SEARCH

SEARCH BY CITATION

Keywords:

  • haematopoietic stem cells;
  • mesenchymal stem cells;
  • skeletal muscle;
  • bone marrow

Summary

  1. Top of page
  2. Summary
  3. Experimental designs and methods
  4. Results
  5. Discussion
  6. References

It has previously been shown that bone marrow cells contribute to skeletal muscle regeneration, but the nature of marrow cell(s) involved in this process is unknown. We used an immunocompetent and an immunocompromised model of bone marrow transplantation to characterize the type of marrow cells participating regenerating skeletal muscle fibres. Animals were transplanted with different populations of marrow cells from Green Fluorescent Protein (GFP) transgenic mice and the presence of GFP(+) muscle fibres were evaluated in the cardiotoxin-injured tibialis anterior muscles. GFP(+) muscle fibres were found mostly in animals that received either CD45(−), lineage(−), c-Kit(+), Sca-1(+) or Flk-2(+) populations of marrow cells, suggesting that haematopoietic stem cells (HSC) rather than mesenchymal cells or more differentiated haematopoietic cells are responsible for the formation of GFP(+) muscle fibres. Mac-1 positive population of marrow cells was also associated with the emergence of GFP(+) skeletal muscle fibres. However, most of this activity was limited to either Mac-1(+) Sca(+) or Mac-1(+)c-Kit(+) cells with long-term haematopoietic repopulation capabilities, indicating a stem cell phenotype for these cells. Experiments in the immunocompromised transplant model showed that participation of HSC in the skeletal muscle fibre formation could occur without haematopoietic chimerism.

Numerous studies have reported that bone marrow cells may participate in the formation of skeletal muscle fibres. Although the frequency of these events is reported by some investigators to be low (0·01–0·1%), others have found that up to 5% of the fibres in a muscle stress model are bone marrow-derived (Brazelton et al, 2003; Corbel et al, 2003; Palermo et al, 2005). Our recent work demonstrated that injury was a critical aspect of this phenomenon and by using appropriate transplant variables, including cell dose, radiation dose, and mode of cell delivery, levels of up to 12% were possible (Abedi et al, 2004, 2005). Some investigators have suggested that marrow cells that have homed to muscle follow a biological progression, first forming satellite cells and then fusing to form mature myofibres (LaBarge & Blau, 2002). However, the same group later suggested that marrow cells directly fused to the muscle fibres (Doyonnas et al, 2004). It is not clear whether all bone marrow cells that participate in muscle regeneration follow this progression or if a proportion of cells fuse directly with existing myofibres without an intermediate transformation. A greater understanding of these variables could elucidate the mechanisms by which marrow cells are recruited, incorporated and reprogrammed during this process, and so aid the design of future experiments for the treatment of muscle diseases, such as the muscular dystrophies.

The haematopoietic system is comprised of many different cell types including mesenchymal cells, endothelial precursor cells, haematopoietic stem cells (HSC) and their progenitors, together with the more differentiated (i.e. lineage positive) cells such as monocytes, lymphocytes, granulocytes and erythrocytes. It is not known which subpopulation of marrow cells participate in muscle regeneration and attempts to identify the cell type has resulted in conflicting results suggesting either haematopoietic stem cells, mesenchymal stem cells, myelomonocytic cells, or even macrophages are responsible for this phenomenon (Camargo et al, 2003; Anjos-Afonso et al, 2004; Doyonnas et al, 2004; Dezawa et al, 2005). To address the controversy in the literature regarding the origin of marrow-derived muscle fibres, we conducted a systematic approach by isolating different subpopulations of marrow cells, based on their surface markers, and studied their participation in the formation of muscle fibres. Subpopulations of marrow cells were isolated using fluorescence-activated cell sorting (FACS)-based fractionation of cells from bone marrow, followed by intravenous injection into the radiated immunocompetent mice or by direct intramuscular injection in immunocompromised host. The latter approach makes it possible to examine more mature cells that are not capable of reconstituting the blood.

The present study found that HSC, rather than other marrow cell types (such as mesenchymal or stromal cells), contributed to muscle fibre formation. We further identified that HSC, and not their differentiated derivatives, have the capacity to be incorporated into skeletal muscle fibres.

Experimental designs and methods

  1. Top of page
  2. Summary
  3. Experimental designs and methods
  4. Results
  5. Discussion
  6. References

Experimental animals

C57BL/6 (H2Kb) mice were purchased from Jackson Laboratories (Bar Harbor, Maine, USA) and Beige severe combined immunodeficient (SCID) mice were purchased from Taconic (Hudson, NY, USA). Mice were certified to be pathogen-free and housed in our animal facility with access to food and water ad libitum. Green Fluorescent Protein (GFP) transgenic breeding pairs ‘C57BL/6-TgN(ACTbEGFP)1Osb’ were purchased from Jackson Laboratories. They were bred in our animal facility by mating the heterozygous GFP(+) animals to C57BL/6. Heterozygous GFP animals were separated from littermates by using an ultra violet light source that caused the heterozygous GFP animals to glow green.

