• Mary B. Goldring,

    Corresponding author
    1. Laboratory for Cartilage Biology, Research Division, The Hospital for Special Surgery, Weill College of Medicine of Cornell University, New York
    • Hospital for Special Surgery, Caspary Research Building, Room 528, 535 East 70th Street, New York, NY 10021.
    Search for more papers by this author
  • Steven R. Goldring

    1. Laboratory for Cartilage Biology, Research Division, The Hospital for Special Surgery, Weill College of Medicine of Cornell University, New York
    Search for more papers by this author


Osteoarthritis (OA) is characterized by degeneration of articular cartilage, limited intraarticular inflammation with synovitis, and changes in peri-articular and subchondral bone. Multiple factors are involved in the pathogenesis of OA, including mechanical influences, the effects of aging on cartilage matrix composition and structure, and genetic factors. Since the initial stages of OA involve increased cell proliferation and synthesis of matrix proteins, proteinases, growth factors, cytokines, and other inflammatory mediators by chondrocytes, research has focused on the chondrocyte as the cellular mediator of OA pathogenesis. The other cells and tissues of the joint, including the synovium and subchondral bone, also contribute to pathogenesis. The adult articular chondrocyte, which normally maintains the cartilage with a low turnover of matrix constituents, has limited capacity to regenerate the original cartilage matrix architecture. It may attempt to recapitulate phenotypes of early stages of cartilage development, but the precise zonal variations of the original cartilage cannot be replicated. Current pharmacological interventions that address chronic pain are insufficient, and no proven structure-modifying therapy is available. Cartilage tissue engineering with or without gene therapy is the subject of intense investigation. There are multiple animal models of OA, but there is no single model that faithfully replicates the human disease. This review will focus on questions currently under study that may lead to better understanding of mechanisms of OA pathogenesis and elucidation of effective strategies for therapy, with emphasis on mechanisms that affect the function of chondrocytes and interactions with surrounding tissues. J. Cell. Physiol. 213:626–634. © 2007 Wiley-Liss, Inc.

A majority of individuals over the age of 65 have radiographic and/or clinical evidence of osteoarthritis (OA). The most frequently affected sites are the hands, knees, hips, and spine. Importantly, the symptoms are often associated with significant functional impairment, as well as signs and symptoms of inflammation, including pain, stiffness, and loss of mobility (Felson, 2006). Anatomic analysis and application of histopathological and imaging techniques have helped to define the natural history of OA with respect to the structural alterations in the articular cartilage. They also have demonstrated that OA is not exclusively a disorder of articular cartilage (Brandt et al., 2006). Multiple components of the joint are adversely affected by OA, including the peri-articular bone, synovial joint lining, and adjacent supporting connective tissue elements. Multiple factors are known to affect the progression of OA, including joint instability and/or malalignment, obesity, increasing age, associated intra-articular crystal deposition, muscle weakness and peripheral neuropathy. These factors can be segregated into categories that include hereditary contributions, mechanical factors, and the effects of aging.

Biomechanical Modulation of Chondrocyte Function: Abnormal Loading on Normal Cartilage or Normal Loading on Abnormal Cartilage?

Current knowledge segregates the risk factors for development of OA into two fundamental mechanisms related either to the adverse effects of “abnormal” loading on normal cartilage or of “normal” loading on “abnormal” cartilage. Aging has been suggested as the primary factor contributing to this “abnormal” state of articular cartilage, although genetic factors that cause disruption of chondrocyte differentiation and function and influence the composition and structure of the cartilage matrix also contribute to abnormal biomechanics, independent of the influence of the aging process.

The articular surface plays an essential role in load transfer across the joint and there is good evidence that conditions that produce increased load transfer and/or altered patterns of load distribution can accelerate the initiation and progression of OA (Roos, 2005). For example, examination of the factors responsible for symptomatic OA 15–22 years after meniscectomy in a Swedish cohort showed that partial resection of the meniscus, which produced only limited disruption in the physiological biomechanical environment of the knee, was associated with a lower rate of radiographic progression than total meniscectomy (Englund and Lohmander, 2004). Also, these studies showed that risk factors for OA initiation and progression were similar to those for idiopathic OA in that systemic host factors and local biomechanical factors interacted and that obesity, female gender, and preexisting early-stage OA contributed significantly to adverse radiographic and clinical OA outcomes (Figure 1).

Figure 1.

Scheme of events involved in the initiation of osteoarthritis (OA) and progression to late stage OA. Potential causative factors are listed in the boxes above and the cellular and morphologial changes are listed below.

