Inhibitors of Mitogen-Activated Protein Kinases Downregulate COX-2 Expression in Human Chondrocytes

Inducible prostaglandin synthase (cyclooxygenase-2, COX-2) is expressed in rheumatoid and osteoarthritic cartilage and produces high amounts of proinflammatory prostanoids in the joint. In the present study we investigated the effects of the inhibitors of mitogen-activated protein kinase (MAPK) pathways Erk1/2, p38, and JNK on COX-2 expression and prostaglandin E2 (PGE2) production in human chondrocytes. Proinflammatory cytokine IL-1β caused a transient activation of Erk1/2, p38, and JNK in immortalized human T/C28a2 chondrocytes and that was followed by enhanced COX-2 expression and PGE2 production. PD98059 (an inhibitor of Erk1/2 pathway) suppressed IL-1-induced COX-2 expression and PGE2 production in a dose-dependent manner, and seemed to have an inhibitory effect on COX-2 activity. SB203580 (an inhibitor of p38 pathway) but not its negative control compound SB202474 inhibited COX-2 protein and mRNA expression and subsequent PGE2 synthesis at micromolar drug concentrations. SP600125 (a recently developed JNK inhibitor) but not its negative control compound N1-methyl-1,9-pyrazolanthrone downregulated COX-2 expression and PGE2 formation in a dose-dependent manner. SP600125 did not downregulate IL-1-induced COX-2 mRNA expression when measured 2 h after addition of IL-1β but suppressed mRNA levels in the later time points suggesting post-transcriptional regulation. Our results suggest that activation of Erk1/2, p38, and JNK pathways belongs to the signaling cascades that mediate the upregulation of COX-2 expression and PGE2 production in human chondrocytes exposed to proinflammatory cytokine IL-1β.


INTRODUCTION
Prostaglandins (PGs) are present in a wide variety of human tissues, where they regulate physiological responses, including vascular tone, blood clotting, kidney function, gastric secretion and reproduction [1]. In arthritis, prostaglandins (especially PGE 2 ) are produced in much higher amounts, and they mediate inflammation, tissue destruction, and inflammatory pain. PGs are synthesized from arachidonic acid by cyclooxygenase (COX) enzymes [2,3]. Two isoforms of COX have been identified: COX-1 is constitutively expressed and produces low Correspondence and reprint requests to Eeva Moilanen, The Immunopharmacology Research Group, Medical School, University of Tampere, 33014 Tampere, Finland; eeva.moilanen@uta.fi This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. physiological levels of prostanoids, whereas the expression of the inducible isoform, COX-2, is increased in response to proinflammatory cytokines or bacterial products [4]. COX-2 is highly expressed in rheumatoid (RA) and osteoarthritic (OA) cartilage [5,6]. Interleukin (IL)-1 is a key cytokine involved in the joint destruction in RA and OA and it has been shown to enhance COX-2 expression in articular chondrocytes [5,7,8].
Mitogen-activated protein kinases (MAPKs) are a family of serine/threonine kinases, that are part of the signal transduction pathways which connect inflammatory and other extracellular signals to intracellular responses, for example, gene expression. The three better characterized MAPK pathways are extracellular signal-regulated kinase 1 and 2 (Erk1/2), p38, and c-Jun N-terminal kinase (JNK). The growth-factor-induced Erk1/2, and the stressactivated p38 and JNK protein kinases are phosphorylated in response to extracellular stimuli at conserved threonine and tyrosine residues and have regulatory functions in inflammation [9,10]. Inhibitors of p38 and JNK are under development for treatment of arthritis and they 2005:5 (2005) have shown efficacy in experimentally induced arthritis and joint pain [11,12]. Inhibition of MAPKs is likely to result in suppression of inflammatory mediators, which in turn leads to the desired therapeutic effects. We hypothesized that one of the inflammatory pathways in the cartilage, that might be down-regulated by MAPK inhibitors, is COX-2-PGE 2 pathway. The aim of the present study was to investigate if inhibitors of JNK (SP600125), p38 (SB203580), and Erk1/2 (PD98059) MAP-kinase pathways downregulate IL-induced COX-2 expression and PGE 2 production in human chondrocytes.

