Osaka, Japan). Fetal bovine serum, horse serum, Ham’s F-10 medium, Eagle’s minimum essential medium, 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide (MTT), and NS398, a selective inhibitor of COX-2, were purchased from Cell Culture Laboratory (Cleveland, OH), Dainippon Pharmaceutical Co. Ltd. (Osaka, Japan), Invitrogen (Carlsbad, CA), Nissui (Tokyo, Japan), Dojindo (Kumamoto, Japan), and Calbiochem (San Diego, CA), respectively. Anti-COX-1, anti-COX-2, anti-I B, anti-phospho-I B, anti-IKK / , and anti-actin antibodies and recombinant I B fusion protein (1–317), anti-PGE2 antibody, and protein A Sepharose 4B were obtained from Santa Cruz Biochemicals (Santa Cruz, CA), Chemicon International (Temecula, CA), and Zymed Laboratories (South San Francisco, CA), respectively. Alkaline phosphatase-conjugated affinity-purified antigoat IgG and horseradish peroxide-conjugated affinity-purified antirabbit IgG were from Rockland (Gilbertsville, PA) and Cell Signaling Technology Inc. (Beverly, MA), respectively. Total RNA extraction and reverse-transcription polymerase chain reaction (RT-PCR) kits, EndoFree Plasmid Maxi Kit, and pRG-TK vector and Dual-Luciferase Reporter Assay System were purchased from Toyobo Engineering (Osaka, Japan), QIAGEN K.K. (Tokyo, Japan), and Promega (Madison, WI), respectively. [H]PGE2 (200 Ci/mmol) and [ P]ATP (5000 Ci/mmol) were from PerkinElmer Life and Analytical Sciences (Boston, MA) and Amersham Biosciences Inc. (Piscataway, NJ), respectively. Other chemicals and drugs were of reagent grade or of the highest quality. Extraction and Isolation of -Mangostin. -Mangostin (Fig. 1A) was obtained from the fruit hull of G. mangostana L. as reported previously (Jefferson et al., 1970). In brief, the dried fruit hull (1 kg) was ground and then extracted with ethanol (10 liters) for 1 h. The ethanol extract was partitioned between ethyl acetate and water to afford an ethyl acetate fraction (87.4 g). The ethyl acetate soluble fraction was dissolved in hexane/ethyl acetate (4:1) and subjected on a silica gel column chromatography (Wakogel C-200, 800 g, a diameter of 11 cm; Wako Pure Chemicals, Tokyo, Japan), and the sample was eluted with a step-wise gradient of hexane/ethyl acetate [3:1 (2 liters), 3:2 (1 liter), 1:1 (1 liter), 2:3 (1 liter), and 3:7 (2 liters)]. After collecting 1400 ml of the first eluent, the subsequent eluent was Fig. 1. Chemical structure of -mangostin (A) and analysis of purified -mangostin by HPLC (B). The elution profile was obtained by monitoring at UV absorbance of 280 nm using an HPLC column (SenshuPak ODS1251-SS, 4.5 250 mm; elution: a linear gradient of 65 to 100% CH3CN in H2O at 1 ml/min before which the column was washed with 65% CH3CN in H2O for 60 min). 668 Nakatani et al. at A PE T Jornals on O cber 4, 2017 m oharm .aspeurnals.org D ow nladed from fractionated (200 ml each). Each fraction was monitored by thin layer chromatography [ODS silica gel; acetonitrile/water (8:2) as the developing solvent], and thereby the -mangostin-rich fractions, which contained -mangostin (Rf 0.49) but not -mangostin (Rf 0.38), were combined. This silica gel column chromatography was performed two times additionally to further obtain the -mangostin fraction. The combined -mangostin fractions (17.8 g) underwent chromatography on a Senshupak PEGSIL ODS column (20 250 mm; Senshu Scientific Co., Tokyo, Japan) and were eluted with methanol/water (4:1 to 1:0) to yield partially purified -mangostin (9.0 g). The partially purified -mangostin fraction was subjected to an Ultra Pack Silica gel column chromatography (50 mm x 300 mm) (Yamazen, Schaumburg, IL) and eluted with hexane/ethyl acetate (1:1) followed by recrystalization in hexane to finally give purified -mangostin (6.0 g). Its purity was more than 90%, as determined by HPLC (SenshuPak ODS-1251-SS, 4.5 250 mm; Senshu Scientific) (Fig. 1B). As shown in Fig. 1B, each of the other constituents was less than 3%. Purified -mangostin was dissolved in dimethyl sulfoxide (DMSO) to make a concentration of 20 mM as a stock solution and diluted to appropriate concentrations before use. Cell Culture. C6 rat glioma cells were grown in Ham’s F-10 medium containing 15% horse serum and 2.5% fetal bovine serum in a 37°C humidified incubator in an atmosphere of 5% CO2 in air (Nakatani et al., 