Effects of antirheumatic treatments on the prostaglandin E2 biosynthetic pathway.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 52, No. 11, November 2005, pp 3439–3447 DOI 10.1002/art.21390 © 2005, American College of Rheumatology Effects of Antirheumatic Treatments on the Prostaglandin E2 Biosynthetic Pathway Marina Korotkova,1 Marie Westman,1 Karina R. Gheorghe,2 Erik af Klint,1 Christina Trollmo,1 Ann Kristin Ulfgren,1 Lars Klareskog,1 and Per-Johan Jakobsson1 Objective. Microsomal prostaglandin E synthase 1 (mPGES-1) is up-regulated in experimental arthritis and markedly expressed in synovial tissue biopsy samples from patients with rheumatoid arthritis (RA). This study was carried out to determine the effects of tumor necrosis factor (TNF) blockers and glucocorticoids on mPGES-1 and cyclooxygenase (COX) expression, as well as biosynthesis of PGE2 in rheumatoid joints. Methods. In vitro effects of TNF blockers and dexamethasone on the PGE2 biosynthetic pathway were examined in RA synovial fluid mononuclear cells (SFMCs) by flow cytometry. PGE2 levels in culture supernatants were measured by enzyme immunoassay. Expression of enzymes responsible for PGE2 synthesis ex vivo was evaluated by immunohistochemistry in synovial biopsy samples obtained from 18 patients before and after treatment with TNF blockers and from 16 patients before and after intraarticular treatment with glucocorticoids. Double immunofluorescence was performed using antibodies against mPGES-1, COX-1, COX-2, and CD163. Results. Double immunofluorescence revealed that mPGES-1 and COX-2 were colocalized in SFMCs as well as in RA synovial tissue cells. The addition of either TNF blockers or dexamethasone suppressed lipopolysaccharide-induced mPGES-1 and COX-2 expression in synovial fluid monocyte/macrophages in vitro and decreased the production of PGE2. Intraarticular treatment with glucocorticoids significantly reduced both mPGES-1 and COX-2 expression in arthritic synovial tissue ex vivo. The number of COX-1– expressing cells in synovial tissue was also significantly decreased by glucocorticoid treatment. In contrast, neither mPGES-1 nor COX-2 expression in synovial tissue was significantly suppressed by anti-TNF therapy. Conclusion. These data are the first to demonstrate the effects of antirheumatic treatments on mPGES-1 expression in RA and suggest that the inhibition of PGE2 biosynthesis, preferably by targeting mPGES-1, might complement anti-TNF treatment for optimal antiinflammatory results in RA. Rheumatoid arthritis (RA) is a severe, chronic disease characterized by systemic and local inflammation leading to joint destruction and functional impairment. The proinflammatory cytokines tumor necrosis factor ␣ (TNF␣) and interleukin-1␤ (IL-1␤) play key roles in initiating and driving RA. These cytokines induce the production of prostaglandin E2 (PGE2), which contributes to several of the pathologic features of RA, such as pain, inflammation, and bone destruction (1–3). In the PGE2 biosynthetic pathway, cyclooxygenase 1 (COX-1) and COX-2 both catalyze the conversion of arachidonic acid into PGH2. Subsequently, terminal PGE synthases catalyze the formation of PGE2 from PGH2. Several PGE synthases have been cloned and characterized, including microsomal PGE synthase 1 (mPGES-1) (4,5), mPGES-2 (6), and cytosolic PGES (cPGES) (7). Importantly, mPGES-1 is strongly induced by proinflammatory stimuli and is functionally linked to Supported by grants from the King Gustaf V 80 Years Foundation, Alex and Eva Wallströms Foundation, Pfizer Inc., the Swedish Society of Medicine, Freemason Lodge “Barnhuset” in Stockholm, Börje Dahlin Foundation, the Swedish Medical Research Council, and the Swedish Rheumatism Association. 1 Marina Korotkova, MD, PhD, Marie Westman, BSc, Erik af Klint, MD, Christina Trollmo, PhD, Ann Kristen Ulfgren, PhD, Lars Klareskog, MD, PhD, Per-Johan Jakobsson, MD, PhD: Karolinska Institutet, Karolinska University Hospital, Stockholm, Sweden; 2Karina R. Gheorghe, MD: Karolinska Institutet, Stockholm, and Karolinska Institutet, Novum, Huddinge, Sweden. Dr. Klareskog has obtained research grants from and worked as a scientific advisor for Wyeth, Schering-Plough, and Abbott. Address correspondence and reprint requests to Marina Korotkova, MD, PhD, Rheumatology Research Laboratory, CMM L8-04, Karolinska University Hospital, Stockholm S-17176, Sweden. E-mail: Marina.Korotkova@cmm.ki.se. Submitted for publication February 9, 2005; accepted in revised form July 27, 2005. 3439 3440 COX-2 in various cells (8–10). It has also been reported to function with COX-1 (9,11). In contrast, mPGES-2 and cPGES are constitutively expressed and therefore are likely to be active during basal PGE2 production (6,7). Although cPGES has been predominantly linked to COX-1 (7), mPGES-2 efficiently couples with both COX-1 and COX-2 (6). Recent studies in rodents have provided convincing evidence for the important role of mPGES-1 in the pathogenesis of inflammatory arthritis (12–14). Furthermore, the induction of both mPGES-1 and COX-2 was observed in primary rheumatoid synovial cells after treatment with proinflammatory cytokines in vitro (10). We have recently demonstrated significant expression of mPGES-1 in synovial tissue biopsy samples from patients with RA (15). Since mPGES-1 represents the downstream and rate-limiting enzyme in the induced state of PGE2 biosynthesis, it may be a therapeutic target in RA. Conventional treatment of RA includes nonsteroidal antiinflammatory drugs (NSAIDs), glucocorticoids, and disease-modifying antirheumatic