Abstract
Bone is a dynamic tissue with self-healing capabilities that allow repair of most fractures with restoration of original architecture. However, large bone defects, such as those caused by tumor resections or severe trauma, do not regenerate spontaneously and represent a major clinical challenge for craniomaxillofacial and orthopedic surgeons. Bone grafts are usually used to manage such conditions. Bone autografts consist mainly of bone extracellular matrix (ECM) and associated cells. Recent research has unveiled many unique characteristics of ECM that play a key role in tissue regeneration. ECM enhances cell recruitment through cell surface receptors, which determine cell-ECM interactions and trigger specific cellular functions such as adhesion, proliferation, and differentiation. Discoidin domain receptor 2 (DDR2) is a collagen-activated receptor tyrosine kinase shown to be essential for skeletal development in humans and mice. Ddr2-deficient mice exhibit dwarfism and defective bone formation in the axial, appendicular and cranial skeletons. However, the role of DDR2 in bone regeneration has not yet been investigated. Here we evaluated the requirement for DDR2 in bone regeneration by using two well-established regeneration models; a calvarial subcritical-defect and tibial fracture. In a calvarial subcritical defect model, we showed that DDR2 is essential for regeneration of a subcritical-size defect. Smallie mice (Ddr2slie/slie), which contain a nonfunctional Ddr2 allele, are unable to heal a subcritical-size (0.5 mm) calvarial defect that, in WT mice, can spontaneously heal within 4 weeks. Also, Ddr2 expression during calvarial bone regeneration was defined using Ddr2-LacZ knock-in mice and β-galactosidase staining. Ddr2 expression, which was restricted to periosteal surfaces of uninjured calvarial bone, greatly expanded with injury. Similar results were seen when the lineage of Ddr2-expressing cells was examined using Ddr2creERT, Ai14 TdTomato mice. Ddr2+ cells and their progeny expanded within the defect three days and two weeks post-surgery. Furthermore, three days post-surgery, Ddr2slie/slie mice showed a significant decrease in cell proliferation in the calvarial defect when compared with WT littermates. Lastly, levels of the preosteoblast markers, Osterix and phosphorylated RUNX2 (S319-P) decreased in Ddr2slie/slie mice, which suggests that osteoblast differentiation was arrested. In the tibial fracture model, we first defined the expression pattern of DDR2 during fracture healing using Ddr2-LacZ knock-in mice and Ddr2-CreERT; Tdtomato mice. LacZ expression was first detected in select regions of the fracture site 2- and 5-days post fracture and expanded throughout the fracture callus after 1.5 and 3 weeks. Similar results were observed in Ddr2-CreERT;Tdtomato mice. Ddr2+ cells and their progeny began to expand in the developing fracture callus 1.5-weeks post-fracture and continued to expand after 3 weeks. Ddr2slie/slie mice exhibited significantly less fracture union than WT mice, and this defect was related to a decrease in cartilage formation as measured by safranin O staining. In addition, mutant mice developed significantly less callus tissue at 6-weeks post-fracture. To examine the role of DDR2 in skeletal progenitor cells (SPCs), we purified PDGFRα+ CD51+ SPCs from bone marrow of Ddr2fl/fl mice using FACS followed by treatment with AdCre. Ddr2 deletion resulted in defective osteoblast differentiation and accelerated adipogenesis. On the other hand, DDR2 overexpression in a mesenchymal cell line (ST2 cells) increased osteoblast differentiation Together, our study demonstrates that DDR2 is necessary for normal calvarial bone regeneration as well as for optimal fracture healing. This requirement may be explained in part by effects of DDR2 on proliferation, SPC function and osteoblast differentiation.