Drs. Liu and Lioté contributed equally to this work.
Proline-rich tyrosine kinase 2 and Src kinase signaling transduce monosodium urate crystal–induced nitric oxide production and matrix metalloproteinase 3 expression in chondrocytes
Article first published online: 9 JAN 2004
Copyright © 2004 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 50, Issue 1, pages 247–258, January 2004
How to Cite
Liu, R., Lioté, F., Rose, D. M., Merz, D. and Terkeltaub, R. (2004), Proline-rich tyrosine kinase 2 and Src kinase signaling transduce monosodium urate crystal–induced nitric oxide production and matrix metalloproteinase 3 expression in chondrocytes. Arthritis & Rheumatism, 50: 247–258. doi: 10.1002/art.11486
- Issue published online: 9 JAN 2004
- Article first published online: 9 JAN 2004
- Manuscript Accepted: 13 OCT 2003
- Manuscript Received: 18 MAR 2003
- Department of Veterans Affairs
- NIH. Grant Numbers: HL-61731, R03-AR-49416-10, P01-AG-07996
- Arthritis Foundation
- University of California, San Diego
- Medicine Education Research Foundation
- Stein Institute for Research on Aging, Assistance Publique Hôpitaux de Paris, Paris 7 University
- Philippe Foundation
Articular deposition of monosodium urate monohydrate (MSU) crystals may promote cartilage and bone erosion. Therefore, the aim of this study was to determine how MSU crystals stimulate chondrocytes.
Nitric oxide (NO) release, and expression of inducible nitric oxide synthase (iNOS) and matrix metalloproteinase 3 (MMP-3) were assessed in cultured chondrocytes treated with MSU. MSU-induced functional signaling by specific protein kinases (p38, Src, and the focal adhesion kinase [FAK] family members proline-rich tyrosine kinase 2 [Pyk-2] and FAK) was also examined using selective pharmacologic inhibitors and transfection of kinase mutants.
MSU induced MMP-3 and iNOS expression and NO release in chondrocytes in a p38-dependent manner that did not require interleukin-1 (IL-1), as demonstrated by using IL-1 receptor antagonist. MSU induced rapid tyrosine phosphorylation of Pyk-2 and FAK, their adaptor protein paxillin, and interacting kinase c-Src. Pyk-2 and c-Src signaling both mediated p38 MAPK activation in response to MSU. Pyk-2 and c-Src signaling played a major role in transducing MSU-induced NO production and MMP-3 expression. But, despite the observed FAK phosphorylation, a selective pharmacologic FAK inhibitor and a FAK dominant-negative mutant both failed to block MSU-induced NO release or MMP-3 expression in parallel experiments.
In chondrocytes, MSU crystals activate a signaling kinase cascade typically employed by adhesion receptors that involves upstream Src and FAK family activation and downstream p38 activation. In this cascade, Pyk-2, Src, and p38 kinases transduce MSU-induced NO production and MMP-3 expression. Our results identify Pyk-2 and c-Src as novel sites for potential therapeutic intervention in cartilage degradation in chronic gout.
In gout, the deposition of monosodium urate monohydrate (MSU) crystals in synovium and cartilage can promote chronic inflammation that may lead to cartilage catabolism and bone erosion (1, 2). MSU crystals stimulate both acute neutrophilic inflammation and chronic synovitis in large part via their capacity to directly activate both resident articular connective tissue cells and mononuclear phagocytes (3–5). Direct activation of cells by MSU crystals can trigger the expression of a broad array of inflammatory mediators, including cyclooxygenase 2 and the cytokines tumor necrosis factor α, interleukin-1 (IL-1), IL-6, and IL-8 (6–10). Also consistent with the remarkable intensity of the clinical and pathologic inflammatory response in gout (1, 3, 4) is the capacity of MSU crystals to activate phagocytes by triggering multiple signal transduction pathways (11–14).