Bone marrow cultures

Dexter cultures were created according to the method of Dexter et al, 1977) with modifications as described previously (Frimberger et al, 2001). Fresh whole bone marrow was harvested by flushing bones from C57BL/6 mice in Fischer medium (USBiological, Swampscott, MA, USA) with 100 U/ml penicillin, 100 μg/ml streptomycin (Invitrogen, Carlsbad, CA, USA), 0·0125 μg/ml fungizone (Invitrogen), 10−7 M hydrocortisone sodium succinate (Upjohn, Kalamazoo, MI, USA), and 20% horse serum (lot AFG5429; Hyclone, Logan, UT, USA). Pooled harvested cells were counted in crystal violet on a Neubauer haemacytometer and suspended at a concentration of 4 × 106 cells/ml, then seeded into either a 25 cm2 (10 ml cell suspension per flask) or 75 cm2 (30 ml cell suspension per flask) vent-cap tissue-culture flask (BD Discovery Labware, Bedford, MA, USA). Cultures were incubated at 33°C in 5% CO2 in air, and fed weekly by removal of one-half the supernatant medium and cells and replacing it with fresh medium (demi-population). At every feeding, the non-adherent cells in the supernatant medium were counted (data not shown) in trypan blue to determine cell viability and so indicate the health of the cultures. The mesenchymal cultures were created by flushing whole bone marrow from freshly harvested C57BL/6 bones into RPMI-1640 medium (Invitrogen) with 100 U/ml penicillin, 100 μg/ml streptomycin, 0·0125 μg/ml fungizone and 20% fetal bovine serum (FBS, Hyclone). The cells were counted and seeded into flasks as described for Dexter culture. The mesenchymal cultures were incubated at 37°C in 5% CO2 in air and maintained in the same manner and schedule as the Dexter cultures.

Bone marrow transplantation

Six- to eight-week-old mice were used as donors or recipients. After killing the donor mice and dissecting out the femur, tibia and pelvic bones, bone marrow was obtained by flushing these bones in phosphate-buffered saline (PBS) with 5% heat-inactivated fetal calf serum (HI-FBS) using a syringe and a 22-gauge needle. After re-suspension in 5% HI-FBS/PBS buffer, cell numbers were counted in crystal violet and viability was assessed by trypan blue staining. Whole bone marrow cells, or selected populations of marrow cells based on their surface markers, were injected intravenously by tail vein into each recipient approximately 4 h after radiation of the recipient mice. The dose of infused cells was different in individual experiments and is detailed in the results section. A photon-producing linear accelerator (Elekta Inc., Norcross, GA, USA) was used for irradiation of animals before each transplant, at 100 cGy/min. Animals received 500 cGy of whole body and 500 cGy of local irradiation to the lower extremity. In the immunocompromised transplant setting, different populations of bone marrow cells were transplanted into non-irradiated recipients via direct injection of the cells into cardiotoxin-injured tibialis anterior muscles 1 d after injury.

Muscle injury

We evaluated the dose of cardiotoxin (Naja mossambica mossambica; Sigma, St Louis, MO, USA) appropriate to induce a visible lysis of muscle fibres in 48 h. Haematoxylin and eosin staining of frozen sections of injected muscle showed that a final concentration of 10 μmol/l cardiotoxin induced 80–90% lysis of muscle fibres in the area of injection. The cardiotoxin, a 500 μmol/l stock solution diluted to 100 μl with PBS, was injected into the anterior tibialis muscle using a 27-gauge needle and a 1-ml syringe. The recipient animals were anesthetized by inhaling halothane prior to injection. The needle was inserted deep into the anterior tibial muscle longitudinally towards the knee from the ankle. Cardiotoxin was injected along the length of the muscle.

Immunofluorescent staining

To evaluate the skeletal muscle cell phenotypes in the regenerating muscle, specimens were collected after the anaesthetized mice were killed by cervical dislocation. Excised muscle specimens were placed in freshly prepared PLP fixative solution (balanced phosphate solution with 2% paraformaldehyde, sodium m-periodate and l-lysine) for 2 h at 4°C, with frequent agitation. Samples were then washed in a 7% sucrose buffer overnight followed by a 15% sucrose buffer wash for 2–3 h, ending with 25% sucrose plus 10% glycerol buffer wash for another 2 h, all at 4°C. They were then rinsed in PBS and embedded in tissue-freezing medium (optimal cutting compound), frozen and stored at −70°C until sectioning. Immunofluorescent staining was performed on 10–16 μm thick cryosections of the muscle. In the case of intracellular antigens, permeabilization was performed with 0·2% Triton X-100/PBS for 20 min. Sections were then blocked for 30 min with 20% normal serum buffer. In the case of the antidystrophin antibody, a MOM kit (Vector Laboratories, Burlingame, CA, USA) was used for blocking. Sections were rinsed with PBS then incubated with antidesmin (Abcam, Cambridge, MA, USA) and antidystrophin (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), biotinylated antimouse CD45 (BD PharMingen, San Diego, CA, USA), or Alexa Fluor 488-conjugated anti-GFP antibodies (Invitrogen) for 2 h at room temperature, followed by 1-h incubations with respective secondaries [Alexa Fluor antirabbit for desmin (Invitrogen), rhodamine antimouse for dystrophin (Abcam) and Alexa Flour streptavidin for CD45 (Invitrogen)]. The results of staining were visualized by fluorescent microscopy (Axioplan 2; Carl Zeiss, Oberko-chen, Germany). To rule out the possibility of autofluorescence being mistaken for GFP positivity, the presence of a true green signal was verified with a dual filter for fluorescein isothiocyanate and Rhodamine. Additional staining with Alexa Fluor 594 conjugated anti-GFP antibody (Invitrogen) was performed in parallel sections to confirm that the presumed GFP signals co localized with anti-GFP antibody staining (Abedi et al, 2004).

