Introduction

The cartilage extracellular matrix (ECM) contains two suprastructures, a collagen meshwork and a proteoglycan gel, providing the tissue with both tensile strength and compressive resistance. These suprastructures are further stabilized by a large number of small proteoglycans and many non-collagenous glycoproteins, like e.g. several members of the thrombospondin family. The continuous degradation of the cartilage ECM is a hallmark of osteoarthritis (OA) and recent studies suggest that ECM proteins and proteolytic fragments released from degenerating cartilage may become biologically active factors and contribute to further joint destruction.

Content

Thrombospondin-5, also referred to as cartilage oligomeric matrix protein (COMP), is ubiquitously expressed in healthy cartilage. In cartilage tissue, COMP interacts with collagens, proteoglycans and minor components (fibronectin, matrilins), most likely stabilizing the ECM. In addition to this structural function, COMP is involved in collagen secretion and acts as a catalyst for collagen fibril formation. Further, it has been shown that COMP can serve as a sensitive marker for the chondrocyte phenotype. In early OA, COMP is degraded but re-expressed at later stages of the disease. COMP is released from osteoarthritic cartilage and therefore widely used as diagnsotic OA biomarker. Indeed, serum levels correlate with both the number of affected joints and the severity of the disease. Using proteomic approaches and mass spectrometry, specific COMP-derived peptides were identified in OA cartilage but not in healthy tissue.

However, the function of COMP in OA tissue as well as the contribution of COMP fragments to the progression of OA is still largely unknown. Therefore, we recombinantly expressed full-length COMP and fragments thereof in HEK293-EBNA cells and purified the proteins from supernatants via affinity chromatography. We treated human primary chondrocytes with these COMP constructs and analyzed the cellular response in detail. We investigated the chondrocyte proliferation, migration, adhesion, differentiation and ECM synthesis. In addition, the activation of specific intracellular signalling pathways was monitored using immunoblotting and the expression of inflammatory cytokines like interleukin-6 and TNF-α was investigated by PCR.

We could show that full-length COMP is able to attract chondrocytes and to mediate cell attachment but does not affect cell proliferation. COMP stimulates the expression of chondrocyte-specific matrix proteins like collagens II, IX and XII and contributes to the stabilization of the chondrocyte phenotype by increasing the ratio of collagen II to collagen I expression. Full-length COMP induces a time- and concentration-dependent phosporylation of the MAP kinases ERK1/2 while JNK and p38 kinase are not activated. COMP fragments seem to have only minor effects. Neither full length COMP nor its fragments increase the expression of inflammatory cytokines.

In summary, we could show that full-length COMP has beneficial effects on chondrocyte behavior and phenotype stabilization. Therefore, its re-expression in later OA stages might be interpreted as an attempt to counteract cartilage degeneration. COMP is degraded during OA progression but the degradation products seem to be rather inert. In summary, full-length COMP might be a promising candidate to be added in autologous chondrocyte implantation approaches or incorporated into scaffolds in cartilage tissue engineering.