drugs (DMARDs). NSAIDs and glucocorticoids are well known to suppress PGE2 generation. NSAIDs specifically inhibit COX activity (16), and treatment with COX inhibitors effectively relieves pain and inflammation. However, these drugs cause serious adverse effects, such as gastrointestinal toxicity (17). Selective inhibition of COX-2 has fewer gastrointestinal side effects, but may lead to increased risk of thrombosis and cardiovascular disease (18,19). Treatment with glucocorticoids efficiently suppresses inflammation in RA, but is also associated with a number of adverse reactions (20). One of the antiinflammatory effects of glucocorticoids is associated with suppression of PGE2 biosynthesis by the inhibition of phospholipase A2 activity (21) and down-regulation of COX-2 expression (22). Marked suppression of COX-2 expression by dexamethasone was observed in freshly explanted RA synovial tissues and cultured synoviocytes (22). Recently, the suppressive effects of dexamethasone on mPGES-1 messenger RNA, protein expression, and enzyme activity in synovial fibroblasts have been demonstrated in vitro (10). Whether mPGES-1 might be an additional target for glucocorticoid antiinflammatory action in RA patients has not been reported. New biologic DMARDs, blockers of TNF, display high clinical efficacy and retard joint damage in patients with RA (23), although in most cases this treatment does not induce complete remission. AntiTNF therapy is based on TNF binding and suppression KOROTKOVA ET AL of inflammation by inhibiting its downstream effects. Anti-TNF therapy down-regulates the cytokine/ chemokine cascade (24–26), reduces inflammatory cell migration into the RA joint (26), and decreases serum matrix metalloproteinase levels (27,28). However, whether anti-TNF therapy can have an effect on the PGE2 biosynthetic pathway has not been investigated to date. In the present study, we examined the effects of TNF blockers (infliximab and etanercept) and glucocorticoids on the prostaglandin E2 biosynthetic pathway. We used both an in vitro system to study the expression of mPGES-1 and related enzymes in synovial fluid mononuclear cells (SFMCs) and an in vivo system in which we analyzed synovial tissue biopsy samples from RA patients obtained before and after therapy with TNF blockers or glucocorticoids. PATIENTS AND METHODS Patients and tissue samples. Eighteen patients who met the American College of Rheumatology (formerly, the American Rheumatism Association) diagnostic criteria for RA were recruited into the study (29). At study entry, all patients had a moderate to high level of disease activity (Disease Activity Score ⬎4.8) (30). In the first group, 8 patients (7 women and 1 man, median age 44 years, range 35–59 years) received a subcutaneous injection of 25 mg etanercept (Wyeth Europa, Maidenhead, UK) twice a week. In 3 patients, etanercept was combined with prednisolone, in 4 patients etanercept was combined with DMARDs and prednisolone, and in 1 patient etanercept was combined with DMARDs alone. All patients except 1 received NSAIDs. None of these 8 patients received intraarticular steroid therapy. Synovial tissue biopsy specimens were obtained during arthroscopy from patients before and a median of 8 weeks (range 7–10 weeks) after initiation of treatment with etanercept. In the second group, 10 patients (7 women and 3 men, median age 55 years, range 25–74 years) received intravenous infusions of infliximab (Schering-Plough, Stockholm, Sweden) at a dose of 3 mg/kg at 0, 2, and 6 weeks. All patients received infliximab in combination with DMARDs, and all patients except 1 were treated with NSAIDs. Five patients also received prednisolone. None of these 10 patients received intraarticular steroid therapy. Synovial tissue biopsy specimens were obtained during arthroscopy from patients before and a median of 10 weeks (range 8–16 weeks) after initiation of treatment with infliximab. Sixteen patients with inflammatory arthritis (8 women and 8 men, median age 41 years, range 20–73 years) received intraarticular injections of 40 mg triamcinolone hexacetonide (Lederspan; Wyeth Lederle, Solna, Sweden). Patients with RA (7 patients) and other inflammatory arthritides (3 patients with monarthritis, 2 with juvenile idiopathic arthritis, 1 with psoriatic arthritis, 1 with spondylarthritis, 1 with polyarthritis, and 1 with oligoarthritis) showed clinical signs of active arthritis, including swelling and pain in the joint, which were reduced in mPGES-1 AND ANTIRHEUMATIC TREATMENT all patients after intraarticular corticosteroid treatment. Nine patients received intraarticular corticosteroids alone, 3 patients received intraarticular corticosteroids combined with DMARDs, 1 patient received intraarticular corticosteroids combined with DMARDs and prednisolone, and 3 patients received intraarticular corticosteroids combined with prednisolone. All patients except 3 received NSAIDs. None of these 16 patients received TNF blockers. Synovial biopsy specimens were obtained during arthroscopy before and 9–12 days after the intraarticular injection. Synovial fluid was collected by aspiration from an additional 8 patients with active RA (4 women and 4 men, median age 55 years, range 38–65 years). All patients received DMARDs, 3 in combination with TNF blockers and prednisolone, and 1 in combination with TNF blockers. All patients except 2 received NSAIDs. This study was approved by the Ethics Committee at the Karolinska University Hospital, Solna, Stockholm, Sweden. Cell isolation. SFMCs were isolated using a discontinuous density