4. Discussion
Although many efforts have been made, the cellular origin of osteoblasts responsible for HO is a long standing question which has yet to be solved [20]. Various candidate cells, both hematopoietic and non-hematopoietic, derived from different germ layers have been shown to potentially contribute to HO [1,3–7,12,21–24]. We have previously demonstrated that circulating osteoblast progenitor cells originating from bone marrow contributed to the development of heterotopic bone induced by BMP-2 [3,4]. These circulating osteoblast progenitor cells are negative for CD45, suggesting a non-hematopoietic origin of HO osteoblasts [4]. Contrary to our findings, others have reported that CD45 positive circulating cells formed HO [5,8,25,26]. Therefore, in this study, we traced hematopoietic cells using CD45-Cre mice and investigated whether hematopoietic derived cells contributed to skeletogenesis, HO and fracture healing. The advantage of using the Cre/loxP system is the ability to trace descendant cells derived from Cre-expressing cells, even if they discontinue expressing Cre. In our mouse model, cells that express CD45 at any developmental stage become DsRed positive and remain positive for DsRed even if CD45 expression turns off. In other words, all DsRed positive cells in our experimental animals originated from CD45 positive hematopoietic cells. Our results showed that over 90% of CD45 positive cells expressed DsRed, validating the efficiency of Cre-mediated recombination in our animal model. Additionally, we combined this Cre/loxP murine system with Col2.3GFP mice which are known to label mature osteoblasts with GFP [17], allowing us to evaluate if hematopoietic derived cells can give rise to mature osteoblasts. Our analyses of these triple transgenic mice demonstrated that hematopoietic derived cells did not differentiate into mature osteoblasts either in developmental skeletogenesis, in HO, or in bone regeneration after fracture. These results were supported by our findings that hematopoietic derived cells did not have the potential to become mature osteoblasts even under in vitro osteogenic conditions. Taken together, the results led us to conclude that hematopoietic derived cells are not responsible for heterotopic bone formation as osteoblasts in murine models. Possible explanations for the discovered inconsistencies with the previous findings by Suda et al. showing the contribution of CD45 positive circulating osteogenic precursor (COP) cells to heterotopic bone formation in patients with FOP [5] could be: 1) species differences between humans and mice; 2) our HO model induced by BMP-2 might not exactly reflect the pathogenesis and pathophysiology of patients with FOP; and 3) imperfection of Cre-mediated recombination, despite over 90% efficiency, might fail to label the CD45 positive COP cells, resulting in the failure to detect the contribution of these cells to HO.