c-Jun/Activator Protein-1 Mediates Interleukin-1β-induced Dedifferentiation but Not Cyclooxygenase-2 Expression in Articular Chondrocytes*

Interleukin (IL)-1β is a major catabolic pro-inflammatory cytokine involved in cartilage destruction-associated processes, such as loss of the differentiated chondrocyte phenotype (dedifferentiation) and inflammation. Here, we investigated the role of c-Jun and activator protein-1 (AP-1) in IL-1β-induced dedifferentiation and cyclooxygenase (COX)-2 expression in primary cultured chondrocytes. IL-1β induced expression and transient phosphorylation of c-Jun in primary cultured chondrocytes. Ectopic expression of c-Jun was sufficient to cause dedifferentiation, whereas expression of dominant negative c-Jun blocked IL-1β-induced dedifferentiation. Interestingly, modulation of c-Jun expression did not affect IL-1β-induced COX-2 expression. Further experiments revealed that c-Jun phosphorylation was mediated by c-Jun N-terminal kinase and was required for IL-1β-induced dedifferentiation but not COX-2 expression. Consistent with its ability to induce phosphorylation of c-Jun, IL-1β caused transient activation of AP-1, which is necessary for IL-1β-induced dedifferentiation. IL-1β treatment suppressed expression of Sox-9, a major transcription factor that regulates type II collagen expression. Inhibition of c-Jun N-terminal kinase or AP-1 reversed IL-1β-induced suppression of Sox-9, and ectopic expression of c-Jun was sufficient to cause suppression of Sox-9. Our results collectively suggest that IL-1β suppresses type II collagen expression in articular chondrocytes by inducing expression and phosphorylation of c-Jun, AP-1 activation, and subsequent suppression of Sox-9.

The development of cartilage, which serves as a template for long bone formation, is initiated by the differentiation of mesenchymal cells into chondrocytes (1). During the process of differentiation, chondrocytes receive and process a complex array of signals from local and systemic factors. One such factor is the transcription factor, activator protein-1 (AP-1), 1 which has an essential function in cartilage and bone development (2). AP-1 is a dimeric transcription factor formed by combinations of Jun (c-Jun, JunB, and JunD), Fos (c-Fos, Fra-1, Fra-2, and FosB), and ATF proteins (ATF-2 and ATF-3) and is involved in diverse biological processes such as differentiation, proliferation, cell survival, and transformation (3,4). The various subunits of AP-1 can be induced at sites of active bone formation in vivo by transforming growth factor-␤, parathyroid hormone, 1,25-dihydroxy vitamin D, and other factors (5).
Several lines of evidence indicate that AP-1 regulates cartilage and bone development. For example, AP-1 activation mediates Wnt regulation of chondrogenesis (6) and regulates hypertrophic maturation of chondrocytes (7). Among the components of AP-1, JunB plays a crucial role in endochondral ossification by regulating the proliferation and function of chondrocytes and osteoblasts (8). It has been shown that c-Jun mediates axial skeletogenesis by regulating notochord survival and intervertebral disc formation (9). Overexpression of c-Fos in ATDC5 cells inhibits chondrocyte differentiation in vitro (10). Ectopic expression of Fos in developing chicken limb buds by retroviral microinjection causes truncation of the cartilage in the long bones of the injected leg; this is due to chondrodysplasia caused by severely retarded differentiation of the proliferating chondrocytes into mature chondrocytes, hypertrophy of chondrocytes, and delays in precartilagenous condensation (11). AP-1 also plays an essential role in the regulation of chondrocyte differentiation by parathyroid hormone-related protein through induction of c-Fos protein expression (12). Genetic studies in mice have provided compelling evidence for the role of AP-1 family members in skeletal development in vivo. For example, transgenic mice overexpressing c-Fos develop chondro-and osteosarcomatous lesions (13,14), whereas knock-out of c-Fos in mice causes osteopetrosis due to an early block of differentiation in the osteoclast lineage (15,16). Another AP-1 family member, ATF-2, also contributes to endochondral ossification; chondrocyte proliferation is reduced in ATF-2-deficient mice, leading to dwarfism and skeletal deformities (17). Additionally, ectopic expression of Jun family members has been shown to perturb chondrocyte maturation (18).