gradient (Ficoll-Paque; Pharmacia, Uppsala, Sweden). Cells were cultured in RPMI 1640 medium supplemented with 100 units/ml penicillin–streptomycin, 2 mM glutamine, 10 mM HEPES (all from Gibco Invitrogen, Lidingo, Sweden), and 5% human pooled serum (blood bank, Karolinska Hospital, Solna, Stockholm, Sweden) in the presence of 100 ng/ml lipopolysaccharide (LPS; Sigma, St. Louis, MO) at 37°C in a humidified atmosphere containing 5% CO2, for 42 hours. Where indicated, the cultures were treated using 10⫺6–10⫺9M dexamethasone (Sigma), 0.1–100 g/ml infliximab, or 0.1–100 g/ml etanercept. Preliminary experiments have shown that these treatments affect mPGES-1 expression in a dose-dependent manner, and maximal effects were observed at concentrations of 10⫺6M dexamethasone, 100 g/ml etanercept, and 100 g/ml infliximab. Unstimulated control cells were always cultured in parallel. Culture supernatants were harvested and stored at ⫺70°C. For immunostaining, SFMCs were cultured in chamber slides (Nalge Nunc International, Naperville, IL) in the presence of 100 ng/ml LPS, fixed with 2% formaldehyde (Sigma) for 20 minutes, and stored at ⫺70°C. Flow cytometric analysis. Staining of SFMCs for flow cytometry was performed after blocking of nonspecific binding with phosphate buffered saline (PBS) supplemented with 5% human serum. Cells were incubated with peridin chlorophyll protein–conjugated mouse monoclonal anti-CD14 antibody (Becton Dickinson, San Jose, CA) for 20 minutes at 4°C, washed with PBS, and fixed with 4% paraformaldehyde (Sigma) for 10 minutes. For intracellular staining, cells were permeabilized using PBS supplemented with 0.1% saponin (Reidel de Haen, Seelze, Germany) and 1% bovine serum albumin (Sigma) for 10 minutes. For intracellular staining of COX-1 and COX-2, cells were incubated with mouse monoclonal fluorescein isothiocyanate–conjugated anti–COX-1 and phycoerythrin-conjugated anti–COX-2 antibodies (Becton Dickinson) or isotype-matched irrelevant antibodies, respectively, for 45 minutes. For intracellular staining of mPGES-1, cells were incubated with rabbit polyclonal antiserum raised against purified human mPGES-1 (15) or isotype-matched irrelevant control for 45 minutes, then with allophycocyaninconjugated goat anti-rabbit IgG (heavy and light chain; 3441 Molecular Probes, Eugene, OR) for 30 minutes. Analyses were performed using a FACSCalibur (Becton Dickinson) and CellQuest software (Becton Dickinson). Scatter properties were used to identify the monocyte population. Quadrants were set on the respective isotype controls. Results are expressed as a percentage of the total number of gated monocytes expressing CD14 and producing the respective enzymes. Measurement of PGE2. The ability of SFMCs to produce PGE2 was assessed in the supernatants of control and LPS-stimulated cells cultured for 42 hours with or without inhibitors. In addition, the conversion of exogenous arachidonic acid (Nu-Check Prep, Elysian, MN) to PGE2 by SFMCs, cultured under the conditions described above, was assessed in 4 patients. For this purpose, cultured SFMCs were harvested, washed twice in calcium-free PBS, assessed for viability, and counted by trypan blue dye exclusion. Cells were then resuspended in PBS supplemented with Ca2⫹ and Mg2⫹. Arachidonic acid was added to a final concentration of 20 M and cells were stimulated with 2 M calcium ionophore A23187 (Calbiochem-Novabiochem, San Diego, CA) for 30 minutes at 37°C. The incubation was terminated by adding 1M HCl to attain a pH of 3. After centrifugation, the supernatants were collected and stored at ⫺80°C until analyzed. In a preliminary experiment, the supernatants were analyzed using a reverse-phase high-performance liquid chromatography (RP-HPLC) system with ultraviolet detection at 195 nm and online radioactivity detection using a Radiomatic Flo-One beta detector (Packard, Downers Grove, IL). The mobile phase was water, acetonitrile, and trifluoroacetic acid (70:30:0.007, by volume), and the eicosanoid compounds were separated using a Nova-Pak C-18 column (Waters, Milford, MA). Prostaglandin products were identified by comparison with the retention time of synthetic standards (Cayman Chemical, Ann Arbor, MI). Aliquots of the supernatants were measured using a PGE2 enzyme immunoassay (EIA) kit (Cayman Chemical). PGE2 concentrations were assayed in duplicate or triplicate and were read against a standard curve. The results are expressed in ng per 1 ⫻ 106 cells. Immunohistochemical analysis. Tissue samples were snap frozen in prechilled isopentane and stored at ⫺70°C until sectioned. Serial cryostat sections (8 m) were fixed with 2% formaldehyde (Sigma) for 20 minutes and stored at ⫺70°C. Sections were incubated with rabbit polyclonal antiserum raised against purified human mPGES-1 (15) or with the following antibodies (all purchased from Cayman Chemical): rabbit polyclonal anti-cPGES, rabbit polyclonal anti–COX-1, and mouse monoclonal anti–COX-2. The staining procedure has been described previously (31). Negative control experiments were performed using isotype-matched irrelevant antibodies and omitting the primary antibodies. Staining in synovial tissue was completely abolished by preincubation of anti– mPGES-1 serum with mPGES-1 protein and preincubation of commercial antibodies with respective blocking peptides (Cayman Chemical), which confirmed the specificity of the staining. Stained sections were examined using a Polyvar II microscope (Reichert-Jung, Vienna, Austria) and photographed with a digital camera (300F; Leica, Cambridge, UK). Positive staining