Rabbit anti-human mPGES-1 polyclonal antibody, rabbit anti-human cPGES polyclonal antibody, rabbit anti-human COX-2 polyclonal antibody, PGE₂, PGH₂, and enzyme-linked immunosorbent assay (ELISA) kits for PGE₂, 6-keto-PGF_1α, and thromboxane (TX)B₂ were all purchased from Cayman Chemical Co. (Ann Arbor, MI). Mouse anti-human COX-1 monoclonal antibody and MK-886 were purchased from Wako Pure Chemical Industries (Osaka, Japan). Goat anti-human COX-2 polyclonal antibody was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG, HRP-conjugated goat anti-mouse IgG, tetramethylrhodamine β-isothiocyanate-conjugated donkey anti-goat IgG, and fluorescein isothiocyanate-conjugated donkey anti-rabbit IgG were obtained from Jackson ImmunoResearch (West Grove, PA). Dulbecco's modified Eagle's medium (DMEM), fetal bovine serum (FBS), and TRIzol reagent were from Invitrogen (Carlsbad, CA). Recombinant human IL-1β, TNF-α, IL-4, IL-6, IL-8, IL-10, and interferon-γ were from BD Biosciences (San Diego, CA). The Megaprime DNA labeling system, [α-³²P]dCTP, nylon membrane (Hybond N+), and enhanced chemiluminescence reagent were purchased from Amersham Pharmacia Biotech (Little Chalfont, Bucks., UK), and poly(vinylidene difluoride) membrane (Immobilon-P) was obtained from Millipore (Bedford, MA).

Cartilage samples were obtained at the time of joint replacement surgery from patients with OA who fulfilled the American College of Rheumatology criteria for this disease 31. These experiments were performed in accordance with a protocol approved by the ethics committee of St Marianna University, and all patients gave written consent to the use of their tissues for this research.

Chondrocytes were prepared as described elsewhere 32. In brief, cartilage samples were minced and digested with 1 mg/ml collagenase in DMEM for 18 h at 37°C, filtered through a nylon mesh, and washed extensively. Then the isolated chondrocytes were seeded at a high density in tissue culture flasks coated with type I collagen and were incubated in DMEM containing 10% FBS, 100 units/ml penicillin, and 100 μg/ml streptomycin at 37°C under an atmosphere of 5% CO₂. At confluence, the cells were detached and passaged, and the second-passage cells were used in subsequent experiments. The chondrocytes' phenotype was assessed by analyzing the expression of type II collagen with the reverse transcription–polymerase chain reaction (RT–PCR) (Fig. 1). OA synovial fibroblasts and normal dermal fibroblasts were prepared as described previously 2533.

Cells were seeded into the wells of a six-well plate coated with type I collagen at a density of 5 × 10⁴ cells/cm² in culture medium containing 1% FBS, and then were cultured under various conditions. RNA from the cells was extracted with TRIzol reagent in accordance with the manufacturer's instructions. Reverse transcription was performed with a SuperScript preamplification system (Invitrogen, Carlsbad, CA), in accordance with the manufacturer's instructions, with 1 μg of total RNA from the cells as a template. Equal amounts of each reverse-transcription product were amplified by PCR with HotStarTaq polymerase (Qiagen, Valencia, CA). After initial denaturation at 95°C for 15 min, PCR involved amplification cycles of 30 s at 95°C, 30 s at 56°C, and 45 s at 72°C, followed by elongation for 5 min at 72°C. The sequences of the primers, the size of the PCR products, and the number of amplification cycles are shown in Table 1. The amplified cDNA fragments were resolved by electrophoresis on 1.5% (w/v) agarose gel, and were detected under ultraviolet after staining of the gel with ethidium bromide.

Cells were seeded into the wells of a six-well plate coated with type I collagen at a density of 5 × 10⁴ cells/cm² in culture medium containing 1% FBS, and were then cultured under various conditions. Northern blotting was performed as described previously 25. Total RNA (4 μg) was extracted from the cells with TRIzol reagent and was subjected to electrophoresis on 1% formaldehyde/agarose gel, followed by transfer to a Hybond N+ membrane and hybridization with cDNA probes that had been labeled with [α-³²P]dCTP by random priming. After hybridization, the membranes were washed and the RNA bands were detected by autoradiography.