In leukocytes, MSU crystals nonspecifically engage plasma membrane proteins, such as the Fc receptor CD16 and the β2 integrin CD11b/CD18 (11). Concordantly, certain downstream signal transduction pathways implicated in leukocyte phagocytosis and adhesion responses have been observed to mediate MSU crystal–induced activation of leukocytes (12). For example, MSU crystal–induced activation of Syk tyrosine kinase, Src family tyrosine kinases, and the ERK-1/2, JNK, and p38 MAPKs modulate crystal-induced phagocyte responses, including oxygen radical generation and IL-8 expression (12–14).
MSU crystals are known to induce several proinflammatory mediators in fibroblasts and osteoblasts (15–18). Yet there is relatively little understanding of the mechanisms by which MSU crystals activate chondrocytes. Our objectives were to test the potential for MSU crystals to directly activate chondrocytes and to define the signaling mechanisms involved. In doing so, we tested the potential capacity for MSU crystals to directly promote cartilage degeneration, with a focus on chondrocyte nitric oxide (NO) generation and expression of matrix metalloproteinase 3 (MMP-3). Significantly, MMP-3 cleaves the nonhelical, noncollagen domains of several cartilage matrix proteins, including type IX collagen, with the major substrates of MMP-3 being proteoglycans and certain latent MMPs (19). Furthermore, excessive production of NO can be triggered by up-regulated expression of inducible NO synthase (iNOS) (20, 21), a major effect of IL-1 in chondrocytes (22). Increased NO generation inhibits chondrocyte proteoglycan synthesis (23), can impair chondrocyte viability (24), and has the potential, through S-nitrosylation of certain MMPs, to enhance the matrix catabolic activity of MMPs (25, 26).
MSU crystals can bind to integrins and can employ integrin-mediated adhesion signaling in more than one cell type (11, 27). Therefore, we explored the potential for MSU crystals to activate chondrocytes through the activity of chondrocyte-expressed members of the FAK family, a group of tyrosine kinases involved in integrin-mediated focal adhesion complex formation and signaling (28, 29). We found that a differential proline-rich tyrosine kinase 2 (Pyk-2) signaling chain mediates MSU crystal–induced NO generation and MMP-3 expression in chondrocytes.
MATERIALS AND METHODS
The pharmacologic MAPK inhibitors PD 98059, JNK inhibitors II and SB 203580, the selective Pyk-2 inhibitor AG-17, and the selective FAK inhibitor AG-82 were obtained from Calbiochem (San Diego, CA), the Src family kinase inhibitor PP1 was obtained from Alexis Biochemicals (San Diego, CA), and recombinant human IL-1β and IL-1 receptor antagonist (IL-1Ra) were purchased from R&D Systems (Minneapolis, MN). MSU crystals were prepared as previously described (30) and suspended and stored in phosphate buffered saline at 25 mg/ml.
Phosphospecific ERK-1/2 (Thr202/Tyr206), JNK (Thr183/Tyr185), and p38 (Thr180/Tyr182), and phosphorylation-independent ERK-1/2, JNK, and p38 antibodies were purchased from Cell Signaling Technology (Beverly, MA). The Pyk-2-Y402, FAK-Y397, and c-Src-Y418 phosphospecific antibodies were obtained from BioSource International (Camarillo, CA). The Pyk-2, FAK, and Src phosphorylation-independent antibodies were purchased from Upstate Biotechnology (Lake Placid, NY). The anti-phosphotyrosine (pY99) monoclonal antibody, horseradish peroxidase–conjugated goat anti-rabbit IgG, and anti-mouse IgG were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Antibody to MMP-3 was obtained from Chemicon (Temecula, CA).
Cell culture and transfection.