Chondrocytes can respond to direct biomechanical perturbation by upregulating synthetic activity or by increasing the production of inflammatory cytokines, which are also produced by other joint tissues. The consensus from in vitro mechanical loading experiments is that injurious static compression stimulates depletion of proteoglycans and damage to the collagen network and decreases the synthesis of cartilage matrix proteins, whereas dynamic compression increases matrix synthetic activity (Guilak et al., 2004). In response to traumatic injury, global gene expression is activated, resulting in increased expression of inflammatory mediators, cartilage-degrading proteinases, and stress response factors (Fitzgerald et al., 2004; Kurz et al., 2005). Chondrocytes have receptors for responding to mechanical stimulation, many of which are also receptors for extracellular matrix (ECM) components (Millward-Sadler and Salter, 2004). Included among these receptors, are several of the integrins, which serve as receptors for fibronectin (FN) and type II collagen fragments. Activation of these receptors can stimulate the production of matrix-degrading proteinases and inflammatory cytokines and chemokines (Pulai et al., 2005). Discoidin domain receptor-2 (DDR-2) represents an additional receptor that binds to and is activated by native type II collagen fibrils and preferentially activates matrix metalloproteinase (MMP)-13 (Xu et al., 2007).

Is OA a Genetic Disease?

There are several lines of evidence indicating that genetic abnormalities can result in earlier onset of OA (Valdes et al., 2006). Results of epidemiological studies, analysis of patterns of familial clustering, twin studies and the characterization of rare genetic disorders suggest that hereditary predisposition is a risk factor. For example, twin studies have shown that the influence of genetic factors may approach 70% in OA that affects certain joints. Candidate gene studies and genome-wide linkage analyses have revealed polymorphisms or mutations in genes encoding ECM and signaling molecules that may determine susceptibility to OA (Loughlin, 2005; Valdes et al., 2007) (Table 1). Interestingly, there is evidence that the genes may operate differently in males and females and at different body sites (Bukulmez et al., 2006). Many of the gene defects that affect the formation of cartilage matrix and patterning of skeletal elements during development result in a variety of congenital cartilage dysplasias. The adverse effects of joint malalignment and congruity, may contribute to the eventual loss of articular cartilage, and in some cases early onset of OA in these individuals (Li et al., 2007).

Table 1. Genes implicated in hand or hip osteoarthritis (OA)
Candidate gene studiesGenome-wide linkage scans
Extracellular matrix molecules
 Type II collagen (COL2A1)COL9A1
 Type IX collagen (COL9A1)Matrilin-3 (MATN3)
 Aggrecan (AGC1)CILP
 Cartilage intermediate layer protein (CILP) 
Signaling molecules
 Vitamin D receptor (VDR)Secreted frizzled related protein-3 (FRZB)
 Estrogen receptor α (ESR1)Bone morphogenetic protein-5 (BMP5)
 IL-1 gene clusterInterleukin-4 receptor (IL4R)
 Bone morphogenetic protein-2 (BMP2)Asporin (ASPN)
 CD36Calmodulin 1 (CALM1)
 Cyclooxygenase (COX) 2BMP2
 Osteoprotegerin (OPG)
 IL-17A & F
 Nuclear co-repressor (NCOR) 2
 Growth and differentiation factor (GDF) 5

Why Is Aging Cartilage Not Necessarily OA Cartilage?

Whereas it is clear that mechanical and genetic factors play major roles in determining the natural history of OA, the primary risk factor for OA is age. Nevertheless, it is important to note that, while changes in the composition and structure of the cartilage matrix are inevitable, the development of OA with aging, while common, is not universal (Carrington, 2005). The changes observed during aging in articular cartilage include softening of the articular surface and decreases in the tensile strength and stiffness of the matrix. These changes in the material properties of cartilage have been attributed to changes in the structural organization of the ECM (Dudhia, 2005; Loeser, 2006; Aigner et al., 2007). There is evidence that the major components of the ECM, type II collagen and proteoglycan, undergo changes in content, composition, and structural organization during the aging process. For example, aggrecan, which is the major cartilage proteoglycan, decreases in molecular size and its content in the ECM diminishes, likely contributing to an alteration in the biomechanical properties of the matrix. In addition, there is evidence of accumulation of advanced glycation end products (AGEs; Verzijl et al., 2003). This process enhances collagen cross-linking and has been proposed as a significant contributing factor to the increases in cartilage stiffness and altered biomechanical properties that have been observed with aging.

The capacity of the chondrocyte to remodel and repair the cartilage ECM diminishes with age and this has been attributed primarily to a decrease in the anabolic capacity of this cell (Dudhia, 2005; Aigner et al., 2007). Unless perturbed, the normal chondrocyte remains in a postmitotic quiescent state throughout life and the proliferative potential decreases with age. This has been attributed to replicative senescence associated with erosion of telomere length (Martin et al., 2004). Although controversial, there appears to be an increase in chondrocyte apoptosis in OA, and there is evidence of chondrocyte cloning that has been interpreted as a localized attempt at tissue regeneration and repair (Horton et al., 2005). It has also been suggested that the accumulation of cartilage matrix proteins in the endoplasmic reticulum and golgi of chondrocytes that have been modified by oxidant stress during aging leads to decreased synthesis of cartilage matrix proteins and eventual chondrocyte apoptosis (Yang et al., 2005).

