Comparative Single Cell Analysis of Side Population (SP) / CD45+ Cells from Marrow and Skeletal Muscle Reveals Evidence of Genuine Stem Cell Function and Multilineage Differentiation Ability in Muscle-Resident Stem Cells.

Murine skeletal muscle harbors hematopoietic stem cells. It has been suggested that these cells of hematopoietic origin have an altered stem cell function possibly due to their inadeqaute environment as compared to marrow-resident stem cells. The comparative quantitative and qualitative analysis of marrow and muscle-resident stem cells at the single cell level has not been performed so far. To this end, we have performed in vitro and in vivo stem cell detection assays using highly purified CD45+ cells, side population (SP) cells and SP/CD45 +cells. Muscle and marrow were found to contain 1–3 % and 0.2– 0.5 % of SP cells, respectively. The frequency of SP/CD45+ phenotype was 0.1–0.4% for the marrow and 0.2–0.5% in the muscle. Hematopoietic clonogenic cell efficiency from total nucleated cells was 1/500 for marrow and 1/10000 for muscle. Clonogenic efficiency of muscle CD45+ cells was about 1/3rd of that of marrow but with preserved erythroid and granulocytic differentiation ability. The use of SP/CD45+ cells from both muscle and marrow allowed an enrichment of clonogenic capacity by 60-fold in marrow and 360-fold in the muscle. In limiting dilution assays performed in MS-5 cells over 5 weeks, LTC-IC frequency was found to be 1/100 for marrow SP/CD45+ cells and 1/550 for muscle SP/CD45+ cells. To determine cloning and differentiation abilities of single SP/CD45+ cells purified from muscle and marrow, we have cultured single FACS-sorted cells in the presence of SCF, l-Flt3, IL-7, IL-11 and / or the OP-9 stroma (which promotes hematopoietic cell differentiation from embryonic stem cells) for 14– 21 days. Single SP/CD45+ cell cloning efficiency was 14% for marrow (109 wells + / 768) and 3% for muscle (42 wells+ / 1248). Despite this difference, single muscle-derived SP/CD45+ cells exhibited very robust proliferative activity, with 8– 13 cell doubling being obtained in 8 days in the presence of either cytokines alone or OP9 cells + cytokines, leading to absolute numbers of up to 60000/well. More importantly, like marrow SP/CD45+ cells, individual muscle-derived cells exhibited multilineage differentiation ability, with evidence of myeloid, B, NK and dendritic cell differentiation at day 14–21. In in vivo reconstitution experiments, the mean % of Ly5.1 chimerism generated after transplantation of of highly purified marrow SP/CD45+ cells ( 30 – 5700 cells/mouse, n=25 mice) and muscle SP/CD45+ cells ( 300–6500 cells / mouse) was 60% and 9 %, respectively (+ 8 months). To determine if this difference could be due to homing characteristics, SP+/CD45+ cells of marrow (300 /mouse) and muscle origin (300–500/mouse) were transplanted in lethally irradiated Ly5.2 mice by intrafemoral injection. In these assays, muscle-derived SP+/CD45+ cells gave rise also to persistent but lower Ly5.1 chimerisms as compared to marrow (+ 3 months). Thus, our results demonstrate that murine skeletal muscle harbors true stem cells with extensive proliferative and multilineage differentiation ability but as compared to marrow this population occurs with lower frequency. This heterogeneity, explaining an apparently reduced stem cell function in vivo, is not due to homing inability. Experiments underway will determine the in vivo potential of single muscle-resident stem cells.