Endothelin-1 in osteoarthritic chondrocytes triggers nitric oxide production and upregulates collagenase production

Human cartilage was obtained with the consent of 12 OA patients (mean ± standard error of the mean age, 58 ± 6 years) undergoing total knee replacement. The Institutional Ethics Committee Board of Notre Dame Hospital in Montreal, Canada approved the study protocol. Tissue specimens were embedded in paraffin, were sectioned and stained with Safranin O and fast green, and were evaluated using the Mankin histological/histochemical scale [17]. Only tissues corresponding to a moderate degree of OA severity (Mankin 3–7) were included in this study. Cartilage was sectioned from the tibial plateaus, rinsed and finely chopped, and the cells released by enzymatic digestion performed as previously described [2, 11]. The cells were seeded in culture flasks at the density of 10⁴ cells/cm² and were grown to confluence in DMEM (Gibco BRL, Burlington, ON, Canada) containing 10% heat-inactivated FCS (Hyclone, Logan, UT, USA) and 1% penicillin/streptomycin (Gibco BRL). Only first-passage-cultured cells were used.

MMP-1 and MMP-13 protein levels were determined in the culture media using specific ELISA assays. The ELISA assay (Amersham Biosciences Corp., Baie d'Urfé, QC, Canada) for MMP-1 specifically detected the total human MMP-1 (i.e. active MMP-1, the latent enzyme and the enzyme complexed with inhibitors such as tissue inhibitor of matrix metalloproteinases 1). The sensitivity of this assay is 1.7 ng/ml, and there is no significant cross-reactivity or interference with MMP-3, MMP-2 and MMP-9. The MMP-13 ELISA assay (R&D Systems Inc., Minneapolis, MN, USA) is a monoclonal polyclonal-based assay specific for both the active and latent MMP-13. Its sensitivity is 0.032 ng/ml, and there is no cross-reactivity with MMP-1, MMP-2, MMP-3, MMP-7, MMP-8, MMP-9 and MT1-MMP. Results are expressed as nanograms per 5 × 10⁵ cells.

MMP-1 production, MMP-13 production and NO production were studied in the absence of and in the presence of ET-1, using various inhibitors: 1 μM SB 202190 (inhibitor of p38 mitogen-activated protein [MAP] kinase), 10 μM PD 98059 (a selective mitogen-activated protein kinase kinase 1/2 [MEK1/2] inhibitor), 100 nM Wortmannin (a phosphatidyl inositol 3 kinase inhibitor), 4 μM KT5720 (a protein kinase A [PKA] inhibitor), or 2 μM LY83583 (an inhibitor of NO-dependent soluble guanylate cyclase inhibitor). All inhibitors were purchased from Calbiochem EDM Biosciences Inc. (San Diego, CA, USA), and the active concentrations chosen are based on the literature or were assayed in preliminary experiments [18, 19]. ET-1 was purchased from (Sigma-Aldrich, Oakville, ON, Canada). Confluent OA chondrocytes were preincubated for 30 min with these inhibitors and then 10 nM ET-1 was added for 24 hours. Following incubation, the MMP-13 and MMP-1 protein levels and NO levels were determined in the media of six independent cultures as described in the following.

Nitrite (NO₂^-), a stable end product of NO, was measured in the media of cultured cells using a spectrophotometric method based on the Griess reaction [20]. To examine the effects of ET-1 on NO production, a dose–response curve was performed by incubating OA chondrocytes for 24 hours with increased concentrations (0–100 nM) of ET-1, or by pretreating with protein kinase inhibitors or a guanylate cyclase inhibitor and ET-1 as already described. NO production was also evaluated in the presence of the iNOS inhibitor L-NIL (L-N⁶ (1-iminoethyl)lysine) (Calbiochem EDM Biosciences Inc.). Chondrocytes were preincubated for 30 min with 0–50 μM L-NIL and were then incubated for 24 hours with 10 nM ET-1. The media were collected and the released NO levels were determined. Results are expressed as nanomoles per 5 × 10⁵ cells ± standard error of the mean or as a percentage of the control cultures.

Confluent OA chondrocytes were incubated in the presence of or in the absence (control) of 10 nM ET-1, and the cells were lysed in 0.2 ml lysis buffer (25 mM HEPES, 5 mM MgCl₂, 1 mM EDTA, 1 mM PMSF, 10 μg/ml pepstatin, 10 μg/ml leupeptin, pH 7.5). The protein concentration of the lysate was determined with the Bradford dye assay (Bio-Rad Laboratories, Hercules, CA, USA). For western blot, 10 μg lysate protein was separated by electrophoresis on a 10% SDS discontinuous gradient polyacrylamide gel. Separated proteins were then transferred electrophoretically onto a nitrocellulose membrane (Hybond C extra; Amersham, Pharmacia Biotech, Chalfont St Giles, UK). The membranes were immersed overnight in the Super Block Blocking buffer (Pierce, Rockford, IL, USA), rinsed and incubated for 24 hours at 4°C with one of the mouse monoclonal primary antibodies (New England Biolabs, Mississauga, ON, Canada) specifically recognizing phosphorylated p38 or total p38 (dilution, 1/1000), phosphorylated p44/42 (dilution, 1/5000), phosphorylated Akt (dilution, 1/2000), phosphorylated stress-activated protein kinase/Jun-N-terminal kinase (SAP/JNK) (dilution, 1/1000), or actin C-terminal fragment (dilution, 1/5000). iNOS was detected with a rabbit polyclonal antibody (dilution, 1/1000; Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA).