What Is the Contribution of Phenotypic Modulation of Chondrocyte Function?

The chondrocyte, the only cell type residing in the adult cartilage matrix, has a low metabolic activity, surviving under relatively hypoxic conditions and in the absence of a vascular supply. This cell, which is ultimately responsible for remodeling and maintaining the structural and functional integrity of the cartilage matrix, possesses little regenerative capacity. During skeletal development, the chondrocyte arises from mesenchymal progenitors and synthesizes the templates, or cartilage anlagen, for the developing limbs in a process known as chondrogenesis (Goldring et al., 2006). Following mesenchymal condensation and chondroprogenitor cell differentiation, the chondrocytes undergo proliferation, terminal differentiation to chondrocyte hypertrophy, and apoptosis in a process termed endochondral ossification whereby the hypertrophic cartilage is replaced by bone. A similar sequence of events occurs in the postnatal growth plate and leads to rapid growth of the skeleton.

In the adult, articular chondrocytes are fully differentiated cells that remain after formation of articular cartilage matrix. The adult chondrocyte plays a critical role in the pathogenesis of OA in responding to adverse environmental stimuli by promoting matrix degradation and downregulating processes essential for cartilage repair. In the absence of disease, the chondrocytes maintain a low turnover rate of replacement of cartilage matrix proteins with a half-life for collagen of greater than 100 years (Verzijl et al., 2000). In contrast, the glycosaminoglycan constituents on the aggrecan core protein are more readily replaced and the half-life of aggrecan subfractions has been estimated in the range of 3–24 years (Maroudas et al., 1998).

In early OA, there is evidence of increased synthetic activity, which is viewed as an attempt to regenerate the matrix with cartilage-specific components, including types II, IX, and XI collagens, aggrecan, and pericellular type IV collagen (Poole et al., 2007). The aberrant behavior of OA chondrocytes is reflected in the appearance of fibrillations, matrix depletion, cell clusters, and changes in quantity, distribution, or composition of matrix proteins (Pritzker et al., 2006). Evidence of phenotypic modulation is reflected in the presence of collagens not normally found in adult articular cartilage, including the hypertrophic chondrocyte marker, type X collagen, as well as other chondrocyte differentiation genes, suggesting a recapitulation of a developmental program (Sandell and Aigner, 2001; Tchetina et al., 2005).

The Role of Inflammation: What Are the Critical Regulatory Molecules and Where Do They Come From?

OA is not considered a classical inflammatory arthropathy, due to the absence of neutrophils in the synovial fluid and the lack of systemic manifestations of inflammation. OA is frequently associated, however, with signs and symptoms of inflammation, including joint pain, swelling and stiffness leading to significant functional impairment and disability (Felson, 2006). Although there remains debate regarding the essential role of synovial inflammation in OA, synovitis involving infiltration of activated B cells and T lymphocytes and overexpression of proinflammatory mediators is common in early and late OA (Benito et al., 2005). Synovial inflammation is a factor that likely contributes to dysregulation of chondrocyte function, favoring an imbalance between the catabolic and anabolic activities of the chondrocyte in remodeling the cartilage ECM (Loeser, 2006).

Evidence from in vivo and in vitro studies indicate that chondrocytes can produce and/or respond to a number of cytokines and chemokines that are present in OA joint tissues and fluids (Figure 2). The relationship between the increased levels of catabolic enzymes and inflammatory mediators such as prostaglandins and nitric oxide and the levels of interleukin-1β (IL-1β) and tumor necrosis factor (TNF)-α in OA synovial fluids and joint tissue is well documented, Although the mechanism by which production of inflammatory mediators is initiated is unclear, abnormal mechanical, and oxidative stresses are probably involved. Chondrocytes in OA cartilage, especially those in clonal clusters, express IL-1, IL-1β converting enzyme (caspase-1), and type 1 IL-1 receptor (IL-1RI). IL-1 is synthesized by chondrocytes at concentrations that are capable of inducing the expression of MMPs, aggrecanases, and other catabolic genes and it colocalizes with TNF-α, MMP-1, 3, 8, and 13, and type II collagen cleavage epitopes in regions of matrix depletion in OA cartilage (Tetlow et al., 2001; Wu et al., 2002). In addition to inducing the synthesis of MMPs and other proteinases by chondrocytes, IL-1 and TNF-α increase the synthesis of prostaglandin E2 (PGE2) by stimulating the expression or activities of cyclooxygenase (COX)-2, microsomal PGE synthase-1 (mPGES-1), and soluble phospholipase A2 (sPLA2), and they up-regulate the production of nitric oxide via inducible nitric oxide synthetase (iNOS, or NOS2). IL-1β also induces other proinflammatory cytokines such as IL-6, leukemia inhibitory factor (LIF), IL-17, and IL-18, and chemokines, including IL-8, and suppresses the expression of a number of genes associated with the differentiated chondrocyte phenotype, including COL2A1 (see for review, Goldring and Berenbaum, 2004; Goldring and Goldring, 2004).

