In both primary and passaged cultures of mMSCs, the unwanted growth of nonmesenchymal cells is observed. Therefore, the isolation of MSCs from mice is far more difficult than that from other species . To resolve this problem, we used an isolation technique that is considered to be specific for separating mMSCs [22–26]. Differentiation assays showed that under appropriate culture conditions, mMSCs of C57BL/10 and mdx origin could differentiate into the adipocyte phenotype. This property is presently considered to be a critical requirement in identifying a putative MSC population [27, 28]. However in this study, with regard to the osteogenesis ability of these mMSCs, the results of RT-PCR and alizarin red staining indicated that the osteogenesis ability of C57BL/10 MSCs was much greater than that of mdx MSCs (Fig. 5). Further research is necessary to clarify this finding.
Immunophenotyping of both types of mMSCs at passages 5, 9, and 19 showed that CD11b, which is a marker for granulocytes, monocytes, and natural killer cells, was absent in mMSCs derived from both mice. In our culture system, there was no need to sort CD11b-positive cells from the medium [29, 30]. In addition, both mMSCs at different passage levels exhibited strong expression of Sca-1.
CD34 is a very interesting stem cell marker. It is expressed on the surface of HSCs, satellite cells, and endothelial progenitors. Its pattern and level of expression in muscle stem cells change as these cells differentiate into myotubes . It is not expressed on the surface of the MSCs of most species [11, 32]. However, some discrepancies were observed between the findings of studies on CD34 expression on the surface of mMSCs [33–36]. In our study, significant differences were observed between the 2 types of mMSCs with regard to the expression of CD34 (p < 0.05). This may result from differences in the niches of C57BL/10 and mdx mMSCs. It can also be attributed to the different behaviors of the 2 types of mMSCs. The mdx MSCs lost their proliferative capability gradually at passage 21 and formed a mesh-like network (Fig. 1I), and most cells decreased in size. These cells comprised a small population (Fig. 4F–G) at passage 19 and showed increased heterochromatin (Fig. 2C). Human MSCs and other cells show different morphological changes when they lose their proliferative ability, and senescent cells are larger and flatter than cells at other growth stages . Compared to the changes observed in mdx MSCs, C57BL/10 MSCs showed persistent proliferative capability and no morphological changes over a period of 5 months. Meirelles et al  showed that C57BL/10 cells maintained homogeneous characteristics over a period of 8 months . The correlation between proliferative ability and CD34 expression suggests that CD34+ cells may divide many more times and proliferate to a greater extent than CD34- cells [38, 39]. In addition, although the colony-forming efficiency of the 2 types of mMSCs in this study was not significantly different, the colony-forming efficiency of CD34+ MSCs was greater than that of CD34- cells (Additional tab. 1). We found that C57BL/10 MSCs (CD34+ MSCs) could form myotubes (Fig. 7A, E) under suitable conditions , whereas mdx MSCs (CD34- MSCs) did not do so under comparable conditions. This also may be associated with differences between the 2 types of MSCs with regard to the expression of CD34 and imply that CD34+ MSCs may exert a greater role in myogenesis than CD34- MSCs [31, 41, 42]. This finding also suggested that CD34+ MSCs in the BM are one of the candidate cells involved in the formation of both myoblasts and myofibers. .
Another explanation for the differences between the 2 types of mMSCs with regard to myogenesis is that the dystrophin protein may play a role in myotube maintenance. The lack of dystrophin can lead to abnormal calcium homeostasis  and elevated calpain proteolysis  in the myotubes, resulting in a decrease in myotube formation.
There are only 5 published studies on the ultratructure of MSCs [28, 46–49]. These focus on the changes in normal MSCs obtained after trypsinization. In our study, cells were scraped from flasks; these cells might better reflect the original cell state than cells treated with trypsin.
Two studies [48, 49] showed that MSCs are "frequently binucleate," but according to a study by Raimondo et al , the 2 nuclei observed in MSCs can be attributed to the irregular shape of the nucleus. Our findings suggested that some binucleate mMSCs indeed existed (Fig. 2B). Three observations supported this: first, the nuclei in the MSCs were longer than those observed in other cells; second, each nucleus had its own nucleolus; and third, the space between the 2 nuclei was large.
Three studies [28, 48, 49] have reported the occurrence of vacuoles in the cytoplasm of MSCs. According to Raimondo et al , in most MSCs, vacuoles are formed as a result of the dilatation of the endoplasmic reticulum and Golgi apparatus. When the vacuoles in the C57BL/10 and mdx mMSCs were compared, the existence of 2 types of vacuoles was noted. The small vacuoles observed in both mMSCs had arisen from the dilatation of the endoplasmic reticulum and Golgi apparatus . Large vacuoles were mainly observed in mdx MSCs in addition to lysosomes. They were formed as a result of cell degeneration, which could be one of the reasons for the loss in the proliferative ability of mdx MSCs.
In conclusion, our data showed that C57BL/10 MSCs (CD34+ MSCs) exhibited stronger proliferative capability and myogenic potential than mdx MSCs (CD34- MSCs). dystrophin protein would play main role in the changes of mouse MSC behavior.
Future research should be directed in 2 ways. First, the relationship between dystrophin and myotube characteristics such as myotube maintenance and myotube contraction should be determined; second, the proliferation and in vivo delivery of CD34+ MSCs should be investigated to clarify their potential application as therapeutic agents in mdx mice.