To isolate primary chondrocytes, articular cartilage obtained from normal bovine knee joints (Animal Technology, Tyler, TX) was minced and placed in bacterial collagenase (2 mg/ml; Worthington Biochemical, Lakewood, NJ) in Dulbecco's modified Eagle's medium (DMEM) high-glucose medium (Omega Scientific, Tarzana, CA) containing 1% fetal calf serum (FCS), 2 mML-glutamine, 100 μg/ml streptomycin, and 100 IU/ml penicillin at 37°C for 16 hours. After straining through a 100-μm nylon cell strainer (Becton Dickinson, Mountain View, CA), cells were pelleted, washed, resuspended, and plated at 70% density in DMEM high-glucose medium containing 10% FCS, 2 mML-glutamine, 100 μg/ml streptomycin, and 100 IU/ml penicillin, and cultured until they were confluent (∼7 days). Cells were then removed using the cell detachment reagent Accutase (Innovative Cell Technologies, La Jolla, CA), transferred to plates coated with poly-(2-hydroxyethyl methacrylate) (poly-HEMA) (Sigma, St. Louis, MO), which was used to maintain chondrocytic phenotype under more physiologic, nonadherent conditions (31). The poly-HEMA–coated plates were prepared as follows. The poly-HEMA that dissolved in 95% ethanol and 5% water was added to the tissue culture plates, and then the plates were left to dry in the tissue culture hood and under ultraviolet light overnight. Cells were cultured for 18 hours on the poly-HEMA–coated plates in the aforementioned medium supplemented with only 1% FCS, prior to stimulation under those conditions with MSU crystals at the indicated concentrations.
For transfection studies in bovine chondrocytes, aliquots of 4 × 105 of cells were plated in 60-mm dishes and allowed to adhere overnight. The next day, cells were transfected using FuGENE 6 and hyaluronidase, as previously described (32). Transfection efficiency was evaluated in control samples via β-galactosidase transfection and staining (32). Myc-tagged wild-type Pyk-2, 2 mutants of Pyk-2 (Pyk-2–F402 and Pyk-2–A457), and an FAK-related nonkinase (FRNK) were expressed in pcDNA3.1 vector (all complementary DNA was a generous gift of Dr. D. Schlaepfer, The Scripps Research Institute, La Jolla, CA). Hemagglutinin-tagged Csk was expressed in pEFneo vector (kindly provided by Dr. T. Hunter, The Salk Institute, La Jolla, CA). Where indicated for signal transduction studies, human T/C28a2 immortalized chondrocytic cells (a previous gift of Dr. Mary Goldring, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA) were cultured in the same medium described above, as previously detailed (33). The T/C28a2 cells were transfected by a previously validated Lipofectamine Plus method (33). Twenty-four hours after the transfection, the medium was replaced with fresh complete DMEM high-glucose medium containing 10% FCS, and the cells were allowed to recover for another 24 hours. Subsequently, the T/C28a2 cells also were cultured on poly-HEMA–coated plates for studies of stimulation by MSU crystals, as described above for bovine chondrocytes.
Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE)/Western blot analyses, in vitro kinase assays.
For analysis of MMP-3 protein expression, secreted proteins were precipitated from the conditioned media using trichloroacetic acid to a final concentration of 15% volume/volume. For cell lysate preparation, cell pellets were resuspended in lysis buffer (20 mM Tris HCl, pH 7.5, 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, 1 mM EGTA, 1 mM sodium orthovanadate, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, 1 mM phenylmethylsulfonyl fluoride, 1 μg/ml pepstatin, aprotinin, and leupeptin). They were then sonicated and incubated on ice for 15 minutes. After centrifugation at 14,000 revolutions per minute for 10 minutes, the supernatants containing cell lysates were collected. Cell lysates (aliquots of 30 μg protein) were separated by 10% SDS-PAGE and then transferred onto nitrocellulose (Bio-Rad, Hercules, CA) and studied by Western blot analysis as previously described, using horseradish peroxidase–conjugated secondary antibody. Immunoreactive products were detected using an enhanced chemiluminescence system (Amersham Pharmacia, Piscataway, NJ). Nonradioactive in vitro kinase assays for activities of ERK-1/2, JNK, and p38 MAPKs were carried out using a kit from Cell Signaling Technology. The kinase reactions were performed using Elk-1, c-Jun, and activating transcription factor 2 (ATF-2) as the substrates for ERK-1/2, JNK, and p38 MAPKs, respectively. The reaction mixture was then separated by SDS-PAGE. The kinase activity was studied by Western blot analysis with phosphospecific antibodies to Elk-1, c-Jun, or ATF-2.
