Activation of Peroxisome Proliferator-activated Receptor γ Inhibits Interleukin-1β-induced Membrane-associated Prostaglandin E2 Synthase-1 Expression in Human Synovial Fibroblasts by Interfering with Egr-1*

Membrane-associated prostaglandin (PG) E2 synthase-1 (mPGES-1) catalyzes the conversion of PGH2 to PGE2, which contributes to many biological processes. Peroxisome proliferator-activated receptor γ (PPARγ) is a ligand-activated transcription factor and plays an important role in growth, differentiation, and inflammation in different tissues. Here, we examined the effect of PPARγ ligands on interleukin-1β (IL-1β)-induced mPGES-1 expression in human synovial fibroblasts. PPARγ ligands 15-deoxy-Δ12,14 prostaglandin J2 (15d-PGJ2) and the thiazolidinedione troglitazone (TRO), but not PPARα ligand Wy14643, dose-dependently suppressed IL-1β-induced PGE2 production, as well as mPGES-1 protein and mRNA expression. 15d-PGJ2 and TRO suppressed IL-1β-induced activation of the mPGES-1 promoter. Overexpression of wild-type PPARγ further enhanced, whereas overexpression of a dominant negative PPARγ alleviated, the suppressive effect of both PPARγ ligands. Furthermore, pretreatment with an antagonist of PPARγ, GW9662, relieves the suppressive effect of PPARγ ligands on mPGES-1 protein expression, suggesting that the inhibition of mPGES-1 expression is mediated by PPARγ. We demonstrated that PPARγ ligands suppressed Egr-1-mediated induction of the activities of the mPGES-1 promoter and of a synthetic reporter construct containing three tandem repeats of an Egr-1 binding site. The suppressive effect of PPARγ ligands was enhanced in the presence of a PPARγ expression plasmid. Electrophoretic mobility shift and supershift assays for Egr-1 binding sites in the mPGES-1 promoter showed that both 15d-PGJ2 and TRO suppressed IL-1β-induced DNA-binding activity of Egr-1. These data define mPGES-1 and Egr-1 as novel targets of PPARγ and suggest that inhibition of mPGES-1 gene transcription may be one of the mechanisms by which PPARγ regulates inflammatory responses.

Prostaglandin (PG) 1 E 2 is an important modulator of many physiological and pathophysiological conditions, including cell growth, vascular homeostasis, inflammation, immune regulation, cancer, and arthritis (1,2). The biosynthesis of PGE 2 requires two enzymes. Cyclooxygenase (COX; also termed PGH synthase) converts arachidonic acid into PGH 2 . Subsequently, PGE synthase (PGES) converts COX-2-derived PGH 2 to PGE 2 . Two isoforms of COX exist, COX-1 and COX-2, with similar enzymatic properties but distinctly different biological functions. COX-1 is expressed in most tissues and is responsible for physiological production of PGs. COX-2, in contrast, is almost undetectable under physiological conditions but is strongly induced in response to pro-inflammatory stimuli, growth factors, and mitogens (1)(2)(3).
At least three distinct PGES isoforms have been identified, including microsomal PGES-1 (mPGES-1), which was originally designated MGST1-L-1 (for membrane-bound GST1like-1) (4, 5), mPGES-2 (6), and cytosolic PGES (cPGES, or the heat shock protein-associated protein p23) (7). cPGES is constitutively and ubiquitously expressed and is preferentially coupled with COX-1, promoting immediate production of PGE 2 (7,8). By contrast, mPGES-1 is markedly up-regulated by pro-inflammatory stimuli and is functionally coupled with COX-2, promoting delayed PGE 2 synthesis (4,5,9). mPGES-2 is ubiquitously expressed in diverse tissues (6); however, its role remains elusive. Studies with deletion of mPGES-1 demonstrate that this isoform is largely responsible for the production of PGE 2 both in vitro and in vivo (10,11). mPGES-1 protein expression is induced in vitro in several cell types after treatment with the pro-inflammatory cytokines, interleukin (IL)-1␤, and tumor necrosis factor (TNF)-␣ and is down-regulated by anti-inflammatory glucocorticoids (5,12,13). Moreover, mPGES-1 was shown to be up-regulated in vivo in animal models of rheumatoid arthritis (9) and lipopolysaccharide-induced pyresis (5). Increased levels of mPGES-1 mRNA and protein were also detected in symptomatic atherosclerotic plaques, as well as various cancer cell lines and car-cinoma (14 -18), suggesting that aberrant expression of this enzyme could contribute to the pathogenesis of these disorders. Therefore, mPGES-1 may constitute a potential target for therapeutic intervention.