Flow cytometry

Blood chimerism was determined by obtaining peripheral blood from each mouse by tail vein venepuncture. Blood samples (50 μl) were incubated for 10 min at room temperature with 1·5 ml ice-cold erythrocyte-lysing solution (150 mmol/l NH4Cl, 10 mmol/l NaHCO3, 1 mmol/l EDTA, pH 7·4), then washed with PBS. The cells were re-suspended in PBS, fixed with 1% paraformaldehyde (Sigma, St Louis, MO, USA) and kept at 4°C until analysis. Peripheral donor chimerism was assessed by FACS. The percentage of GFP (+) cells was calculated by dividing the total number of GFP (+) cells by the total number of cells in the tissue section, after subtracting the background.

Cell separation

Bone marrow was isolated from iliac bones, femurs, and tibiae of 6- to 8-week-old GFP transgenic mice. Bone marrow cells were incubated with anti CD45-allophycocyanin (APC), anti c-Kit APC and anti-Sca-1-biotin (followed by streptavidin APC), all from BD PharMingen (San Diego, CA, USA), for 30 min. Cells that were positive and negative for individual markers were then sorted into different tubes with a high speed MoFlo cell sorter(Dako Cytomation, Fort Collins, CO, USA). For lineage negative separation, a low-density fraction (<1·077 g/cm2) was isolated on Nycoprep 1·077A (Accurate Chemical and Scientific Corporation, Westbury, NY, USA). These cells were lineage-depleted using magnetic beads from Lineage Depletion Kit (Miltenyi Biotec Inc., Auburn, CA, USA). The cells were washed and counted after depletion.

Counting and statistics

To count GFP (+) muscle fibres, six sections, each 16 μm thick and 200–300 μm apart, were prepared for each muscle specimen and the number of GFP (+) muscle fibres in each section was counted. GFP (+) muscle fibres were previously defined according to their characteristic morphology and double staining for antidesmin and antidystrophin antibodies (Abedi et al, 2004). These fibres were negative for CD45. We also counted three random high-power fields (HPF) in each section, counted the total number of HPFs for each section, and then calculated the total estimated muscle fibres (i.e. average of cell number in HPF multiplied by the number of HPFs per section). The ratio between GFP (+) muscle fibres and the estimated total number of muscle fibres was determined. We used the non-parametric Wilcoxon rank-sum test for comparison, and the trend test developed by Cuzick for testing trend (Cuzick, 1985). The level of significance was set at 0·05 (two-sided). Data are presented as the mean ± one standard error of the mean (SEM; Cuzick, 1985).

Results

  1. Top of page
  2. Summary
  3. Experimental designs and methods
  4. Results
  5. Discussion
  6. References

To identify the cells that were incorporated into regenerating muscle fibres, we compared different populations of marrow cells to each other:

Haematopoietic versus non-haematopoietic

Most investigators agree that CD45 (+) cells are haematopoietic in origin and that non-haematopoietic cells, with the exception of the cells from erythroid lineage, do not express CD45. Therefore, to determine whether haematopoietic cells contribute to the formation of muscle fibres, we separated GFP (+) whole bone marrow cells based on their CD45 antigen positivity. Equal numbers of either GFP(+)CD45(+) or GFP(+)CD45(−) populations were given intravenously to cohorts of C57BL/6 mice after 500 cGy of whole body irradiation and 500 cGy of lower extremity irradiation. Seven days later, the animals were subjected to cardiotoxin injury in their right tibialis anterior muscle and the same muscle was analysed a month later. The vast majority of GFP (+) fibres were seen in animals that had been transplanted with CD45 (+) marrow cells (Fig 1).

image

Figure 1.  Haematopoietic cells contribute to skeletal muscle fibre formation: (A–D) GFP(+), CD45 positive and negative cells were separated by cell sorting (A) and equal numbers of each population were transplanted to cohorts of C57BL/6 mice (5 × 106 cells/mouse). One month after cardiotoxin injury to the tibialis anterior muscle, the percentage of GFP(+) fibres in regenerated skeletal muscle was determined. Data represent analysis of a minimum of six sections of each tibialis anterior muscles of six CD45(+) or seven miceCD45(−) from two separate experiments. Panel C shows a digital micrograph of a frozen section of skeletal muscle from a CD45(+) transplanted animal. Red is dystrophin staining and green shows GFP(+) muscle fibres (arrows). Original magnification × 20. (D) Percentage of GFP+ muscles fibres in animals transplanted with an equal number of lineage(−) or lineage(+) GFP(+) cells (5 × 106 cells/mouse. n = 12 mice for lin(−) group and 10 mice for lin(+) group from 3 separate experiments). Bars represent mean ± SEM. *P-value < 0·05.

