Identification in human osteoarthritic chondrocytes of proteins binding to the novel regulatory site AGRE in the human matrix metalloprotease 13 proximal promoter.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 54, No. 8, August 2006, pp 2471–2480 DOI 10.1002/art.21961 © 2006, American College of Rheumatology Identification in Human Osteoarthritic Chondrocytes of Proteins Binding to the Novel Regulatory Site AGRE in the Human Matrix Metalloprotease 13 Proximal Promoter Zhiyong Fan,1 Ginette Tardif,1 Christelle Boileau,1 Joseph P. Bidwell,2 Changshan Geng,1 David Hum,1 Alexander Watson,1 Jean-Pierre Pelletier,1 Martin Lavigne,3 and Johanne Martel-Pelletier1 Objective. Matrix metalloprotease 13 (MMP-13) plays a major role in osteoarthritic (OA) processes. We previously identified the AG-rich element (AGRE) regulatory site (GAAAAGAAAAAG) in the proximal promoter of this gene. Electrophoretic mobility shift assays (EMSAs) done with nuclear extracts from OA chondrocytes showed the presence of 2 AGRE protein–binding complexes, the formation of which depended on the pathophysiologic state (high or low) of the cells; the low OA (L-OA) chondrocytes have low MMP-13 basal levels and high interleukin-1␤ (IL-1␤) inducibility, and the high OA (H-OA) chondrocytes have high MMP-13 basal levels and low IL-1␤ inducibility. In this study, we sought to determine the importance of individual AGRE bases in promoter activity and to identify AGRE binding proteins from L-OA and H-OA chondrocyte complexes. Methods. Promoter activity was determined following transient transfection into human OA chondrocytes. AGRE binding proteins were identified by mass spectroscopy. Results. Individual mutations of the AGRE site differentially modulated promoter activity, indicating that the intact AGRE site is required for optimal MMP-13 expression. Damage-specific DNA binding protein 1 (DDB-1) was identified in the L-OA chondrocyte–binding complex. EMSA experiments performed with the mutation of the left AGRE site (GTGCTGAAAAAG) and nuclear extracts of L-OA chondrocytes reproduced the pattern seen in the H-OA chondrocytes. Mass spectroscopy identified p130cas as one of the proteins in this complex. Supershift experiments showed the presence of p130cas and nuclear matrix transcription factor 4 (NMP-4) in the wild-type AGRE/H-OA chondrocyte complex. Conclusion. These data suggest that the binding of p130cas and NMP-4 to the AGRE site regulates MMP-13 expression and may trigger the change in human chondrocytes from the L-OA state to the H-OA state. Matrix metalloprotease 13 (MMP-13; collagenase 3) is a potent protease capable of degrading a wide range of collagenous and noncollagenous extracellular matrix macromolecules, including types I, II, III, IV, IX, X, and XI collagens (with higher activity for type II collagen), gelatin, laminin, tenascin, aggrecan, fibrinogen, and connective tissue growth factor (1–8). In contrast to many human MMPs, the distribution pattern of MMP-13 is restrictive in normal tissues and selective in tissues with pathologic conditions. The physiologic expression of MMP-13 is limited to situations in which rapid and effective remodeling of collagen is required, such as in fetal bone development and postnatal bone remodeling (9,10). In pathologic conditions, MMP-13 is expressed at sites of excessive degradation of the extra- Supported by a grant from the Canadian Institutes of Health Research (CIHR) and by a student bursary funded in part by an unrestricted grant from Amgen Canada and in part by a grant from the CIHR/MENTOR training program. 1 Zhiyong Fan, MD, MSc, Ginette Tardif, PhD, Christelle Boileau, PhD, Changshan Geng, MD, MSc, David Hum, MSc, Alexander Watson, MSc, Jean-Pierre Pelletier, MD, Johanne MartelPelletier, PhD: Centre Hospitalier de l’Université de Montréal, Hôpital Notre-Dame, Montreal, Quebec, Canada; 2Joseph P. Bidwell, PhD: Indiana University School of Medicine, Indianapolis; 3Martin Lavigne, MD: Hôpital Maisonneuve-Rosemont, Montreal, Quebec, Canada. Address correspondence and reprint requests to Johanne Martel-Pelletier, PhD, Osteoarthritis Research Unit, Centre Hospitalier de l’Université de Montréal, Hôpital Notre-Dame, 1560 rue Sherbrooke Est, Montreal, Quebec H2L 4M1, Canada. E-mail: email@example.com. Submitted for publication September 28, 2005; accepted in revised form April 6, 2006. 2471 2472 cellular matrix, such as in cartilage from patients with osteoarthritis (OA) and patients with rheumatoid arthritis (RA) (2–4,11,12), in RA synovium (13), and in various carcinomas (14). MMP-13 expression is modulated by numerous agents, including proinflammatory cytokines (interleukin-1␤ [IL-1␤], tumor necrosis factor ␣, IL-17), growth factors (transforming growth factor ␤, hepatocyte growth factor), fibronectin fragments, and mechanical stimuli (2,15–20). The signals triggered by these factors are relayed to the nucleus and activate different transcription factors. Many binding sites for transcription factors have been identified in the human MMP-13 promoter (21–23). The activator protein 1 (AP-1), Ets/ polyomavirus enhancer 3 (PEA-3), and osteoblastspecific element 2 (OSE-2) sites are present in the proximal promoter. Functional analysis of the promoter revealed that the AP-1 site is essential for both basal and proinflammatory cytokine–induced transcription and that the PEA-3 site exerts a cooperative effect (17,18). The OSE-2 site, which binds core-binding factor ␣1 (CBF␣1), is required for IL-1␤ induction of MMP-13 in stably transfected chondrocytic cells (24). Recently, we identified in the human MMP-13 proximal promoter a novel regulatory element GAAAAGAAAAAG, designated AG-rich element (AGRE), that appears to be involved in basal expression of MMP-13 (25). Interestingly, the AGRE site is not found as such in the proximal promoter sequence of other human MMP genes, although the proximal promoter region of some MMPs contains AGRE-like