In addition to the regulation of chondrocyte differentiation and cartilage development, it has been suggested that AP-1 may regulate destruction of arthritic cartilage. Arthritis is associated with perturbation of chondrocyte homeostasis and is characterized by loss of the differentiated phenotype (dedifferentiation), apoptotic cell death, stimulation of matrix metalloproteinases, and inflammation (19). Interleukin (IL)-1␤ which is produced by chondrocytes, synovial fibroblasts, and inflammatory cells, is a major catabolic pro-inflammatory cytokine involved in cartilage destruction (20). IL-1 ␤ causes chondrocyte dedifferentiation by suppressing cartilage-specific type II collagen and the onset of fibroblastic type I collagen expression. IL-1 ␤ also induces expression of cyclooxygenase (COX)-2, a critical pro-inflammatory enzyme that converts arachidonic acid to prostaglandins, which have been implicated in the pain and inflammation of arthritic disease (21,22). Inhibition of IL-1␤-stimulated AP-1 down-regulates matrix metalloproteinase gene expression in articular chondrocytes (23). In addition, overexpression of c-Fos inhibits proteoglycan synthesis in articular chondrocytes (24), and COX-2 expression is c-Jun-dependent in chondrocytic cells (25). These observations together support the possibility that AP-1 is involved in cartilage destruction.
Although AP-1 appears to play an essential role in chondrocyte differentiation and cartilage formation during embryonic development, the exact roles and underlying molecular mechanisms of AP-1 and its components in the maintenance of chondrocytic differentiation and cartilage homeostasis are not clearly understood. In this study we investigated the role of c-Jun and AP-1 in IL-1␤-induced chondrocytic dedifferentiation and COX-2 expression. We found that increased expression of c-Jun and its phosphorylation by c-Jun N-terminal kinase (JNK) triggered activation of AP-1, leading to IL-1␤-induced dedifferentiation of primary cultured articular chondrocytes. In contrast, we found that IL-1␤-induced COX-2 expression was independent of c-Jun/AP-1 signaling.

MATERIALS AND METHODS
Culture and Treatment of Rabbit Articular Chondrocytes-Articular chondrocytes were isolated from cartilage slices of 2-week-old New Zealand White rabbits by enzymatic digestion, as previously described (26). Briefly, cartilage slices were enzymatically dissociated in 0.2% collagenase type II (381 units/mg solid, Sigma) in Dulbecco's modified Eagle's medium (Invitrogen). Individual cells were suspended in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) bovine calf serum, 50 g/ml streptomycin, and 50 units/ml penicillin. Cells were plated on culture dishes at a density of 5 ϫ 10 4 cells/cm 2 . The medium was replaced every 2 days, and cells reached confluence after approximately 5 days. For experiments, 3.5-day cell cultures were treated with IL-1␤, retinoic acid (RA), or epidermal growth factor (EGF) as indicated. For inhibitor studies, chemical inhibitors were added 1 h before IL-1␤ treatment. SP600125 (Calbiochem) was used to inhibit JNK (27), whereas N-acetyl-L-cysteine (NAC) (28) or nordihydroguaiaretic acid (NDGA) (29) were used to inhibit AP-1 activity. Dedifferentiation of primary cultured chondrocytes was determined by examining the suppression of type II collagen and the onset of type I collagen expression with reverse transcription (RT)-PCR or Western blot analysis with a mouse monoclonal anti-type II collagen antibody (MAB8887, Chemicon, Temecula, CA).