Cell culture
Immortalized human T/C28a2 chondrocytes [13] were grown in Dulbecco's modified Eagle's medium (Cambrex Bioproducts Europe, Verviers, Belgium) and Ham's F-12 medium (Gibco, Paisley, Scotland) (1:1, v/v). Culture media contained 10% heat-inactivated fetal bovine serum, 100 U/mL penicillin, 100 µg/mL streptomycin, and 250 ng/mL amphotericin B (all from Gibco, Paisley, Scotland). Cells were seeded on 24-well plates for prostaglandin E 2 measurements and on 6-well plates for Western blot and RT-PCR. Cell monolayers were grown for 72 h to confluence before the experiments were started and the compounds of interest were added in fresh medium.

Prostaglandin E 2 assays
At the indicated time points, the culture medium was collected for prostaglandin E 2 (PGE 2 ) measurement. PGE 2 concentrations were determined by radioimmunoassay using reagents from the Institute of Isotopes (Budapest, Hungary).

Western blot analysis
At the indicated time points, cells were rapidly washed with ice-cold PBS and solubilized in cold lysis buffer containing 10 mM Tris base, 5 mM EDTA, 50 mM NaCl, 1% Triton X-100, 0.5 mM phenylmethylsulfonyl fluoride, 2 mM sodium orthovanadate, 10 µg/mL leupeptin, 25 µg/mL aprotinin, 1.25 mM NaF, 1 mM sodium pyrophosphate, and 10 mM n-octyl-β-D-glucopyranoside. After incubation for 20 min on ice, lysates were centrifuged (14 500 g for 10 min), and supernatants were mixed in a ratio of 1:4 with SDS loading buffer (62.5 mM Tris-HCl, pH 6.8, 10% glycerol, 2% SDS, 0.025% bromophenol blue, and 5% β-mercaptoethanol) and boiled for 5 min. Protein concentrations in the samples were measured by the Coomassie blue method [14]. After boiling for 5 min, equal aliquots of protein (20 µg) were loaded on a 10% SDS-polyacrylamide electrophoresis gel and electrophoresed for 4 h at 100 V in a buffer containing 95 mM Tris-HCl, 960 mM glycine, and 0.5% SDS. After electrophoresis, the proteins were transferred to Hybondenhanced chemiluminescence nitrocellulose membrane (Amersham, Buckinghamshire, UK) with semidry blotter at 2.5 mA/cm 2 for 60 min. After transfer, the membrane was blocked in TBS/T (20 mM Tris-base pH 7.6, 150 mM NaCl, 0.1% Tween-20) containing 5% nonfat milk for 1 h at room temperature and incubated overnight at 4 • C with COX-2, JNK, p38, Erk1/2, phospho-specific JNK, phospho-specific p38, or phospho-specific Erk1/2 antibodies in TBS/T containing 5% nonfat milk. Thereafter the membrane was washed 4 times with TBS/T for 5 min, incubated with secondary antibody coupled to horseradish peroxidase in the blocking solution for 0.5 h at room temperature, and washed four times with TBS/T for 5 min. Bound antibody was detected using SuperSignal West Pico chemiluminescent substrate (Pierce, Cheshire, UK) and FluorChem 8800 imaging system (Alpha Innotech Corp, San Leandro, Calif). The quantitation of the chemiluminescent signal was carried out with the use of FluorChem software version 3.1.