Download figure to PowerPoint

Lineage positive versus lineage negative cells

Based on the level of differentiation, bone marrow cells are classically defined as lineage positive (cells with different differentiation markers) and lineage negative (cells without those markers). A Miltenyi Cell Separation kit was used to separate lineage-positive and lineage-negative GFP (+) marrow cells and equal numbers of these cells were transplanted intravenously into cohorts of irradiated C57BL/6 mice followed by cardiotoxin injury to the tibialis anterior muscle. Four weeks later, analysis of muscle samples showed that the origin of the majority of GFP (+) fibres (more than double) were from lineage-negative marrow cells (Fig 1D).

Stem cells versus differentiated cells

Common markers for HSC (i.e. Sca-1, c-Kit, endoglin, CD34, Flk-1 and Flk-2) were used to sort the populations of marrow cells, with and without stem cell potential. In each group, equal numbers of positive and negative cells were injected intravenously to cohorts of irradiated mice followed by cardiotoxin injury to the tibialis anterior muscle 7 d later. GFP (+) muscle fibres were found only in Sca(+)and c-Kit(+) populations, compared with Sca(−) and c-Kit(−) groups (Fig 2A and B). In addition, significantly more GFP(+) muscle fibres were found in animals that were transplanted with endoglin(+) or Flk-2(+) marrow populations compared with the animals that received the cells without these surface markers (Fig 2C and D). Other markers, such as Flk-1 or CD34, were not able to predict marrow to muscle conversion (data not shown).

image

Figure 2.  Marrow cells with stem cell markers are responsible for GFP(+) muscle fibres. Equal numbers of different populations of GFP(+) marrow cells were transplanted intravenously to cohorts of C57BL/6 mice. GFP (+) muscle fibres were evaluated in regenerative muscles of transplanted animals. Left panels represents the FACS plots used for the sorting of each population. Middle panels show digital micrographs of frozen sections of tibialis anterior muscle stained for GFP (green) and dystrophin (red). Arrows indicate GFP(+) muscle fibres. Original magnification × 20. Right panels represent the percentage of GFP(+) muscle fibres based on the analysis of a minimum of 6 sections of tibialis anterior muscles of different cohorts of mice from separate experiments [Sca(+) n = 9, Sca(−) n = 15 from three separate experiments] [c-Kit(+) n = 13, c-Kit(−) n = 15 from three separate experiments] [Flk-2(±) n = 9 from two separate experiments] [endoglin(±) n = 5 from two separate experiments]. Bars represent mean ± SEM. *P value < 0·05.

Download figure to PowerPoint

Stem cells versus progenitor cells

Other investigators have suggested that marrow progenitors, specifically myelomonocytic cells, contribute to the formation of skeletal muscle fibres. To study the role of progenitor cells, we compared the Sca(+) c-Kit(+) population of marrow cells (containing HSCs and progenitors) with Sca(−)c-Kit(+) population (containing progenitor cells but no HSC (Doyonnas et al, 2004) Transplantation of equal numbers of these cells to cohorts of irradiated C57BL/6 mice, followed by cardiotoxin injury to the tibialis anterior muscle, showed that only cells with stem cell potential, i.e. Sca(+)c-Kit(+) cells, were incorporated into regenerating muscle (Fig 3).

image

Figure 3.  Marrow stem cells, and not their progenitors, are responsible for the appearance of GFP(+) muscle fibres. Equal numbers of GFP(+) marrow progenitors or marrow stem cells were transplanted intravenously to cohorts of C57BL/6 mice followed by cardiotoxin injury to their tibialis anterior muscles 1 week later. One month after injury, the regenerated muscles were analysed for the presence of GFP (+) muscle fibres. The dot plot (left) shows a sample of FACS plots used for sorting c-Kit(+)Sca(+) and c-Kit(+)Sca(−) cells. Charts on the right represent the analysis of a minimum of six sections of each tibialis anterior muscle of cohorts of mice. [c-Kit(+)Sca(+) and c-Kit(+)Sca(−) n = 9 from two separate experiments] [lin(−)c-Kit(+)Sca(+) n = 6, lin(-)c-Kit(+)Sca(+) n = 4 from one experiment]. Bars represent mean ± SEM. *P-value < 0·05.

