Postnatal Osterix but not DMP1 lineage cells significantly contribute to intramembranous ossification in three preclinical models of bone injury.

Murine models of long-bone fracture, stress fracture, and cortical defect are used to discern the cellular and molecular mediators of intramembranous and endochondral bone healing. Previous work has shown that Osterix (Osx(+)) and Dentin Matrix Protein-1 (DMP1(+)) lineage cells and their progeny contribute to injury-induced woven bone formation during femoral fracture, ulnar stress fracture, and tibial cortical defect repair. However, the contribution of pre-existing versus newly-derived Osx(+) and DMP1(+) lineage cells in these murine models of bone injury is unclear. We addressed this knowledge gap by using male and female 12-week-old, tamoxifen-inducible Osx Cre_ERT2 and DMP1 Cre_ERT2 mice harboring the Ai9 TdTomato reporter allele. To trace pre-existing Osx(+) and DMP1(+) lineage cells, tamoxifen (TMX: 100 mg/kg gavage) was given in a pulse manner (three doses, 4 weeks before injury), while to label pre-existing and newly-derived lineage Osx(+) and DMP1(+) cells, TMX was first given 2 weeks before injury and continuously (twice weekly) throughout healing. TdTomato positive (TdT(+)) cell area and cell fraction were quantified from frozen histological sections of injured and uninjured contralateral samples at times corresponding with active woven bone formation in each model. We found that in uninjured cortical bone tissue, Osx Cre_ERT2 was more efficient than DMP1 Cre_ERT2 at labeling the periosteal and endosteal surfaces, as well as intracortical osteocytes. Pulse-labeling revealed that pre-existing Osx(+) lineage and their progeny, but not pre-existing DMP1(+) lineage cells and their progeny, significantly contributed to woven bone formation in all three injury models. In particular, these pre-existing Osx(+) lineage cells mainly lined new woven bone surfaces and became embedded as osteocytes. In contrast, with continuous dosing, both Osx(+) and DMP1(+) lineage cells and their progeny contributed to intramembranous woven bone formation, with higher TdT(+) tissue area and cell fraction in Osx(+) lineage versus DMP1(+) lineage calluses (femoral fracture and ulnar stress fracture). Similarly, Osx(+) and DMP1(+) lineage cells and their progeny significantly contributed to endochondral callus regions with continuous dosing only, with higher TdT(+) chondrocyte fraction in Osx(+) versus DMP1(+) cell lineages. In summary, pre-existing Osx(+) but not DMP1(+) lineage cells and their progeny make up a significant amount of woven bone cells (particularly osteocytes) across three preclinical models of bone injury. Therefore, Osx(+) cell lineage modulation may prove to be an effective therapy to enhance bone regeneration.