What makes the permanent articular cartilage permanent?


Articular cartilage functions by providing a specialized mechanically competent extracellular matrix to withstand load bearing, thus protecting the underlying bones and allowing their near friction-free articulation in the joints. The resident chondrocytes maintain this function throughout life and do so by retaining a stable phenotype that resists hypertrophy and vascular invasion from the bone. This articular phenotype is distinct from the chondrocytes that drive endochondral ossification, which undergo hypertrophy and apoptosis followed by vascularization and bone formation (1). Unfortunately, stem cells used for cartilage repair seem to have a similar fate, making the repair tissue inadequate for normal joint function (2). How the articular chondrocyte avoids this and maintains its specialized phenotype is one of the fundamentally important issues in skeletal biology, yet the mechanisms remain unclear. Recent studies, however, have begun to provide some much needed insights.

Although different growth factors (including transforming growth factor β family members, bone morphogenetic proteins, and fibroblast growth factor family members) are important in skeletal development, there is relatively little evidence that these molecules are endogenously produced in physiologically significant amounts by adult articular cartilage. However, Klinger and colleagues, whose article appears elsewhere in this issue of Arthritis & Rheumatism (3), have identified a critical role for secreted matrix protein chondromodulin 1 in stabilizing the chondrocyte phenotype and inhibiting vascular invasion and endochondral ossification in stem cell–mediated articular cartilage repair (3). In their study, using a miniature pig model of cartilage repair, chondromodulin 1 was overexpressed in osteochondral progenitor cells (cells infected with adeno-associated virus vectors carrying chondromodulin 1 complementary DNA [AAV-Chm-1]), or AAV-Chm-1 vectors were directly administered to cartilage defects undergoing microfracture to induce repair. In both cases of chondromodulin 1 treatment, elaboration of a type II collagen–rich matrix was seen at 6 weeks, and most importantly, the tissue resisted calcification and vascular invasion over an extended experimental period (6 months). When progenitor cells were administered to cartilage defects with intact subchondral bone, although little calcification was observed, the repair tissue was fibrocartilaginous, with strong staining for type I collagen and less staining for type II collagen compared to the equivalent chondromodulin 1 treatment.

Chondromodulin 1 has been shown to have antiangiogenic properties (4), but its mechanism of action is far from fully elucidated. Since Klinger and colleagues found no effect of chondromodulin 1 overexpression on VEGF mRNA levels, it appears not to inhibit this key angiogenic factor, at least in vitro. However, the situation in vivo is liable to be far more complex, and it will be of great interest to assess the levels of VEGF in chondromodulin 1–treated cartilage repair tissue. Klinger and colleagues also suggest that chondromodulin 1 prevents chondrocyte hypertrophy through suppression of type X collagen in cultured osteochondral progenitor cells. This in vitro finding must now be investigated in vivo, and it will also be of great interest to investigate if runt-related transcription factor 2 (RUNX-2) is down-regulated by chondromodulin 1, since this transcription factor plays an important role in endochondral ossification through regulation of type X collagen and induction of VEGFA (5, 6).

The role of SOX9 in this process should also be investigated further. SOX9 has been shown to be essential for early events in cartilage differentiation (7) and for expression of the cartilage-specific matrix genes in human articular chondrocytes (8). Moreover, mutations in SOX9 cause the severe skeletal abnormality of campomelic dysplasia (9). Interestingly, however, SOX9 is greatly down-regulated as chondrocytes undergo hypertrophy in the growth plate during endochondral ossification (10). This raises the question: Is maintenance of SOX9 expression key to maintenance of the permanent articular cartilage? In a very elegant series of experiments, Hattori and colleagues (11) made some progress in addressing this issue. Through the development and use of mice misexpressing Sox9 specifically in hypertrophic chondrocytes (under the control of a BAC-Col10a1 promoter), they demonstrated that Sox9 is an important negative regulator of cartilage vascularization and endochondral ossification (11). Furthermore, they showed by in situ hybridization and real-time polymerase chain reaction that Vegfa, Mmp13, and osteopontin were all down-regulated in hypertrophic chondrocytes that misexpress Sox9, thus indicating that terminal differentiation was inhibited. Vegfa was also shown to be negatively regulated by direct binding of Sox9 to the Vegfa promoter.