Immunofluorescence Microscopy-Immunofluorescence microscopy was used to determine the expression patterns of type I collagen, type II collagen, COX-2, c-Jun, and phosphorylated c-Jun (pc-Jun) in primary cultured articular chondrocytes. Briefly, primary cultured chondrocytes were fixed with 3.5% paraformaldehyde in phosphate-buffered saline for 10 min at room temperature. Cells were permeabilized and blocked in phosphate-buffered saline containing 0.1% Triton X-100 and 5% fetal calf serum for 30 min. Fixed cells were washed with phosphatebuffered saline and incubated for 1 h with the following primary antibodies: mouse monoclonal anti-murine collagen-I (SP1.D8, Developmental Studies Hybridoma Bank, University of Iowa), mouse monoclonal anti-chicken collagen-II (Chemicon, MAB8887), rabbit polyclonal anti-human c-Jun (sc-1694, Santa Cruz Biotechnology Inc., Santa Cruz, CA), mouse monoclonal anti-mouse c-Jun (J31920, BD Transduction Laboratories), mouse monoclonal anti-phosphorylated c-Jun at serine 63 (Santa Cruz, sc-822), or rabbit polyclonal anti-murine COX-2 (160106, Cayman Chemical, Ann Arbor, MI). The cells were washed and incubated with rhodamine-or fluorescein-conjugated secondary antibodies, washed again, and then observed under a standard fluorescence microscope. Cell nuclei were identified with 4,6-diamidino-2-phenylindole staining.
Immunoprecipitation-Chondrocytes were lysed in Nonidet P-40 lysis buffer (50 mM Tris, pH 8.0, 1% Nonidet P-40, 150 mM NaCl) containing inhibitors of proteases and phosphatases as described above. After incubation on ice for 30 min, the lysates were centrifuged at 13,000 ϫ g for 10 min at 4°C to remove cell debris. Supernatant containing 1000 g of protein was precleared by incubating with 25 l of protein A-Sepharose for 1 h. After centrifugation, proteins in the supernatant were incubated with 2 g of antibody against p300 (Santa Cruz, sc-585). The immune complex was precipitated by the incubating with 25 l of protein A-Sepharose for 2 h at 4°C. After washing with lysis buffer, the immunocomplex was fractionated by SDS-polyacrylamide gel electrophoresis. Immunoprecipitated p300 and associated Sox-9 or c-Jun was detected by Western blotting.

Construction of Expression Vectors and Reporter
Genes-A cDNA for wild-type c-Jun was RT-PCR-amplified from ICR mouse mRNA with specific primers (sense 5Ј-CAG GAT CCG TTC TAT GAC TGC AAA GAT G-3Ј and antisense 5Ј-CAG AAT TCA GCC CTG ACA GTC TGT TCT C-3Ј) designed to introduce BamH1 and EcoR1 restriction sequences at the 5Ј and 3Ј ends of the amplified fragment, respectively. The resulting cDNA was cloned into the BamH1 and EcoR1 sites of the vector, pcDNA3.1(ϩ) (Invitrogen). Sox-9 expression vector was constructed in pcDNA3.1 vector by using RT-PCR-amplified Sox-9 cDNA from newborn mouse rib chondrocytes with specific primers (sense 5Ј-CGG GAT CCG CCA CCA TGA ATC TCC TGG ACC-3Ј and antisense 5Ј-CGG AAT TCC TCA AGG TCG AGT GAG CTG T-3Ј). A SalI-SphI DNA fragment (about 8.23 kilobase) of the rabbit COL2A1 gene was cloned in pBluescript II phagemid vector (Stratagene). This clone contains promoter region, exon-1, intron-1, exon-2, and part of intron-2 (from Ϫ3591 to ϩ4639). The cloned DNA fragment was digested with NaeI to produce a 3.5-kilobase fragment that contains promoter region, exon-1, and ϳ60% of the first intron (from Ϫ1239 to ϩ2260). The included intron-1 contains the enhancer element that binds to Sox-9 transcription factor (31). The 3.5-kilobase COL2A1 promoter gene was inserted into the SmaI site of the pGL3-basic luciferase vector (Promega, Madison, WI) and was designated as Coll II-Pro/Enh gene. A Sox-9 reporter gene, which contains the 48-bp Sox-9 binding site in the first intron of human COL2A1 (31), was constructed by inserting a chemically synthesized 48-bp sequence into pGL3-promoter luciferase reporter vector (Promega). The reporter plasmids were sequenced to confirm the accuracy of the constructions.
