15-Deoxy-Δ12,14-prostaglandin J2 Inhibits Bay 11-7085-induced Sustained Extracellular Signal-regulated Kinase Phosphorylation and Apoptosis in Human Articular Chondrocytes and Synovial Fibroblasts*

We have previously shown that nuclear factor-κB inhibition by adenovirus expressing mutated IκB-α or by proteasome inhibitor increases human articular chondrocytes sensibility to apoptosis. Moreover, the nuclear factor-κB inhibitor BAY11-7085, a potent anti-inflammatory drug in rat adjuvant arthritis, is itself a proapoptotic agent for chondrocytes. In this work, we show that BAY 11-7085 but not the proteasome inhibitor MG-132 induced a rapid and sustained phosphorylation of extracellular signal-regulated kinases (ERK1/2) in human articular chondrocytes. The level of ERK1/2 phosphorylation correlated with BAY 11-7085 concentration and chondrocyte apoptosis. 15-Deoxy-Δ12,14-prostaglandin J2 (15d-PGJ2) and its precursor prostaglandin (PG) D2 but not PGE2 and PGF2α rescued chondrocytes from BAY 11-7085-induced apoptosis. 15d-PGJ2 markedly inhibited BAY 11-7085-induced phosphorylation of ERK1/2. BAY 11-7085 also induced ERK1/2 phosphorylation and apoptosis in human synovial fibroblasts, and these reactions were down-regulated by 15d-PGJ2. Further analysis in synovial fibroblasts showed that only molecules that suppressed BAY 11-7085-induced phosphorylation of ERK1/2 (i.e. 15d-PGJ2, PGD2, and to a lesser extent, MEK1/2 inhibitor UO126, but not prostaglandins E2 and F2α or peroxisome proliferator-activated receptor-γ agonist ciglitazone) were able protect cells from apoptosis. These results suggested that the antiapoptotic effect of 15d-PGJ2 on chondrocytes and synovial fibroblasts might involve inhibition of ERK1/2 phosphorylation.

Joint diseases such as osteoarthritis (OA) 1 and rheumatoid arthritis lead to chondrocyte cell death and irreversible joint damage. Synovial tissue proliferation is one of the major detrimental events in rheumatoid disease. Cartilage is a highly specialized tissue that relies on only one cell type, the chondrocyte, and has very limited ability to regenerate (1). Chondrocyte survival is thus crucial for the preservation of cartilage and joint function (2).
Many biological functions of the proinflammatory cytokines interleukin-1␤ and tumor necrosis factor-␣ are mediated by nuclear factor-B (NF-B) or mitogen-activated protein kinases, and deregulation of both pathways is involved in the pathogenesis of rheumatoid arthritis and OA (3). Because NF-B controls the expression of pro-inflammatory molecules, such as cytokines, chemokines, nitric oxide, cyclo-oxygenase 2, metalloproteinases, and adhesion molecules (4), its inhibition by the super-repressor IB, by NF-B decoys (5), or by BAY11-7085 (6) is effective as treatment in animal model of experimentally induced arthritis. NF-B activity, however, also has a protective role against apoptosis of primary (7) and dedifferentiated (8) chondrocytes. Although the proteasome and NF-B inhibitor MG-132 merely sensitized chondrocytes to induced apoptosis, NF-B inhibitor BAY 11-7085 was able to cause chondrocyte apoptosis directly (7). We thus searched for additional pathways triggered by BAY 11-7085 and for an antiinflammatory molecule able to protect chondrocytes from BAY 11-7085-induced apoptosis.
In this study, we showed that, in human primary articular chondrocytes, BAY 11-7085 but not MG-132 induced a sustained ERK1/2 phosphorylation. Furthermore, we showed that BAY 11-7085-induced ERK1/2 phosphorylation as well as BAY 11-7085-induced chondrocyte apoptosis can be markedly inhibited by pretreatment with 15-deoxy-⌬ 12,14 -prostaglandin J2 (15d-PGJ2). However, we also showed that 15d-PGJ2 has a pronounced antiapoptotic effect on synovial fibroblasts, suggesting that 15d-PGJ2 could also have pro-inflammatory effects on the cells from human joint.