was indicated as brown deposits. The positive staining was assessed quantitatively using computer-assisted image analysis and expressed as percentages of the total area of counterstained tissue. Double immunofluorescence staining 3442 KOROTKOVA ET AL was performed using anti-human mPGES-1 antiserum and mouse monoclonal anti-human COX-2 (Cayman Chemical), anti-human COX-1 (Wako, Neuss, Germany), or anti-human CD163 (BerMac3 clone; Dako, Glostrup, Denmark) antibodies. The staining procedure has been described previously (15). Briefly, after blocking with an avidin–biotin kit (Vector, Peterborough, UK), the sections were incubated overnight with primary antibodies. Thereafter, the sections were incubated with biotinylated goat anti-rabbit IgG (heavy and light chain; Vector), followed by incubation with streptavidin-conjugated fluorophore Alexa Fluor 488 (Molecular Probes, Leiden, The Netherlands). After blocking using the avidin–biotin kit, the sections were incubated with biotinylated horse anti-mouse IgG (heavy and light chain; Vector), followed by incubation with streptavidin-conjugated fluorophore Alexa Fluor 546 (Molecular Probes). Some sections were also incubated with 4⬘,6-diamidino-2-phenylindole (KPL, Gaithersburg, MD) for 5 minutes. Statistical analysis. Data were analyzed by Friedman’s test for repeated measures, followed by Wilcoxon’s signed rank test and Bonferroni correction for multiple comparisons. P values less than 0.05 were considered significant. RESULTS Induction of mPGES-1 in SFMCs. SFMCs were isolated and cultured in chamber slides in the presence of LPS. LPS-induced expression of mPGES-1 was detected in SFMCs with macrophage-like morphology (Figure 1A). Double immunofluorescence revealed the expression of mPGES-1 in cells expressing CD163, a member of the scavenger receptor family, which is highly specific for cells of the mononuclear phagocyte lineage (Figure 1B). In CD163⫹ SFMCs, no expression of mPGES-1 was evident (Figure 1C). Expression of mPGES-1 in synovial fluid monocytes was also evaluated by flow cytometric analysis (Figure 2). In monocytes that had not been exposed to LPS, mPGES staining was detected at a low level, while stimulation with LPS substantially increased mPGES-1 expression in CD14⫹ monocytes (Figure 2A). LPSinduced mPGES-1 expression in CD14⫹ monocytes was almost completely suppressed by dexamethasone. In addition, both infliximab and etanercept reduced the percentage of CD14⫹ monocytes that expressed mPGES-1 (Figure 2A). In order to clarify a functional coupling between mPGES-1 and COX-1 or COX-2, we analyzed the expression of these enzymes in synovial fluid monocytes. Figure 2B shows coordinated upregulation of mPGES-1 and COX-2 in CD14⫹ monocytes by LPS and suppression by dexamethasone. In addition, the percentage of CD14⫹ monocytes stained for mPGES-1, as well as for COX-2, was similarly decreased after treatment with infliximab or etanercept. Figure 1. Induction of microsomal prostaglandin E synthase 1 (mPGES-1) in synovial fluid mononuclear cells (SFMCs) by lipopolysaccharide. A, Green immunofluorescence staining of mPGES-1– positive SFMCs (original magnification ⫻ 250). B and C, Double fluorescence staining, showing mPGES-1–positive (green), CD163⫹ (red), and double-stained (yellow) SFMCs (4⬘,6-diamidino-2phenylindole–counterstained, original magnification ⫻ 500). D, Double fluorescence staining, showing mPGES-1–positive (green), cyclooxygenase 2 (COX-2)–positive (red), and double-stained (yellow) SFMCs (original magnification ⫻ 500). In contrast, the percentage of CD14⫹ monocytes that expressed COX-1 was not increased by LPS stimulation or suppressed by treatment with dexamethasone or TNF blockers. Double immunofluorescence confirmed coexpression of mPGES-1 and COX-2 in SFMCs stimulated with LPS (Figure 1D). PGE2 generation by SFMCs. As demonstrated, 42 hours after LPS treatment both mPGES-1 and COX-2 were induced in SFMCs. PGE2 accumulation in the supernatant at this point was enhanced in LPStreated cells (mean ⫾ SEM 8.6 ⫾ 2.7 ng/106 cells) (n ⫽ 8 patients) compared with control cells, which mainly expressed COX-1 (mean ⫾ SEM 0.7 ⫾ 0.2 ng/106 cells). Treatment with dexamethasone or TNF blockers prevented the accumulation of PGE2 (Figure 3A). In comparison, Figure 3B shows the results after incubation of these cells, treated as above, with exogenous arachidonic acid and calcium ionophore for 30 minutes. In this short time frame, the pattern of PGE2 release was similar to that for endogenous production and accumulation of PGE2 in culture medium after 42 hours (Figure 3A). Metabolic profiling was also performed by incubation of SFMCs with exogenous 14C-labeled arachidonic acid, followed by RP-HPLC analyses with online radioactivity detection. In LPS-treated cells, the predominant radio- mPGES-1 AND ANTIRHEUMATIC TREATMENT 3443 Figure 2. Coordinate expression of mPGES-1 and COX-2 in the monocyte/macrophage population of SFMCs in patients with rheumatoid arthritis. A, Flow cytometric analysis revealed the expression of mPGES-1 in CD14⫹ monocytes a, without stimulation and b–f, after stimulation with lipopolysaccharide (LPS) and treatment with dexamethasone (c), infliximab (d), or etanercept (e), compared with irrelevant antibody control (f). Results are expressed as the percentage of the total number of gated monocytes expressing CD14 and producing mPGES-1. B, Flow cytometric analysis results indicating the effects of treatment with dexamethasone (Dex), infliximab (Inf), and etanercept (Eta) on enzyme expression in the LPS-induced monocyte/ macrophage