Assays for NO production, iNOS expression, MMP-3 messenger RNA (mRNA), and protein expression.
NO production was measured as the concentration of nitrites in conditioned media by the Griess reaction (34) using NaNO2 as the standard. The iNOS, MMP-13, and IL-1β transcripts were evaluated by reverse transcriptase–polymerase chain reaction (RT-PCR) in studies in which RNA was isolated from bovine chondrocytes using TRIzol (Life Technologies, Grand Island, NY). Aliquots of 600 ng of total RNA were reverse-transcribed, as previously described (31). PCR reactions were performed for 30 cycles: 95°C for 5 minutes, 95°C for 1 minute, 55°C for 1 minute, 72°C for 1 minute, and 72°C for 5 minutes. The bovine iNOS primers were 5′-TAGAGGAACATCTGGCCAGG-3′ (sense) and 5′-TGGCAGGGTCCCCTCTGATG-3′ (antisense), which amplified a 372-bp product. The MMP-3 primers were 5′-CCCTCCAGAACCTGGGAC-3′ (sense) and 5′-ATAAAAGAACCCAAATTCTTCAAAA-3′ (antisense), which amplified a 505-bp product. The bovine IL-1β primers were 5′-TACCTGAACCCATCAACGAAA-3′ (sense) and 5′-GATGAATGAAAGGATGCCCTC-3′ (antisense), which amplified a 275-bp product. The housekeeping gene L30 primers were as previously described (35). Densitometric analysis of RT-PCR gel bands was performed as previously described (36).
The data are expressed as the mean ± SD. Statistical analyses were performed by Student's 2-tailed t-test.
MSU crystal–induced iNOS gene expression, NO release, and MMP-3 expression in chondrocytes.
In these studies, the cells were grown in monolayer culture in medium supplemented with 10% FCS for a week after isolation and, 18 hours prior to stimulation, were transferred to more physiologic, nonadherent culture conditions using poly-HEMA–coated plates (31) and minimal serum supplementation with 1% FCS. Under these conditions, MSU crystals induced iNOS gene expression and NO release in normal bovine articular chondrocytes incubated with the crystals for 24 hours (Figure 1A). The crystal-dose dependence of these effects reached a plateau between 0.5–1 mg/ml (Figure 1A), consistent with cytotoxic effects of escalating doses of MSU crystals in other cells (37). We demonstrated time-dependent induction of iNOS expression by MSU crystals, with a plateau at 24 hours (Figure 1B). Thus, in further evaluating chondrocyte activation by MSU crystals, 0.5 mg/ml MSU crystals was principally used and cells were studied 24 hours after the addition of MSU crystals.
Constitutive MMP-3 expression at the mRNA level was observed in bovine chondrocytes and was not substantially enhanced by MSU crystals (Figure 1). The transcription inhibitor actinomycin D had no significant effect on MMP-3 mRNA expression (data not shown). In contrast, MSU crystal–induced iNOS expression was markedly inhibited by actinomycin D under the same conditions, consistent with crystal-induced regulation of iNOS but not MMP-3 at the transcription level (data not shown).
Under the conditions used, MSU crystals induced IL-1β mRNA expression in bovine chondrocytes, as assessed by RT-PCR (Figure 2, top). Because IL-1 can induce NO production, MMP-3 expression (38, 39), as well as IL-1β mRNA accumulation (40) in chondrocytes, bovine chondrocytes were treated with IL-1Ra for 1 hour before stimulation with MSU crystals or IL-1β. IL-Ra markedly inhibited IL-1β expression, iNOS expression, and NO release induced by IL-1β treatment but not by incubation with MSU crystals. In addition, IL-1Ra markedly inhibited MMP-3 induction by IL-1β, but only slightly inhibited the induction of MMP-3 by MSU crystals (Figure 2). Because MSU crystal–induced NO production and MMP-3 expression were not simply attributable to indirect effects of IL-1β in chondrocytes, functional implications of MSU crystal–induced signal transduction were examined.