Transfection and Reporter Gene Assay-The expression vectors and reporter genes were transfected into articular chondrocytes (day 2.5 cultures with ϳ60% confluency) by using the Lipofectamine reagent (Invitrogen) as previously described (32). The transfected cells were maintained in complete medium for 48 h and used for further analysis as indicated in each experiment. For the reporter gene assay, Coll II-Pro/Enh or Sox-9 reporter genes (2 g) were co-transfected with pCMV-␤-galactosidase expression vector (0.3 g) as an internal control of transfection efficiency with or without indicated amount of Sox-9 expression vector. After incubation of the transfected cells in complete medium for 24 h, the cells were treated with IL-1␤ for an additional 24 h. Luciferase activity was normalized against ␤-galactosidase activity.
AP-1 Activity Assay-The DNA binding activity of AP-1 was determined using an AP-1 enzyme-linked immunosorbent assay kit essentially as instructed by the manufacture (Active Motif North America, Carlsbad, CA). Briefly, chondrocytes treated with various reagents as specified in each experiment were lysed in 10 mM Hepes buffer, pH 7.9, containing 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol, and inhibitors of proteases as described above. After the addition of 0.6% (v/v) Nonidet P-40, the cells were incubated for 15 s on ice and then centrifuged at 13,000 ϫ g for 30 s at 4°C. The pellet was suspended in the supplied nuclear lysis buffer and centrifuged at 13,000 ϫ g for 10 min at 4°C. Nuclear protein (10 g) was loaded into the 96-wells of an enzyme-linked immunosorbent assay plate precoated with an oligonucleotide containing the sequence 5Ј-TGAGTCAG-3Ј and incubated for 1 h at room temperature. Mutated c-Jun oligonucleotides supplied in the kit were used as specificity controls. AP-1 binding to the nucleotide was detected with an anti-phospho-c-Jun antibody and horseradish peroxidase-conjugated secondary antibody followed by colorimetric analysis.

c-Jun Mediates IL-1␤-induced Dedifferentiation but Not COX-2 Expression in Articular
Chondrocytes-IL-1␤ is a major pro-inflammatory cytokine involved in cartilage destruction processes, such chondrocyte dedifferentiation and inflammatory responses. As shown in Fig. 1, RT-PCR analysis revealed that IL-1␤ treatment of primary cultured articular chondrocytes led to suppression of cartilage-specific type II collagen expression and induction of fibroblastic type I collagen expression (Fig. 1, A and  B), two hallmarks of chondrocyte dedifferentiation. Immunofluorescence microscopy further indicated that IL-1␤ treatment dramatically increased the number of type I collagen-expressing cells with a concomitant decrease in type II collagen-expressing cells (Fig. 1C). Western blot analysis (Fig. 1, A and B) also revealed that IL-1␤ treatment induced expression of COX-2, a primary mediator of cartilage inflammation. Expressed COX-2 is localized mainly in perinuclear region (Fig. 1C), which is similar to the report by Kojima et al. (33). These results are consistent with the accepted function of IL-1␤ in dedifferentiation and induction of COX-2 expression in chondrocytes.
In an attempt to elucidate the role of c-Jun in IL-1␤-induced dedifferentiation and COX-2 expression, we first determined the expression, phosphorylation, and subcellular localization of c-Jun in IL-1␤-treated chondrocytes. As shown in Fig. 2, A and B, IL-1␤ treatment caused transient increases in the expression levels (both transcript and protein) and phosphorylation of c-Jun in a dose-and time-dependent manner. Induction of c-Jun expression and its phosphorylation were detectable within 30 min after IL-1␤ treatment and peaked at 1 h. Immunofluorescence microscopy revealed that the expressed c-Jun was localized mainly in the chondrocytic nuclei (Fig. 2C), which is consistent with its function as a component of the AP-1 transcription factor.