population of SFMCs. Results are expressed as the mean and SEM percentage of the total number of gated monocytes expressing CD14 and producing the respective enzymes: a, mPGES-1 (n ⫽ 8), b, COX-2 (n ⫽ 4), and c, COX-1 (n ⫽ 4). See Figure 1 for other definitions. active product (Figure 3B, part C) eluted at the time corresponding to PGE2 standard (Figure 3B, part A), while control cells produced significantly less PGE2 (Figure 3B, part B). These data are consistent with the EIA results. Expression of mPGES-1 and COX in synovial tissue. COX-1–positive staining was observed in the synovial lining layer cells and in the synovial sublining mononuclear and fibroblast-like cells (Figure 4A). Strong mPGES-1 staining was detected in synovial lining cells and in scattered macrophage- and fibroblast-like cells in the synovial sublining (Figure 4B). The distribution pattern of COX-2–positive cells in the synovial lining layer and in the synovial sublining (Figure 4C) was similar to that of mPGES-1–positive cells (Figure 4D). However, mPGES-1 staining was observed in vessel endothelial cells in only some biopsy specimens (5 of 16), while COX-2 staining was detected in endothelial cells in the majority of synovial specimens (15 of 16). As demonstrated with double immunofluorescence, intensive mPGES-1 and COX-2 staining colocalized in a considerable number of cells in the synovial lining layer (Figure 5A) and in the synovial sublining (Figure 5B). In contrast, double immunofluorescence showed no increased staining for COX-1 in synovial cells expressing mPGES-1 (Figures 5C and D). 3444 Figure 3. Prostaglandin E2 (PGE2) generation in synovial fluid mononuclear cell (SFMC) cultures stimulated with lipopolysaccharide (LPS) and treated with dexamethasone (Dex), infliximab (Inf), or etanercept (Eta) compared with unstimulated cells. A, Accumulation of PGE2 over the entire treatment period. Values are the mean and SEM (n ⫽ 8). B, Conversion of exogenous arachidonic acid to PGE2 by SFMCs, cultured under the conditions described above, washed, and incubated with arachidonic acid and calcium ionophore A23187 for 30 minutes. Values are the mean and SEM (n ⫽ 4). The boxed area shows the prostaglandin profile for the control and LPS-stimulated SFMCs incubated with 14C-labeled arachidonic acid. The chromatogram at 195 nm represents the retention times for PGF2␣, PGE2, and PGD2 standards (marked I, II, and III, respectively) (A) and the detection of radioactive prostaglandin products formed in control and LPSstimulated SFMCs (B and C). Effects of anti-TNF therapy on mPGES-1 and COX-2 expression in synovial tissue. COX-2– and mPGES-1–positive cells were present in the synovial tissue biopsy samples from all patients, although signif- Figure 4. Brown immunoperoxidase staining of A, COX-1, B and D, mPGES–positive cells, and C, COX-2 in representative synovial tissue sections from patients with rheumatoid arthritis. See Figure 1 for definitions. (Hematoxylin counterstained; original magnification ⫻ 250.) KOROTKOVA ET AL Figure 5. Coexpression of mPGES-1 and COX-2 in synovial tissue sections from patients with rheumatoid arthritis. A and B, Double immunofluorescence staining shows mPGES-1–positive (green), COX-2–positive (red), and double-stained (yellow) cells in A, the synovial lining layer and B, the synovial sublining. C and D, There was no increase in COX-1 staining (red) in mPGES-1–positive (green) cells in C, the synovial lining layer or D, the synovial sublining. See Figure 1 for definitions. (Original magnification ⫻ 250 in A and C; ⫻ 500 in B and D.) icant interindividual variations in the extent of the staining were observed. After therapy with etanercept, the mPGES-1 staining was higher in 2 patients, lower in 4 patients, and unchanged in 2 patients. The COX-2– positive area was increased in 2 patients, decreased in 4 patients, and unchanged in 2 patients (Figure 6A). After treatment with infliximab, the expression of mPGES in synovial tissue was unchanged in 2 patients, increased in 4 patients, and decreased in 4 patients. COX-2 expression in synovial tissue after treatment with infliximab was lower in 1 patient, higher in 5 patients, and unchanged in 4 patients (Figure 6B). The differences in the expression of mPGES-1 as well as COX-2 in synovial tissue biopsy specimens from RA patients before and after treatment with TNF blockers were not significant. Effects of intraarticular steroid therapy on the expression of mPGES-1 and related enzymes in synovial tissue. After treatment with locally administered steroids, the expression of mPGES-1 in synovial tissue was significantly reduced (P ⬍ 0.005), while the expression of housekeeping cPGES was not significantly changed (Figure 6C). The area stained positive for COX-1 and COX-2 in synovial tissue was significantly lower after treatment with intraarticular steroids (P ⬍ 0.05) (Figure 6C). mPGES-1 AND ANTIRHEUMATIC TREATMENT Figure 6. Expression of microsomal prostaglandin synthase 1 (mPGES-1) and related enzymes in synovial tissue of patients with rheumatoid arthritis before and after therapy with A, etanercept, B, infliximab, or C, local intraarticular steroids. Values are expressed as the mean and SEM percentage of the total area of counterstained tissue. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.005. COX-2 ⫽ cyclooxygenase 2; cPGES ⫽ cytosolic PGES. DISCUSSION In patients with RA, PGE2 levels in synovial fluid are markedly elevated (32,33), and both activated synovial tissue cells and recruited synovial fluid cells might contribute to the release of PGE2 into the synovial fluid. We have previously reported that mPGES-1 is strongly expressed in synovial tissue from patients with RA, specifically in synovial macrophages and fibroblasts (15). Here we analyzed the expression and codistribution of mPGES-1 and cyclooxygenases (COX-1 and COX-2) in both synovial fluid cells and synovial tissues from patients with RA. We also studied the effects of different antirheumatic therapies (TNF blockade and locally administered steroids) on mPGES-1 and cyclooxygenase expression in these patients in vitro and in vivo. Expression of mPGES-1 and COX-2 in SFMCs was low under basal conditions and substantially increased after 42 hours of in vitro culture in the presence of LPS (Figures 1A and 2B). Double immunofluorescence staining revealed that mPGES-1 was expressed in the cells of the mononuclear phagocyte lineage (Figure 1B). Flow cytometric analysis demonstrated that COX-2 was induced by LPS 3445 in ⬃35% of the monocytes, whereas 25% expressed mPGES-1. Approximately 25–30% of the monocytes under these culture conditions coexpressed COX-2 and mPGES-1. In comparison, COX-1 was expressed in 15–20% of these cells with or without LPS activation. Treatment of synovial fluid cells with either TNF blockers or dexamethasone significantly reduced the number of cells expressing mPGES-1 and COX-2, but not COX-1. In accordance with this, a strong correlation of PGE2 formation with the expression profiles of COX-2 and mPGES-1 was observed. Under these experimental conditions, the induction of PGE2 release by LPS seems to be mediated through a TNF-dependent pathway. LPS is known to increase the release of both TNF␣ and IL-1␤ by monocytes (34), and both cytokines are known to induce COX-2 and mPGES-1 in synovial fibroblasts and chondrocytes (10,35). Thus, our results suggest that TNF blockers also prevent the production of IL-1␤, as has been previously demonstrated (36). In RA synovial tissue biopsies, up-regulation of COX-2 was demonstrated in infiltrating mononuclear cells, vascular endothelial cells, synovial lining cells, and sublining fibroblast-like cells (22,37). Although COX-1 expression has also been detected in synovial tissue from patients with RA (22,37), the role of COX-1 in RA has not been elucidated. Recently, we have shown strong mPGES-1 staining in RA synovial lining cells, in sublining macrophage- and fibroblast-like cells, and, in a few patients, in vascular endothelial cells (15). In the present study, we found that the distribution pattern of mPGES1–positive cells in RA synovial tissue largely reflected the distribution of COX-2–positive cells. Moreover, using double immunofluorescence, we have directly shown for the first time the colocalization of mPGES-1 and COX-2 in RA synovial lining layer and sublining cells. The observed coexpression of these 2 enzymes in RA synovial tissue cells could account for the significant increase of PGE2 formation, which contributes to inflammatory and destructive processes. Endothelial cells may also contribute to PGE2 biosynthesis under proinflammatory conditions. For example, endothelial cells of the blood–brain barrier in vivo, as well as human umbilical vein endothelial cells and human rheumatoid synovium microvessel endothelium in vitro, have been demonstrated to produce PGE2 after induction with IL-1␤ (38–40). Consistent with this, we observed endothelial cell staining of mPGES-1 in approximately one-third of the study patients, and this finding might be important in, for example, inflammatory angiogenesis. However, the significance of endothelial mPGES-1 expression is presently unclear, and many 3446 factors could be involved in explaining the result, such as treatments, disease stage, etc. Endothelial cells with the capacity to produce PGE2 are likely to constitute an additional cell target for treatment of inflammatory diseases and fever. Intraarticular treatment with glucocorticoids efficiently reduces clinical signs of active arthritis, such as swelling and joint pain. Consistent with our in vitro results discussed above, locally administered steroids significantly decreased both mPGES-1 and COX-2 expression in synovial tissue from patients with inflammatory arthritis, including RA (Figure 6C). This suggests that mPGES-1 might be an additional target for glucocorticoid antiinflammatory action in RA. Intriguingly, COX-1 staining was also significantly down-regulated after glucocorticoid treatment. Although COX-1 is generally considered a constitutively expressed enzyme involved in cell homeostasis, recent studies have shown that it might be induced by different stimuli, for instance by vascular endothelial growth factor (VEGF) in endothelial cells (41,42) and by IL-1␣ in synovial fibroblasts (43). VEGF accumulates in SF and is expressed in RA synovial tissue (44). A suppressive effect of glucocorticoids on VEGF expression has been shown in rheumatoid synovial cells (45). Thus, glucocorticoids might affect COX-1 expression in RA synovial tissue indirectly by suppression of VEGF or other unknown factors. Although treatment of RA with TNF blockers leads to strong relief of pain and inflammation by a number of mechanisms, these drugs do not induce complete remission. Interestingly, in contrast to our in vitro data, analysis of synovial tissue from RA patients before and after treatment with etanercept or infliximab revealed that expression of COX-2 and mPGES-1 was not significantly down-regulated by TNF blockade (Figures 6A and B). This suggests that either the TNF blocking effects are not optimal in vivo or, more likely, that