7 ± 15.0 and 17.4 ± 8.0% respectively). The GFP+ fraction in the adjacent tissue of the skin was significantly larger than in the adjacent tissue of the mucoperiosteum (p = 0.004). The fraction of myofibroblasts (Fig. 3B) in the mucoperiosteal wounds (46.4 ± 23.8%) was

larger than in the adjacent tissue (0.69 ± 0.53%; p = 0.002) but also larger than in the skin wounds (7.3 ± 7.1%; p = 0.012). In contrast, the fraction of myofibroblasts in skin wounds and adjacent tissue was similar. The fraction of GFP-positive myofibroblasts ( Fig. 3C) in the mucoperiosteal wounds (4.6 ± 3.0%) was larger than in the adjacent tissue (0 ± 0%; p = 0.023), which was not the case in skin wounds and adjacent tissue. The fraction of activated fibroblasts (Fig. 3D) in the skin wounds (78.5 ± 4.7%) was slightly larger than in the Bleomycin mw adjacent tissue (64.6 EGFR inhibitor ±7.4%, p = 0.010). The slight difference in the mucoperiosteum was not significant. The fraction of GFP-positive activated fibroblasts tended to be larger in both types of wound tissues than in the adjacent tissues (

Fig. 3E). The fraction of macrophages (Fig. 3F) was not significantly different in all tissues. The mucoperiosteal adjacent tissue (7.5 ± 5.7%) and the skin adjacent tissue (16.1 ± 6.2%) contained similar numbers of macrophages. No significant differences were found in the fraction of GFP-positive macrophages (Fig. 3G). We hypothesized that more BMDCs are recruited to quickly healing tissues such as the oral mucosa than

to more slowly healing tissues such as the skin. This was based on earlier data obtained from regenerating endometrium of the human uterus where up to 48% of the epithelial cells are derived from the bone marrow.26 However, later it was shown in some mouse models that the contribution was far less.27 This is probably due to differences in the process of endometrial regeneration between humans and rodents. Previous studies indicate that about 14% of the cells in skin wounds in mice are derived from the bone marrow, and that this is increased by wounding.28 Our data show that about from 8% of the cells in mucoperiosteal wounds is recruited from the bone marrow, which is about 10 times higher than in the normal adjacent tissue. In contrast, the recruitment of BMDCs to skin wounds and the adjacent normal tissue is comparable, but about twofold larger than in mucoperiosteal wounds. Moreover, the total population of BMDCs in normal skin is about 25 times larger than in normal mucoperiosteum. Our data indicate that, in the mucoperiosteum, BMDCs are preferentially recruited to the wound but not in the skin. Alternatively, BMDCs recruitment in skin wounds might have peaked earlier than two weeks after wounding as reported in a mouse model.28 The long-term contribution of BMDCs, however, was similar to our findings. In the light of tissue remodelling and scarring, this might be the more relevant population.