Highly Purified Hematopoietic Stem Cells Convert to Cells with Myoblast Characteristics before Forming Myotubes.

We have previously shown robust conversion of marrow cells to skeletal muscle (up to 12 percent) in a regenerative environment. We have also recently shown the presence of marrow derived muscle colonies. Evaluation of GFP fluorescent intensity of fibers in individual colonies and comparing them with the single fibers outside the colonies strongly suggests a clonal origin for these colonies. Cotransplantation of marrow cells from Yellow and Cyan Fluorescent Protein transgenic mice also supported this theory. We studied the type of cells responsible for marrow to muscle conversion in an immunocompetent and an immunocomprosied model of marrow transplant. The choice of two models was to clarify any bias related to immunologic reactivity of GFP positive cells that has been well documented in the literature. In the first model, different population of GFP positive marrow cells were transplanted intravenously, after sublethal doses of radiation into C57BL/6 mice. One week after transplant their tibialis anterior muscle were injured with cardiotoxin and their muscles were evaluated for GFP positive, dystrophin positive and CD45 negative muscle fibers. In the second model, different population of GFP positive marrow cells were injected directly into tibilias anterior of Bege/Scid mice one day after cardiotoxin injury. Similar to the first model, GFP positive muscle fibers were evaluated 4 weeks post injury. For each population of marrow cells 3 to 6 animals were transplant with cells either positive or negative for the specific surface markers. Our data demonstrates that conversion potential of marrow cells to skeletal muscle fibers is highly enriched in a CD45 positive, c-Kit positive, Flk-2 positive and Sca-1 positive population. Furthermore, CD34 negative, CXCR4 negative and lineage negative cells have higher potential to participate in marrow to muscle conversion. This indicates that under the conditions of our experiments, mesenchymal stem cells are not the origin of marrow derived GFP muscle fiber. To follow the fate of marrow cells homed into muscle we transplanted GFP marrow cells into radiated C57BL/6 mice and at different timepoints before and after injury, harvested tibialis anterior muscle of these animals. Isolation of mononuclear cells from these samples revealed a new GFP+ CD45 negative population which was very rare at 1 week but more abundant at 4 weeks after transplant. Real time PCR analysis of these cells showed multiple stem cells markers including Sca-1, endoglin, c-Kit and CD34. They did not express any RNA markers for satellite cells or myoblasts (pax 7, MyoD, Myf 5) or hematopoietic cells (CD45, CD11b). We next injured GFP transplanted animals with cardiotoxin and isolated the same CD45-GFP+ population 3 days after injury. Real time PCR analysis at this point showed RNA activity for myoblast including Myf5, MyoD and desmin. Immunofluorescent staining of Lin-, Sca+ cKit+ marrow cells from GFP transgenic mice injected into cardiotoxin injured tibialis anterior muscle of Bege Scid mice confirmed our RT-PCR data and showed the presence of MyoD, Myf5, Myogenin and desmin in donor derived cells in recipient muscle. These data suggests that marrow derived mononuclear cells with hematopoietic stem cell phenotype in the recipient muscle lose their hematopoietic characteristics and obtain myoblast phenomenon after being exposed to injury.