other mechanisms operate in sustaining the inflammation independently of TNF. For instance, it has been shown that short-term therapy with infliximab reduced synovial TNF␣ expression in RA patients, while expression of IL-1␤ was not significantly changed (25). Moreover, in mice with experimental arthritis, synovial inflammation was reduced by anti-TNF treatment and almost completely blocked by a combination of antiTNF and anti–IL-1 treatments (46). In primary cultures of synovial cells, IL-1␤ seems to be more potent in inducing mPGES-1 compared with TNF␣ (10). Thus, TNF blocking agents may not prevent the IL-1␤– dependent induction of mPGES-1 and PGE2 release, KOROTKOVA ET AL which explains the limited therapeutic effects of TNF blockade in RA. In conclusion, our results demonstrate that local treatments with glucocorticoids led to a significant down-regulation of mPGES-1 as well as COX-1 and COX-2. In contrast, systemic anti-TNF therapy was found to be insufficient to suppress the PGE2 pathway in vivo, albeit not in vitro. Thus, the development of specific mPGES-1 inhibitors is encouraging since they might complement TNF blockers for optimal antiinflammatory results in RA. ACKNOWLEDGMENT The authors would like to thank Associate Professor R. A. Harris for linguistic advice. REFERENCES 1. Portanova JP, Zhang Y, Anderson GD, Hauser SD, Masferrer JL, Seibert K, et al. Selective neutralization of prostaglandin E2 blocks inflammation, hyperalgesia, and interleukin 6 production in vivo. J Exp Med 1996;184:883–91. 2. Ben-Av P, Crofford LJ, Wilder RL, Hla T. Induction of vascular endothelial growth factor expression in synovial fibroblasts by prostaglandin E and interleukin-1: a potential mechanism for inflammatory angiogenesis. FEBS Lett 1995;372:83–7. 3. Robinson DR, McGuire MB, Levine L. Prostaglandins in the rheumatic diseases. Ann N Y Acad Sci 1975;256:318–29. 4. Jakobsson PJ, Thoren S, Morgenstern R, Samuelsson B. Identification of human prostaglandin E synthase: a microsomal, glutathione-dependent, inducible enzyme, constituting a potential novel drug target. Proc Natl Acad Sci U S A 1999;96:7220–5. 5. Thoren S, Weinander R, Saha S, Jegerschold C, Pettersson PL, Samuelsson B, et al. Human microsomal prostaglandin E synthase-1: purification, functional characterization, and projection structure determination. J Biol Chem 2003;278:22199–209. 6. Murakami M, Nakashima K, Kamei D, Masuda S, Ishikawa Y, Ishii T, et al. Cellular prostaglandin E2 production by membranebound prostaglandin E synthase-2 via both cyclooxygenases-1 and -2. J Biol Chem 2003;278:37937–47. 7. Tanioka T, Nakatani Y, Semmyo N, Murakami M, Kudo I. Molecular identification of cytosolic prostaglandin E2 synthase that is functionally coupled with cyclooxygenase-1 in immediate prostaglandin E2 biosynthesis. J Biol Chem 2000;275:32775–82. 8. Thoren S, Jakobsson PJ. Coordinate up- and down-regulation of glutathione-dependent prostaglandin E synthase and cyclooxygenase-2 in A549 cells: inhibition by NS-398 and leukotriene C4. Eur J Biochem 2000;267:6428–34. 9. Murakami M, Naraba H, Tanioka T, Semmyo N, Nakatani Y, Kojima F, et al. Regulation of prostaglandin E2 biosynthesis by inducible membrane-associated prostaglandin E2 synthase that acts in concert with cyclooxygenase-2. J Biol Chem 2000;275: 32783–92. 10. Stichtenoth DO, Thoren S, Bian H, Peters-Golden M, Jakobsson PJ, Crofford LJ. Microsomal prostaglandin E synthase is regulated by proinflammatory cytokines and glucocorticoids in primary rheumatoid synovial cells. J Immunol 2001;167:469–74. 11. Dieter P, Scheibe R, Jakobsson PJ, Watanabe K, Kolada A, Kamionka S. Functional coupling of cyclooxygenase 1 and 2 to discrete prostanoid synthases in liver macrophages. Biochem Biophys Res Commun 2000;276:488–92. mPGES-1 AND ANTIRHEUMATIC TREATMENT 12. Claveau D, Sirinyan M, Guay J, Gordon R, Chan CC, Bureau Y, et al. Microsomal prostaglandin E synthase-1 is a major terminal synthase that is selectively up-regulated during cyclooxygenase-2dependent prostaglandin E2 production in the rat adjuvantinduced arthritis model. J Immunol 2003;170:4738–44. 13. Trebino CE, Stock JL, Gibbons CP, Naiman BM, Wachtmann TS, Umland JP, et al. Impaired inflammatory and pain responses in mice lacking an inducible prostaglandin E synthase. Proc Natl Acad Sci U S A 2003;100:9044–9. 14. Kamei D, Yamakawa K, Takegoshi Y, Mikami-Nakanishi M, Nakatani Y, Oh-Ishi S, et al. Reduced pain hypersensitivity and inflammation in mice lacking microsomal prostaglandin E synthase-1. J Biol Chem 2004;279:33684–95. 15. Westman M, Korotkova M, af Klint E, Stark A, Audoly LP, Klareskog L, et al. Expression of microsomal prostaglandin E synthase 1 in rheumatoid arthritis synovium. Arthritis Rheum 2004;50:1774–80. 16. Vane JR. Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nat New Biol 1971;231:232–5. 17. Crofford LJ. Clinical experience with specific COX-2 inhibitors in arthritis [review]. Curr Pharm Des 2000;6:1725–36. 18. Stichtenoth DO, Frolich JC. The second generation of COX-2 inhibitors: what advantages do the newest offer? [review]. Drugs 2003;63:33–45. 19. Juni P, Nartey L, Reichenbach S, Sterchi R, Dieppe PA, Egger M. Risk of cardiovascular events and rofecoxib: cumulative meta-analysis. Lancet 2004;364:2021–9. 20. Lundberg IE, Grundtman C, Larsson E, Klareskog L. Corticosteroids: from an idea to clinical use [review]. Best Pract Res Clin Rheumatol 2004;18:7–19. 21. Parente L, Solito E. Annexin 1: more than an anti-phospholipase protein [review]. Inflamm Res 2004;53:125–32. 