Injury to articular cartilage is strongly associated with the development of osteoarthritis (OA). Gross damage to the articular surface is known to occur following intraarticular fracture and is a predictable risk factor for the development of secondary OA (1), perhaps accounting for up to 10% of OA cases (for review, see ref.2). Excessive wear and tear, through overuse, age, joint instability, malalignment, and obesity, account for most of the other cases and are possibly induced by chronic low-grade microscopic tissue injury.

How joint tissue responds to mechanical disruption has been the subject of intense research interest over the years. In 1963, Meachim (3) performed a comprehensive study of the effects of superficial scarification in rabbit articular cartilage to determine whether injury per se led to cartilage degeneration. He found an early increase in metachromasia around superficial chondrocytes near the cuts, with subsequent loss of interterritorial proteoglycan. At 16 weeks, cell clusters formed, especially in deep layers associated with the cuts. He concluded that these changes were consistent with those observed in OA and that the chondrocytes developed a reactive phenotype in order to try to replenish their damaged matrix (3). Similar studies by Mankin showed that chondrocyte proliferation occurred adjacent to the cut surface. He concluded that the cells of articular cartilage were responsive to injury and were metabolically active (for review, see ref.4).

Since the 1960s, a number of observations have been made relating to the response of articular cartilage to injury in vitro. Chondrocyte death, a mixture of necrosis and apoptosis, was described by Tew et al (5) following trephine injury of cartilage explants. This response was greater following blunt, rather than sharp, injury (6). Impact loading, a model of high-velocity mechanical injury to cartilage, was shown to induce cell death by apoptosis (7), as well as reversible suppression of proteoglycan and protein synthesis (8). Impact loading of cartilage explants also triggered migration of cells through the damaged matrix to the articular surface (9), suggesting activation and recruitment of intrinsic repair cells. We have previously described activation of the 3 MAPKs upon explantation of articular cartilage from porcine joints, which results in induction of interleukin-1 protein (10). Activation of the ERK was due to release of fibroblast growth factor 2 (FGF-2), which is bound to perlecan in the pericellular matrix of cartilage (11, 12).

In this issue of Arthritis & Rheumatism, Dell'Accio and colleagues (13) describe a microarray analysis of genes up-regulated by articular cartilage explant injury. Cartilage explants were rested for 1 week following explantation to allow for changes that occur upon initial dissection and were then recut. Gene analysis was performed on cut and uncut explants 24 hours later. The authors make a number of interesting observations. One of the most significant of these is the strong up-regulation of Wnt-16. The Wnt gene was first defined as a protooncogene, but is now recognized to be one of a family of ∼19 related molecules (14). Wnt ligands bind to cell surface Frizzled receptors as well as to the coreceptors low-density lipoprotein receptor–related protein 5 (LRP-5) and LRP-6. This results in activation of the canonical pathway of Wnt signaling, mediated by stabilization and nuclear translocation of β-catenin (15). In development, Wnt signaling is essential for chondrocyte differentiation during endochondral ossification (16).

Wnt activity is negatively regulated by dickkopf-1, which inhibits signaling by binding to LRP-6 (17), and the family of antagonists that bind directly to Wnt ligands, which include secreted Frizzled-related proteins such as FRZB, Wnt inhibitory factor 1, and Cerberus (for review, see ref.18). One exciting recent discovery was that of another potential Wnt antagonist, Klotho. Klotho has been shown to bind to and neutralize Wnt proteins in culture, and mice deficient in Klotho had evidence of increased Wnt activity, which led to increased cell senescence and produced an “aging” phenotype (19).

Wnt family members have emerged over recent years as potential key players in the development of OA. Genetic studies have revealed functional variants within the FRZB gene that are associated with hip OA in women (20) and with generalized OA (21). Loughlin et al (20) demonstrated that an Arg342Gly substitution in FRZB led to diminished Wnt neutralizing activity in vitro, suggesting that loss of function might predispose to OA. These results were corroborated in the FRZB-knockout mouse, which was recently described by Lories et al (22). This mouse developed accelerated experimental OA and was also noted to have increased cortical bone thickness. A bone phenotype is also apparent in a polymorphism in Klotho, which has been associated with hand OA, particularly where osteophytes are a prominent clinical feature (23), and following inhibition of dickkopf-1, which leads to osteophyte formation and suppression of bone resorption in murine inflammatory arthritis (24).