Download figure to PowerPoint

True macrophages do not incorporate into regenerating fibres

Macrophages, perhaps more than any other haematopoietic cells, have the ability to fuse not only to each other but to other neighbouring cells (De Baetselier et al, 1984; Han et al, 2000; Yagi et al, 2006). On the contrary, some researchers have suggested fusion as the main mechanism for the putative marrow to muscle conversion (Camargo et al, 2004; Willenbring et al, 2004). To study the potential role of macrophages in the formation of marrow-derived muscle fibres, we used Mac-1 as a common marker for monocytes and macrophages. Interestingly, transplantation of either Mac-1 (−) or Mac-1(+) cells, administrated intravenously into irradiated mice, produced GFP (+) fibres in cardiotoxin-injured tibialis anterior muscles. However, analysis of the Mac-1(+) cell subpopulations, both by high proliferative potential (HPP) and engraftment assays, demonstrated the existence of a population of Mac1(+)Sca(+) and Mac1(+)c-Kit(+) cells with stem cell potentials, capable of long-term engraftment (T.J. Zomorodian, D.A. Greer, K.D. Wood, K. Johnson, S. Mclean and M. Abedi, in preparation). When equal numbers of Mac-1(+) c-Kit(+) or Mac-1(+) c-Kit(−) cells from GFP transgenic mice were transplanted intravenously into irradiated C57BL/6 mice, GFP(+) fibres in the regenerating muscle were only found in animals that were transplanted with Mac-1(+)c-Kit(+) cells. The Mac-1(+)c-Kit(−) populations did not produce any GFP (+) fibres (Fig 4). Experiments with Mac-1(+)Sca(+) and Mac-1(+)Sca(−) populations demonstrated similar results (Fig 4).

image

Figure 4.  A subpopulation of Mac-1+ cells with stem cell markers is responsible for the appearance of GFP(+) fibres in skeletal muscle. Left panels represents the FACS plots used for sorting of Mac-1(+) cells, with and without stem cell markers. Digital micrographs in the middle show frozen sections of regenerative tibialis anterior muscle stained for GFP (green) and dystrophin (red). Arrows show GFP(+) muscle fibres. Original magnification × 20. Bar graphs on the right represent the percentage of GFP(+) muscle fibres based on the analysis of a minimum of six sections of tibialis anterior muscles, from different cohorts of mice, from two to three separate experiments [Mac-1(+)c-Kit(+) n = 12; Mac-1(+)c-Kit(−) n = 10; Mac-1(+)Sca(+) n = 11; Mac-1(+)Sca(−) n = 12 each from two separate experiments]. Bars represent mean ± SEM. *P-value < 0·05.

Download figure to PowerPoint

Experiments in immunocompromised mice hosts

Previous studies and our own experience have shown that GFP is immunogenic (Rosenzweig et al, 2001; Steinbauer et al, 2003; Inoue et al, 2004). Therefore, whole body irradiation or other myeloablative and immunosuppressive approaches are needed to prevent rejection of the cells following transplantation. Comparison between the haematopoietic and muscle chimerism [GFP (+) peripheral blood and GFP(+) muscle fibres] in animals transplanted with different populations of marrow cells showed that GFP(+) muscle fibres were seen only in transplants with populations of marrow cells that resulted concurrenly in haematopoietic chimerism (Fig 5). This may suggest that haematopoietic engraftment is necessary for the appearance of GFP(+) muscle fibres. Alternatively, it may simply reflect that bone marrow stem cells are capable of both muscle and muscle chimerism in the recipient, and haematopoietic chimerism is not a necessary step for the presence of these fibres. To further explore these issues, we used an immunocompromised model of transplantation, where no irradiation was needed for the transplant and the recipients could not reject the donor cells. Tibialis anterior muscles of Beige SCID mice were injured with intramuscular injection of cardiotoxin 1 d before cell transplantation. Harvested marrow cells from GFP (+) transgenic mice were separated, based on their surface markers, via immunofluorescent staining followed by cell sorting as described earlier. For each surface marker, either positive or negative cells were directly injected into cardiotoxin injected tibialis anterior muscle of different cohorts of mice. Muscle samples were harvested four weeks after injury and were analysed for the presence of GFP (+) muscle fibres (see Methods section for details). Peripheral blood chimerism was checked for the presence of GFP (+) haematopoietic cells. The experiments confirmed our previous data showing that the lineage (−), c-Kit(+), Sca(+) or endoglin(+) cells resulted in significantly higher numbers of GFP (+) muscle fibres than the lineage (+), c-Kit (−), Sca-1 (−) or endoglin (−) populations (Fig 6). It further confirmed our data that stem cells, not progenitor cells or macrophages, incorporate into regenerating muscle fibres (Fig 6).

image

Figure 5.  Peripheral blood chimerism in animals transplanted with different populations of marrow cells. Different populations of marrow cells from GFP transgenic mice were transplanted to irradiated C57BL/6 mice after whole body irradiation. Four weeks after transplant, GFP chimerism in peripheral blood was evaluated. The number of cells transplanted was 5 × 106 for whole bone marrow and lineage (±) cells and 5 × 105 cells for c-Kit(±) and Sca-1(±) cells. Bars represent mean ± SEM