Peroxisome proliferator-activated receptors (PPARs) are a family of ligand-activated transcription factors belonging to the nuclear receptor superfamily (19). To date, three PPAR subtypes have been identified: PPAR␣, PPAR␤/␦, and PPAR␥. PPAR␣ is highly expressed in the liver, heart, kidney, and intestinal mucosa, where its regulates lipid metabolism. PPAR␥ is predominantly expressed in adipose tissue and regulates adipocyte differentiation. PPAR␣ is activated by eicosanoids, fatty acids, and the hypolipidemic drug Wy14643 (selective for PPAR␣). PPAR␥ is activated by the prostaglandin D 2 metabolite 15-deoxy-⌬ 12,14 -PGJ 2 (15d-PGJ 2 ) and synthetic anti-diabetic thiazolidinedione drugs (e.g. troglitazone) (19). There is accumulating evidence that PPAR␣ and PPAR␥ are implicated as important regulators of immune and inflammatory responses. PPAR␣ activation inhibits inflammatory mediators release from several cell types (20 -22). In addition, PPAR␣-deficient mice exhibit exacerbated inflammatory responses (23). PPAR␥ activation results in the inhibition of various inflammatory events, such as the production of IL-1␤, TNF-␣, and IL-6 in monocytes/macrophages as well as the proliferation and the production of IL-2 by T lymphocytes (24 -26). Moreover, we have observed that PPAR␥ ligands can suppress the expression of the inducible nitric-oxide synthase (iNOS), MMP-13, and COX-2 in human chondrocytes and the expression of MMP-1 in human synovial fibroblasts (19,27,28). These actions of PPAR␥ ligands were proven through repression of activities of many transcription factors, including nuclear factor-B (NF-B), activator protein 1, signal transducers and activators of transcription, and nuclear factor of activated T cells (19, 24 -28).
To date, limited information is available on the effect of PPAR␥ ligands on the induction of mPGES-1. Here, we analyze the effect of two PPAR␥ ligands, 15d-PGJ 2 and troglitazone, on IL-1␤induced-mPGES-1 expression in human synovial fibroblasts and investigate the mechanisms underlying this regulation.

EXPERIMENTAL PROCEDURES
Materials-Human recombinant IL-1␤ was obtained from R&D Systems Inc. 15d-PGJ 2 , troglitazone (TRO), Wy14643, GW9226, and enzyme immunoassay reagents for PGE 2 assays were purchased from Cayman Chemical. Aprotinin, leupeptin, pepstatin, and phenylmethylsulfonyl fluoride (PMSF) were from Sigma-Aldrich. Dulbecco's modified Eagle's medium (DMEM), penicillin and streptomycin, fetal calf serum (FCS), and TRIzol reagent were supplied by Invitrogen. [ 32 P]ATP was from Amersham Biosciences. Plasmid DNA was prepared using a kit from Qiagen. FuGENE 6 transfection reagent was from Roche Applied Science. The luciferase reporter assay system was from Promega. All other chemicals were purchased from either Fisher Scientific or Bio-Rad. Anti-mPGES-1 antibody was from Cayman Chemical, whereas anti-cPGES antibodies were from Cayman Chemical or Affinity BioReagents. Antibodies against Egr-1, PPAR␥, and ␤-actin were purchased from Santa Cruz Biotechnology Inc. Polyclonal rabbit anti-mouse IgG coupled with horseradish peroxidase and polyclonal goat anti-rabbit IgG with horseradish peroxidase were from Pierce.