Figure 2.

The mediators and cellular changes involved in cartilage destruction in OA. The destruction of cartilage in OA is considered to be a chondrocyte-mediated process, in part. The normally quiescent chondrocytes, as well as the synovial cells, respond to repetitive excess mechanical loading via stress-induced intracellular signals that mediate the production of cytokines, chemokines, other inflammatory mediators, and cartilage-degrading proteinases. Matrix degradation products can feedback upregulate these cellular events. Anabolic factors, including BMPs and TGFβ, may also be upregulated and participate in osteophyte formation. In addition to matrix loss, evidence of earlier changes, such as chondrocyte proliferation (cloning), phenotypic modulaton (hypertrophy), and increased cartilage calcification (tidemark duplication), and microfractures with blood vessel invasion from the subchondral bone (angiogenesis), can be observed in late OA samples obtained from patients after total joint replacement. Many of these events can be observed in animal models, although tidemarks are not evident in mouse joints, where growth plate closure does not occur.

Chondrocytes express several chemokines, as well as chemokine receptors that could participate in the induction of cartilage catabolism (Borzi et al., 2004). IL-17, a T-lymphocyte product, also stimulates the production of other proinflammatory cytokines and has effects on chondrocytes similar to IL-1 (Lubberts et al., 2005). Many of these factors synergize with one another in promoting chondrocyte catabolic responses. For example, oncostatin M produces mild catabolic responses in chondrocytes through a gp130 receptor, but synergizes strongly with IL-1 or TNF-α (Barksby et al., 2006). The upregulation by IL-1β of COX2, MMP13, and NOS2 gene expression in chondrocytes and other cell types is mediated by the induction and activation a number of transcription factors, including NF-κB, C/EBP, AP-1, and ETS family members, which are involved generally in stress- and inflammation-induced signaling.

Catabolism Versus Anabolism: What Gets Turned on First?

The effects of abnormal mechanical loading and synovial inflammation likely contribute to dysregulation of chondrocyte function, favoring disequilibrium between the catabolic and anabolic activities of the chondrocyte in remodeling the cartilage ECM. The association of increased production of proteinases, including the metalloproteinases (MMPs), MMP-1, MMP-3, MMP-8, MMP-13, and the aggrecanases, particularly ADAMTS-5, with cartilage damage has been established (Cawston and Wilson, 2006; Plaas et al., 2007). Local loss of proteoglycans and cleavage of type II collagen occurs initially at the cartilage surface resulting in an increase in water content and loss of tensile strength in the cartilage matrix as the lesion progresses. There is evidence, however, of compensatory increases in type II collagen synthesis in deeper regions of the articular cartilage (Poole et al., 2007). Genomic and proteomic analyses of global gene expression in cartilage have confirmed the increased levels of type II collagen (COL2A1) mRNA levels in early OA cartilage (Bau et al., 2002; Hermansson et al., 2004; Aigner et al., 2006). The increased levels of anabolic factors such as bone morphogenetic protein-2 (BMP-2) and inhibin βA/activin, members of the TGF-β superfamily (Fukui et al., 2003; Nakase et al., 2003; Hermansson et al., 2004), as well as prostaglandins (Tchetina et al., 2006), suggest a possible mechanism. Nevertheless, Aigner et al. (2006) have shown that expression of the type II collagen gene (COL2A1) is suppressed in upper zones of OA cartilage with progressing matrix destruction, whereas global COL2A1 gene expression is increased in late-stage OA cartilage compared to normal and early degenerative cartilage. Importantly, once the cartilage is severely degraded the chondrocyte is unable to replicate the complex arrangement of collagen laid down during development.

What Is the Role of the Subchondral Bone and Does OA Pathogenesis Start or Progress There?

In addition to the progressive loss of articular cartilage, OA is characterized by increased subchondral plate thickness, formation of new bone at the joint margins (osteophytes) and the development of subchondral bone cysts (Hill et al., 2001; Buckland-Wright, 2004; Burr, 2004). Also, at the junction of the articular hyaline cartilage and adjacent subchondral bone, there is evidence of vascular invasion and advancement in the zone of calcified cartilage in the region of the so-called tidemark that further contributes to a decrease in articular cartilage thickness (Lane et al., 1977; Burr and Schaffler, 1997). A recent study showed that angiogenesis in the osteochondral junction is independent of synovial angiogenesis and synovitis, but is associated with cartilage changes and clinical disease activity (Walsh et al., 2007). These structural alterations in the articular cartilage and peri-articular bone may lead to modification of the contours of the adjacent articulating surfaces (Radin and Rose, 1986; Coats et al., 2003; Bullough, 2004; Messent et al., 2006). The accompanying alterations in subchondral bone remodeling and modulus may contribute further to the development of an adverse biomechanical environment and enhance the progression of the articular cartilage deterioration.