sequences. In a previous study (26), we identified 2 subpopulations of OA chondrocytes: low (L), which have a low basal level of MMP-13 and high IL-1␤ inducibility, and high (H), which have a high basal level of MMP-13 and low IL-1␤ inducibility (26). We have found, by electrophoretic mobility shift assay (EMSA) experiments, that different proteins bind to the AGRE site and form 2 different migrating complexes, depending on whether the nuclear proteins are extracted from OA chondrocytes of the slow-migrating complex (L) or the fastmigrating complex (H) state (25). In this study, we examined MMP-13 regulation at the AGRE site and identified AGRE binding proteins from L-OA and H-OA chondrocyte complexes. We found that the intact AGRE site is required for optimal expression of MMP-13. Damage-specific DNA binding protein 1 (DDB-1) was identified in the AGRE complex formed in L-OA chondrocyte complexes, and p130cas and nuclear matrix transcription factor 4 (NMP-4) were identified in the H-OA chondrocyte complexes. These FAN ET AL findings are important and may explain the differential regulation of MMP-13 expression in human OA, in addition to providing a new basis for the rationalization of a therapeutic strategy. MATERIALS AND METHODS Specimen selection. Articular cartilage samples (n ⫽ 36) from femoral condyles and tibial plateaus and synovial membrane samples (n ⫽ 25) were obtained from OA patients (mean ⫾ SD age 71 ⫾ 9 years) who were undergoing total knee arthroplasty. All patients had been evaluated by a certified rheumatologist and diagnosed as having OA according to the American College of Rheumatology criteria (27). The specimens represented moderate-to-severe OA, as defined according to macroscopic criteria (28). At the time of surgery, the patients had symptomatic disease requiring medical treatment (acetaminophen, nonsteroidal antiinflammatory drugs, or selective cyclooxygenase 2 inhibitors). The institutional Ethics Committee Board of the Hôpital Notre-Dame approved the use of the human articular tissues. Cell culture. Chondrocytes and synovial fibroblasts were released from full-thickness strips of cartilage and synovial membrane, respectively, followed by sequential enzymatic digestion at 37°C as previously described (2). The chondrocytes were extracted from whole OA cartilage. The isolated cells were seeded at high density (105 cells/cm2) in tissue culture flasks and cultured to confluence in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fetal calf serum and an antibiotics mixture (100 units/ml of penicillin base and 100 g/ml of streptomycin base) (all from GibcoBRL Life Technologies, Burlington, Ontario, Canada) at 37°C in a humidified atmosphere of 5% CO2/95% air. To ensure phenotype, only first-passage cultured chondrocytes were used. Synovial fibroblasts were used at the second or third passage. Identification of L-OA and H-OA chondrocytes. L-OA and H-OA chondrocytes were identified as described previously (18,26), by incubating the cells for 48 hours in the presence or absence of IL-1␤ (100 pg/ml). The MMP-13 released into the culture medium was quantified by a specific enzyme-linked immunosorbent assay (Amersham Biosciences, Baie d’Urfé, Quebec, Canada). Chondrocytes with low basal levels of MMP-13 and a ⬎2-fold increase following treatment with IL-1␤ were classified as L-OA. Those with high basal levels of MMP-13 and a ⬍2-fold increase following treatment with IL-1␤ were classified as H-OA. Of note, this classification discriminates specimens from different patients but not cells from the same patient. The proportion of L-OA pattern chondrocytes to H-OA pattern chondrocytes is ⬃10:1. Plasmids and site-directed mutagenesis. Plasmid –175Luc consists of the first proximal 175 bp of the human MMP-13 promoter containing the OSE-2, AGRE, PEA-3, and AP-1 sites as well as the TATA box cloned into the pGL3 luciferase reporter plasmid (Promega, Madison, WI) (Figure 1). Plasmid –137Luc is identical to plasmid –175Luc except that it has only the first proximal 137 bp and lacks the OSE-2 site (Figure 1). The human MMP-1 promoter includes basepairs –512 to ⫹63 cloned into the PXP2 luciferase vector (kindly provided IDENTIFICATION OF AGRE BINDING PROTEINS IN HUMAN OA CHONDROCYTES Figure 1. Proximal promoter sequence of the human matrix metalloprotease 13 (MMP-13) gene. The transcription start site (⫹1) is indicated by an arrow. The locations of the binding sites are numbered from the transcription start site and are underlined. The MMP-13 promoter sequence was linked to the luciferase reporter gene to construct plasmid –175Luc. Plasmid –137Luc is identical to plasmid –175Luc, except that it terminates at the base indicated by the asterisk. STAT ⫽ signal transducer and activator of transcription; OSE-2 ⫽ osteoblast-specific element 2; AGRE ⫽ AG-rich element; PEA-3 ⫽ polyomavirus enhancer 3; AP-1 ⫽ activator protein 1. by Dr. Herman Cheung, University of Miami School of Medicine, Miami, FL) (29). A 501-bp fragment of the MMP-2 proximal promoter was generated by polymerase chain reaction using human chromosomal DNA as substrate and the primers 5⬘-CCCAGGAGCTCTCTGTCCC-3⬘ (sense), and 5⬘-GTAAGATCTGGATGCAGCGGAAACAAG-3⬘ (antisense). A 420-bp fragment of the MMP-14 promoter was similarly generated with the primers 5⬘-ACCATCCCACACTCTGAGCTCCTC-3⬘ (sense), and 5⬘-CTTAGATCTTTCTTCTGCTTAGTCGGCG-3⬘ (antisense). Both sense primers were designed with a Sac I site and antisense primers with a Bgl II site to facilitate cloning into the pGL-3 basic luciferase vector (Promega). The mutations were done using a QuickChange SiteDirected Mutagenesis kit (Stratagene, La Jolla, CA) as described by the manufacturer. All mutated constructs were verified by DNA sequencing using the BigDye Terminator v3.1 Cycle Sequencing kit (Applied Biosystems, Foster City, CA). Transient transfection assays. Transient transfection assays were performed by the calcium phosphate coprecipitation