To examine whether increased c-Jun expression was required for IL-1␤-induced dedifferentiation and COX-2 expression, we induced chondrocytes to express ectopic wild-type or dominant negative c-Jun and examined the effects of IL-1␤ treatment in these cells. As shown in Fig. 3A, overexpression of wild-type c-Jun was associated with a reduction in type II collagen expression level but did not appear to alter COX-2 expression. Double staining for type II collagen and c-Jun in chondrocytes transfected with wild-type c-Jun indicated that Expression levels of type I collagen (COL1A1) and type II collagen (COL2A1) were determined by RT-PCR, and COX-2 levels were determined by Western blotting. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and ERK were used as loading controls. C, primary cultured chondrocytes were treated with 5 ng/ml IL-1␤ for 24 h. Expression of type I collagen, type II collagen, and COX-2 were determined by immunofluorescence microscopy (upper panels). Cells were identified by 4,6-diamidino-2-phenylindole staining of nuclei (lower panels). The data show the results of a typical experiment from four independent experiments. cells expressing ectopic c-Jun were negative for type II collagen staining (Fig. 3B). Ectopic expression of c-Jun also potentiated the IL-1␤-induced blockade of type II collagen expression (Fig.  3C). However, cells expressing ectopic c-Jun did not show induction of COX-2 expression (Fig. 3B) nor was IL-1␤-induced COX-2 expression altered in these cells (Fig. 3C). The role of c-Jun was further investigated by ectopic expression of dominant negative c-Jun prior to IL-1␤ treatment. As shown in Fig.  4A, dominant negative c-Jun expression reduced IL-1␤-induced phosphorylation of c-Jun and reversed the IL-1␤-induced suppression of type II collagen expression. Immunofluorescence microscopy revealed that cells expressing dominant negative c-Jun expressed high levels of type II collagen (Fig. 4B, upper  panel), indicating that dominant negative c-Jun blocked IL-1␤induced suppression of type II collagen expression. Consistent with the results shown in Figs. 1C and 2C, levels of COX-2 in IL-1␤-treated chondrocytes were heterogeneous, and the expressed COX-2 was localized mainly in nuclear envelope. As shown in Fig. 4B, lower panel, cells overexpressing dominant negative c-Jun did not show significantly altered expression pattern of COX-2. These results collectively suggest that c-Jun expression is sufficient to cause abrogation of type II collagen expression, whereas it is not involved in the induction of COX-2 expression in articular chondrocytes.

c-Jun Phosphorylation by JNK and Activation of AP-1 Mediates IL-1␤-induced Dedifferentiation of Chondrocytes-Having
established that expression of c-Jun is necessary for IL-1␤-induced dedifferentiation of chondrocytes, we next examined whether this process also requires the phosphorylation of c-Jun. We first examined the involvement of JNK in IL-1␤-induced c-Jun phosphorylation, because JNK is one of the major protein kinases responsible for phosphorylating c-Jun. Western blot analysis of JNK phosphorylation (which represents JNK activation) revealed that IL-1␤ treatment caused a transient activation of JNK that apparently preceded the expression and phosphorylation of c-Jun (Fig. 5A). Pretreatment of cells with a JNK inhibitor (SP600125) blocked IL-1␤-induced JNK activation, as evi-denced by inhibition of its phosphorylation (Fig. 5B). More specifically, JNK inhibition by treatment with SP600125 did not affect IL-1␤-induced c-Jun expression but dose-dependently blocked c-Jun phosphorylation (Fig. 5B). Consistent with the above results, inhibition of JNK in IL-1␤-treated chondrocytes inhibited the suppression of type II collagen expression but did not affect COX-2 expression (Fig. 5B). Immunofluorescence microscopy also demonstrated that pretreatment with SP600125 blocked IL-1␤-induced c-Jun phosphorylation and inhibited suppression of type II collagen expression but did not block induction of COX-2 expression (Fig. 5C).