Although chondromodulin 1 was not investigated in that study, it is very tempting to speculate that SOX9 regulation of chondromodulin 1 plays an important role in maintaining articular cartilage, in particular by inhibiting vascularization. Shukunami and colleagues (12) performed some very informative localization studies of chondromodulin 1 in the developing mouse and chick (12). They found that onset of chondromodulin 1 expression was associated with chondrogenesis during mouse embryonic development, being highly expressed in the avascular regions of the developing cartilage. Chondromodulin 1 strongly colocalized with the main cartilage matrix marker and key Sox9 target gene, Col2a1, in the developing chick wing. Most interestingly, chondromodulin 1 was also detected in the avascular mesenchyme in chick embryos, including the developing heart, and colocalized strongly with Sox9 in the avascular endocardial cushions that form the valve structures. Indeed, several of the key Sox9 target genes, which are critical to cartilage function (e.g., Col2a1), have also been detected in the heart valves (13).

Recently, it was shown that Sox9 function is key in regulating normal heart valve function in adult mice, and its reduction (by crossing loxP-flanked Sox9 with Col2a1-cre mice) led to calcification of the heart valves (the main contributor to heart valve malfunction) (14). Ectopic calcification was also observed by direct knockdown of Sox9 in heart valves in explant experiments; and increased Runx2 and osteopontin expression was detected. Hence, the parallels with development and maintenance of articular cartilage are most striking, with both tissue types requiring SOX9 levels to be maintained in order to prevent calcification and maintain normal tissue structure and, hence, function.

When one considers the function of the heart valves, these parallels are perhaps not quite so surprising. Similar to articular cartilage, the heart valves must develop a specialized extracellular matrix to resist the harsh mechanical environment in which they function throughout life. For both tissue types, this environment also excludes the possibility of the development of delicate blood vessels; therefore, avascularity is a functional requirement, and chondromodulin 1 seems likely to play a key role in this process.

In human articular chondrocytes, chondromodulin 1 has been shown to be highly SOX9-dependent, and furthermore, chondromodulin 1 expression was shown to be up-regulated in response to hypoxic conditions, which may more closely mimic those found in vivo in the avascular human articular cartilage (15). In fact, this hypoxic up-regulation of chondromodulin 1 in isolated human articular chondrocytes was largely prevented by depletion of SOX9 using RNA interference. Hence, chondromodulin 1 is most likely indirectly up-regulated by hypoxia, i.e., via SOX9, which is up-regulated under hypoxic conditions by hypoxia-inducible factor 2α (8). Whether SOX9 directly binds and regulates chondromodulin 1 is an important unresolved question. However, this pathway provides one plausible mechanism by which this avascular tissue maintains its function throughout life, i.e., the chronic hypoxia in the cartilage helps maintain SOX9 expression levels that prevent vascular invasion (through, e.g., chondromodulin 1 induction) in addition to direct inhibition of terminal differentiation and subsequent calcification of the tissue as occurs, for example, in the murine growth plate when Sox9 levels greatly decrease prior to hypertrophy (10).

The role of mechanical factors is also likely to be important in the maintenance of articular cartilage. For example, parathyroid hormone–related protein (PTHrP), which is critical to regulation of the growth plate, has been shown to be up-regulated in response to mechanical loading in cultured chondrocytes (16). Furthermore, a role has been suggested for load-induced expression of PTHrP in the maintenance of articular cartilage following experiments in a mouse model of joint loading and unloading (17). Much work remains to be done to investigate the relevant mechanosensitive mechanisms participating in articular cartilage homeostasis and specifically prevention of hypertrophy and vascularization; a fruitful area of future investigation will be the intersection of hypoxia and mechanically regulated pathways. Such insights are needed not only to further our basic understanding of articular cartilage biology, but also to help develop novel and improved strategies for cartilage repair and treatments for degenerative diseases such as osteoarthritis.


Dr. Murphy drafted the article, revised it critically for important intellectual content, and approved the final version to be published.