Specimen Selection and Cell Culture-HSFs were isolated from synovial membranes obtained from osteoarthritic (OA) patients undergoing total knee joint replacement. All OA patients were evaluated by a certified rheumatologist and diagnosed on criteria developed by the American College of Rheumatology Diagnostic Subcommittee for OA (29). Briefly, synovial fibroblasts were released by sequential enzymatic digestion with 1 mg/ml Pronase (Roche Applied Science) for 1 h, followed by a 6-h incubation with 2 mg/ml collagenase (Type IA, Sigma) at 37°C in DMEM supplemented with 10% heat-inactivated FCS, 100 units/ml penicillin, and 100 g/ml streptomycin. Cells were incubated for 1 h at 37°C in tissue culture flasks (Primaria 3824, Falcon, Lincoln Park, NJ) allowing the adherence of non-fibroblastic cells possibly pres-ent in the cell preparation. In addition, flow cytometric analysis (Epic II, Coulter, Miami, FL) using the anti-CD14 (fluorescein isothiocyanate) antibody confirmed that no monocyte/macrophages were present in the synovial fibroblast preparation. The cells were seeded in tissue culture flasks and cultured until confluence in DMEM supplemented with 10% FCS and antibiotics at 37°C in a humidified atmosphere of 5% CO 2 /95% air. Only cells between passages 3 and 7 were used. PGE 2 Assays-At the end of the incubation period, the culture medium was collected and stored at Ϫ80°C. Levels of PGE 2 were determined using a PGE 2 enzyme immunoassay kit from Cayman Chemical. The detection limit and sensitivity was 9 pg/ml. All assays were performed in duplicate.
Real-time Quantitative PCR-Quantitative PCR analysis was performed in a total volume of 50 l containing cDNA template, 200 nM of sense and antisense primers, and 25 l of SYBR® Green master mix (Qiagen). Incorporation of SYBR® Green dye into PCR products was monitored in real time using a Gene Amp 5700 sequence detector (Applied Biosystems) allowing determination of the threshold cycle (C T ) at which exponential amplification of PCR products begins. After incubation at 95°C for 10 min to activate the AmpliTaq Gold enzyme, the mixtures were subjected to 40 amplification cycles (15 s at 95°C for denaturation and 1 min for annealing and extension at 60°C). After PCR, dissociation curves were generated with one peak, indicating the specificity of the amplification. A threshold cycle (C T value) was obtained from each amplification curve using the software provided by the manufacturer (Applied Biosystems). Data were expressed as -fold changes relative to control conditions (unstimulated cells) using the ⌬⌬C T method as detailed in the manufacturer's guidelines (Applied Biosystems). A ⌬C T value was first calculated by subtracting the C T value for the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase from the C T value for each sample. A ⌬⌬C T value was then calculated by subtracting the ⌬C T value of the control from the ⌬C T value of each treatment. The -fold changes compared with the control (unstimulated cells) were then determined by raising 2 to the ⌬⌬C T power. Each PCR reaction generated only the expected specific amplicon as shown by the melting-temperature profiles of the final product and by gel electrophoresis of test PCR reactions. Each PCR was performed in triplicate on two separate occasions from at least three independent experiments.
Plasmids and Transient Transfection-The human mPGES-1 promoter construct (Ϫ538/Ϫ28) was kindly provided by Dr. Terry J. Smith (University of California, Los Angeles) (13). The human expression vectors for wild type and dominant negative PPAR␥ were a kind gift from Dr. Krishna K. Chatterjee (University of Cambridge, Cambridge, UK) (30). The human Egr-1 expression vector and the pEgr-1Mutx3-TK-Luc reporter construct were generously provided by Dr. Yuqing E. Chen (Morehouse School of Medicine, Atlanta, GA) (31). A ␤-galacto-sidase reporter vector under the control of SV40 promoter (pSV40-␤galactosidase) was from Promega.
Transient transfection experiments were performed using FuGENE 6 (1 g of DNA:4 l of FuGENE 6) (Roche Applied Science) according to the manufacturer's recommended protocol. Briefly, HSFs were seeded and grown to 50 -60% confluence. The cells were transfected with 1 g of the reporter construct and 0.5 g of the internal control pSV40-␤galactosidase. In cotransfection experiments the amount of transfected DNA was kept constant by using a corresponding empty vector. Six hours later, the medium was replaced with DMEM containing 1% FCS. The next day, the cells were treated for another 14 h with or without IL-1␤ in the absence or presence of 15d-PGJ 2 or TRO. After harvesting, luciferase activity was determined and normalized to ␤-galactosidase activity (27).
