Osteoarthritis (OA) is a common and disabling disease, but the mechanisms that drive the disease are unclear. Although the etiology is likely multifactorial, increasing evidence has suggested that the synovium is involved in the induction of cartilage degradation during OA and RA development [10
]. The results of this and other studies [10
] suggest that the chemokine SDF-1 plays an important role in the development of OA. A complementary expression pattern exists between SDF-1 and CXCR4; the synovium produces SDF-1, whereas its receptor, CXCR4 is preferentially expressed by articular chondrocytes [7
]. Kanbe et al.
] showed that human chondrocytes expressed functional chemokine receptors and released MMP-3 and MMP-13 in response to SDF-1 [10
]. In organ culture, we showed that high concentrations of SDF-1 (250 ng/ml), comparable to levels observed in the synovial fluid of osteoarthritic knees (> 200 ng/ml), can readily penetrate the articular cartilage. This suggests that SDF-1 synthesized by synovial cells can diffuse freely into the adjacent cartilage. Previous studies showed that SDF-1 binds to glycosaminoglycans in the extracellular matrix or on the cell surface [26
], which may stabilize SDF-1 and result in its accumulation around chondrocytes. Pericellular accumulation of SDF-1 around OA chondrocytes, in which CXCR4 is upregulated [15
] (Figure ), could heighten the response to SDF-1 and induce cartilage matrix degradation.
With Safranin-O staining, we demonstrated that the cartilage explants treated with SDF-1 for as little as 2 days had significantly less proteoglycan (PG) content than did untreated explants, whereas GAG levels were higher in the media of explants treated with SDF-1. This result is consistent with the findings of previous studies in which severe OA is associated with a decrease in PG content [29
]. Enlarged and empty lacunae were also observed in the cartilage treated with SDF-1, consistent with previous studies linking OA to decreased chondrocyte numbers and empty lacunae [31
In our investigation, the cartilage-degrading enzyme MMP-13 was upregulated with SDF-1 treatment. This is consistent with previous studies, which showed that SDF-1 increases MMP-3 and MMP-13 levels in a dose-dependent manner [9
]. We also demonstrated that the SDF-1-induced increases of MMP-13 levels can be disrupted by blocking the SDF-1 pathway with either anti-CXCR4 antibodies or siRNA directed against the CXCR4. Based on these findings, we propose that in OA, the synovium produces high levels of SDF-1 which induces matrix degradation via the release of MMP-13. Loss of the matrix disrupts the microenvironment surrounding chondrocytes, which leads to cell dysfunction and death. The cycle between matrix degradation and chondrocyte loss ultimately results in gross changes in cartilage characteristics of OA.
To test whether disruption of SDF-1 signaling could attenuate OA pathogenesis in vivo
, we treated the primary Hartley guinea pig OA model with AMD3100, a specific nonpeptide CXCR4 chemokine receptor antagonist [35
]. Dunkin Hartley guinea pigs develop spontaneous OA of the knee at around age 9 months, and the gross lesions are invariably present at 12 months [30
]. The data of body weight indicated the animal tolerated AMD3100 well at the dose of 160 μg/per day, a dose that has been reported safe in animals and humans [37
]. After 3 months of treatment, the gross and histology observations displayed less cartilage damage compared with the primary OA group and the sham group. The cartilage damage severity quantified by the Mankin score further indicates that AMD3100 treatment has the lowest Mankin score, almost half that of the primary OA group. Thus, the OA severity was attenuated by the injection of AMD3100.
Elevated concentrations of SDF-1 in synovial fluid have been observed in OA patients [10
] and the guinea pig OA model [22
]. Our data demonstrate that this pathologic elevation can be inhibited by blocking SDF-1/CXCR4 signaling with AMD3100. The regulation of SDF-1 expression is not fully understood. One study found that hypoxic conditions in the synovium of rheumatoid arthritis may induce production of SDF-1, contributing to the persistence of synovitis [40
]. Other recent studies showed that interleukin-1 appears to induce SDF-1 expression in human subacromial bursa [41
], and that SDF-1 expression in bursal cells can be inhibited by steroid and nonsteroidal antiinflammatory agents [42
]. These studies suggest that SDF-1 levels are likely associated with the progression of inflammation. In this study, blockage of the SDF-1/CXCR4 pathway by AMD3100 reduced the level of IL-1, which indirectly reduced SDF-1 level in synovial fluid through the IL-1-SDF-1 regulation pathway.
AMD3100 also reduced the level of GAG, MMP-13, MMP-1, and IL-1β in synovial fluid and IL-1β in serum. Matrix metalloproteinases (MMPs) play an important role in the extracellular matrix degradation [43
]. Previous work showed that MMPs are regulated by the SDF-1/CXCR4 axis in OA and growth-plate chondrocytes [10
]. In the present study, AMD3100 inhibited MMP-1 and MMP-13 release in OA animals. Thus this association raises the intriguing possibility that AMD3100 may reduce MMPs expression through the SDF-1/CXCR4 axis. IL-1β is one of the important factors in OA pathogenesis [44
] causing collagen and aggrecan breakdown [45
]. Targeted reduction of IL-1β via RNAi in guinea pig chondrocytes showed beneficial effect on the OA pathogenesis [46
]. In our study, the IL-1β level in synovial fluid is much higher than in serum, which is consistent with previous reports [47
]. Our study shows that AMD3100 decreases the IL-1β levels in both synovial fluid and serum, which may reduce its harmful effects on OA progression.
Our research suggests that chemokines, along with proinflammatory cytokines, play critical roles in the pathogenesis of OA. We demonstrated that SDF-1/CXCR4 signaling directly induces cartilage matrix degradation via MMP-13 release, and that this destruction can be inhibited by blocking the SDF-1/CXCR4 pathway. The loss of matrix integrity directly compromises the mechanical properties of cartilage and may act as an accelerating force in the progression of OA. We also demonstrated that AMD3100 decreases the level of several OA-related factors in our animal model, attenuating the severity of primary OA. The age-related tissue wear results in fibrillation of the articular surface. This mechanical damage of the cartilage surface allows increasing amounts of inflammatory chemokines, such as SDF-1, to infiltrate into the cartilage. Once in the matrix, SDF-1 can then interact with CXCR4 on the cell surface, resulting in the release of MMPs and subsequent cartilage degeneration. Addition of AMD3100 blocked the binding of SDF-1 with CXCR4, removing the stimulatory trigger for the release of matrix-degrading enzymes, such as MMPs. Thus, further damage of the joint was suppressed.