Changes in the mineral content and thickness of the calcified cartilage and the associated tidemark advancement may be related to the localization of COL10A1, MMP-13, and Runx2 in the deep zone of OA cartilage, where the chondrocytes may attempt a defective repair response by recapitulation of the hypertrophic phenotype (Aigner et al., 2004; Wang et al., 2004). Although receptor activator of NFκB (RANK), a member of the TNF receptor family, RANK ligand (RANKL), and the soluble receptor, osteoprotegerin (OPG), which regulate osteoclast differentiation and activity, are expressed in adult articular chondrocytes, a direct action in cartilage has not been identified (Komuro et al., 2001). Interestingly, Runx2-dependent expression of RANKL is present in the hypertrophic chondrocytes at the boundary next to the calcifying cartilage in the developing growth plate (Kishimoto et al., 2005). Although inhibition of RANKL expression does not block cartilage destruction directly, at least in inflammatory models, indirect effects may occur through protection of the bone (Pettit et al., 2001).

Support for the concept of modified biological activities in the subchondral bone of OA patients is provided by studies identifying polymorphisms in the gene encoding asporin, which inhibits cartilage anabolism by binding to TGF-β (Kizawa et al., 2005), and in FRZB, which encodes secreted frizzled-related protein 3 (sFRP3) (Loughlin et al., 2004; Lane et al., 2006). Members of the sFRP family, including sFRP3, are glycoproteins that antagonize the signaling of Wnt ligands through frizzled membrane-bound receptors. Since activation of β-catenin in mature cartilage cells stimulates hypertrophy, matrix mineralization, and expression of VEGF, ADAMTS5, MMP-13, and several other MMPs (Tamamura et al., 2005), defective inhibition of Wnt signaling due to FRZB polymorphisms may disrupt normal homeostasis resulting in abnormal cartilage and bone metabolism. The growth arrest and DNA damage inducible protein, GADD45β, which mediates cell survival and the expression of Mmp13 and Col10a1 in hypertrophic chondrocytes in the growth plate (Ijiri et al., 2005), also may function as a mediator that modulates chondrocyte phenotype in OA cartilage, where it is expressed in chondrocytes clusters and deep zone OA chondrocytes (Ijiri et al., 2007).

What Is the Source of Pain in the OA Joint and Can Anti-Pain Therapies Also Modify Structure?

In addition to the potentially adverse effects of synovial inflammation on chondrocyte function, the synovial products may also contribute to the symptoms of pain, which is the most prominent but least well-studied feature of OA. In general, painful mechanical stimuli are detected by types III and IV afferent (sensory) nerves that are located in the joint capsule, ligaments, periosteum, and subchondral bone (see for reviews, Niissalo et al., 2002; Schaible et al., 2002; McDougall, 2006a). These pain-sensing fibers, termed nociceptors, have a high threshold of activation. They release into the local microenvironment and respond to neuropeptides such as substance P, calcitonin gene related peptide (CGRP), neuropeptide Y, and vasoactive intestinal peptide (VIP). Movement of the joint induces the opening of mechano-gated ion channels located in the terminals of the sensory nerves resulting in depolarization and nerve firing. The action potentials are propagated to the central nervous system that translates the electrical activity into mechanosensation. When the limits of physiologic joint movement are exceeded, then the nerve firing dramatically increases and the central nervous system interprets these signals as pain. Developing models for assessing the pathogenetic factors and mechanisms responsible for joint pain in OA has been very challenging because of the subjective nature of pain and the complex role of cognitive interpretation of sensory information. Recent studies by Schuelert and McDougall, (2006) have shown that the mechanosensory nerves become sensitized in animals with experimental OA, resulting in an increase in afferent firing rate even in response to physiological joint motion. They also implicated VIP as a locally produced sensitizing agent that enhances pain perception. In more recent studies, employing the Dunkin Hartley guinea pig model of spontaneous OA, they observed that afferent firing for a given level of joint torque was significantly enhanced in older compared to younger animals (McDougall, 2006b). They speculated that joint nociception might be age-dependent, although they could not exclude the possibility that the more advanced OA in the older animals may have contributed to the enhanced sensitivity. Although the normal joint may respond predictably to painful stimuli, there is often a poor correlation between apparent joint disease and perceived pain in chronic arthritis. Thus, it will be difficult to assess the efficacy of structure-modifying drugs if they do not modify pain perception.

Are There Validated Approaches for Early Diagnosis Using Biomarkers and Non-Invasive Imaging Techniques?