MAPK p38 in MSU crystal–induced NO production and MMP-3 expression.
MSU crystals induced activities of the MAPKs ERK-1/2, JNK, and p38 kinase in chondrocytes, as assessed by in vitro kinase assays (Figure 3A). MSU crystal–induced phosphorylation of ERK-1/2, JNK, and p38 MAPKs was confirmed to be inhibited by respective pharmacologic inhibitors for each MAPK pathway (Figure 3B). Each inhibitor at the concentration used was tolerated by the chondrocytes without loss of viability and was verified to be selectively inhibitory for the appropriate MAPK signaling pathway by Western blot analysis for stimulated phosphorylation of individual MAPKs (results not shown). Under these conditions, the ERK-1/2 pathway inhibitor PD 98059 exerted minimal inhibitory effects on iNOS expression, NO release, and MMP-3 expression induced by MSU crystals (Figure 4). In contrast, the p38 pathway inhibitor SB 203580 markedly inhibited iNOS expression, NO release, and MMP-3 protein expression. The JNK pathway inhibitor JNK inhibitor II also inhibited iNOS expression and NO release, but to a lesser extent than did SB 203580. JNK inhibitor II failed to inhibit MMP-3 protein expression. Thus, p38 pathway signaling mediated both MSU crystal–induced NO production and MMP-3 expression in chondrocytes. Next, potential signaling events upstream of p38 that MSU crystals used to induce NO generation and MMP-3 expression were investigated.
MSU crystal–induced activation of FAK, Pyk-2, and c-Src kinases and the adaptor protein paxillin in chondrocytes.
Consistent with previous observations in leukocytes (11–13), MSU crystals induced rapid tyrosine phosphorylation of multiple proteins in chondrocytes (Figure 5A). The array of tyrosine-phosphorylated proteins in chondrocytes included the FAK family members FAK and Pyk-2 (Figure 5B), which were recently demonstrated to be expressed by chondrocytes (28) and are known to interact with c-Src and the adaptor protein paxillin (29). Concurrent phosphorylation of c-Src and paxillin in response to MSU crystals in primary articular chondrocytes was demonstrated (Figure 5B).
To test the roles of FAK and Pyk-2 signaling in the induction of NO production and MMP-3 expression by MSU crystals, AG-82 and AG-17 were used, which were confirmed (results not shown) to exhibit preferential inhibitory effects for FAK and Pyk-2 (41, 42), respectively. AG-17, but not AG-82, markedly inhibited MSU crystal–induced iNOS expression, NO production, and MMP-3 expression (Figure 6A). To further address the results obtained with pharmacologic inhibitors, transient transfections with kinase mutants to FAK and Pyk-2 were performed in primary bovine chondrocytes. The recently validated FuGENE 6–hyaluronidase transfection methods for primary bovine chondrocytes were used (32), as described in Materials and Methods, via control β-galactosidase transfection and staining, to yield a transfection efficiency of ≥25% in this study.
Primary chondrocytes were transfected with HA-tagged FRNK, a truncated, inactivate form of FAK that serves as a dominant-negative regulator of FAK signaling (43). In addition, Myc-tagged wild-type Pyk-2, and Pyk-2-F402 and Pyk-2-A457 were transfected, which have point mutations at the autophosphorylation site and in the kinase domain, respectively, associated with dominant-negative functions (44). Expression of the epitope-tagged transfected proteins was verified by SDS-PAGE/Western blot analyses using antibodies against Myc or hemagglutinin (results not shown). Under these transfection conditions, both of the aforementioned Pyk-2 mutants significantly inhibited MSU crystal–induced iNOS expression, NO release, and MMP-3 protein expression in comparison with empty vector control transfection or expression of wild-type Pyk-2 (Figure 6B). Although MSU crystals had induced rapid phosphorylation of FAK, the FAK mutant FRNK had no significant effect on MSU crystal–induced NO release and MMP-3 expression (Figure 6B).