method as previously described (18,21,25). Chondrocytes were seeded at 80% confluency and transfected with a DNA– calcium preparation consisting of 125 mM CaCl2 and 3 g of the plasmids. Plasmid pCMV-␤-galactosidase (VR1412; Vical, San Diego, CA) was used to monitor transfection efficiency. The cells were incubated for 48 hours, after which they were lysed in Reporter lysis buffer (Promega). Luciferase activity was measured with the Lumat LB 9507 luminometer (EG&G Berthold, Bad Wildbad, Germany). Total protein was quantitated by the bicinchoninic acid method (Pierce, Rockford, IL). Promoter activity was calculated as relative luciferase units per microgram of protein and expressed as a percentage of the control, which was assigned a value of 100%. EMSA and supershift assay. EMSA and supershift assays were performed as previously described (18,25). Nuclear proteins (10 g) from human OA chondrocytes were incubated with wild-type and mutated AGRE-containing oligonucleotides end-labeled with ␥32P-ATP. Binding complexes were resolved on nondenaturing 6% polyacrylamide gels, after 2473 which the gels were fixed, dried, and exposed to Kodak XAR-5 film (Eastman Kodak, Rochester, NY). The antibodies used in the supershift reactions were 0.5 g of mouse monoclonal anti-human p130cas (no. 610272; BD Transduction Laboratories, Mississauga, Ontario, Canada) and 2 l of rabbit polyclonal anti-rat NMP-4 (30). DNA affinity chromatography. Magnetic beads (BioMag Nuclease-Free Streptavidin; Qiagen, Valencia, CA) were coupled to wild-type or mutated (GAAGAGACAAAG [boldfacing indicates mutated bases]) 5⬘-biotinylated AGREcontaining oligonucleotides. The double-stranded oligonucleotides were coupled to magnetic beads (100 pmoles/5 mg, in 20 mM Tris HCl, 1 mM EDTA, and 0.02% Triton X-100, pH 7.8), incubated for 15 minutes at room temperature, and washed 3 times with the above buffer. Since preliminary experiments showed that nuclear extracts from all human OA synoviocytes induce the same slow-migrating protein complex seen in the normal and L-OA chondrocytes (25), OA synoviocytes were then used as the source of nuclear extracts for this experiment. Nuclear extracts were incubated with the wild-type and mutated AGRE oligonucleotides/magnetic beads in a Tris buffer (20 mM Tris HCl, 0.5M NaCl, pH 8.0) at 4°C for 30 minutes. The solution was washed extensively with the buffer to remove unbound proteins, heated at 56°C for 5 minutes to release the bound proteins, and the beads were removed with a magnetic field. The samples were desalted, concentrated with an Amicon membrane (Amicon, Stonehouse, UK), precipitated using a 2D Cleaners kit from Amersham Biosciences, dried, and recovered in Destreak Rehydration buffer (Amersham Biosciences). The proteins were separated on 2-dimensional electrophoresis gel using a SureLock system kit (Invitrogen, San Diego, CA). The proteins were first focused on pH 3–10NL strips for a total of 2,200 volts/hour. The strips were then reduced with 2% dithiothreitol, alkylated with 2% iodoacetamide, and equilibrated with sodium dodecyl sulfate. The second dimension was done on a 10% acrylamide Bis–Tris precast gel at 200 volts for 50 minutes. The gels were fixed, stained with silver nitrate (31), and scanned and analyzed with ImageMaster software (Amersham Biosciences). Protein patterns obtained from wild-type and mutated oligonucleotides were compared and were analyzed by mass spectroscopy. Extraction of proteins from EMSA complexes. Protein/ AGRE binding reactions were electrophoresed as described above. The gel was sliced in the zone containing the EMSA binding complex, digested by trypsin (6 ng/l for 5 hours), and processed on the MassSpec robotic workstation (PerkinElmer, Boston, MA). Peptides were extracted in 0.5% formic acid and 9% acetonitrile and analyzed by mass spectrometry (Q-Trap 4000 ion trap) using BioAnalyst 1.4 software (both from Applied Biosystems). The results were submitted to the Mascot search engine (Matrix Science, Boston, MA) for identification and compared with the National Center for Biotechnology Information database. Statistical analysis. Values are expressed as the mean ⫾ SEM. When applicable, statistical analysis was performed using Student’s 2-tailed t-test. P values less than 0.05 were considered significant. 2474 FAN ET AL Figure 2. Effect of different AG-rich element (AGRE) mutations on matrix metalloprotease 13 (MMP-13) promoter activity in the low subpopulation (low basal level of MMP-13 and high interleukin-1␤ inducibility) of human osteoarthritic (OA) chondrocytes (L-OA chondrocytes). Plasmid –175Luc containing wild-type (WT) or mutated (M) AGRE sites was transfected into L-OA chondrocytes (n ⫽ 7–11), and promoter activity was measured by luciferase activity. The promoter activity of the wild-type AGRE-containing plasmid was given an arbitrary value of 100%. Values are the mean and SEM. The mutated bases are indicated in boldface. P values are versus the wild-type, as determined by Student’s 2-tailed t-test. RESULTS Effect of AGRE mutations on basal MMP-13 promoter activity. The AGRE site was mutated at different bases to identify those involved in promoter activity. Plasmid –175Luc was mutated at its AGRE site, L-OA chondrocytes were transfected, and promoter activity was determined. The data showed that most AGRE mutations induced a significant decrease in MMP-13 promoter activity (Figure 2). The highest decrease (90% [P ⬍ 0.0001]) was seen with the mutation in the left side GAAGAGAAAAAG (the M-2 mutation). Mutations in the right side (GAAAAGAAGAAG and GAAAAGACAAAG [the M-5 and M-3 mutations, respectively]) resulted in decreases of 46% (P ⬍ 0.03) and 75% (P ⬍ 0.0001), respectively. Mutations in both sides (GAAGAGACAAAG [the M-1 mutation]) resulted in a 74% decrease in promoter activity (P ⬍ 0.0001). The mutation GAAAAGCAAAAG (the M-4 mutation) reduced promoter activity by 21%. Surprisingly, only the mutation GTGCTGAAAAAG (the M-6 mutation) resulted in a significant increase in promoter activity (P ⬍ 0.0001). Activity of AGRE-like sequences in the promoters of other human MMPs. To determine whether the AGRE site was unique to the human MMP-13 promoter, we searched the GenBank database for its pres- ence in the proximal promoter sequence (up to 400 bp upstream of the reported transcription start site) of several human MMP genes. We identified AGRE-like sequences in some promoters. The single GAAAA motif was found in the promoters of human MMP-1 (⫺237 to ⫺233 bp) (32), MMP-3 (⫺251 to ⫺247 bp) (33), MMP-7 (⫺126 to ⫺122 bp) (34), and MMP-12 (3 copies at positions ⫺409 to ⫺405, ⫺329 to ⫺325, and ⫺305 to ⫺301 bp) (35). The sequence AAAAG was found in MMP-1 (⫺311 to ⫺307 bp) (32), MMP-3 (⫺301 to ⫺297 bp) (33), MMP-7 (⫺89 to ⫺85 bp) (34), MMP-11 (⫺81 to ⫺77 bp) (36), MMP-8 (2 copies at ⫺190 to ⫺186 and ⫺213 to ⫺209 bp) (GenBank accession no. AF059679), MMP-9 (2 copies at ⫺434 to ⫺430 and ⫺506 to ⫺502 bp) (37), and MMP-2 promoters (3 copies at ⫺160 to ⫺156, ⫺140 to ⫺136, and ⫺130 to ⫺126 bp) (38). The sequence CAAAAAG (⫺207 to ⫺201 bp) was found in the MMP-14 promoter (39). To determine whether these motifs have the same transcriptional role as the whole AGRE site, we mutated the AGRE-like sequences in the proximal promoters of MMP-1, MMP-2, and MMP-14 and transfected L-OA chondrocytes with the mutated or wild-type plasmids. The data showed that their mutation did not significantly affect promoter activities, although a slight decrease was noted for the second mutation of the MMP-14 promoter (Figure 3). IDENTIFICATION OF AGRE BINDING PROTEINS IN HUMAN OA CHONDROCYTES Figure 3. Activities of wild-type and mutated AG-rich element–like sequences in the promoters of different matrix metalloproteases (MMPs). The wild-type (WT) and mutated (MUT) plasmids were constructed as described in Materials and Methods, transfected into the low subpopulation (low basal level of MMP-13 and high interleukin-1␤ inducibility) of osteoarthritic (OA) chondrocytes (n ⫽ 5–12), and promoter activity was measured by luciferase activity. Promoter activity of the wild-type plasmids was given an arbitrary value of 100%. The mutated bases are indicated in boldface. Values are the mean and SEM. Figure 4. Effect of wild-type (WT) (ACCACAAACCACA) and mutated (MUT) (ACCGCAAACCGCA) osteoblast-specific element 2 (OSE-2) sites on matrix metalloprotease 13 (MMP-13) promoter activity in the low subpopulation (low basal level of MMP-13 and high interleukin-1␤ inducibility) of human osteoarthritic (OA) chondrocytes (L-OA chondrocytes). Plasmids –175Luc, which contain permutations of wild-type and mutated AR-rich element (AGRE) and OSE-2 sites, were transfected into L-OA chondrocytes (n ⫽ 10). Plasmid –137Luc, which does not have the OSE-2 site, was included as a reference. Promoter activities were measured by luciferase activity. The promoter activity of wild-type –175Luc was given an arbitrary value of 100%. Values are the mean and SEM. P values are versus the wild-type, as determined by Student’s t-test. 2475 2476 Figure 5. Representative electrophoretic mobility shift assay (EMSA) patterns of AG-rich element (AGRE)–binding complexes of osteoarthritic (OA) chondrocytes. A, EMSA performed with the oligonucleotide containing the wild-type (WT) AGRE site (GAGGGAAAAGAAAAAGTCGCC) and nuclear extracts from the low subpopulation (low basal level of matrix metalloprotease 13 [MMP-13] and high interleukin-1␤ [IL-1␤] inducibility) and the high subpopulation (high basal level of MMP-13 and low IL-1␤ inducibility) of OA chondrocytes (L-OA and H-OA chondrocytes, respectively), as described in Materials and Methods. As previously reported (25), the L-OA chondrocytes have a slow-migrating band (L), and the H-OA chondrocytes have a fast-migrating band (H); ns indicates a nonspecific band. B, EMSAs (n ⫽ 4) performed with wild-type or mutated (M) AGREcontaining oligonucleotides and nuclear proteins from L-OA chondrocytes. C, Sequences of the wild-type and mutated AGRE sites. The mutated bases are indicated in boldface. Effect of OSE-2 mutation on basal MMP-13 promoter activity. To evaluate a possible interaction between the proteins binding the AGRE and OSE-2 sites, the –175Luc plasmid was mutated in the OSE-2 site either alone (ACCGCAAACCGCA) or conjointly with AGRE (GAAGAGACAAAG). For comparison purposes, the wild-type –137Luc plasmid (Figure 1) was also included. L-OA chondrocytes were transfected with wild-type and mutated plasmids (Figure 4). The results showed that the mutation in the OSE-2 site alone did not significantly affect MMP-13 promoter activity, but that the AGRE mutation, whether alone or in conjunction with OSE-2, significantly decreased promoter activity (by 70% [P ⬍ 0.0001] and 43% [P ⬍ 0.01], respectively). The promoter activity of –137Luc was slightly lower than that of –175Luc, but the difference was not significant. Protein-binding potential of mutated AGRE sites. The AGRE site was mutated at different bases to identify those involved in the formation of proteinbinding complexes. EMSA experiments were performed FAN ET AL with oligonucleotides containing wild-type and mutated AGRE sites and nuclear extracts from L-OA chondrocytes. The results showed that the migration pattern of the complex associated with H-OA chondrocytes could be reproduced with oligonucleotides containing the AGRE site mutated in its left side and nuclear extracts from L-OA chondrocytes (Figure 5); this fast-migrating complex was more pronounced with the mutation GTGCTGAAAAAG (the M-6 mutation) than with the mutation GAAGAGAAAAAG (the M-2 mutation). This binding complex was specific and depended upon the presence of the right side sequence of AGRE, since the mutation GTGCTGACAAAG abolished the formation of this complex (data not shown). The GTGCTGAAAAAG/binding complex was excised from the gel, and the proteins were analyzed by mass