2002). PGE2 Assay. For PGE2 assay, C6 cells were seeded onto 12-well plates at a cell density of 1.0 10 cells per well. Two days after cell seeding, cells were subjected to the experiment. Cells were cultured for 18 h in culture medium containing vehicle (0.1% DMSO), -mangostin, or NS398, a selective inhibitor of COX-2. The conditioned medium was acidified to pH 4.0 by the addition of 1 N HCl, and PGE2 was extracted twice with ethyl acetate. After ethyl acetate was evaporated under a stream of N2 gas, the sample was dissolved in 10 mM Tris-HCl, pH 7.6. PGE2 was determined by radioimmunoassay, as described previously (Nakatani et al., 2002). Cell Viability Assay. Cell viability for C6 cells was measured using the MTT method as reported previously (Taglialatela et al., 1997). Cells were seeded onto 96-well plates at a cell density of 1.0 10 cells per well and cultured with vehicle or -mangostin. Twentyfour hours later, cells were subjected to MTT assay. Product formation was monitored by reading the absorbance at 595 nm using a microplate reader (model 450; Bio-Rad, Hercules, CA). Semi Quantitative RT-PCR. For RT-PCR, C6 cells were seeded onto six-well plates at a cell density of 2.0 10 cells per well. Two days later, cells were pretreated with vehicle or -mangostin for 1 h and were incubated subsequently with or without 10 g/ml LPS for 1 h. The total RNA from cells was prepared by using a total RNA extraction kit. Both the COX-1 and COX-2 mRNA levels were semiquantitatively assayed using an RT-PCR kit as reported previously (Yang et al., 1998) as follows: The sense primers (5 -CTG CTG AGA AGG GAG TTC AT-3 , 602–621 of rat COX-1 cDNA; and 5 -ACA CTC TAT CAC TGG CAT CC-3 , 1229–1248 of rat COX-2 cDNA) and the antisense primers (5 -GTC ACA CAC ACG GTT ATG CT-3 , 9811000 of rat COX-1 cDNA; and 5 -GAA GGG ACA CCC TTT CAC AT-3 , 11794-1813 of rat COX-2 cDNA) were complementary to the conserved regions of the cDNAs (Yang et al., 1998). The cDNA fragments were amplified 32 cycles (94°C for 60 s, 62°C for 60 s, and 72°C for 60 s). It was demonstrated that this condition for RT-PCR quantitatively yielded PCR product by our preliminary experiments (data not shown). Glyceraldehyde-3-phosphate dehydrogenase mRNA was used as an internal control for the present semiquantitative RTPCR. PCR products for COX-1 and COX-2 mRNAs were separated by 2% agarose gel electrophoresis, detected by ethidium bromide staining, and subjected to quantitative analysis using an image scanner (Foto/ Eclipse; Fotodyne, Hartland, WI). Furthermore, after purification of the PCR products by electrophoresis and filtration, the nucleotide sequences were determined by the dideoxy nucleotide chain termination method to verify that these PCR products are derived from COX-1 and COX-2 mRNAs. Immunoblotting. For immunoblotting, C6 cells were seeded onto six-well plates at a cell density of 2.0 10 cells per well. Two days after seeding, the cells were incubated with vehicle or -mangostin for 1 h at 37°C. After cells were incubated with or without 10 g/ml LPS for an additional 1 h, the medium was aspirated. The cells were lysed by addition of SDS-PAGE sample buffer (187.5 mM Tris-HCl, 6% SDS, 30% glycerol, and 15% 2-mercaptoethanol, pH 6.8). These protein samples were boiled for 5 min, subjected to SDS-PAGE (8–11% gel), and then transferred electrically onto polyvinylidene difluoride membranes (Immobilon-P; Millipore Corporation, Bedford, MA) by the semidry blotting method. The blots were incubated with 2% bovine serum albumin in Tris-buffered saline containing 0.05% Tween 20 at 25°C for 2 h and incubated with goat anti-COX-1 antibody (0.1 g/ml), goat anti-COX-2 (0.1 g/ml), rabbit anti-I B (0.2 g/ml), rabbit anti-phospho-I B (0.2 g/ml), or rabbit anti-actin (0.2 g/ml) antibody at 25°C for 2 h. The blots were washed several times and incubated with a 1:1000 to 2000 dilution of alkaline phosphatase-conjugated affinity-purified anti-goat IgG or horseradish peroxide-conjugated affinity-purified anti-rabbit IgG in Tris-buffered saline/Tween 20 containing 2% bovine serum albumin at 4°C overnight. Immunoreactive signals were visualized by incubation of the blots with chemiluminescence assay reagents followed by exposing them to Hyperfilm ECL (Amersham Biosciences). In Vitro IKK Assay. After 10 min of treatment with 10 