Next, the role of c-Src in the induction of NO production and MMP-3 expression by MSU crystals was tested. PP1, a selective pharmacologic inhibitor for Src family tyrosine kinases, inhibited MSU crystal–induced NO production and MMP-3 expression (Figure 7A). Thus, primary chondrocytes were transfected with hemagglutinin-tagged Csk, a native, C-terminal Src kinase that negatively regulates Src tyrosine kinase function (45). Expression of the hemagglutinin-tagged Csk was verified by Western blot analysis using antibody to hemagglutinin (results not shown). Under these transfection conditions, MSU crystal–induced iNOS expression, NO production, and MMP-3 expression were significantly inhibited by Csk (Figure 7B).
Pyk-2 and Src family kinase signaling can mediate downstream activation of MAPKs (44, 46, 47). Thus, we examined if upstream Pyk-2 and Src tyrosine kinases transduced downstream p38 pathway activation in response to MSU crystals. Significant changes in p38 MAPK phosphorylation were below limits of detection in primary bovine knee chondrocytes transfected with Pyk-2 and Csk constructs using the methods described above (results not shown). Therefore, human immortalized chondrocytic T/C28a2 cells were used in a lipofectamine-based method (33) to obtain a doubling of transfection efficiency (to ≥50%) relative to primary bovine knee chondrocytes transfected as described above (33). T/C28a2 cells transfected with Pyk-2 and Csk constructs were first verified to express each epitope-tagged protein by SDS-PAGE/Western blot analyses (results not shown). Under these conditions, MSU crystal–induced phosphorylation of p38 MAPK was consistently inhibited by both Pyk-2 mutants (Pyk-2-F402 and Pyk-2-A457), as well as by Csk in T/C28a cells stimulated under nonadherent culture conditions with only 1% FCS supplementation of medium (Figure 8). Therefore, MSU crystal–induced activation of Pyk-2 and Src family tyrosine kinases transduced p38 pathway activation in chondrocytes (Figure 9).
Tophaceous MSU crystal deposits can develop in both the synovium and cartilage in gout and promote cartilage and bone destruction. Furthermore, renal failure, allopurinol hypersensitivity, cyclosporine treatment, and other factors render some cases of gout refractory to successful long-term management with currently avail-able antihyperuricemics (48). In addition, radiologic changes in established tophaceous gout can progress despite adequate lowering of serum uric acid levels (49), reflecting the capacity of deposited urate crystals to promote connective tissue degradation. Hence, the basic mechanisms by which MSU crystals induce inflammatory mediators and MMPs in resident articular connective tissue cells were studied. Chondrocyte stimulation was used as a model system to yield insight into signal transduction pathways relevant to development of cartilage degradation in chronic tophaceous gout.
Within minutes of addition to chondrocytes, MSU crystals triggered stress kinase signaling, consistent with direct activation of the cells by the crystals. Within hours, MSU crystals induced iNOS expression, NO production, and MMP-3. IL-1 up-regulates iNOS expression, NO generation, and expression of certain MMPs (50, 51), and IL-1β was induced by MSU crystals in chondrocytes. Thus, IL-1 and potentially other cytokines could amplify the essential capacity of MSU crystals to activate chondrocytes. But in this study, the induction of NO generation and MMP-3 expression in chondrocytes by MSU crystals did not require IL-1.