spectroscopy. Two peptides matched the protein p130cas, a docking protein known to bind the transcrip- Figure 6. Supershift assay pattern of AG-rich element (AGRE)– protein complexes of osteoarthritic (OA) chondrocytes. Nuclear extracts from the high subpopulation (high basal level of matrix metalloprotease 13 and low interleukin-1␤ inducibility) of OA chondrocytes (H-OA chondrocytes) were incubated with the wild-type (WT) AGRE-containing oligonucleotide (GAGGGAAAAGAAAAAGTCGCC) in the absence (WT) or presence of specific antibodies against nuclear matrix transcription factor 4 (NMP-4) and p130cas. Competition (Comp) was performed by the addition of a 50-fold molar excess of the cold AGRE oligonucleotide. Arrows indicate the supershifted bands; H indicates the fast-migrating complex of the H-OA chondrocytes; ns indicates a nonspecific band. IDENTIFICATION OF AGRE BINDING PROTEINS IN HUMAN OA CHONDROCYTES tion factor NMP-4/CIZ, which binds the consensus site (G/C)AAAAA (40), a sequence present in the right side of the AGRE site. Identification of AGRE binding proteins in the complex formed between the wild-type AGRE site and H-OA chondrocyte proteins. The previous experiment showed that p130cas was present in the complex formed by a mutated AGRE site and nuclear extracts from L-OA chondrocytes. To confirm that p130cas was present in the wild-type AGRE/H-OA chondrocyte complex, a specific p130cas antibody was added to the H-OA chondrocyte EMSA reaction mixture. Because p130cas is a cofactor of NMP-4, we also used a specific antibody to test for the presence of NMP-4 in the H-OA complex. As illustrated in Figure 6, both NMP-4 and p130cas antibodies supershifted the AGRE–protein H-OA complex. The specificity was verified by the addition of a 50-fold molar excess of the cold AGRE oligonucleotide (Figure 6). Identification of AGRE binding proteins in the complex formed between the wild-type AGRE site and L-OA chondrocyte proteins. Identification of the proteins formed in the L-OA chondrocyte/AGRE complex was performed by DNA affinity chromatography with nuclear extracts incubated with magnetic beads bound to the wild-type or mutated (GAAGAGACAAAG; the M-1 mutation) AGRE-containing oligonucleotides. The protein DDB-1 was identified following 2-dimensional gel electrophoresis and mass spectroscopy. DISCUSSION In our previous study (25), we showed that the AGRE site in the proximal promoter of MMP-13 plays an important role in its regulation. The objective of this study was to complement the previously acquired data and to identify proteins that bind to this site. Most mutations of the different bases of the AGRE site in the –175Luc plasmid resulted in decreased promoter activities, suggesting that the whole AGRE site is required for full MMP-13 basal activity and that the proteins binding the site have a stimulatory role in basal transcription under the conditions used. The fact that the whole AGRE site needs to be intact for optimal MMP-13 activity is supported by the results obtained by the mutation of the AGRE-like sequences present in the proximal promoter of other MMPs. We found that the mutated sequences did not significantly alter basal promoter activity, although we cannot rule out a role of these sites in conditions in which MMPs are induced. The only mutation that resulted in increased promoter activity was the M-6 mutation, which totally 2477 abolished the left side of AGRE and resulted in the strongest H-OA type binding, as determined by EMSA. This was unexpected, since the M-2 mutation, which also has the intact right side with the consensus NMP-4 binding site, decreased rather than increased promoter activity. It should be noted, however, that neither the wild-type AGRE nor the M-2 mutation showed an H-OA type binding as strong as the M-6 mutation. In these last 2 cases, the NMP-4 binding site is intact; however, the binding of other proteins present in higher amounts or having greater affinity for the site may impede the binding of NMP-4. The M-6 mutation created an optimal binding site for NMP-4/p130cas, leading to an up-regulation of promoter activity. The L-OA proteins seem to have a more potent stimulatory effect than the H-OA proteins, with most AGRE mutations reducing activity by 5–10-fold. L-OA proteins likely bind either or both sides of AGRE: mutations in the left side (M-2), the right side (M-3 and M-5), or both sides (M-1) all reduce promoter activity. Thus, the H-OA chondrocytes may result from the strong binding of NMP-4/p130cas to AGRE, resulting in a 2-fold stimulation of transcription, in contrast to the L-OA state, where the binding by the L-OA proteins would contribute to a higher degree of basal transcription. The stimulating capacities of the H-OA and L-OA binding proteins may be different, or alternatively, the H-OA complex, in addition to NMP-4/p130cas, may recruit a cofactor that could limit the stimulatory effect. These results showing that the AGRE binding proteins stimulate transcription contrast with our previous study (25), in which we showed that the AGRE mutation or deletion resulted in a significant increase in basal MMP-13 transcription. However, the plasmid used in the previous study was smaller by 38 bp. The extra DNA in plasmid –175Luc contains an OSE-2 site as well as a putative STAT site (Figure 1), which has yet to be described or analyzed with respect to its involvement in MMP-13 transcription. Since the mutation of the OSE-2 site does not seem to make a significant difference in basal promoter activity, it is possible that the STAT site may play a role in MMP-13 transcription. It has previously been reported that interferon-␥–induced phosphorylation of STAT-1 inhibits MMP-13 expression in transformed human epidermal keratinocytes (41). It remains to be determined, however, whether this putative STAT site is involved in basal MMP-13 modulation in human chondrocytes. The mutation of the OSE-2 site, located 9 bp