Cells (at a density of 5 × 10⁴ cells/cm²) were cultured under various conditions in medium containing 1% FBS. Subsequently, the cells were lysed in Tris-buffered saline (TBS) containing 0.1% sodium dodecyl sulfate (SDS), and the protein content of the lysates was determined with the bicinchoninic acid protein assay reagent (Pierce, Rockford, IL), with bovine serum albumin as the standard. Cell lysates were adjusted to 5 μg of protein and were then applied to SDS-polyacrylamide gels (10–20%) for electrophoresis, as reported previously 25. Next, the proteins were electroblotted onto Immobilon-P poly(vinylidene difluoride) membranes with a semidry blotter (Atto, Tokyo, Japan). After the membranes had been blocked in 10 mM TBS containing 0.1% Tween-20 (TBS-T) and 5% skimmed milk, the primary antibody (rabbit anti-human mPGES-1 antibody, rabbit anti-human cPGES antibody, mouse anti-human COX-1 antibody, or rabbit anti-human COX-2 antibody) was added at a dilution of 1:500 (mPGES-1 and cPGES) or 1:200 (COX-1 and COX-2) in TBS-T, and incubation was performed for 1.5 h. After the membranes had been washed with TBS-T, the secondary antibody (HRP-conjugated goat anti-rabbit antibody or HRP-conjugated goat anti-mouse antibody) was added (at a dilution of 1:10,000 or 1:5,000, respectively, in TBS-T) and incubation was performed for 1 h. After further washing with TBS-T, protein bands were detected with an enhanced chemiluminescence Western blot analysis system.

Immunofluorescent microscopy was performed as reported previously 25. In brief, cells were seeded at a density of 2.5 × 10⁴ cells/cm² onto culture slides coated with type I collagen and were incubated with or without 1 ng/ml IL-1β for 48 h. After removal of the supernatants, the cells were fixed with 4% (w/v) paraformaldehyde in phosphate-buffered saline (PBS) for 30 min at 4°C. Then the cells were permeabilized with 0.1% Triton-X in PBS for 3 min. After blocking with 1% (w/v) bovine serum albumin in PBS for 1 h at 20–25°C, rabbit anti-human mPGES-1 polyclonal antibody and goat anti-human COX-2 polyclonal antibody (premixed so that the final dilutions were, respectively, 1:100 and 1:250 in PBS containing 1% albumin) were added and incubation was done for 1.5 h at 20–25°C. After washing with PBS, tetramethylrhodamine β-isothiocyanate-conjugated and fluorescein isothiocyanate-conjugated secondary antibodies were added at 1:50 dilution in PBS containing 1% albumin, and incubated for 1 h. Then the glass slides were sealed under coverslips using Vectashield Mounting Medium (Vector Laboratories, Burlingame, CA) and examined under an IX71 microscope connected to a Cool SNAP HQ digital camera (Olympus, Tokyo, Japan).

Cells stimulated with IL-1β or TNF-α were washed with DMEM and incubated for a further 30 min with 3 μM arachidonic acid, as reported previously 2534. The culture medium was then harvested, and the concentrations of PGE₂, 6-keto-PGF_1α (a stable metabolite of PGI₂), and TXB₂ (a stable metabolite of TXA₂) were measured by ELISA.

PGES activity was measured by assessment of the conversion of PGH₂ to PGE₂, as reported previously 2535. In brief, cells (at a density of 5 × 10⁴ cells/cm²) were cultured under various conditions in medium containing 1% FBS. The cells were then scraped off the dish and lysed by sonication (twice for 10 s each at 1 min intervals) in 100 μl of 1 M Tris–HCl (pH 8.0). After centrifugation of the lysate at 15,000 g for 1 min at 4°C, the supernatant was used to measure PGES activity. For assessment of PGES activity, an aliquot of each supernatant equivalent to 50 μg of protein was incubated with 2 μg of PGH₂ for 30 s at 24°C in 0.1 ml of 1 M Tris–HCl containing 2 mM glutathione, in the absence or presence of MK-886. The reaction was terminated by the addition of 100 mM FeCl₂; the reaction mixture was then left to stand at 20–25°C for 15 min. After centrifugation of the reaction mixture, the PGE₂ concentration in the supernatant was measured with an ELISA kit.