MSU crystal–induced iNOS gene expression was regulated at the transcription level in chondrocytes. MSU crystals induce NF-κB activation in monocytic cells (14). The iNOS gene promoter contains functional binding sites for NF-κB that mediate the induction of iNOS expression by IL-1 in chondrocytes (52). In addition, certain pharmacologic inhibitors of NF-κB block MSU crystal–induced iNOS expression and NO release in chondrocytes (Liu R, et al: unpublished observations). But the extent and fundamental mechanisms of NF-κB activation by MSU crystals in chondrocytes remain to be defined.
MSU crystals did not up-regulate MMP-3 expression at the transcription level in chondrocytes, but did stimulate MMP-3 translation. Interestingly, NO generation became markedly up-regulated at approximately the same time (16–48 hours) as marked MMP-3 expression up-regulation in response to MSU crystals. The observation that the NOS inhibitor NG-monomethyl-L-arginine suppresses MSU crystal–induced MMP-3 expression at the protein level (Liu R, et al: unpublished observations) additionally suggests a significant regulatory role of NO in the capacity of MSU crystals to induce MMP-3.
MSU crystals concurrently activated the ERK-1/2, JNK, and p38 MAPK signaling pathways in chondrocytes. Significantly, the ERK-1/2 signaling pathway is essential for the induction of MMP-1 and MMP-3 by basic calcium phosphate (BCP) crystals in adherent human fibroblasts (53). But here, in MSU crystal–stimulated chondrocytes studied under nonadherent conditions, the p38 pathway was the principal MAPK cascade involved in the induction of both NO production and MMP-3 expression. Although the JNK pathway also mediated the induction of NO production, it was not essential in the induction of MMP-3, for reasons that will be of interest to determine. MSU crystals generally have substantially more inflammatory potential than BCP crystals deposited in the joint. Differences between sequelae of MSU and BCP crystal deposition in the joint might partially reflect differences in crystal-induced MAPK signaling.
MSU crystals stimulated phosphorylation of the FAK family members FAK and Pyk-2 in chondrocytes, but only activation of Pyk-2 mediated activation of p38 MAPK in response to MSU crystals in chondrocytes. FAK and Pyk-2 are highly similar in their sequences and structural configuration, their sites for potential phosphorylation, and their capacity to bind components of focal adhesions, including paxillin (29, 54), which binds both FAK members and certain α integrin cytoplasmic domains (55). FAK and Pyk-2 have both been implicated in promoting signal transduction events leading to MAPK activation (29, 54). Furthermore, both FAK and Pyk-2 can become tyrosine-phosphorylated in response to integrin-mediated and non-integrin–mediated stimuli (29, 54). However, distinct properties of FAK and Pyk-2 have been recognized. These include FAK localization primarily to focal adhesions in adherent cells, whereas Pyk-2 is mainly diffused throughout the cytoplasm and is concentrated in the perinuclear region (54). Differential subcellular localization of FAK and Pyk-2 may account for the failure of Pyk-2 to correct the migration defect in FAK knockout fibroblasts (56). Significantly, the kinase domain of Pyk-2 has also been reported to be more catalytically active than that of FAK and this correlates with greater tyrosine phosphorylation of substrates by Pyk-2 (57). Furthermore, Pyk-2 activation is stimulated by a variety of extracellular agonists that increase intracellular Ca2+ concentration. In contrast, FAK activation is principally a feature of integrin-mediated adhesion signaling (54), and integrin clustering is intimately involved in the induction of tyrosine phosphorylation of FAK (54).