Download figure to PowerPoint

image

Figure 6.  Direct transplantation of bone marrow cell populations to the tibialis anterior of immunocompromised mice. Different populations of GFP(+) marrow cells were transplanted directly into cardiotoxin injured tibialis anterior muscles of Beige SCID recipients. Four weeks after transplant, GFP(+) muscle fibres in the regenerating skeletal muscles were evaluated. The number of cells injected between the groups were different. However, for each marker an equal number of cells with or without that specific marker were injected. Data represent the percentage of GFP(+) muscle fibres based on the analysis of a minimum of 6 sections of tibialis anterior muscles of different cohorts of mice from one or two separate experiments. [total number of animals per population: lin(−) n = 4, lin(+) n = 5; Mac-1(+) n = 4, Mac-1(−) n = 3; Sca(+) n = 8, Sca(−) n = 8; c-Kit(+) n = 3, c-Kit(−) n = 3; c-Kit(+)Sca(+) n = 6, c-Kit(+)Sca(−) n = 9; Endoglin(+) n = 3, Endoglin(−) n = 3; Dexter's Culture n = 9, Mesenchymal Culture n = 6). Bars represent mean ± SEM.

Download figure to PowerPoint

To study the role of marrow-derived mesenchymal cells in the appearance of GFP(+) muscle fibres, we established both Dexter's cultures and more conventional mesenchymal cultures. The mesenchymal culture consisted mostly of macrophages and stromal cells and there was very minimal, if any, haematopoietic activity. Dexter's cultures, in addition to stromal cells and macrophages, have a noticeable level of haematopoietic activity. Cultures were grown for 3 weeks before the collection of adherent cells. Tibialis anterior muscles of Beige SCID mice were injured 1 d prior to transplantation of cells. Following direct injection of 106 cells from either Dexter's cultures or mesenchymal cultures into each tibialis anterior muscle of cohorts of mice, most of the GFP(+) muscle fibres were seen in muscles that received cells from Dexter cultures (Fig 6). Finally, analysis of haematopoietic chimerism, 4 weeks after the direct injection of cells from different marrow populations of marrow cells into cardiotoxin-injured muscles of Beige/SCID mice, did not show any peripheral blood chimerism in any of the recipients. This demonstrated that haematopoietic chimerism is not necessary for marrow to muscle conversion.

Discussion

  1. Top of page
  2. Summary
  3. Experimental designs and methods
  4. Results
  5. Discussion
  6. References

This study found that a population of bone marrow stem cells directly participates in the regeneration of muscle fibres. We could not show any such participation from haematopoietic progenitor cells, lineage-differentiated cells or mesenchymal cells. The present study did not address the mechanism of marrow to muscle conversion but focussed on the type of marrow cells responsible for this process, regardless of the mechanism (fusion or direct differentiation of marrow cells to satellite cells), because participation of marrow cells in muscle regeneration would still be scientifically and clinically significant. Furthermore, we have previously shown that, by appropriate conditioning of the host, increasing numbers of marrow cells participated in formation of muscle fibres (Abedi et al, 2005). However, for practical reasons, we used a more simplified protocol (with lower yields for GFP+ muscle fibres) to characterize the type of marrow cells.

In recent years, multiple reports have shown that marrow cells are involved in the regeneration of skeletal muscle fibres. In these experiments, the investigators used one of several cell tracking methods (i.e. transgenic mice expressing fluorescent proteins, male donors to female recipients, or beta galactosidase transgenic mice) to follow the fate of the donor marrow cells in the recipient mice. The type of marrow cells responsible for this phenomenon, remains unknown. While some researchers showed a mesenchymal origin for marrow-derived muscle fibres (Anjos-Afonso et al, 2004; Dezawa et al, 2005), others have suggested that haematopoietic cells are responsible for regeneration of muscle fibres (Camargo et al, 2003; Doyonnas et al, 2004). Characterization of mesenchymal cells has proved difficult and, at least in the mouse, there are no specific mesenchymal markers, although a combination of both positive and negative markers have been used for their characterization (Krebsbach et al, 1999; Alhadlaq & Mao, 2004). Still, most investigators agree that mesenchymal cells are CD45 (−), c-Kit (−) and Flk-2 (−). Our data showed that CD45(+), c-Kit(+) and Flk-2(+) cells were responsible for the appearance of GFP(+) muscle fibres in recipient mice. Furthermore, transplantation of cells from mesenchymal culture versus cells from the long-term haematopoietic Dexter's culture, shows that the majority of GFP(+) muscle fibres came from the latter. Although it is always difficult to exclude contamination of the transplanted cells with a rare mesenchymal stem cell, our data suggests that, at least in the context of our experimental setting, haematopoietic cells have more potential than mesenchymal cells to be involved in the regeneration of muscle fibres.

It has recently been shown that a single haematopoietic stem cell transplanted to a sublethally-irradiated recipient can give raise to a progeny that reconstitutes the blood and integrates into regenerating myofibres (Camargo et al, 2003; Corbel et al, 2003). However, their data do not show that HSC directly participate in muscle regeneration. In fact, Doyonnas et al (2004) proposed that it is not the marrow stem cells, but a subpopulation of progenitor cells (myelomonocytic cells) that are responsible for this phenomenon. Our results showed that the majority of GFP(+) muscle fibres came from undifferentiated HSC. Progenitor cells, as well as terminally differentiated cells, did not result in GFP(+) muscle fibres. Our experiments in the immunocompromised mouse model (i.e. Beige SCID) also clearly showed that the same populations of marrow stem cells, when directly injected into injured muscle, can result in GFP(+) muscle fibres. This occurred in the absence of any further haematopoietic differentiation of stem cells and haematopoietic chimerism.