Nuclear Extract Preparation and Electrophoretic Mobility Shift Assay-Nuclear extracts were prepared as previously described (28). Briefly, HSFs were washed in ice-cold phosphate-buffered saline and gently scrapped in ice-cold hypotonic buffer containing 10 mM HEPES-KOH, pH 7.9, 10 mM KCl, 1.5 mM MgCl 2 , 0.5 mM DTT, 1 mM PMSF, 1 mM Na 3 VO 4 , and 10 g/ml each of aprotinin, leupeptin, and pepstatin. The cells were allowed to swell on ice, and the nuclei were recovered by brief centrifugation. The pellets were resuspended in high salt buffer containing 20 mM HEPES, pH 7.9, 420 mM NaCl, 1.2 mM MgCl 2 , 0.5 mM DTT, 0.2 mM EDTA, 25% glycerol, 0.5 mM PMSF, 1 mM Na 3 VO 4 , and 10 g/ml each of aprotinin, leupeptin, and pepstatin, followed by incubation on ice for 20 min. The nuclear extracts were recovered by centrifugation, and protein concentration was determined by the method of Bradford (Bio-Rad).
A synthetic double-stranded oligonucleotide, corresponding to the Egr-1 motifs in the human mPGES-1 promoter (5Ј-GTGGGG-CGGGGCGTGGGCGGTGCT-3Ј), was end-labeled by T4 polynucleotide kinase in the presence of [␥-32 P]ATP. The mutant competitor oligonucleotide had the following sequence with a 4-bp substitution (underlined): 5Ј-GTGGTTCGGGGCGTGTTCGGTGCT-3Ј. The binding buffer consisted of 10 mM Tris-HCl, pH 7.5, 50 mM NaCl, 0.5 mM DTT, 0.5 mM EDTA, 1 mM MgCl 2 , 4% glycerol, and 2.5 g of poly-(dI-dC). Binding reactions were conducted with 5 g of nuclear extract and 100,000 cpm 32 P-labeled oligonucleotide probe at 22°C for 20 min in a final volume of 10 l. In supershift assays, the antibody to Egr-1 (1 g/reaction) was incubated with the reaction mixture for 1 h at 4°C before the addition of 32 P-labeled oligonucleotide. In cold competition assays, 100-fold molar excess of cold wild-type or mutant oligonucleotide was used. Binding complexes were resolved on nondenaturating 6% polyacrylamide gel electrophoresis in a Tris borate buffer system, after which the gels were fixed, dried, and subjected to autoradiography.
Statistical Analysis-All results were calculated as the mean Ϯ S.E. of independent experiments. Statistics were analyzed using Student's 2-tailed t test. p values of Ͻ0.05 were considered significant.
Activation of PPAR␥ Inhibits IL-1␤-induced mPGES-1 Expression at the Transcriptional Level-To further elucidate the mechanism responsible for the changes in amounts of mPGES-1 protein, we measured the steady-state level of mPGES-1 mRNA by quantitative reverse transcription-PCR. Treatment with IL-1␤ (100 pg/ml) enhanced the expression of mPGES-1 mRNA (Fig. 3A). Both 15d-PGJ 2 and TRO dose-dependently suppressed IL-1␤-induced mPGES-1 mRNA expression (Fig. 3A). In contrast, and in agreement with the data in Fig. 2C, the PPAR␣-specific ligand Wy14643 had no significant effect on IL-1␤-induced mPGES-1 mRNA expression (Fig. 3A). As expected, the level of cPGES mRNA was not altered as a consequence of treatment with IL-1␤ alone or in combination with either PPAR␥ ligands (Fig. 3B). Thus, the level of mPGES-1 and cPGES mRNA expression mirrors the pattern of their respective protein expression.