RNA extraction and real-time RT-PCR
At the indicated time points, cell monolayers were rapidly washed with ice-cold PBS, and cells were homogenized using QIAshredder (QIAGEN, Valencia, Calif). RNA extraction was carried out with the use of RNeasy kit for isolation of total RNA (QIAGEN). Total RNA (25 ng) was reverse-transcribed to cDNA using TaqMan reverse transcription reagents and random hexamers (Applied Biosystems, Foster City, Calif). cDNA obtained from the RT reaction (amount corresponding to approximately 1 ng of total RNA) was subjected to PCR using TaqMan Universal PCR Master Mix and ABI PRISM 7000 Sequence detection system (Applied Biosystems). The primer and probe sequences and concentrations were optimized according to manufacturer's guidelines in TaqMan Universal PCR Master Mix Protocol part number 4304449 revision C and they were 5 -CAACTCTATATTGCTGGAACATGGA -3 (human COX-2 forward primer, 300 nM), 5 -TGGAAGCCTGT-GATACTTTCTGTACT-3 (human COX-2 reverse primer, 300 nM), 5 -TCCTACCACCAGCAACCCTGCCA-3 (human COX-2 probe containing 6-FAM as 5 -reporter dye  Figure 1. The effects of IL-1β on JNK, p38, and Erk1/2 MAPK activation in human T/C28a2 chondrocytes. The chondrocytes were stimulated with IL-1β (100 pg/mL). Incubations were terminated at the indicated time points. Two parallel immunoblots were run from same cell lysates using antibodies against the Thr-183/Tyr-185, Thr-180/Tyr-182, and Thr-202/Tyr-204 phosphorylated (ie, activated) JNK (p-JNK), p38 (p-p38), and Erk1/2 (p-Erk1/2) and against total JNK, p38, and Erk1/2. The experiment was repeated three times with similar results. and TAMRA as 3 -quencher, 150 nM). Human β-actin was obtained from TaqMan Human β-actin Reagents kit (Applied Biosystems), containing VIC as 5 -reporter dye and TAMRA as 3 -quencher. PCR reaction parameters were as follows: incubation at 50 • C for 2 min, incubation at 95 • C for 10 min, and thereafter 40 cycles of denaturation at 95 • C for 15 s and annealing and extension at 60 • C for 1 min. Each sample was determined in duplicate.
A standard curve method was used to determine the relative mRNA levels as described in the Applied Biosystems User Bulletin: a standard curve for each gene was created using RNA isolated from IL-1β-stimulated human T/C28a2 chondrocytes. Isolated RNA was reversetranscribed and dilution series of cDNA ranging from 1 pg to 10 ng were subjected to real-time PCR. The obtained threshold cycle values were plotted against the dilution factor to create a standard curve. Relative mRNA levels in test samples were then calculated from the standard curve.

Statistics
Results are expressed as the mean ± SEM. Statistical significances were calculated by analyses of variance supported by the Dunnett's multiple comparisons test. Differences were considered significant at P < .05.

IL-1β-induced COX-2 expression and PGE 2 production in human T/C28a2 chondrocytes
IL-1β enhanced COX-2 expression in a concentrationdependent manner, being detectable at 10 pg/mL and increasing up to 1000 pg/mL (Figure 2a). Radioimmunoassay of prostaglandin E 2 (PGE 2 ) in the culture medium was carried out to investigate PGE 2 production. IL-1β induced PGE 2 production in a concentration-dependent manner. Increased PGE 2 production was detected at 10 pg/mL of IL-1β and was maximal at 100 pg/mL remaining elevated up to 1000 pg/mL (Figure 2b).

SP600125, SB203580, and PD98059 suppressed IL-1β-induced PGE 2 production in human T/C28a2 chondrocytes
Inhibitors of JNK (SP600125), p38 (SB203580), and Erk1/2 (PD98059) reduced IL-1β-induced PGE 2 production in T/C28a2 chondrocytes in a concentrationdependent manner (Figures 3a, 3b, 3c). In the further studies, SP600125, SB203580, and PD98059 were added in T/C28a2 chondrocyte cultures at the same time or 6 h after IL-1β. In contrast to the inhibitory effect seen when added at the same time, SP600125 (10 µM) and SB203580 (1 µM) did not inhibit PGE 2 production when added 6 h after IL-1β. PD98059 (10 µM) inhibited the production of PGE 2 also when added 6 h after IL-1β, but the inhibition was notably less than when added to cells at the same time as the stimulus (Figure 3g).

SP600125, SB203580, and PD98059 inhibited COX-2 mRNA expression in human T/C28a2 chondrocytes
We used real-time RT-PCR to investigate the effects of SP600125, SB203580, and PD98059 on COX-2 mRNA expression. IL-1β induced transient COX-2 expression that peaked 4 h after addition of IL-1. SB203580 and PD98059 reduced IL-1β-induced COX-2 mRNA expression significantly when measured either 2 h or 8 h after IL-1β stimulation. In contrast, SP600125 had no marked effect on IL-1β-induced COX-2 mRNA expression at the 2 h time point, whereas the level of COX-2 mRNA was reduced by about 75% at the 8 h time point (Figure 4).