EXPERIMENTAL PROCEDURES
Chondrocyte and Synovial Fibroblast Isolation-Human cartilage was obtained post-mortem from patients who had not been admitted to the hospital for joint diseases (used for experiments presented in Figs. [1][2][3], and from OA patients during joint replacement (used for experiments presented in Fig. 4). Cartilage was cut in 1-2-mm 3 explants and digested at 37°C with gentle agitation with, in succession, 0.5 mg/ml hyaluronidase (Sigma-Aldrich) for 30 min, 1 mg/ml pronase (Merck KGaA) for 1 h, and 0.8 mg/ml collagenase IA (Sigma-Aldrich) for 16 -20 h. After digestion, cells were filtered through 70 M nylon membrane (Falcon; BD Biosciences), washed and seeded in Dulbecco's modified Eagle's medium (DMEM) lacking phenol red (Cambrex Bio Science Walkersville, Inc., Walkersville, MD) and supplemented with 10% fetal calf serum (FCS). Only primary chondrocytes were used (i.e. all experiments began on the first or second day after chondrocyte isolation and continued for a maximum of 48 h). * This work was supported by the FIRS credit 4774 (CHU Sart-Tilman). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Synovial membranes were obtained from OA patients during joint replacement. For cell isolation, synovium was finely cut and digested 4 or 16 h with 0.5 or 1 mg/ml collagenase IA (Sigma-Aldrich), respectively. Synovial fibroblasts were filtered through 70 M nylon membrane (Falcon), washed, and seeded in DMEM (Cambrex Bio Science Walkersville, Inc.) supplemented with 10% FCS. For experiments, synovial fibroblasts (passages 2-6) were seeded in DMEM supplemented with 1% FCS. Primary endothelial cells from human umbilical vein were purchased from Cambrex Bio Science Walkersville, Inc., and maintained in EGM2 medium (Cambrex Bio Science Walkersville, Inc.).
Survival Assay-Cell survival was measured as mitochondrial NADH/NADPH-dependent dehydrogenase activity, resulting in the cellular conversion of methyltetrazolium salt MTS (Promega, Madison, WI) into a soluble formazan dye (9). An electron-coupling agent, phenazine methosulfate, was obtained from Sigma-Aldrich. Colorimetric measurement of formazan dye was performed at 490 nm.
Annexin V Labeling-Chondrocytes were seeded in 8-well chamber slides (Nalge Nunc International, Naperville, IL) and pretreated with 15d-PGJ2 for 24 h. BAY 11-7085 was then added for an additional 24 h. Cell supernatant was carefully discarded and chondrocytes labeled with annexin V-FITC using Annexin V fluos staining kit (Roche, Mannheim, Germany). Cell images were captured under fluorescence microscope.

Bay 11-7085 but Not MG-132 Induces Sustained ERK1/2 Phosphorylation in Human Chondrocytes-We and others
showed previously that inhibition of NF-B by adenovirusexpressing mutated IB-␣ or by proteasome inhibitor increases human chondrocyte sensitivity to apoptosis (7,8). However, only BAY11-7085, not MG-132 or adenovirus-expressing mutated IB-␣, is itself a proapoptotic agent for chondrocytes. Thus, we searched for a signaling pathway distinct from NF-B that was involved in BAY 11-7085-induced chondrocyte apoptosis. Results showed that in isolated human articular chondrocytes BAY 11-7085 (Fig. 1A), but not MG-132 (Fig. 1B), induced rapid and sustained ERK1/2 phosphorylation.
15d-PGJ2 was also shown to have proapoptotic effect in several cell types: colon cancer cells (30), lung cancer cell (31), endothelial cells (11), and granulocytes (32). We also confirm here that 15d-PGJ2 induces apoptosis in endothelial cells, and our preliminary results showed that 15d-PGJ2 induces apoptosis in HCT-116 cancer cell line (data not shown).
It was shown previously that 15d-PGJ2 has anti-inflammatory effects on both human chondrocytes (33,34) and rheumatoid synovial fibroblasts (35), mainly because of its inhibition of NF-B (33,34). The results presented in this work suggest that although 15d-PGJ2 may have a cartilage-protective effect, the antiapoptotic effect of 15d-PGJ2 on synovial fibroblasts, despite suppression of endothelial cell growth (11), may favor inflammation in vivo. However, although our results showed that 15d-PGJ2 has a strong antiapoptotic effect on synovial fibroblasts from osteoarthritis patients, others reported that 15d-PGJ2 can induce apoptosis in synovial fibroblasts from patients with rheumatoid arthritis (36) and can suppress adjuvant-induced arthritis in rats. The effect of 15d-PGJ2 on synovial fibroblasts may therefore be disease-dependent, and we are currently testing this hypothesis.