Following incubation with primary antibody, membranes were carefully washed and reincubated for 1 hour at 4°C with a second antibody (anti-rabbit IgG). Anti-mouse horseradish peroxidase-conjugated IgG (dilution, 1/25,000) was used for the detection of the monoclonal antibody, and sheep anti-rabbit horseradish peroxidase-conjugated IgG (dilution, 1/40,000) was used for the polyclonal antibody. Detection was performed using the Super Signal Ultra Western blot chemiluminescence system (Pierce) [11].

Apoptosis was investigated in OA chondrocytes cultured on Lab-Tec chamber slides (Nalge Nunc International, Naperville, IL, USA). At confluence, the cells were rinsed and incubated at 37°C for 72 hours in DMEM containing 2.5% heat-inactivated FCS in the absence of or in the presence of 10 nM human recombinant ET-1. Apoptotic cells were detected by in situ staining using the TUNEL method (Trevigen Inc., Gaithersburg, MD, USA). Both pro-apoptotic Bad and anti-apoptotic Bcl2 proteins were determined by immunocytochemical detection using specific anti-Bad and anti-Bcl2 antibodies (Upstate Biotechnology, Lake Placid, NY, USA). The results are expressed as the mean percentage of positively stained cells according to a previously published method [21, 22].

The effects of ET-1 and those of various inhibitors on MMP-1 production and MMP-13 production are shown in Fig. 1. At 10 nM ET-1 the production of both enzymes was significantly increased (P < 0.005). SB202190, a p38 inhibitor, completely suppressed the ET-1-stimulated production of both enzymes, whereas the phosphatidyl inositol 3 kinase inhibitor Wortmannin and the PKA inhibitor KT5720 partially but significantly (P < 0.01) decreased the level of MMP-13 only. Interestingly, the most potent inhibitor of MMP-1 and MMP-13 production was LY83583, an inhibitor of NO-dependent soluble guanylate cyclase and of cGMP. This agent not only suppressed the ET-1-induced stimulation, but also decreased the level of both enzymes below the basal level: a significant difference was found for both MMP-13 and MMP-1 when compared with the ET-1 stimulation (P < 0.005) and for MMP-13 when compared with the control (P < 0.05). Although a decrease in MMP-13 was noted with the MEK1/2 kinase inhibitor PD98059 at the concentration tested, it did not reach statistical significance. With this inhibitor, no effect was found on MMP-1 production.

The effects of ET-1 on NO release and on iNOS expression are shown in Fig. 2. Figure 2a shows that ET-1 greatly stimulated NO production and was released in a concentration-dependent manner. Incubation with increasing concentrations of ET-1, from 0.1 to 100 nM, augmented almost 12-fold the linear accumulation of NO. To determine the mechanism involved in the ET-1-induced NO production, the effects of the major intracellular signalling pathways were investigated. Figure 2b shows that the ET-1-induced NO release was significantly inhibited by p38 inhibition and prevented by KT5720, a PKA inhibitor. No significant effect was noted for MEK1/2 inhibition by PD98059 and by Wortmannin. Moreover, the guanylate cyclase inhibitor LY83583 reduced the NO production as significant differences were found when compared with either the ET-1 stimulation (P < 0.05) or with the control (P < 0.05), and this inhibitor also decreased both the endogenous and ET-1-induced iNOS level (Fig. 2d). The ET-1-induced NO release occurs via iNOS as shown in Figure 2c: complete inhibition of iNOS by 50 μM allosteric iNOS inhibitor L-NIL, as expected, almost completely inhibited NO release. Figure 2d shows the effects of various inhibitors on iNOS expression, as determined by western blot analysis of cell extracts. The 24-hour incubation of cells with ET-1 results in an increase of iNOS protein (Fig. 2d, lane 2). The ET-1-induced iNOS protein expression was completely suppressed by SB202190 and LY83583, and was partially suppressed by Wortmannin and KT5720. PD98059 had no effect.

Figure 3a–d show the effects of ET-1 on the phosphorylation of p38, Akt, p44/42 and SAP/JNK kinases as detected by western blot of cell extracts. ET-1 at 10 nM induced p38, Akt, p44/42, and SAP/JNK phosphorylation in a time-ordered manner. For p38, the maximal effect following cell exposure to ET-1 was obtained at 10 min. For Akt, the maximal effect was observed at 2 min of cell exposure and this effect persisted during 30 min, followed by a decline at 45 min. At this time (45 min), both p38 kinase and Akt phosphorylated forms were diminished. The maximal effect was obtained at 15 min for p44/42 kinase and at 45 min for SAP/JNK. The SAP/JNK phosphorylated forms were not detected at 60 min, whereas that of p44/42 decreased but was still present even at 60 min.

As ET-1 induces NO release and because the accumulation of NO causes apoptosis, we explored this potential effect. OA chondrocytes incubated in the absence of (control) or in the presence of ET-1 (10 nM) for 72 hours showed that ET-1 did not affect apoptosis (TUNEL reaction; data not shown) or the production of either anti-apoptotic Bcl2 or pro-apoptotic Bad proteins. A similar percentage of positively stained cells was found for Bcl2 (42.8 ± 5.1% and 43.2 ± 4.3% for the control and for ET-1, respectively) and for Bad (10.1 ± 3.8% and 9.5 ± 2.1%, respectively).