DISCUSSION
In the present study, we found that inhibitors of JNK, p38, and Erk1/2 pathways downregulate IL-1-induced COX-2 expression and PGE 2 production in human chondrocytes.
Consistently with earlier findings using primary articular chondrocytes [17,18], our results show that IL-1β causes a rapid activation of JNK, p38, and Erk1/2 MAP kinases in immortalized human T/C28a2 chondrocytes. These events were followed by enhanced COX-2 expression and subsequent PGE 2 production. Inhibition of JNK activity by SP600125, p38 activity by SB203580, and Erk1/2 activity by PD98059 resulted in a reduction in the amount of PGE 2 produced. However, when SP600125 and SB203580 were added 6 h after IL-1β, they did not affect PGE 2 production. These findings suggest that SP600125 and SB203580 did not inhibit COX-2 activity, but rather reduced the expression of COX-2. PD98059 inhibited the production of PGE 2 also when added 6 h after IL-1β, but the inhibition was notably smaller than in those experiments where PD98059 was added at the same time as IL-1β. This suggests that PD98059 may have also some inhibitory effect on cyclooxygenase activity in activated chondrocytes, as has been earlier reported in arachidonic-acid-stimulated human platelets [19]. Western blot analysis showed that all the three inhibitors (SP600125, SB203580, and PD98059) caused also a concentration-dependent reduction in COX-2 protein levels in IL-1-treated chondrocytes.
In the real-time RT-PCR studies, SP600125 had practically no effect on IL-1β-induced COX-2 mRNA expression in human T/C28a2 chondrocytes when measured 2 h after IL-1β, whereas when measured 8 h after the addition of IL-1β, a significant reduction in the levels of COX-2 mRNA was observed in the presence of SP600125. These results suggest that inhibition of JNK pathway does not affect the early events in IL-1-induced COX-2 expression but it may regulate the process at post-transcriptional level.
SB203580 and PD98059 had a significant effect on COX-2 mRNA expression when measured 2 h after addition of IL-1β. Our results are consistent with previous reports showing that p38 and Erk1/2 MAP kinase pathways are involved in the cellular events leading to upregulation of COX-2 gene transcription in human monocytes and in RAW264 macrophages [20,21,22,23,24]. In addition, p38 has also reported to stabilize COX-2 mRNA in mammary carcinoma cells, in HeLa-TO cells, and another chondrocyte cell line [25,26,27].
We have earlier shown that, in J774 macrophages, SP600125, SB203580, and PD98059 do not have an effect on nuclear translocation and DNA binding activity of NF-κB [28,29,30] which plays a role in stimulating COX-2 expression [31]. In addition to NF-κB, the expression of COX-2 is regulated by other factors including NF-IL6 and AP-2 [32,33]. Further studies are needed to determine the molecular mechanisms which mediate the effects of JNK, p38, and Erk1/2 inhibitors on COX-2 expression.
Inhibitors of p38 and JNK are under development for treatment of arthritis and they have been shown to have antiinflammatory and antierosive effects in experimentally induced arthritis, and to relief inflammatory pain [10,11]. The inhibition of MAP kinases and  Human chondrocytes were incubated with IL-1β (100 pg/ml) and with or without SP600125, SB203580, and PD98059. Incubations were terminated at the indicated time points, and the extracted total RNA was subjected to real-time RT-PCR. COX-2 mRNA levels were normalized against β-actin mRNA. Mean ± SEM, n = 6, * * indicates P < .01 as compared with cells treated with IL-1β only.
subsequent inhibition of the synthesis of a number of important proinflammatory cytokines like IL-1, TNF-α, IL-6, IL-8, and matrix metalloproteases has been identified as a mechanism contributing to the antiinflammatory activity of these compounds [10,11,12,15,34,35]. The present results show that inhibitors of JNK, p38, and Erk1/2 MAP kinases also downregulate COX-2 expression and PGE 2 production in human chondrocytes which is likely involved in the mechanisms of their therapeutic effects in arthritis and other inflammatory diseases.