upstream of the AGRE site, did not affect basal MMP-13 promoter activity. The transcription factor 2478 CBF␣1, which binds OSE-2, is known to be involved in chondrocyte and osteoblast differentiation and bone formation, and it is also up-regulated in fibrillated OA cartilage (42,43). However, the increased expression of CBF␣1 in OA cartilage does not seem to be sufficient to affect MMP-13 expression without growth factor activation (43). Moreover, and consistent with our data, there was no significant increase in MMP-13 in serum-free culture medium derived from human chondrocytes infected with the recombinant CBF␣1 adenovirus. In another study (23), increased promoter activity was observed when HeLa cells were cotransfected with a CBF␣1 expression plasmid and an OSE-2–containing MMP-13 promoter plasmid. However, there was no difference in the basal activity between MMP-13 promoter plasmids with and those without the OSE-2 site, which is consistent with our results. Although CBF␣1 is important in developmental processes and is involved in the stimulatory pathways of some factors, it may not be activated or expressed at a level high enough to affect basal MMP-13 expression in human OA chondrocytes. MMP-13 promoter activity was reduced by AGRE mutations in the plasmid –175Luc. The mutated sites were also found to have different protein-binding capacities in vitro. An interesting result was that we were able to reproduce the H-OA binding complex with L-OA nuclear extracts and the AGRE site mutated in its left side. This suggests that the proteins that form the H-complex are present in the L-OA chondrocytes but are not capable of binding to AGRE. It is possible that the concentration of L-proteins is much higher in L-OA chondrocytes than in H-OA chondrocytes and that the presence of the L-proteins on the AGRE site prevents the H-proteins from binding. Although we have identified DDB-1 by affinity chromatography, it is likely that other proteins are involved in the L-OA complex. DDB-1 (127 kd) and DDB-2 (48 kd) are part of the DDB complex that binds ultraviolet light–damaged DNA in mammalian cells (44). DDB is not known to bind the specific AGRE sequence but has been reported to function as a transcriptional partner of the transcription factor E2F1, which suggests a role of DDB in the cell cycle (45). It remains to be determined whether DDB-1 can form a complex with other transcription factors and transactivate MMP-13 expression. Although not related to the AGRE site, another DNA damage–inducible protein, growth arrest and DNA damage–induced 45␤, was shown to stimulate MMP-13 promoter activity in chondrocytes (46). We have identified the proteins p130cas and FAN ET AL NMP-4 in the H-complex. The p130cas protein does not bind DNA directly but is known to interact with the transcription factor NMP-4 (40). This is the first report of the implication of p130cas as a possible transcription modulator. The p130cas protein is a docking protein found in focal adhesion points, and it is involved in cellular processes such as migration, proliferation, and apoptosis (47–49). The CIZ protein was originally identified in rat cells through its interaction with p130cas (40). It is a nucleocytoplasmic shuttling protein that binds the consensus sequence G/CAAAAA. Overexpression of this protein up-regulates MMP-1, MMP-3, and MMP-7 transcription (40). At about the same time, NMP-4 complementary DNA splice variants were identified, also from rat cells, and some of these variants were identical to the CIZ proteins (30). Shah et al (50) showed that NMP-4/CIZ binds the homopolymeric (dA-dT) element within the MMP-13 promoter and promotes basal expression in rat cells. The regulatory role of NMP-4 in MMP-13 expression may differ in rat and human cells. The NMP-4 binding site in the rat MMP-13 promoter does not contain an AGRE site per se, but does contain the sequence GAAAAAAAAAA, to which NMP-4 can bind; this sequence is located near an OSE-2 site in the same region of the AGRE site of the human promoter. Although a similar sequence is also found in the human MMP-13 promoter (CAAAAAAAAAAA), which could possibly bind NMP-4, it is located approximately 170 bp upstream of the AGRE site (21) and is not included in plasmid –175Luc. In this study, we found that NMP-4 was part of the H-complex formed with AGRE. Unlike the rat MMP-13 promoter, the human promoter may offer 2 levels of regulation of MMP-13 by NMP-4: one at the sequence similar to the rat binding site and the other at the AGRE site where 2 different sets of proteins (L and H) would act. The L-OA and H-OA chondrocytes may represent cells at different stages of the disease process. Although the initiating event responsible for the 2 different populations is not yet known, the H-OA chondrocytes may represent cells in a later stage of the disease, during which there is a greater need for MMP-13 in the remodeling/repair processes. NMP-4 and p130cas may be involved in the switch from L-OA to H-OA chondrocytes. Because of increased synthesis and/or activation, these proteins would replace the L-proteins on the AGRE site. Their presence on this site may lead to increased basal levels but limit IL-1␤ inducibility, perhaps by interfering with proteins binding IDENTIFICATION OF AGRE BINDING PROTEINS IN HUMAN OA CHONDROCYTES at the AP-1 site. NMP-4 has been reported to interfere with transcription induction. In rodent osteoblast cells, it suppresses bone morphogenetic protein 2–induced expression of alkaline phosphatase, osteocalcin, and type I collagen genes (51) and parathyroid hormone–induced increase in MMP-13 (50). The study of the regulation of NMP-4 and p130cas in OA chondrocytes is ongoing. In summary, the AGRE site and its binding proteins play important roles in the regulation of MMP-13 and may be involved in the change of OA chondrocytes from the L state to the H state. Data from this study offer a better understanding of the regulation of MMP-13, and the knowledge acquired will open up potential novel avenues in the development of therapeutic strategies targeting MMP-13. ACKNOWLEDGMENTS We are grateful to Herman Cheung, PhD (University of Miami, Miami, FL) for providing the MMP-1 promoter construct and to Pierre Pépin, MSc (Sheldon Biotechnology Centre, Montreal, Quebec, Canada) for assistance in the protein sequencing. We also thank Santa Fiori and Virginia Wallis for assistance in the manuscript preparation. REFERENCES 1. Freije JM, Diez-Itza I, Balbin M, Sanchez LM, Blasco R, Tolivia J, et al. Molecular cloning and expression of collagenase-3, a novel human matrix metalloproteinase produced by breast carcinomas. J Biol Chem 1994;269:16766–73. 