Because the above results suggest that JNK-mediated c-Jun phosphorylation is necessary for IL-1␤-induced chondrocytic dedifferentiation and c-Jun is a component of the AP-1 transcription factor, we next examined whether AP-1 activity is involved in IL-1␤-induced dedifferentiation of chondrocytes. AP-1 activity (i.e. the DNA binding activity of c-Jun) was determined using a highly specific and sensitive AP-1 enzymelinked immunosorbent assay kit. We found that treatment with IL-1␤ caused a transient increase in the transcriptional activity of AP-1; this effect peaked at 1 h after IL-1␤ treatment (Fig.  6A) and was dose-dependent (Fig. 6B). As expected, inhibition of JNK by SP600125 treatment blocked IL-1␤-induced AP-1 activation (Fig. 6C). To further examine the role of AP-1 activity, we pretreated chondrocytes with NAC or NDGA to inhibit AP-1 activity. NDGA inhibits lipoxygenase and AP-1 (29). NAC, an antioxidant and precursor of glutathione, is known to inhibit NFB and AP-1 (28). Pretreatment with NAC or NDGA blocked IL-1␤-induced activation of AP-1 transcriptional activity in a dose-dependent manner (Fig. 6D). NAC and NDGA also blocked c-Jun phosphorylation and reversed IL-1␤-induced suppression of type II collagen expression but not induction of COX-2 expression (Fig. 6, E and F). Taken together, the above results suggest that JNK-mediated c-Jun phosphorylation and stimulation of AP-1 transcriptional activity mediates IL-1␤induced suppression of type II collagen expression but not induction of COX-2 expression in articular chondrocytes.
c-Jun/AP-1 Regulates Sox-9 Expression-In an attempt to elucidate the regulatory mechanism of c-Jun/AP-1 in suppression of type II collagen expression, we examined the role of c-Jun/AP-1 in expression of Sox-9, a major transcription factor that regulates type II collagen (COL2A1) expression. IL-1␤ treatment reduced the transcript and protein levels of Sox-9 (Fig. 7A), and this effect was blocked by inhibition of JNK with SP600125 or by inhibition of AP-1 activity with NAC or NDGA (Fig. 7B). Additionally, ectopic expression of c-Jun was sufficient to reduce transcript and protein levels of Sox-9 (Fig. 7C), which is consistent with the reduction of type II collagen expression levels (Fig. 3A). To further clarify the relationship between c-Jun/AP-1 and Sox-9 in the regulation of type II collagen expression, we induced chondrocytes to express ectopic Sox-9 and examined the effects of IL-1␤ treatment in these cells on expression and promoter activity of COL2A1. We constructed two reporter genes for this purpose. One is Coll II-Pro/Enh gene, which contains promoter region, exon-1, and ϳ60% of the first intron. The included intron-1 contains enhancer element that binds to Sox-9 transcription factor (31). The other is a Sox-9 reporter gene, which contains only the 48-bp Sox-9 binding site in the first intron of human COL2A1 (Fig. 7E). As determined by reporter gene assay and Western blot analysis, IL-1␤ significantly reduced the promoter activity of COL2A1 (Fig. 7F), transcriptional activity of Sox-9 (Fig. 7G), and expression of type II collagen transcript and protein (Fig.  7D). The effects of IL-1␤ were completely blocked by the ectopic expression of Sox-9 as shown in Fig. 7, D, F, and G. The results collectively suggest that expression and phosphorylation of c-Jun and subsequent stimulation of AP-1 transcriptional activity suppresses Sox-9 expression, which in turn leads to inhibi- tion of type II collagen expression.
We next examined whether CBP/p300 is involved in the regulation of type II collagen expression based on the observations that this molecule interacts with both c-Jun and Sox-9 and that the interactions are necessary for the action of these transcription factors (34,35). IL-1␤ treatment, which causes induction of c-Jun expression and suppression of Sox-9 expression, did not alter the protein levels of p300 (Fig. 8, A and B). As determined by co-immunoprecipitation assay, c-Jun is associated with p300 in cells treated with IL-1␤ for 1 or 24 h (Fig. 8C, upper panels). p300 also interacted with Sox-9 both in control and IL-1␤-treated (for 1 h) cells. Co-immunoprecipitation of Sox-9 and p300 was not detected in cells treated with IL-1␤ for 24 h (Fig. 8C, lower panels) probably because of the downregulation of Sox-9 by IL-1␤ treatment. Taken together, our results suggest that suppression of Sox-9 expression rather than regulation of Sox-9 activity by p300 is associated with inhibition of type II collagen expression.