Cells were seeded at a density of 5 × 10⁴ cells/cm² in culture medium containing 1% FBS and were incubated with IL-1β for 48 h. In experiments involving treatment with MK-886, it was added 30 min before IL-1β. The culture supernatant was then harvested and the PGE₂ concentration was measured with an ELISA kit.

Tissue sections were incubated for 5 min with 3% (v/v) H₂O₂. After blocking with 1% (w/v) bovine serum albumin in PBS for 30 min at 20–25°C, rabbit anti-human mPGES antibody or goat anti-human COX-2 antibody (diluted 1:100 or 1:200, respectively, in PBS containing 1% albumin) was added and incubation was performed for 1.5 h at 20–25°C. After being washed, the sections were incubated with the biotinylated secondary antibody, followed by streptavidin–HRP complex and 3,3'-dimaminobenzidine tetrahydrochloride in accordance with the manufacturer's protocol (LSAB staining kit; Dako, Carpinteria, CA). As adsorption controls, primary antibodies against mPGES-1 and COX-2 were used that had been previously reacted with mPGES-1 (10 μg/ml) or COX-2 peptide (10 μg/ml), respectively. In addition, rabbit anti-human mPGES-1 was replaced with normal rabbit IgG as a negative control.

Results are expressed as means ± SD for triplicate experiments, and representative results from three independent patients are shown. Statistical analysis for triplicate experiments was performed with Tukey's multiple comparison test; P < 0.05 was considered significant.

Morphology and collagen mRNA expression were examined for assessment of the chondrocyte phenotype. The morphology of chondrocytes used in this study was obviously different from that of synovial fibroblasts (Fig. 1a). Type I collagen mRNA was expressed in chondrocytes as well as synovial fibroblasts and dermal fibroblasts. In contrast, expression of type II collagen mRNA was detected only in chondrocytes, as shown in Fig. 1b. Moreover, expression of type II collagen mRNA as well as type I collagen mRNA in chondrocytes was maintained after stimulation of IL-1β or TNF-α for 12 h (Fig. 1c).

The expression of mRNA for mPGES-1 and COX-2 in chondrocytes from patients with OA was assessed by RT–PCR after the cells had been treated with various cytokines for 12 h. As shown in Fig. 2a, TNF-α or IL-1β induced the expression of mPGES-1 and COX-2 mRNA in chondrocytes. In contrast, IL-4, IL-6, IL-8, IL-10, and interferon-γ all had no effect on the expression of mPGES-1 or COX-2 mRNA.

We next examined the induction of mRNA for PGES and COX isoenzymes by IL-1β or TNF-α in chondrocytes from four patients with OA (Fig. 2b). Expression of mPGES-1 and COX-2 mRNA was significantly induced by IL-1β or TNF-α in chondrocytes from all four patients. In contrast, mRNA for mPGES-2, cPGES, and COX-1 was constitutively expressed in unstimulated chondrocytes, and exposure to IL-1β or TNF-α had little effect on the expression of these mRNAs. The pattern of mRNA expression for each enzyme was similar in all four patients.

Subsequently, we examined the effect of various concentrations of IL-1β and TNF-α on mPGES-1 and COX-2 mRNA by Northern blot analysis. As shown in Fig. 3, expression of mPGES-1 and COX-2 mRNA was induced by IL-1β in a concentration-dependent manner. TNF-α also induced expression of mPGES-1 mRNA, but had a weaker effect than IL-1β. The effect of TNF-α on the expression of COX-2 mRNA was also weaker than that of IL-1β.

Expression of mPGES-1 and COX-2 protein in chondrocytes stimulated with IL-1β or TNF-α was investigated by Western blot analysis and immunofluorescent microscopy (Fig. 4). As shown in Fig. 4a, both mPGES-1 and COX-2 protein were induced by IL-1β and TNF-α, similar to the results for mRNA expression shown in Fig. 3. Colocalization of these enzymes was observed in the perinuclear region of IL-1β-stimulated chondrocytes (Fig. 4b).