Whereas pain represents the most prominent clinical manifestation associated with OA, the structural deterioration that characterizes this joint disorder represents the major factor contributing to disability and impaired quality of life. With increasing knowledge of the composition of the cartilage matrix, molecular markers in body fluids have been identified for monitoring changes in cartilage metabolism and for assessing joint damage in arthritis (Charni-Ben Tabassi and Garnero, 2007; Rousseau and Delmas, 2007). Monoclonal antibodies have been developed that recognize products of proteoglycan or collagen degradation (catabolic epitopes) or synthesis of newly synthesized matrix components (anabolic neo-epitopes), which represent attempts to repair the damaged matrix. Such epitopes can be detected in the synovial fluids and sera of patients with OA, and the synovial fluid-to-serum ratio has been suggested as a potential diagnostic indicator of cartilage turnover. The degradation of aggrecan in cartilage has been characterized using antibodies 846, 3B3(−) and 7D4 that detect chondroitin sulfate neoepitopes, 5D4 that detects keratan sulfate epitopes, and the VIDIPEN and NITEGE antibodies that recognize aggrecanase and MMP cleavage sites, respectively, within the interglobular G1 domain of aggrecan.

Similarly, it is possible to monitor the synthesis of type II collagen by measuring serum and synovial fluid levels of the carboxyl-terminal propeptide (CPII), and urinary excretion of hydroxylysyl pyridinoline cross-links may indicate collagen degradation. Specific antibodies that recognize epitopes on denatured type II collagen at the collagenase cleavage site are promising diagnostic reagents. These include the C2C (Col2-3/4CLong mono) and C1,2C (COL2-3/4Cshort) epitopes, which have been used to detect cleavage of the triple helix of type II collagen in experimental models and human OA cartilage. The ratios of these markers to the synthetic marker, CPII, are associated with a greater likelihood of radiological progression in OA patients (Cahue et al., 2007). Cartilage oligomeric matrix protein (COMP), chitinase 3-like protein 1 (CH3L1; also known as YKL-40 or HC-gp39), and cartilage-derived retinoic acid sensitive protein (CD-RAP) are also under investigation as OA biomarkers. Such biomarker assays have been used as research tools and are currently under development as diagnostic tools for monitoring cartilage degradation or repair in OA patient populations and for assessment of treatment effects in large cohorts. Although a single marker may not be sufficient, it may be possible eventually to utilize a combination of biomarkers that may discriminate between different stages of OA in different populations. The major challenge, however, will be to apply such biomarkers to the diagnosis and monitoring of OA in individual patients.

The introduction of magnetic resonance imaging (MRI) techniques in recent years has begun to address the need for noninvasive techniques for monitoring the effects of therapeutic interventions on the structural and functional properties of cartilage and bone (Burstein and Gray, 2006; Majumdar et al., 2006; Ding et al., 2007). MRI imaging techniques have been used to successfully quantify articular cartilage morphology, volume and thickness and to identify focal defects. In addition, the technology can be adapted to characterize changes in the composition of the ECM. This information is of particular interest since cartilage thinning in OA is preceded by modification of the composition and structural organization of the collagen-proteoglycan matrix. Using MRI, it is possible to detect these changes in matrix composition as alterations in molecular relaxation times T2 and T1rho, as well as in the uptake of contrast agents such as Gd-DTPA. In addition, alterations in the subchondral bone marrow accompany both OA and injury, and these changes can be detected by an increase in the signal intensity in the bone marrow on fat-saturated T2-weighted images. The presence of the so-called bone marrow edema has been of particular interest, since this finding is associated with the progression and severity of OA (Felson et al., 2001). Recently, investigators have explored the use of MRI to monitor the results of stem cell regeneration treatment strategies by detecting transplanted cells after labeling with super paramagnetic iron oxide (SPIO) particles (Majumdar et al., 2006). The cells metabolize the SPIO contrast agent, increasing their magnetic susceptibility, and as a result the cells appear as hypo-intense regions permitting their tissue detection and localization. Such techniques hold promise for tracking cell survival and outcomes in models of cartilage regeneration.

Are Tissue Engineering and Gene Therapy Viable Approaches in the Foreseeable Future?

Major challenges for cartilage repair are the restoration of the three-dimensional collagen structure and integration of the newly synthesized matrix with the resident tissue. Current procedures for cartilage repair include joint lavage, tissue debridement, microfracture of the subchondral bone, and the transplantation of autologous or allogeneic osteochondral grafts, in addition to the ultimate therapy of total joint replacement (reviewed in Hunziker, 2002). These procedures may lead to the formation of fibrous tissue, chondrocyte death, and further cartilage degeneration and thus have variable success rates. Autologous chondrocyte transplantation has been used to repair small cartilage lesions in young adults with sports injuries, but the outcomes are variable and the defects frequently are repaired with fibrocartilage. Novel approaches have been explored, therefore, to promote differentiation of the autologous chondrocytes by gene transfer ex vivo of anabolic factors such as BMPs, TGF-β, and IGF-1 before implantation (Evans, 2005; Lories and Luyten, 2005).