g/ml LPS, C6 cells were washed with phosphate-buffered saline, lysed with ice-cold lysis buffer (2 mM EGTA, 150 mM NaCl, 2 mM dithiothreitol, 1 mM p-amidinophenyl methanesulfonyl fluoride hydrochloride, 10 g/ml leupeptin, 10 g/ml aprotinin, 1 mM Na3VO4, and 10 mM Tris-HCl, pH 7.5), and sonicated to prepare cell extract. IKK proteins were prepared by immunoprecipitation as follows: the cell extract was incubated with 6 g of anti-IKK / antibody at 4°C for 4 h, and the immunocomplex was recovered using protein A Sepharose 4B beads. The IKK protein-bound beads were then washed three times with lysis buffer and were aliquoted to five reaction tubes, including 25 l of the following kinase reaction mixture: 10 mM MgCl2 6H2O, 0.1 mM Na3VO4, 2 mM dithiothreitol, 5 mM -glycerolphosphate, [ -P]ATP, and 25 mM Tris-HCl, pH 7.5. After a 10-min incubation with or without tested concentrations of -mangostin at 30°C, 1 g of I B was added as a substrate to each reaction tube, and the reaction mixtures were further incubated for 30 min. The reaction was terminated by the addition of SDS-PAGE sample buffer and fractionated by SDS-PAGE. Phosphorylated I B (1–317) was visualized as radioluminogram and quantitatively analyzed with the use of Molecular Imager (GS363; Bio-Rad). Transient Transfection and Dual Luciferase Assay. C6 cells were plated at a cell density of 3.0 10 cells per well on a 24-well plate or 1.2 10 cells per well on a 48-well plate. Two days later, the cells were transfected using LipofectAMINE 2000 in serum-free medium according to the manufacturer’s recommended method. Cells plated onto 24-well plates were subjected to transfection with 0.475 g/well pNFB-Luc, a firefly luciferase reporter construct containing five repeated NFB–responsive elements, or dN-Luc, a reporter plasmid that is deficient in the repeated NFB–responsive elements in the pNFB-Luc (Hirai et al., 1994). Cells plated onto 48-well plates were also subjected to transfection with 0.4 g/well of phPES2( 327/ 59)-Luc, a firefly luciferase reporter construct containing the human COX-2 gene promoter fragment including NFB–responsive element ( 223/ 214) (Inoue et al., 1995). In addition, to normalize transfection efficiency, C6 cells were cotransfected with 0.025 g of a Renilla reniformis luciferase control vector (phRG-TK). After transfection, cells were cultured in the culture medium for 13 or 20 h and then preincubated with vehicle or -mangostin for 3 h. Cells were incubated with or without 1 g/ml LPS for an additional 13 or 18 h, and then the cells were harvested. Determination of both the firefly and R. reniformis luciferase activities was performed using a MiniLumat LB 9506 (Berthold Technologies, Bad Wildbad, Germany) with Dual-Luciferase Reporter Assay System (Promega). -Mangostin Inhibits IKK Activity and COX-2 Gene Expression 669 at A PE T Jornals on O cber 4, 2017 m oharm .aspeurnals.org D ow nladed from Relative luciferase activity represents the ratio of the activity of firefly luciferase to that of R. reniformis luciferase. Animals. 7-week-old male Wistar rats (weighing 150–170 g) were obtained from Charles River Japan (Yokohama, Japan). All experimental procedures were approved by the Laboratory Animal Care and Use Local Committee of Graduate School of Pharmaceutical Sciences, Tohoku University, and were in accordance with the principles and guidelines on animal care of Tohoku University. Paw Edema. Paw edema was induced on the left hind paw in each rat by a subplantar injection of 75 l of sterile saline (0.9% NaCl) containing 1% carrageenan. -Mangostin at doses ranging from 1 to 30 mg/kg or the vehicle control (DMSO) was given i.p. 30 min before the carrageenan injection. According to the procedure described previously (Planas et al., 1995), the presence of edema was assessed by measuring the volume of the left hind paw before (V0) and 5 h after carrageenan injection (V5). The increase in volume in the inflamed paw was obtained by subtracting the volume measured before the carrageenan injection from the observed value at 5 h and expressed as a percentage: % edema [(V5 V0)/V0] 100. Data Analysis. IC50 values were calculated from nonlinear regression analysis of the data. The data are expressed as the means S.E.M., and a significant difference (P 0.05) was analyzed by analysis of variance.