To determine whether the regulation of IL-1␤-induced mPGES-1 mRNA by 15d-PGJ 2 and TRO occurred at the level of transcription, we carried out transient transfection studies. Synovial fibroblasts were transfected with a human mPGES-1 promoter (Ϫ538 to Ϫ28) region/luciferase reporter gene construct and stimulated with IL-1␤ in the absence or presence of PPAR␥ ligands. As shown in Fig. 4, IL-1␤ increased the luciferase activity of the mPGES-1 promoter, and this activation was dose-dependently reduced by 15d-PGJ 2 (Fig. 4A). Similarly, treatment with TRO prevented IL-1␤-mediated activation of the mPGES-1 promoter (Fig. 4B). Thus, PPAR␥ ligands suppress IL-1␤-induced mPGES-1 promoter activity, suggesting that PPAR␥ ligands exert their inhibitory effects on mPGES-1 expression through a transcriptional mechanism.
Suppression of mPGES-1 Expression by 15d-PGJ 2 and TRO Is Mediated by PPAR␥-PPAR␥ ligands were reported to exert their transcriptional effects through PPAR␥-depend- and were subsequently stimulated with IL-1␤ (100 pg/ml) for 18 h. Western blot analysis revealed that GW9662 dose-dependently relieved the suppressive effect of 15d-PGJ 2 (Fig.  6A) and TRO (Fig. 6B) on IL-1␤-induced mPGES-1 protein expression. GW9662 on its own had no significant effect on mPGES-1 expression (Fig. 6, last three lanes). As expected, the level of cPGES protein expression was not altered by these treatments (Fig. 6, lower panels). Taken together, these results suggest that 15d-PGJ 2 and TRO inhibit IL-1␤-induced mPGES-1 expression at the transcriptional level in a PPAR␥-dependent mechanism.
PPAR␥ Activation Inhibits Transcriptional Activation by Egr-1-The transcription factor Egr-1 had been shown to play a crucial role in the transcription of mPGES-1 (32, 33). Therefore, we investigated the effect of 15d-PGJ 2 and TRO on Egr-

FIG. 3. Effect of PPAR agonists on IL-1␤-induced mPGES-1 mRNA expression in HSFs.
Confluent HSFs were treated with increasing concentrations of 15d-PGJ 2 , TRO, or Wy14643 for 30 min before incubation in the absence or the presence of 100 pg/ml IL-1␤ for 12 h. Total RNA was isolated; cDNA was synthesized; and mPGES-1, cPGES, and glyceraldehyde-3-phosphate dehydrogenase mRNAs were quantified using real-time quantitative PCR. The results are expressed as -fold changes, considering 1 as the value of untreated cells. All experiments were performed in triplicate, and negative controls without template RNA were included in each experiment. The results are the mean Ϯ S.E. of three independent experiments. *, p Ͻ 0.05; compared with cells treated with IL-1␤ alone (control). Fig.  7, the activity of the mPGES-1 promoter was enhanced by cotransfection with a human Egr-1 expression plasmid (bar 1 versus bar 2). However, the activation of the mPGES-1 promoter by Egr-1 was significantly attenuated by cotransfection with the human PPAR␥ expression plasmid (bar 2 versus bar 3). The activity of mPGES-1 promoter was also reduced by either 15d-PGJ 2 (5 M) (bar 2 versus bar 4) or TRO (10 M) (bar 2 versus bar 5). Moreover, the suppressive effect of PPAR␥ was further enhanced in the presence of 15d-PGJ 2 (bar 3 versus bar 6) or TRO (bar 3 versus bar 7).

1-mediated activation of mPGES-1 promoter. As shown in
Next, we sought to confirm that the inhibition of Egr-1 transcriptional activity is essential in the suppression of mPGES-1 by PPAR␥. To this end, we analyzed the effect of PPAR␥ on the transcriptional activation of a synthetic luciferase reporter construct containing three tandem repeats of the putative Egr-1 binding sequence, pEgr-1 ϫ 3-TK-Luc (31). As shown in Fig.  7B, overexpression of the human Egr-1 cDNA induced a robust increase in the transcriptional activity of the above construct (bar 1 versus bar 2). This activation was attenuated by cotransfection with the human PPAR␥ expression plasmid (bar 2 ver-sus bar 3). 15d-PGJ 2 (bar 2 versus bar 4) and TRO (bar 2 versus bar 5) also reduced the transcriptional activity induced by Egr-1. Again, the suppressive effect of PPAR␥ was further enhanced in the presence of 15d-PGJ 2 (bar 3 versus bar 6) or TRO (bar 3 versus bar 7). Taken together, these data suggest that PPAR␥ activation inhibits mPGES-1 promoter activation by interfering with the Egr-1 transcriptional activity.