Our results underscore recent observations indicating that FAK and Pyk-2, which have similar structures and potential functions, nevertheless differ with respect to modes of activation, cellular localization, catalytic activity, and substrate specificity (54, 56, 57). One or more of these differences may explain the preferential requirement of Pyk-2 in MSU crystal–induced NO and MMP-3 expression. In this context, the capacity of MSU crystals to stimulate increased cytosolic Ca2+ is well recognized in cells other than chondrocytes (58). MSU crystals also transduce secretory responses and stimulation of the respiratory burst partly through integrin-mediated signaling in cells other than chondrocytes (11). Furthermore, normal human and bovine chondrocytes express certain integrins, including α1β1, α5β1, and αVβ5 (59, 60), and ligand-induced α5β1 integrin engagement promotes MAPK activation, IL-1 expression, NO generation, and MMP-13 expression in chondrocytes (61–63). Thus, the mechanisms used by MSU crystals to activate FAK and Pyk-2, although not specifically tested in this study, might include integrin crosslinking and clustering and cytosolic calcium mobilization. Whether p38 pathway signaling and NO generation amplify and sustain FAK family kinase activation in response to MSU crystals also remains to be determined.
MSU crystals induced tyrosine phosphorylation of c-Src tyrosine kinase in bovine chondrocytes. Moreover, MSU crystal–induced NO production and MMP-3 expression were inhibited by the pharmacologic Src family tyrosine kinase inhibitor PP1 and by overexpression of the native Src family kinase inhibitor Csk. In addition, activation of Src family tyrosine kinases was implicated in MSU crystal–induced activation of p38 MAPK. It is noteworthy that c-Src can mediate tyrosine phosphorylation of Pyk-2 and paxillin (29, 55). Moreover, Pyk-2 contains binding sites for Src family tyrosine kinases and paxillin (29, 54, 55), and MSU crystals stimulated paxillin phosphorylation in chondrocytes. Hence, we speculate that the observed MSU crystal–induced phosphorylation of c-Src, Pyk-2, and paxillin by MSU crystals in chondrocytes is coordinated and that paxillin could be among the regulators of MSU crystal–induced NO production and MMP-3 expression.
In conclusion, MSU crystals directly activate FAK and Pyk-2 in chondrocytes, and Pyk-2 and c-Src signaling transduces activation of p38 MAPK, thereby promoting induction of iNOS expression, NO production, and MMP-3 expression. Our results suggest a novel model for MSU crystal–induced promotion of cartilage matrix degradation in chronic tophaceous gout (schematized in Figure 9). In addition, the differential effects of Pyk-2 activation, relative to FAK activation on crystal-induced NO generation and MMP-3 expression, suggest that crystal-induced chondrocyte Pyk-2 activation could trigger expression of genes that promote cartilage matrix catabolism. Hence, chondrocyte Pyk-2 activation may provide a selective intervention site to preserve cartilage integrity in chronic gout. Interestingly, FAK expression is nearly ubiquitous, but Pyk-2 expression is more restricted (29, 54). As such, studies of Pyk-2 function have previously focused principally on cells of the central nervous system and of hematopoietic lineages (29, 54). The findings of this study underscore the functional significance of Pyk-2 expression by articular chondrocytes and also reinforce the recently advanced notion that chondrocyte Pyk-2 activation modulated by cartilage matrix remodeling (63) is a potential therapeutic target in arthritis.
We gratefully acknowledge the assistance of Karen Thiesfeld in the preparation of the manuscript.
- 1Gout. In: KlippelJH and DieppeP, editors. Slide Atlas of Rheumatology. London: Gower Medical Publishing; 1985., , , .
- 14Extracellular signal-regulated kinase 1/extracellular signal-regulated kinase 2 mitogen-activated protein kinase signaling and activation of activator protein 1 and nuclear factor κB transcription factors play central roles in interleukin-8 expression stimulated by monosodium urate monohydrate and calcium pyrophosphate crystals in monocytic cells. Arthritis Rheum 2000; 43: 1145–55., , , , , .
- 62Fibronectin fragments and blocking antibodies to α2β1 and α5β1 integrins stimulate mitogen-activated protein kinase signaling and increase collagenase 3 (matrix metalloproteinase 13) production by human articular chondrocytes. Arthritis Rheum 2002; 46: 2368–76., , .
- 63Fibronectin-fragment activation of Pyk-2 mediates integrin signals regulating collgenase-3 expression by chondrocytes. Arthritis Rheum 2002; 46(Suppl): S372., , .