2. Reboul P, Pelletier JP, Tardif G, Cloutier JM, Martel-Pelletier J. The new collagenase, collagenase-3, is expressed and synthesized by human chondrocytes but not by synoviocytes: a role in osteoarthritis. J Clin Invest 1996;97:2011–19. 3. Mitchell PG, Magna HA, Reeves LM, Lopresti-Morrow LL, Yocum SA, Rosner PJ, et al. Cloning, expression, and type II collagenolytic activity of matrix metalloproteinase-13 from human osteoarthritic cartilage. J Clin Invest 1996;97:761–8. 4. Billinghurst RC, Dahlberg L, Ionescu M, Reiner A, Bourne R, Rorabeck C, et al. Enhanced cleavage of Type II collagen by collagenases in osteoarthritic articular cartilage. J Clin Invest 1997;99:1534–45. 5. Knauper V, Cowell S, Smith B, Lopez-Otin C, O’Shea M, Morris H, et al. The role of the C-terminal domain of human collagenase-3 (MMP-13) in the activation of procollagenase-3, substrate specificity, and tissue inhibitor of metalloproteinase interaction. J Biol Chem 1997;272:7608–16. 6. Fosang AJ, Last K, Knauper V, Murphy G, Neame PJ. Degradation of cartilage aggrecan by collagenase-3 (MMP-13). FEBS Lett 1996;380:17–20. 7. Hashimoto G, Inoki I, Fujii Y, Aoki T, Ikeda E, Okada Y. Matrix metalloproteinases cleave connective tissue growth factor and reactivate angiogenic activity of vascular endothelial growth factor 165. J Biol Chem 2002;277:36288–95. 8. Hiller O, Lichte A, Oberpichler A, Kocourek A, Tschesche H. Matrix metalloproteinases collagenase-2, macrophage elastase, collagenase-3, and membrane type 1-matrix metalloproteinase impair clotting by degradation of fibrinogen and factor XII. J Biol Chem 2000;275:33008–13. 2479 9. Johansson N, Saarialho-Kere U, Airola K, Herva R, Nissinen L, Westermarck J, et al. Collagenase-3 (MMP-13) is expressed by hypertrophic chondrocytes, periosteal cells, and osteoblasts during human fetal bone development. Dev Dyn 1997;208:387–97. 10. Stahle-Backdahl M, Sandstedt B, Bruce K, Lindahl A, Jimenez MG, Vega JA, et al. Collagenase-3 (MMP-13) is expressed during human fetal ossification and re-expressed in postnatal bone remodeling and in rheumatoid arthritis. Lab Invest 1997;76:717–28. 11. Bau B, Gebhard PM, Haag J, Knorr T, Bartnik E, Aigner T. Relative messenger RNA expression profiling of collagenases and aggrecanases in human articular chondrocytes in vivo and in vitro. Arthritis Rheum 2002;46:2648–57. 12. Moldovan F, Pelletier JP, Hambor J, Cloutier JM, Martel-Pelletier J. Collagenase-3 (matrix metalloprotease 13) is preferentially localized in the deep layer of human arthritic cartilage in situ: in vitro mimicking effect by transforming growth factor ␤. Arthritis Rheum 1997;40:1653–61. 13. Moore BA, Aznavoorian S, Engler JA, Windsor LJ. Induction of collagenase-3 (MMP-13) in rheumatoid arthritis synovial fibroblasts. Biochim Biophys Acta 2000;1502:307–18. 14. Pendas AM, Uria JA, Jimenez MG, Balbin M, Freije JP, LopezOtin C. An overview of collagenase-3 expression in malignant tumors and analysis of its potential value as a target in antitumor therapies. Clin Chim Acta 2000;291:137–55. 15. Forsyth CB, Pulai J, Loeser RF. Fibronectin fragments and blocking antibodies to ␣2␤1 and ␣5␤1 integrins stimulate mitogenactivated protein kinase signaling and increase collagenase 3 (matrix metalloproteinase 13) production by human articular chondrocytes. Arthritis Rheum 2002;46:2368–76. 16. Loeser RF, Forsyth CB, Samarel AM, Im HJ. Fibronectin fragment activation of proline-rich tyrosine kinase PYK2 mediates integrin signals regulating collagenase-3 expression by human chondrocytes through a protein kinase C-dependent pathway. J Biol Chem 2003;278:24577–85. 17. Tardif G, Reboul P, Dupuis M, Geng C, Duval N, Pelletier JP, et al. Transforming growth factor-␤ induced collagenase-3 production in human osteoarthritic chondrocytes is triggered by Smad proteins: cooperation between activator protein-1 and PEA-3 binding sites. J Rheumatol 2001;28:1631–9. 18. Benderdour M, Tardif G, Pelletier J, Di Battista JA, Reboul P, Ranger P, et al. Interleukin-17 (IL-17) induces collagenase-3 production in human osteoarthritic chondrocytes via an AP-1 dependent activation: differential activation of AP-1 members by IL-17 and IL-1␤. J Rheumatol 2002;29:1262–72. 19. Guevremont M, Martel-Pelletier J, Massicotte F, Tardif G, Pelletier JP, Ranger P, et al. Human adult chondrocytes express hepatocyte growth factor (HGF) isoforms but not HGF: potential implication of osteoblasts on the presence of HGF in cartilage. J Bone Miner Res 2003;18:1073–81. 20. Yokota H, Goldring MB, Sun HB. CITED2-mediated regulation of MMP-1 and MMP-13 in human chondrocytes under flow shear. J Biol Chem 2003;278:47275–80. 21. Tardif G, Pelletier JP, Dupuis M, Hambor JE, Martel-Pelletier J. Cloning, sequencing and characterization of the 5⬘-flanking region of the human collagenase-3 gene. Biochem J 1997;323:13–6. 22. Pendas AM, Balbin M, Llano E, Jimenez MG, Lopez-Otin C. Structural analysis and promoter characterization of the human collagenase-3 gene (MMP-13). Genomics 1997;40:222–33. 23. Jimenez MJ, Balbin M, Lopez JM, Alvarez J, Komori T, LopezOtin C. Collagenase 3 is a target of Cbfa1, a transcription factor of the runt gene family involved in bone formation. Mol Cell Biol 1999;19:4431–42. 