PPAR␥ Ligands Inhibit Egr-1 Binding Activity-Egr-1 has been shown to bind to the GC box of the mPGES-1 promoter (32,33). To establish whether IL-1␤ could induce binding of Egr-1 to the mPGES-1 promoter and whether this was altered by PPAR␥ ligands, we investigated the effect of 15d-PGJ 2 and TRO on the binding activity of Egr-1 using nuclear extracts from HSFs and a radiolabeled oligonucleotide corresponding to the Egr-1 binding sites in the mPGES-1 promoter. As shown in Fig. 8A, the binding of Egr-1 was strongly induced by IL-1␤ (lane 1 versus lane 2). When the cells were treated with 15d-PGJ 2 (lanes 3-5) or TRO (lanes 6 -8) the formation of the Egr-1-DNA complex decreased in a dose-dependent manner. This binding was specific, because it could be completely abolished by coincubation with a 100-fold molar excess of unlabeled probe (lane 9). Coincubation, with a 100-fold molar excess of the mutant probe, did not affect Egr-1 DNA-binding activity (lane 10). The specificity of this interaction was further observed by the supershift assays, showing a further retardation in the electrophoretic mobility of the Egr-1-DNA complex in the presence of a specific anti- Egr-1 antibody (lane 11). These results suggest that PPAR␥ ligands inhibit IL-1␤-induced mPGES-1 expression by reducing Egr-1 DNA-binding activity to the promoter sequence.
To determine whether the reduction of Egr-1 DNA-binding activity by PPAR␥ ligands in HSFs was due to inhibition of Egr-1 expression, we examined the effects of 15d-PGJ 2 and TRO on IL-1␤-induced Egr-1 protein expression. The cells were pretreated with increasing concentrations of 15d-PGJ 2 or TRO prior to stimulation with IL-1␤. In quiescent HSFs, protein levels of Egr-1 were very low. Treatment with IL-1␤ (100 pg/ml) caused a robust induction of Egr-1. Interestingly, neither 15d-PGJ 2 nor TRO altered IL-1␤-induced Egr-1 (Fig. 8B). These data suggest that 15d-PGJ 2 and TRO are not general inhibitors of IL-1␤-induced gene expression and that the inhibitory effect of PPAR␥ ligands on mPGES-1 expression does not involve inhibition of Egr-1 protein expression. DISCUSSION An expanding body of evidence indicates that PPAR␥ and its ligands play an important role in the regulation of multiple inflammatory processes (19, 24 -28). In the present study, we have extended these observations by showing that both natural and synthetic PPAR␥ ligands inhibit IL-1␤-induced mPGES-1 expression in HSFs. Furthermore, we elucidate the molecular mechanism underlying this effect. We demonstrate that this suppressive effect is transcriptional and PPAR␥-dependent. Moreover, PPAR␥ activation inhibited the transcriptional and DNA binding activities of Egr-1. Taken together, our results reveal a novel function of PPAR␥, further supporting its role in the control of inflammatory responses.