The capacity of several BMPs to stimulate the synthesis of the cartilage matrix proteins by adult articular chondrocytes and to promote cartilage repair in various models of focal cartilage defects provides the basis for their use in cartilage repair strategies. BMP-2 and BMP-7 (osteogenic protein-1/OP-1) are currently approved for multiple indications in the area of bone fracture repair and spinal fusion. However, consistent with their roles in vivo in promoting endochondral ossification, BMPs may promote chondrocyte hypertrophy in repair models, and they may serve as more effective anabolic factors for cartilage repair in juveniles where chondroprogenitors are available. BMPs and TGF-β also have been implicated in the formation of osteophytes, which exhibit features consistent with endochondral bone formation, and it has been suggested that this process represents an attempt at bone repair. Several studies have shown that injection of free TGF-β or adenovirus-mediated delivery of TGFβ promotes fibrosis and osteophyte formation, while stimulating proteoglycan synthesis in cartilage. Based on these observations, local application of molecules that block endogenous TGFβ signaling, such as the soluble form of TGFβRII, inhibitory SMADs, or the physiological antagonist, latency-associated peptide-1 (LAP-1), have been proposed as a means for blocking osteophyte formation (Blaney Davidson et al., 2007).

The feasibility of using mesenchymal stem cells (MSCs) from bone marrow or other tissue sites is also the subject of current research. Bone marrow-derived MSCs have the capacity to differentiate into cartilage, bone, fat, and muscle cells (Barry and Murphy, 2004; Caplan, 2005). The fates of these cells and patterning within tissues are determined by specific cell–cell and cell–matrix interactions that are controlled by the same growth and differentiation factors that regulate different stages of cartilage development (Magne et al., 2005; Goldring et al., 2006). Many members of the BMP/TGF-β family are able to induce chondrogenic differentiation of mesenchymal progenitor cells in vitro (Huang et al., 2005; Palmer et al., 2005). Another strategy for improving cartilage tissue engineering is the transduction of the chondrogenic transcription factor Sox9, alone or together with L-Sox5 and Sox6, into MSCs ex vivo or into joint tissues in vivo to induce cartilage formation (Ikeda et al., 2004; Tew et al., 2005).

Findings that MSCs are available in adult tissues, including muscle, adipose tissue and synovium in addition to bone marrow, has prompted further investigations in animal models to determine the potential of these cells to participate in the repair of skeletal tissues (Goldring, 2006; Kuroda et al., 2006). Despite intensive investigation of cartilage repair strategies and the increased understanding of the cellular mechanisms involved, many issues remain to be resolved. These include the fabrication and maintenance of the repair tissue in the same zonal composition as the original cartilage; the recruitment and maintenance of cells with an appropriate chondrocyte phenotype; and integration of the repair construct with the surrounding cartilage matrix (Kuo et al., 2006).

What Are the Lessons From Animal Models?

The classical models of OA have generally involved induction of OA-like changes in a number of animal species, including rats, rabbits, dogs, and sheep, using surgical techniques to produce joint instability, although spontaneous models also exist (Smith and Little, 2007). More recently, transgenic or knockout mouse models with defects in the expression of transcription factors, MMPs, angiogenic factors, or ECM proteins have provided insight into the mechanisms that control cartilage development and, in some cases, OA pathology (Table 2) (see for reviews, Helminen et al., 2002; Young, 2005; Goldring, 2007; Li et al., 2007). Deficiencies in TGF-β signaling molecules, type II TGFβ receptor, TgfβRII, or Smad3, and vascular endothelial growth factor (VEGF) or VEGF receptors impair hypertrophic chondrocyte differentiation and endochondral ossification, often resulting in cartilage changes in postnatal mice and OA-like cartilage pathology. Although disruption of BMP signaling by ablation of Bmpr1a results in early embryonic death, postnatal conditional knockout results in OA-like cartilage degeneration. Mutations in the Ank gene encoding a multipass transmembrane protein and knockout of nucleotide pyrophosphatase phosphodiesterase-1 (Nnp1), whose gene products control intracellular pyrophosphate levels cause progressive or early onset OA associated with increased chondrocyte hypertrophy and inappropriate mineralization of the cartilage matrix.

Table 2. Transgenic and knockout mouse models and susceptibility to OA
GeneGene defect or modificationFeatures
  1. For more details, see Bendele (2001), Helminen et al. (2002), Saamanen and Vuorio (2004), Kamekura et al. (2005, 2006), Ameye and Young (2006), Glasson et al. (2007), Goldring (2007), and Smith and Little (2007).