Other investigators have suggested that macrophages fusing to other cells resulted in the formation of marrow-derived tissues. For example, fusion of macrophages to the recipient hepatocytes brought about the regeneration of liver from marrow cells (Rizvi et al, 2006 and Willenbring et al, 2004). This argument was based on the use of Mac-1 for the definition of macrophages. Our results showed that surface markers, like Mac-1, could be misleading since a subpopulation of the Mac-1(+) cells, harbouring stem cells markers such as Sca-1 and c-Kit, were able to engraft and repopulate the haematopoietic system in long term experiments and produce high proliferative colonies in HPP assays. In our experiments, only Mac-1 (+) cells with stem cell potential, and not the true macrophages, were responsible for the appearance of GFP (+) muscle fibres.

The discrepancies between our data and those of other investigators results primarily from differences in experimental design. For example, the experiments by Doyonnas et al (2004) were performed by direct injection of different populations of marrow cells into muscle of immunocompetent mice without any systemic irradiation or immunosuppression, while our experiments were performed either after systemic irradiation or in immunosuppressed hosts. Immunoreactivity of GFP (Rosenzweig et al, 2001; Steinbauer et al, 2003; Inoue et al, 2004) may result in different outcomes between the two sets of experiments. In our experience, the infusion of up to 100 × 106 bone marrow cells from GFP transgenic mice (C57BL/6 background) into non-irradiated C57BL/6 mice produced minimal or no chimerism, suggesting the strong immunoreactivity of GFP (data is not shown). Other investigators have shown that mesenchymal cells could integrate into the muscle fibres (Anjos-Afonso et al, 2004; Dezawa et al, 2005) while our experiments showed very little evidence for that. Our previous data showed that marrow stem cells had a tendency to avidly bind to marrow stromal cells (where multiple washes failed to separate them from the stromal cells; Frimberger et al, 2001) and, at least in short term cultures, switching the mesenchymal culture medium (RPMI with 20% fetal calf serum) to a Dexter's culture system will frequently result in the recovery of haematopoiesis in the cultures, suggesting the presence of dormant HSC. One can speculate that the presence of these rare HSCs in the mesenchymal cell preparations might account for some of the reports that showed mesenchymal cells as the origin of marrow-derived muscle fibres. Alternatively, our experiments do not rule out the possibility that, with different experimental designs, mesenchymal cells may also show differentiation to muscle fibres as they do in in-vitro models.

The process that leads to the participation of marrow cells in regeneration of muscle fibres remains unclear. Whether these cells become satellite cells, assume a different stem cell phenotype such as pericytes or mesangioblasts, directly fuse with regenerating myofibres, or employ a combination of some or all of these pathways remains to be determined. Further lineage studies using tools such as genetic markers will be required to define the intermediate cell types that may be involved in appearance of marrow-derived muscle fibres.

Acknowledgements

We would like to acknowledge Dr Nicola Koutab and his staff for the help with FACS analysis. We also thank Dr Elaine Bearer for help with imaging and microscopy. This publication was made possible by Grant Number 1P20 RR018757-01 from the National Center for Research Resources (NCRR), and by Grant Number 1KO8 HL072332-01 from NHLBI, both are components of the National Institutes of Health (NIH). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of NCRR or NIH.

Disclaimers

None.