Several lines of evidence indicate that the inhibitory effects of 15d-PGJ 2 and TRO on IL-1␤-induced mPGES-1 expression are likely to act through PPAR␥ activation. First, treatment with the specific PPAR␣ activator, Wy14653, had no effect on IL-1␤-induced mPGES-1 expression. Second, overexpression of PPAR␥ suppressed transcriptional activation of the mPGES-1 promoter, which was enhanced by the addition of PPAR␥ ligands. Third, overexpression of a dominant negative form of PPAR␥ relieved the inhibitory effect of PPAR␥ ligands on the mPGES-1 promoter activation. Finally, pretreatment with an irreversible pharmacological PPAR␥ antagonist, GW9662, overcame the inhibitory effect of PPAR␥ ligands on mPGES-1 protein expression. However, inhibition of PPAR␥ (via GW9662 or a DN), almost completely restored the suppressive effect of TRO, whereas the suppressive effect of 15d-PGJ 2 was only partially restored, suggesting that 15d-PGJ 2 can activate other PPAR␥-independent signaling pathways to inhibit mPGES-1 expression. In this context, several studies reported that 15d-PGJ 2 inhibits many inflammatory responses by mechanisms that are independent of PPAR␥, such as the expression of iNOS in microglial cells and astrocytes (34), the ␤ 2 -integrin-dependent oxidative burst in human neutrophils (35) and the expression of CD95 ligand in T lymphocytes (36). In addition, it was demonstrated that 15d-PGJ 2 inhibits NF-B signaling at different levels, including modification of I-B kinase activity, which reduces the NF-Bp65 nuclear translocation, and by direct modification of the DNA binding domain of NF-Bp50 (37,38). Finally, Chawla et al. (39) examined inflammatory responses in macrophages derived from PPAR␥ Ϫ/Ϫ embryonic stem cells and reported that PPAR␥ ligands still repress LPSinduced iNOS and COX-2 expression. Elucidation of PPAR␥independent mechanisms of PPAR␥ ligands needs further study.
The transcriptional induction of mPGES-1 is controlled primarily by Egr-1 through two Egr-1 binding motifs identified in the proximal promoter region of the mPGES-1 region (32,33). We hypothesized that inhibition of Egr-1 activity by PPAR␥ could be the mechanism by which PPAR␥ exerts its repressive effect on mPGES-1 transcription. By using reporter gene assays, we found that Egr-1 indeed activated the mPGES-1 promoter, and this activation was reduced by cotransfection with an expression vector for PPAR␥. Moreover, 15d-PGJ 2 and TRO inhibited Egr-1-mediated mPGES-1 promoter activation and this inhibition was further enhanced in the presence of a PPAR␥ expression plasmid. PPAR␥ activation also inhibited Egr-1-induced activation of a synthetic luciferase reporter construct containing three tandem repeats of Egr-1 motif, suggesting that PPAR␥ inhibits Egr-1 transcriptional activity in a promoter-independent manner. This is the first evidence that PPAR␥ activation inhibits Egr-1 transcriptional activity in HSFs. In EMSA and supershift assays, we observed that PPAR␥ ligands reduced DNA binding of Egr-1 to a radiolabeled oligonucleotide corresponding to the Egr-1 binding sites in the mPGES-1 promoter. Altogether, these results strongly suggest that PPAR␥-mediated repression of mPGES-1 results from decreased Egr-1 binding activity.
Several mechanisms can explain the repression of Egr-1 activities by PPAR␥. One possibility is that PPAR␥ activation suppresses Egr-1 expression. Indeed, PPAR␥ ligands were reported to inhibit hypoxia-induced Egr-1 expression in mononuclear phagocytes (40). However, in our study, 15d-PGJ 2 and TRO had no effect on IL-1␤-induced Egr-1 expression in HSFs, suggesting that PPAR␥ ligands inhibit Egr-1 transcriptional and DNA binding activities in HSFs by distinct mechanisms. A second mechanism could be competition between PPAR␥ and Egr-1 for binding to response elements. This possibility is probably unlikely, because: (i) EMSA analysis showed no binding of PPAR␥ to an oligonucleotide corresponding to the Egr-1 binding sites in the mPGES-1 promoter (data not shown); (ii) the human mPGES-1 promoter construct used in this study contains no consensus PPRE sequence; and (iii) PPAR␥ activation inhibited Egr-1 transcriptional activity in a promoter-independent manner. Alternatively, PPAR␥ may inhibit Egr-1 activity by directly binding to Egr-1 and inhibiting its DNA binding and/or transcriptional