To determine the profile of PG synthesis in OA chondrocytes, cells were exposed to various concentrations of IL-1β (0.01–10 ng/ml) or TNF-α (0.01–10 ng/ml) for 24 h and the conversion of exogenous arachidonic acid to PGE₂, 6-keto-PGF_1α (a stable metabolite of PGI₂), and TXB₂ (a stable metabolite of TXA₂) was measured by ELISA. As shown in Fig. 5a, PGE₂ synthesis was markedly increased by IL-1β or TNF-α (to a smaller extent) in a concentration-dependent manner. The pattern of PGE₂ production was similar to results shown in Fig. 2. In contrast, the changes in 6-keto-PGF_1α (Fig. 5b) and TXB₂ (Fig. 5c) after the addition of IL-1β or TNF-α were almost negligible compared with the effect on PGE₂.

The expression of mRNA (Fig. 6a) and protein (Fig. 6b) for PGES and COX isoenzymes at various times after stimulation with IL-1β was assessed by Northern and Western blotting, respectively. Expression of mPGES-1 mRNA showed a gradual increase, reaching a maximum after 24–48 h of exposure to IL-1β. Expression of mPGES-1 protein showed a slight increase at 12 h, and then gradually increased until at least 72 h. In contrast, COX-2 mRNA increased rapidly after the start of IL-1β stimulation, and its expression remained maximal until 48 h. The COX-2 protein level also showed a rapid increase within 6 h and expression was maximal between 12 and 48 h; it decreased again at 72 h. Expression of mPGES-1 was delayed compared with that of COX-2. In contrast, the levels of cPGES and COX-1 mRNA and protein were not affected by IL-1β.

To investigate whether PGES enzymatic activity was upregulated by IL-1β, we measured the conversion of PGH₂ to PGE₂ in lysates prepared from IL-1β-stimulated chondrocytes at various times (Fig. 7a). PGES activity increased gradually until 72 h after the start of IL-1β stimulation. These results matched the pattern of mPGES-1 protein expression (Fig. 6b).

It was reported that MK-886 could inhibit mPGES-1 activity in mPGES-1 transfected cells 28. To evaluate the contribution of mPGES-1 to the IL-1β-induced increase of PGES activity in chondrocytes, we examined the effect of MK-886 on PGES activity in chondrocyte lysates. As shown in Fig. 7b, PGES activity in lysates from unstimulated chondrocytes was very low, and it was not affected by MK-886 alone. In contrast, the increased level of PGES activity in lysates prepared from IL-1β-stimulated chondrocytes was inhibited by MK-886 in a concentration-dependent manner. These results suggested that the induction of PGES activity by IL-1β in chondrocytes was dependent mainly on the enzymatic activity of inducible mPGES-1 rather than on constitutive mPGES-2 and cPGES.

We next examined the effect of MK-886 on endogenous PGE₂ production in IL-1β-stimulated chondrocytes. Chondrocytes were incubated with IL-1β for 48 h in the absence or presence of MK-886. As shown in Fig. 8, endogenous PGE₂ production by OA chondrocytes was significantly increased after stimulation with IL-1β. When chondrocytes were cultured with IL-1β in the presence of MK-886, there was significant inhibition of PGE₂ production in a concentration-dependent manner. Microscopy of the cells revealed severe toxicity of MK-886 at 30 μM (data not shown).

Articular cartilage obtained from patients with OA was fixed in formaldehyde, embedded in paraffin, and cut into serial sections that were immunostained with antibody against mPGES-1 or COX-2. Chondrocytes in the articular cartilage from patients with OA were stained by the antibody against mPGES-1 (Fig. 9a). When the absorption control using mPGES-1 peptide was tested (Fig. 9b), mPGES-1 immunostaining was almost completely abolished. In addition, no mPGES-1 immunostaining was observed when rabbit anti-human mPGES-1 was replaced by normal rabbit IgG (Fig. 9c). COX-2 immunostaining was also positive in the chondrocytes (Fig. 9d), whereas the absorption control using COX-2 peptide (Fig. 9e) showed no COX-2 immunostaining.