Involvement of c-Jun/AP-1 in RA-and EGF-induced Dedifferentiation of Chondrocytes-Last, we examined whether c-Jun/ AP-1 is a common mediator of chondrocyte dedifferentiation by examining c-Jun and AP-1 during chondrocyte dedifferentiation induced in other ways, including treatment with RA and EGF, or a serial monolayer subculture. Similar to the dedifferentiation of chondrocytes caused by IL-1␤ treatment, RA-or EGF-induced dedifferentiation, which was demonstrated by the suppression of type II collagen expression, was associated with increased expression of c-Jun, phosphorylation of c-Jun and JNK, and inhibition of Sox-9 expression (Fig. 9A and B). In contrast, subculture-induced dedifferentiation of chondrocytes was not associated with changes in the expression levels or phosphorylation of c-Jun and JNK (Fig. 9C), suggesting that c-Jun/AP-1 is involved in the dedifferentiation of chondrocytes caused by IL-1␤, RA, and EGF but not by serial monolayer culture. DISCUSSION IL-1␤ is a pro-inflammatory cytokine that induces several mediators of cartilage degradation and plays a pivotal role in the pathogenesis of arthritis (20). IL-1␤ causes dedifferentiation of chondrocytes by suppressing type II collagen expression and inducing expression of types I and III collagen, which contribute to the destruction of arthritic cartilage. Articular chondrocytes express low levels of c-Jun, which are dramatically induced by IL-1␤. Our gain-of-function and loss-of-function studies clearly indicated that increased expression of c-Jun is sufficient to abrogate type II collagen expression. Our results further indicated that IL-1␤ suppresses type II collagen expression in articular chondrocytes by inducing expression of c-Jun and its phosphorylation by JNK, AP-1 activation, and subsequent suppression of Sox-9, as depicted in Fig. 10. Therefore, unlike the requirement of Jun-containing AP-1 complexes in cartilage formation during embryonic development (8,9), our current study suggests that low expression levels and activity of Jun are required for the maintenance of homeostasis in developed articular cartilage.
We also demonstrated that phosphorylation of c-Jun by JNK mediates IL-1␤-induced cessation of type II collagen expression. It has been shown that in chondrocytes, IL-1␤ activates all subgroups of the mitogen-activated protein kinases, including extracellular signal-regulated protein kinase (ERK), p38 kinase, and JNK. Activation of these mitogen-activated protein kinase subgroups has been shown to activate the AP-1 transcription factor; activation of ERK and JNK was reported to stimulate c-Fos and c-Jun, respectively (23, 36 -38). Here, we observed that treatment of chondrocytes with IL-1␤ led to activation of ERK, p38 kinase, and JNK (data not shown). Among the mitogen-activated protein kinase subtypes, inhibition of JNK blocked the IL-1␤-induced effects on c-Jun phos- . WB, Western blotting. Alternatively, chondrocytes were transfected with the indicated amounts of expression vector carrying wild-type c-Jun and were cultured for 36 h (C). E, two reporter genes, Coll II-Pro/Enh, which contains promoter region, exon-1, and ϳ60% of the first intron including Sox-9 binding site of rabbit COL2A1, and Sox-9 reporter gene, which contains the 48-bp Sox-9 binding site in the first intron of human COL2A1, were constructed in pGL3-basic luciferase vector and pGL3-promoter luciferase reporter vector, respectively. D, F, and G, chondrocytes were transfected with empty vector (D), Coll II-Pro/Enh reporter gene (F), or Sox-9 reporter gene (G) with or without the indicated amount of Sox-9 expression vector. After culture in complete medium for 24 h, the cells were exposed to 5 ng/ml IL-1␤ for an additional 24 h. Levels of Sox-9 and type II collagen were determined by Western blotting and RT-PCR (D). COL2A1 promoter activity (F) and transcriptional activity of Sox-9 (G) were determined by reporter gene assay. Data are presented as the results of a typical experiment or the mean values with S.D. from four independent experiments. phorylation and AP-1 activation (Figs. 5 and 6), indicating that IL-1␤-induced JNK activation phosphorylates c-Jun in articular chondrocytes. It is well known that phosphorylation of c-Jun by JNK is required for the formation of a transcriptionally active c-Jun complex (4). Here, our observation that inhibition of JNK blocked the DNA binding activity of c-Jun further indicates that c-Jun phosphorylation is necessary for the formation of an active AP-1 complex (Fig. 6). The binding partner for phosphorylated c-Jun in the active AP-1 complex is currently unknown. AP-1 consists of a variety of dimers composed of Jun and Fos proteins. Members of the Fos protein family can only heterodimerize with members of the Jun family, whereas Jun proteins can both homodimerize and heterodimerize with other Jun or Fos family members to form transcriptionally active complexes. Based on the dimer protein pairings, tran-scription can be either positively or negatively modulated (3). Therefore, it is possible that phosphorylated c-Jun may homodimerize or heterodimerize with other Jun or Fos members to form an active AP-1 complex.
