The discoidin domain receptors DDR and DDR
The discoidin domain receptors, DDR1 and DDR2, are receptor tyrosine kinases that are activated by different types of collagen (Shrivastava et al., 1997, Vogel et al., 1997). Both receptors interact with a number of fibrillar collagens, and DDR1, but not DDR2, is also activated by the network-forming collagen IV. Interaction of the DDRs with collagen leads to receptor autophosphorylation, the fist step in transmembrane signalling. The DDRs are unusual in that collagen-induced autophosphorylation is very slow and sustained (∼hours), compared to the much more rapid response of typical receptor tyrosine kinases to their ligands (∼seconds to minutes). The DDRs are widely expressed in human and mouse tissue, with DDR1 mainly found in epithelial cells (Alves et al., 1995), but also on leukocytes (Yoshimura et al., 2005), and DDR2 mainly expressed in mesenchymal cells (Alves et al., 1995). Both DDRs are expressed early in embryonic development and are found in many adult tissues, with high levels of DDR1 found in lung, kidney, breast, and sodium fluoride tissue, while the mesenchymal DDR2 shows highest levels in skeletal muscle, skin, kidney and lung tissue (reviewed in Vogel et al., 2006). The generation of DDR1 −/− mice resulted in viable, but smaller animals, the smaller size being a result of defective mammary gland development of female mice resulting in a failure to lactate (Vogel et al., 2001). DDR2 −/− mice, on the other hand, exhibited a real growth defect with shortened long bones (Labrador et al., 2001). Evidence from in vitro studies and the DDR1 −/− and DDR2 −/− mice shows that the DDRs regulate cell proliferation, adhesion and motility, and control remodelling of the extracellular matrix by influencing the expression and activity of matrix metalloproteinases (Hou et al., 2001, Labrador et al., 2001, Olaso et al., 2001, Olaso et al., 2002, Vogel et al., 2001, Ferri et al., 2004). The DDRs are associated with a growing number of human diseases, including fibrotic diseases of the lung, kidney and liver, atherosclerosis, osteoarthritis, as well as several types of cancer (reviewed in Vogel et al., 2006).
The homologous DDRs are composed of an N-terminal discoidin homology domain, followed by a sequence of ∼220 amino acids unique to the DDRs, a transmembrane domain, a large juxtamembrane domain, and a conserved cytoplasmic tyrosine kinase domain. In a previous study, we demonstrated that DDR activation by collagen I is a consequence of direct DDR-collagen interaction via the discoidin domain, and we mapped the collagen I binding site on the DDR2 discoidin domain (Leitinger, 2003). In another study, we mapped the DDR2 binding site on collagen II to the collagen II D2 period (Leitinger et al., 2004).
In this study, we sought to define whether collagen X is a ligand for the DDRs. We were particularly interested in the interaction between DDR2 and collagen X, because DDR2 is known to function in cartilage: DDR2 −/− mice have shorter long bones due to reduced chondrocyte proliferation (Labrador et al., 2001); DDR2 is associated with osteoarthritis in a mouse model of the disease (Xu et al., 2005); and in our previous study, we found that the cartilage-specific collagen II was a much better ligand for DDR2 than for DDR1 (Leitinger et al., 2004). Whether DDR1 has any function in cartilage is currently not known. We found that collagen X is primarily a ligand for DDR2. The characterisation of collagen X receptors and their respective binding site/s is essential for our understanding of how collagen X regulates chondrocyte metabolism and will provide important clues to the mechanism and functional significance of cell adhesion to collagen X.
Discussion Despite nearly two decades of investigations, the functional role of collagen X deposition in the growth plate ECM remains ill-defined. Collagen X null-mutation studies have provided support to the notion that the collagen X network is important in maintaining the compositional and organisational integrity of the ECM, but also pointed to the involvement of the hypertrophic ECM in hematopoiesis (Kwan et al., 1997, Gress and Jacenko, 2000). With its unique temporal and spatial expression pattern and its localisations in both the pericellular and interterritorial matrices, we hypothesised that the pericellular network of collagen X provides an important structural link between hypertrophic chondrocytes and the ECM. Cell signalling events following the binding of collagen X to the respective receptors are important mediators of processes in EO such as cell maturation and matrix turnover. This notion is supported by our recent report of the interaction between hypertrophic chondrocytes with purified collagen X via the collagen binding integrin α2β1 (Luckman et al., 2003).