22. Crofford LJ, Wilder RL, Ristimaki AP, Sano H, Remmers EF, Epps HR, et al. Cyclooxygenase-1 and -2 expression in rheumatoid synovial tissues: effects of interleukin-1␤, phorbol ester, and corticosteroids. J Clin Invest 1994;93:1095–101. 23. Klareskog L, van der Heijde D, de Jager JP, Gough A, Kalden J, Malaise M, et al. Therapeutic effect of the combination of etanercept and methotrexate compared with each treatment alone in patients with rheumatoid arthritis: double-blind randomised controlled trial. Lancet 2004;363:675–81. 24. Charles P, Elliott MJ, Davis D, Potter A, Kalden JR, Antoni C, et al. Regulation of cytokines, cytokine inhibitors, and acute-phase proteins following anti-TNF-␣ therapy in rheumatoid arthritis. J Immunol 1999;163:1521–8. 25. Ulfgren AK, Andersson U, Engstrom M, Klareskog L, Maini RN, Taylor PC. Systemic anti–tumor necrosis factor ␣ therapy in rheumatoid arthritis down-regulates synovial tumor necrosis factor ␣ synthesis. Arthritis Rheum 2000;43:2391–6. 26. Taylor PC, Peters AM, Paleolog E, Chapman PT, Elliott MJ, McCloskey R, et al. Reduction of chemokine levels and leukocyte traffic to joints by tumor necrosis factor ␣ blockade in patients with rheumatoid arthritis. Arthritis Rheum 2000;43:38–47. 27. Brennan FM, Browne KA, Green PA, Jaspar JM, Maini RN, Feldmann M. Reduction of serum matrix metalloproteinase 1 and matrix metalloproteinase 3 in rheumatoid arthritis patients following anti-tumour necrosis factor-␣ (cA2) therapy. Br J Rheumatol 1997;36:643–50. 28. Catrina AI, Lampa J, Ernestam S, af Klint E, Bratt J, Klareskog L, et al. Anti-tumour necrosis factor (TNF)-␣ therapy (etanercept) down-regulates serum matrix metalloproteinase (MMP)-3 and MMP-1 in rheumatoid arthritis. Rheumatology (Oxford) 2002;41: 484–9. 29. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24. 3447 30. Van der Heijde DM, van ’t Hof MA, van Riel PL, Theunisse LM, Lubberts EW, van Leeuwen MA, et al. Judging disease activity in clinical practice in rheumatoid arthritis: first step in the development of a disease activity score. Ann Rheum Dis 1990;49:916–20. 31. Ulfgren AK, Lindblad S, Klareskog L, Andersson J, Andersson U. Detection of cytokine producing cells in the synovial membrane from patients with rheumatoid arthritis. Ann Rheum Dis 1995;54: 654–61. 32. Punzi L, Schiavon F, Cavasin F, Ramonda R, Gambari PF, Todesco S. The influence of intra-articular hyaluronic acid on PGE2 and cAMP of synovial fluid. Clin Exp Rheumatol 1989;7: 247–50. 33. Hishinuma T, Nakamura H, Sawai T, Uzuki M, Itabash Y, Mizugaki M. Microdetermination of prostaglandin E2 in joint fluid in rheumatoid arthritis patients using gas chromatography/ selected ion monitoring. Prostaglandins Other Lipid Mediat 1999; 58:179–86. 34. Bonder CS, Davies KV, Liu X, Hertzog PJ, Woodcock JM, Finlay-Jones JJ, et al. Endogenous interferon-␣ production by differentiating human monocytes regulates expression and function of the IL-2/IL-4 receptor ␥ chain. Cytokine 2002;17:187–96. 35. Kojima F, Naraba H, Miyamoto S, Beppu M, Aoki H, Kawai S. Membrane-associated prostaglandin E synthase-1 is upregulated by proinflammatory cytokines in chondrocytes from patients with osteoarthritis. Arthritis Res Ther 2004;6:R355–65. 36. Brennan FM, Chantry D, Jackson A, Maini R, Feldmann M. Inhibitory effect of TNF ␣ antibodies on synovial cell interleukin-1 production in rheumatoid arthritis. Lancet 1989;2:244–7. 37. Siegle I, Klein T, Backman JT, Saal JG, Nusing RM, Fritz P. Expression of cyclooxygenase 1 and cyclooxygenase 2 in human synovial tissue: differential elevation of cyclooxygenase 2 in inflammatory joint diseases. Arthritis Rheum 1998;41:122–9. 38. Ek M, Engblom D, Saha S, Blomqvist A, Jakobsson PJ, EricssonDahlstrand A. Inflammatory response: pathway across the bloodbrain barrier. Nature 2001;410:430–1. 39. Caughey GE, Cleland LG, Penglis PS, Gamble JR, James MJ. Roles of cyclooxygenase (COX)-1 and COX-2 in prostanoid production by human endothelial cells: selective up-regulation of prostacyclin synthesis by COX-2. J Immunol 2001;167:2831–8. 40. Szczepanski A, Moatter T, Carley WW, Gerritsen ME. Induction of cyclooxygenase II in human synovial microvessel endothelial cells by interleukin-1: inhibition by glucocorticoids. Arthritis Rheum 1994;37:495–503. 41. Bryant CE, Appleton I, Mitchell JA. Vascular endothelial growth factor upregulates constitutive cyclooxygenase 1 in primary bovine and human endothelial cells. Life Sci 1998;62:2195–201. 42. Murphy JF, Fitzgerald DJ. Vascular endothelial growth factor induces cyclooxygenase-dependent proliferation of endothelial cells via the VEGF-2 receptor. FASEB J 2001;15:1667–9. 43. Onodera M, Horiuchi Y, Nakahama K, Muneta T, Mano Y, Morita I. Induction of cyclooxygenase-1 in cultured synovial cells isolated from rheumatoid arthritis patients. Inflamm Res 2004;53: 217–22. 44. Fava RA, Olsen NJ, Spencer-Green G, Yeo KT, Yeo TK, Berse B, et al. Vascular permeability factor/endothelial growth factor (VPF/VEGF): accumulation and expression in human synovial fluids and rheumatoid synovial tissue. J Exp Med 1994;180:341–6. 45. Nagashima M, Yoshino S, Aono H, Takai M, Sasano M. Inhibitory effects of anti-rheumatic drugs on vascular endothelial growth factor in cultured rheumatoid synovial cells. Clin Exp Immunol 1999;116:360–5. 46. Zwerina J, Hayer S, Tohidast-Akrad M, Bergmeister H, Redlich K, Feige U, et al. Single and combined inhibition of tumor necrosis factor, interleukin-1, and RANKL pathways in tumor necrosis factor–induced arthritis: effects on synovial inflammation, bone erosion, and cartilage destruction. Arthritis Rheum 2004;50: 277–90.