Extracellular matrix
Col2a1Heterozygous knockout or mutationOA-like changes during aging
Col9a1Transgenic truncationMild chondrodysplasia; OA
Col9a1KnockoutEarly onset OA
Col11a1Cho/+; spontaneous deletionOA-like changes during aging
AggrecanCmd/+; spontaneous deletionOA-like changes during aging
FibromodulinKnockoutTendon mineralization; OA
Fibromodulin/biglycanDouble knockoutEctopic mineralization; early-onset OA
Matrilin-3KnockoutOA-like changes with aging
Proteinases and inhibitors
Mmp9KnockoutExacerbation of surgically induced OA
Mmp13Postnatal transgenicIncreased susceptibility to OA
Adamts5KnockoutDecreased susceptibility to OA
Timp3KnockoutIncreased susceptibility to OA
Cytokines and related molecules
IL-1βKnockoutExacerbates OA in STR/ORT mice; protects against surgically induced OA
IL-6KnockoutSevere spontaneous OA with subchondral bone sclerosis in aging males but not in females
ICEKnockoutExacerbates OA in STR/ORT mice
MK2KnockoutMore severe in surgically induced OA
NOS2KnockoutReduced proteoglycan depletion and restoration of IGF-1 responsiveness
Growth factor signaling
Bmpr1aPostnatal conditional knockoutOA-like cartilage degeneration
TgfβRIITransgenic dominant-negative truncationEnhanced chondrocyte hypertrophy and progressive skeletal degeneration with OA
Smad3Targeted disruptionEnhanced chondrocyte hypertrophy and OA-like cartilage erosion
Ankank/ank; homozygous truncation mutationEarly onset OA associated with crystal deposition
Npp1KnockoutOA associated with crystal deposition
α1 integrinKnockoutAccelerated cartilage degradation
Runx2Conditional knockoutEnhance cartilage loss and increased osteophyte formation in surgical OA

A number of studies have shown that developmental deficiencies or mutations in cartilage matrix genes, including Col2a1, Col19a1, Col11a1, aggrecan, matrilin-3, or fibromodulin alone or together with biglycan, also lead to degenerative cartilage changes in the adult mice (Saamanen and Vuorio, 2004; Young, 2005; Li et al., 2007) (Table 2). Timp3 deficiency or postnatal overexpression of constitutively active Mmp13 also promotes OA-like pathology (Neuhold et al., 2001; Sahebjam et al., 2007). Thus, the loss or mutation of a single gene that is involved in synthesis or remodeling of the cartilage matrix may lead to a cascade of disruption of other genes and activities in chondrocytes, contributing collectively to alterations in the architecture of the cartilage and bone, joint instability, and OA-like pathology. Many of these genetically modified mice exhibit chondrodysplasias associated with defects in skeletal growth and joint congruity with phenotypes that correlate with human diseases, and the OA-like changes become more prevalent with aging and (Helminen et al., 2002; Li et al., 2007). However, single gene defects do not model all aspects of human OA.

Surgical models have been employed to permit more rapid evaluation of the consequences of genetic modification in mice (Bendele, 2001; Kamekura et al., 2005; Ameye and Young, 2006; Glasson et al., 2007). Of the proteinases that degrade aggrecan, only ADAMTS5 is associated with increased susceptibility to OA, as shown in Adamts5-deficient mice. Runx2-deficient mice exhibit decreased cartilage loss and osteophyte formation in response to surgical OA (Kamekura et al., 2006). Knockout of IL-1β is also protective against OA induced by destabilization of the medial meniscus, whereas OA is exacerbated in mice with deficiency in Mmp9 or Mk2 (mitogen activated kinase 2). According to Glasson et al. (2007), the severity of the OA due to surgical instability induced by destabilization of the medial meniscus depends upon the wild type mouse strain, the 129/SvEv strain developing the most severe OA, followed by C57BL/10, C57BL/6, FVB/N, and DBA, and male mice develop more severe OA than females. Although excellent for establishing proof-of-principle and for examining the time course of progression under different regimens, no experimental animal model is totally equivalent to OA in humans.

Conclusion: Future Directions for Research

OA is the most common joint disorder, affecting large segments of the population and leading to significant disability and impaired quality of life. An understanding of the cellular processes that regulate the functional activities of chondrocytes in both physiological and pathological conditions is essential to the development of more effective strategies for treating patients with OA and altering the natural history of this disorder. Despite the large amount of new information gleaned from in vitro and in vivo studies on chondrocyte responses to catabolic, inflammatory, and mechanical stimuli, there remain challenges to understanding how to modify the complex processes involved in OA pathogenesis. Although OA is defined as a cartilage disease, it is also necessary to consider the contributions of synovial inflammation and subchondral bone changes. Cytokines, acting individually or in networks, profoundly influence cellular responses in joint tissues, modifying both catabolic and anabolic activities. The adult chondrocyte maintains homeostasis under physiological conditions, but undergoes phenotypic modulation without achieving the precise temporal and spatial transitions that occurred when the cartilage was formed originally during development. Since OA progression cannot be halted if early events are not prevented, the current challenge is to use our understanding of the mechanisms regulating chondrocyte responses to develop new strategies for early diagnosis. Alternative approaches at later stages include cartilage tissue engineering enhanced by gene therapy with anabolic factors and MSCs. More information is also needed regarding the origin of pain and related symptoms in OA, so that more effective treatment strategies can be developed. Recent understanding of how the adult articular chondrocyte functions within its unique environment and application of knowledge of how the cartilage was formed during development will enable early diagnosis and the design of new therapies that will alter the course of the disease.


Research related to this work was supported by National Institutes of Health grants R01-AG022021 and R01-AR45378 and Biomedical Science Grants from the Arthritis Foundation.