References

  1. Top of page
  2. Summary
  3. Experimental designs and methods
  4. Results
  5. Discussion
  6. References
  • Abedi, M., Greer, D.A., Colvin, G.A., Demers, D.A., Dooner, M.S., Harpel, J.A., Weier, H.U., Lambert, J.F. & Quesenberry, P.J. (2004) Robust conversion of marrow cells to skeletal muscle with formation of marrow-derived muscle cell colonies: a multifactorial process. Experimental Hematology, 32, 426434.
  • Abedi, M., Greer, D.A., Foster, B.M., Colvin, G.A., Harpel, J.A., Demers, D.A., Pimentel, J., Dooner, M.S. & Quesenberry, P.J. (2005) Critical variables in the conversion of marrow cells to skeletal muscle. Blood, 106, 14881494.
  • Alhadlaq, A. & Mao, J.J. (2004) Mesenchymal stem cells: isolation and therapeutics. Stem Cells and Development, 13, 436448.
  • Anjos-Afonso, F., Siapati, E.K. & Bonnet, D. (2004) In vivo contribution of murine mesenchymal stem cells into multiple cell-types under minimal damage conditions. Journal of Cell Science, 117, 56555664.
  • Brazelton, T.R., Nystrom, M. & Blau, H.M. (2003) Significant differences among skeletal muscles in the incorporation of bone marrow-derived cells. Developmental Biology, 262, 6474.
  • Camargo, F.D., Green, R., Capetanaki, Y., Jackson, K.A. & Goodell, M.A. (2003) Single hematopoietic stem cells generate skeletal muscle through myeloid intermediates. Nature Medicine, 9, 15201527.
  • Camargo, F.D., Chambers, S.M. & Goodell, M.A. (2004) Stem cell plasticity: from transdifferentiation to macrophage fusion. Cell Proliferation, 37, 5565.
  • Corbel, S.Y., Lee, A., Yi, L., Duenas, J., Brazelton, T.R., Blau, H.M. & Rossi, F.M. (2003) Contribution of hematopoietic stem cells to skeletal muscle. Nature Medicine, 9, 15281532.
  • Cuzick, J.A. (1985) Wilcoxon-type test for trend. Statistics in Medicine, 4, 8790.
  • De Baetselier, P., Roos, E., Brys, L., Remels, L., Gobert, M., Dekegel, D., Segal, S. & Feldman, M. (1984) Nonmetastatic tumor cells acquire metastatic properties following somatic hybridization with normal cells. Cancer and Metastasis Reviews, 3, 524.
  • Dexter, T.M., Allen, T.D. & Lajtha, L.G. (1977) Conditions controlling the proliferation of haemopoietic stem cells in vitro. Journal of Cellular Physiology, 91, 335344.
  • Dezawa, M., Ishikawa, H., Itokazu, Y., Yoshihara, T., Hoshino, M., Takeda, S., Ide, C. & Nabeshima, Y. (2005) Bone marrow stromal cells generate muscle cells and repair muscle degeneration. Science, 309, 314317.
  • Doyonnas, R., LaBarge, M.A., Sacco, A., Charlton, C. & Blau, H.M. (2004) Hematopoietic contribution to skeletal muscle regeneration by myelomonocytic precursors. Proceedings of the National Academy of Sciences of the United States of America, 101, 1350713512.
  • Frimberger, A.E., Stering, A.I. & Quesenberry, P.J. (2001) An in vitro model of hematopoietic stem cell homing demonstrates rapid homing and maintenance of engraftable stem cells. Blood, 98, 10121018.
  • Han, X., Sterling, H., Chen, Y., Saginario, C., Brown, E.J., Frazier, W.A., Lindberg, F.P. & Vignery, A. (2000) CD47, a ligand for the macrophage fusion receptor, participates in macrophage multinucleation. Journal of Biological Chemistry, 275, 3798437992.
  • Inoue, S., Tahara, K., Kaneko, T., Ajiki, T., Takeda, S., Sato, Y., Hakamata, Y., Murakami, T., Takahashi, M., Kaneko, M. & Kobayashi, E. (2004) Long-lasting donor passenger leukocytes after hepatic and intestinal transplantation in rats. Transplant Immunology, 12, 123131.
  • Krebsbach, P.H., Kuznetsov, S.A., Bianco, P. & Robey, P.G. (1999) Bone marrow stromal cells: characterization and clinical application. Critical Reviews in Oral Biology and Medicine, 10, 165181.
  • LaBarge, M.A. & Blau, H.M. (2002) Biological progression from adult bone marrow to mononucleate muscle stem cell to multinucleate muscle fiber in response to injury. Cell, 111, 589601.
  • Palermo, A.T., Labarge, M.A., Doyonnas, R., Pomerantz, J. & Blau, H.M. (2005) Bone marrow contribution to skeletal muscle: a physiological response to stress. Developmental Biology, 279, 336344.
  • Rizvi, A.Z., Swain, J.R., Davies, P.S., Bailey, A.S., Decker, A.D., Willenbring, H., Grompe, M., Fleming, W.H. & Wong, M.H. (2006) Bone marrow-derived cells fuse with normal and transformed intestinal stem cells. Proceedings of the National Academy of Sciences of the United States of America, 103, 63216325.
  • Rosenzweig, M., Connole, M., Glickman, R., Yue, S.P., Noren, B., DeMaria, M. & Johnson, R.P. (2001) Induction of cytotoxic T lymphocyte and antibody responses to enhanced green fluorescent protein following transplantation of transduced CD34(+) hematopoietic cells. Blood, 97, 19511959.
  • Steinbauer, M., Guba, M., Cernaianu, G., Kohl, G., Cetto, M., Kunz-Schughart, L.A., Geissler, E.K., Falk, W. & Jauch, K.W. (2003) GFP-transfected tumor cells are useful in examining early metastasis in vivo, but immune reaction precludes long-term tumor development studies in immunocompetent mice. Clinical and Experimental Metastasis, 20, 135141.
  • Willenbring, H., Bailey, A.S., Foster, M., Akkari, Y., Dorrell, C., Olson, S., Finegold, M. & Fleming, W.H. (2004) Grompe Myelomonocytic cells are sufficient for therapeutic cell fusion in liver. Nature Medicine, 10, 744748. Epub 2004 Jun 13.
  • Yagi, M., Miyamoto, T., Toyama, Y. & Suda, T. (2006) Role of DC-STAMP in cellular fusion of osteoclasts and macrophage giant cells. Journal of Bone and Mineral Metabolism, 24, 355358.