Although the mechanisms by which AP-1 activity abrogates type II collagen expression remain to be more clearly elucidated, our results suggest that suppression of Sox-9 expression by AP-1 activity contributes to the inhibition of type II collagen expression. This conclusion was further supported by the observation that ectopic expression of Sox-9 blocked IL-1␤-induced inhibition of type II collagen expression. Sox-9 is a master transcription factor responsible for controlling the differentiation of mesenchymal cells into chondrocytes. Sox-9 binds to the chondrocyte-specific enhancer element in the gene for proalpha1 (II) collagen (COL2A1) (39). Murakami et al. (40) revealed that IL-1␤ suppresses Sox-9 expression, prompting us to speculate that c-Jun/ AP-1-mediated IL-1␤-induced suppression of Sox-9 expression results in abrogation of type II collagen expression. One of interesting findings of this study is that c-Jun/AP-1 signaling appears to mediate dedifferentiation of chondrocytes caused by IL-1␤, RA, or EGF, whereas dedifferentiation of chondrocytes caused by the serial monolayer subculture on plastic was found to be independent of c-Jun/AP-1. Indeed, it has been suggested that activation of the Sp-1 transcription factor is associated with subculture-induced dedifferentiation of chondrocytes (41), suggesting that Sp-1 and AP-1 may regulate different pathways of chondrocyte dedifferentiation.
In contrast to the regulation of type II collagen expression, our results indicate that IL-1␤-induced induction of COX-2 expression is independent of c-Jun/AP-1 activity. However, a previous study showed that c-Jun/JNK regulates shear-induced COX-2 expression in the T/C28a2 chondrocytic cell line (25). This discrepancy may be due to differences in the cell FIG. 8. p300 is not associated with suppression of type II collagen expression. A and B, chondrocytes were treated with 5 ng/ml IL-1␤ for the indicated period (A) or for 24 h with the indicated amounts of IL-1␤ (B). Levels of p300, Sox-9, and c-Jun were determined by Western blotting. ERK was used as the loading control. C, chondrocytes were untreated as a control (Con) or treated with 5 ng/ml IL-1␤ for 1 or 24 h. Binding of p300 with c-Jun or Sox-9 was determined by immunoprecipitation (IP) of p300 and Western blotting (WB) of c-Jun or Sox-9. The data show the results of a typical experiment from at least four independent experiments. types and/or the applied extracellular stimuli. Although inhibition of JNK did not affect IL-1␤-induced COX-2 expression in this study, we observed that IL-1␤-induced COX-2 expression was blocked by the inhibition of ERK with PD98059, inhibition of p38 mitogen-activated protein kinase with SB203580, or inhibition of nuclear factor B with SN50 peptide treatment (data not shown). This suggests that the ERK, p38 kinase, and nuclear factor B pathways regulate IL-1␤-induced COX-2 expression, whereas IL-1␤-induced inhibition of type II collagen expression is regulated by the JNK and c-Jun/AP-1 pathway.
In summary, our results collectively indicate that c-Jun expression and its phosphorylation by JNK as well as AP-1 activation and subsequent suppression of Sox-9 play essential roles in IL-1␤-induced suppression of type II collagen expression in articular chondrocytes. Therefore, it appears that the c-Jun/AP-1 transcription factor regulates not only chondrocyte differentiation and cartilage formation during embryonic development but also cartilage destruction via chondrocyte dedifferentiation.