activity. In this context, PPAR␥ has FIG. 8. PPAR␥ ligands inhibit DNA-binding activity of Egr-1. A, confluent HSFs were pretreated with increasing concentrations of 15d-PGJ 2 or TRO for 4 h, followed by the addition of IL-1␤ (100 pg/ml) for 1 h. Nuclear extracts (5 g) were incubated with a 32 P-labeled oligonucleotide containing the two Egr-1 binding sites of the mPGES-1 promoter. Specificity of binding was confirmed using 100-fold molar excess of unlabeled oligonucleotides containing wild type (wt) or mutated (mt) Egr-1 binding sites. Positions of Egr-1-DNA complex (Egr-1), nonspecific binding (NS), and supershifted band (SS) are indicated. A representative result of four independent experiments is shown. B, confluent HSF were treated with increasing concentrations of 15d-PGJ 2 or TRO or for 4 h, followed by the addition of IL-1␤ (100 pg/ml) for 1 h. Nuclear extracts were prepared and analyzed for Egr-1 protein by Western blotting. This blot is representative of similar results obtained from three independent experiments. been shown to inhibit NF-B, NF-AT, and SP-1 transcriptional activity through mechanisms that involve protein-protein interactions (26,41,42). Finally, PPAR␥ can attenuate Egr-1 activities by competing for general transcriptional coactivators such as CREB-binding protein (CBP/p300). CBP/p300 interacts with PPAR␥ and positively regulates PPAR␥-dependent gene transcription (43,44). Importantly, CBP/p300 also interacts with Egr-1 and enhances its transcriptional activity (45,46). Thus, the sequestering of limiting amounts of CBP/p300 by activated PPAR␥ could account for the transcriptional repressive effect of PPAR␥ ligands on Egr-1 activities and mPGES-1 transcription. This is corroborated by our finding that overexpression of a PPAR␥ mutant lacking transcriptional coactivator recruitment activity overcomes the inhibitory effect of PPAR␥ ligands on mPGES-1 promoter activity.
In addition to Egr-1, the mPGES-1 promoter contains binding sites for transcription factors (AP-1 and SP-1) (32,47) known to associate and/or to be down-regulated by PPAR␥ (24,25,42). Although the role of those elements in IL-1␤-induced mPGES-1 transcription is still unknown, we can not exclude the possibility that PPAR␥ interaction with these transcription factors may be involved in the repression of mPGES-1 expression. The ability of PPAR␥ to repress Egr-1-mediated transcription may be of relevance for other inflammatory genes. Indeed, and as stated above, PPAR␥ ligands were reported to inhibit the expression of IL-1␤, TNF-␣, IL-2, and several chemokines. Interestingly, Egr-1 activation is involved in the transcriptional activation of these genes (48 -50). Therefore, it is possible that PPAR␥-mediated Egr-1 repression may be part of the mechanisms by which PPAR␥ down-regulates these genes. Moreover, it was recently reported that Egr-1 positively regulates expression of PPAR␥. This is consistent with the suggestion that up-regulation of PPAR␥ expression by Egr-1 may constitute a negative feedback mechanism by which Egr-1 inhibits expression and/or signaling pathways of pro-inflammatory mediators.
Several studies demonstrated that PPAR␥ ligands attenuate inflammation in vivo in animal models of experimental allergic encephalomyelitis (51), inflammatory bowel disease (52), lupus nephritis (53), atherosclerosis (54), and arthritis (55,56). Thus, inhibition of mPGES-1 could be part of a mechanism by which PPAR␥ inhibits inflammatory responses in vivo. In addition to inflammation, increased expression of mPGES-1 may have important consequences in other pathological conditions. For instance, increased expression of mPGES-1 has been described in symptomatic atherosclerotic plaques and various carcinoma and cancer cell lines (14 -18). PPAR␥ ligands may therefore have clinical application not only in chronic inflammatory conditions but neoplastic diseases as well.
In conclusion, we show for the first time that PPAR␥ activation suppresses mPGES-1 expression via negative interference with Egr-1. This novel function of PPAR␥ ligands further supports the role of PPAR␥ in inflammation and suggests that the modulation of this gene expression by PPAR␥ ligands may constitute an additional therapeutic tool to take into account for the treatment and/or prevention of inflammatory and neoplastic diseases.