24. Mengshol JA, Vincenti MP, Brinckerhoff CE. IL-1 induces collagenase-3 (MMP-13) promoter activity in stably transfected chondrocytic cells: requirement for Runx-2 and activation by p38 MAPK and JNK pathways. Nucleic Acids Res 2001;29:4361–72. 25. Benderdour M, Tardif G, Pelletier JP, Dupuis M, Geng C, 2480 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. Martel-Pelletier J. A novel negative regulatory element in the human collagenase-3 proximal promoter region. Biochem Biophys Res Commun 2002;291:1151–9. Tardif G, Pelletier JP, Dupuis M, Geng C, Cloutier JM, MartelPelletier J. Collagenase 3 production by human osteoarthritic chondrocytes in response to growth factors and cytokines is a function of the physiological state of the cells. Arthritis Rheum 1999;42:1147–58. Altman R, Asch E, Bloch D, Bole G, Borenstein D, Brandt K, et al. Development of criteria for the classification and reporting of osteoarthritis: classification of osteoarthritis of the knee. Arthritis Rheum 1986;29:1039–49. Mankin HJ, Dorfman H, Lippiello L, Zarins A. Biochemical and metabolic abnormalities in articular cartilage from osteo-arthritic human hips. II. Correlation of morphology with biochemical and metabolic data. J Bone Joint Surg Am 1971;53:523–37. Rutter JL, Benbow U, Coon CI, Brinckerhoff CE. Cell-type specific regulation of human interstitial collagenase-1 gene expression by interleukin-1 ␤ (IL-1 ␤) in human fibroblasts and BC-8701 breast cancer cells. J Cell Biochem 1997;66:322–36. Thunyakitpisal P, Alvarez M, Tokunaga K, Onyia JE, Hock J, Ohashi N, et al. Cloning and functional analysis of a family of nuclear matrix transcription factors (NP/NMP4) that regulate type I collagen expression in osteoblasts. J Bone Miner Res 2001;16: 10–23. Shevchenko A, Wilm M, Vorm O, Mann M. Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels. Anal Chem 1996;68:850–8. Angel P, Baumann I, Stein B, Delius H, Rahmsdorf HJ, Herrlich P. 12-O-tetradecanoyl-phorbol-13-acetate induction of the human collagenase gene is mediated by an inducible enhancer element located in the 5⬘-flanking region. Mol Cell Biol 1987;7:2256–66. Quinones S, Saus J, Otani Y, Harris ED Jr, Kurkinen M. Transcriptional regulation of human stromelysin. J Biol Chem 1989; 264:8339–44. Gaire M, Magbanua Z, McDonnell S, McNeil L, Lovett DH, Matrisian LM. Structure and expression of the human gene for the matrix metalloproteinase matrilysin. J Biol Chem 1994;269: 2032–40. Belaaouaj A, Shipley JM, Kobayashi DK, Zimonjic DB, Popescu N, Silverman GA, et al. Human macrophage metalloelastase. Genomic organization, chromosomal location, gene linkage, and tissue-specific expression. J Biol Chem 1995;270:14568–75. Anglard P, Melot T, Guerin E, Thomas G, Basset P. Structure and promoter characterization of the human stromelysin-3 gene. J Biol Chem 1995;270:20337–44. Huhtala P, Tuuttila A, Chow LT, Lohi J, Keski-Oja J, Tryggvason K. Complete structure of the human gene for 92-kDa type IV collagenase: divergent regulation of expression for the 92- and 72-kilodalton enzyme genes in HT-1080 cells. J Biol Chem 1991; 266:16485–90. Bian J, Sun Y. Transcriptional activation by p53 of the human type FAN ET AL 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. IV collagenase (gelatinase A or matrix metalloproteinase 2) promoter. Mol Cell Biol 1997;17:6330–8. Lohi J, Lehti K, Valtanen H, Parks WC, Keski-Oja J. Structural analysis and promoter characterization of the human membranetype matrix metalloproteinase-1 (MT1-MMP) gene. Gene 2000; 242:75–86. Nakamoto T, Yamagata T, Sakai R, Ogawa S, Honda H, Ueno H, et al. CIZ, a zinc finger protein that interacts with p130cas and activates the expression of matrix metalloproteinases. Mol Cell Biol 2000;20:1649–58. Ala-aho R, Johansson N, Grenman R, Fusenig NE, Lopez-Otin C, Kahari VM. Inhibition of collagenase-3 (MMP-13) expression in transformed human keratinocytes by interferon-␥ is associated with activation of extracellular signal-regulated kinase-1,2 and STAT1. Oncogene 2000;19:248–57. Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G. Osf2/ Cbfa1: a transcriptional activator of osteoblast differentiation. Cell 1997;89:747–54. Wang X, Manner PA, Horner A, Shum L, Tuan RS, Nuckolls GH. Regulation of MMP-13 expression by RUNX2 and FGF2 in osteoarthritic cartilage. Osteoarthritis Cartilage 2004;12:963–73. Dualan R, Brody T, Keeney S, Nichols AF, Admon A, Linn S. Chromosomal localization and cDNA cloning of the genes (DDB1 and DDB2) for the p127 and p48 subunits of a human damagespecific DNA binding protein. Genomics 1995;29:62–9. Hayes S, Shiyanov P, Chen X, Raychaudhuri P. DDB, a putative DNA repair protein, can function as a transcriptional partner of E2F1. Mol Cell Biol 1998;18:240–9. Ijiri K, Zerbini LF, Peng H, Correa RG, Lu B, Walsh N, et al. A novel role for GADD45␤ as a mediator of MMP-13 gene expression during chondrocyte terminal differentiation. J Biol Chem 2005;280:38544–55. Sakai R, Iwamatsu A, Hirano N, Ogawa S, Tanaka T, Mano H, et al. A novel signaling molecule, p130, forms stable complexes in vivo with v-Crk and v-Src in a tyrosine phosphorylation-dependent manner. EMBO J 1994;13:3748–56. Kim W, Kook S, Kim DJ, Teodorof C, Song WK. The 31-kDa caspase-generated cleavage product of p130cas functions as a transcriptional repressor of E2A in apoptotic cells. J Biol Chem 2004;279:8333–42. O’Neill GM, Fashena SJ, Golemis EA. Integrin signalling: a new Cas(t) of characters enters the stage. Trends Cell Biol 2000;10: 111–9. Shah R, Alvarez M, Jones DR, Torrungruang K, Watt AJ, Selvamurugan N, et al. Nmp4/CIZ regulation of matrix metalloproteinase 13 (MMP-13) response to parathyroid hormone in osteoblasts. Am J Physiol Endocrinol Metab 2004;287:E289–96. Shen ZJ, Nakamoto T, Tsuji K, Nifuji A, Miyazono K, Komori T, et al. Negative regulation of bone morphogenetic protein/Smad signaling by Cas-interacting zinc finger protein in osteoblasts. J Biol Chem 2002;277:29840–6.