Along with Wnt-16 gene induction, Dell'Accio and colleagues show that explant injury leads to suppression of FRZB and increased expression of a number of known Wnt-regulated genes, such as osteoprotegerin. Using immunohistochemistry, they demonstrate Wnt-16 and β-catenin expression in injured cartilage explants as well as OA tissue, confirming that this pathway is active in human disease. Together, these observations suggest that Wnt activity may be important in OA, possibly driving bone changes in disease, such as osteophyte formation.

Are the injury-regulated genes identified by Dell'Accio and colleagues similar to those found to have altered expression in OA tissue? Three microarray studies of OA surgical specimens have been published (25–27), and 1 recent study has been performed on rat articular cartilage 4 weeks following surgical joint destabilization, i.e., at an early stage of OA (28). Figure 1 summarizes the key up- and down-regulated genes identified in these studies and shows the gene overlap between studies. What is perhaps most surprising is the apparent diversity of genes expressed in the 3 studies of late-stage OA. Indeed, other than type I collagen, no genes were strongly up- or down-regulated in all 3 groups. This is most likely due to the heterogeneous nature of diseased tissue, as well as the fact that tissue sampling from severely affected and less affected areas varied between the studies.

thumbnail image

Figure 1. Comparison of genes found to be up- or down-regulated in late-stage osteoarthritis (OA) (Aigner 2006 [25], Aigner 2001 [26], and Sato 2006 [27]), early OA in a rat model (Appleton 2007 [28]), and cartilage injury (Dell'Accio 2008 [13]). Genes shown are those that were significantly decreased (arrows) or increased compared with the study control and were selected based on their putative role in disease. Most genes were up- or down-regulated >2-fold compared with controls, apart from ADAMTS-5 (1.7-fold). Parentheses indicate genes that are represented elsewhere on the figure, but where diagrammatic overlap was not possible. CILP = cartilage intermediate-layer protein; IL-1R2 = interleukin-1 receptor type II; Col V = type V collagen; TIMP-1 = tissue inhibitor of metalloproteinases 1; SPARC = secreted protein, acidic and rich in cysteine; MMP-3 = matrix metalloproteinase 3; FGFR-1 = fibroblast growth factor receptor 1; FRIT2 = Frizzled-related FRZB 2; BMP-2 = bone morphogenetic protein 2; OPG = osteoprotegerin; VEGF = vascular endothelial growth factor; ASPN = asporin; TGFβ1 = transforming growth factor β1; NGF = nerve growth factor; IGFBP-3 = insulin-like growth factor binding protein 3; PGES = prostaglandin E synthase; WISP-2 = Wnt-induced signaling protein 2; TLR-2 = Toll-like receptor 2; sFRP-4 = secreted Frizzled receptor protein 4.

Download figure to PowerPoint

It is not a great surprise that late-stage OA gene expression does not closely resemble that induced by cartilage injury. Genes predominantly expressed in OA tissue were those that one might consider to be “repair response genes” and included genes coding for matrix proteins, such as types I, II, III, and VI collagen, perlecan, fibronectin, fibromodulin, osteonectin, and biglycan. Down-regulation of degradative enzymes such as matrix metalloproteinase 2 (MMP-2), MMP-13 (27), and MMP-3 (25, 26) was observed.

It is perhaps intuitive that the biggest observed overlap should be between injury and “early OA” in the rat. A number of genes were up-regulated in those studies (13, 28), including inhibin A, transforming growth factor β2, and insulin-like growth factor binding protein 3, and both showed a significant reduction in FGF receptor 2 expression. The rat study (28) was particularly informative, because it was the first to reveal increased expression of ADAMTS-5, one of the key aggrecan-degrading enzymes implicated in disease (29, 30). It also confirmed increased expression of MMP-13, which is regarded as a key collagenase in OA and which has previously been detected in OA tissue (31, 32), but not up-regulated in OA microarray analyses. Genes generally considered to be downstream of intracellular inflammatory pathways, such as prostaglandin E synthase, CCL2, and tumor necrosis factor–stimulated gene 6, were also strongly up-regulated in early OA in the rat. It is tempting to speculate that these genes were not detected in the study by Dell'Accio et al because the investigators looked at the response to recutting rested cartilage explants and not at explantation of cartilage. In our own studies, we have observed that inflammatory signaling pathways, such as JNK and p38, that are activated upon cartilage explantation are not activated in response to recutting (11).

Overall, these approaches to understanding basic chondrocyte biology lead us toward a better understanding of tissue processes and the signaling pathways that are activated in disease. Further investigation into the pathways revealed in this injury microarray is likely to be highly informative.


  1. Top of page