SEARCH

SEARCH BY CITATION

Keywords:

  • Muscle regeneration;
  • Myf5;
  • Satellite cells;
  • Stem cells;
  • Muscle differentiation

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

The myogenic factor Myf5 defines the onset of myogenesis in mammals during development. Mice lacking both Myf5 and MyoD fail to form myoblasts and are characterized by a complete absence of skeletal muscle at birth. To investigate the function of Myf5 in adult skeletal muscle, we generated Myf5 and mdx compound mutants, which are characterized by constant regeneration. Double mutant mice show an increase of dystrophic changes in the musculature, although these mice were viable and the degree of myopathy was modest. Myf5 mutant muscles show a small decrease in the number of muscle satellite cells, which was within the range of physiological variations. We also observed a significant delay in the regeneration of Myf5 deficient skeletal muscles after injury. Interestingly, Myf5 deficient skeletal muscles were able to even out this flaw during the course of regeneration, generating intact muscles 4 weeks after injury. Although we did not detect a striking reduction of MyoD positive activated myoblasts or of Myf5-LacZ positive cells in regenerating muscles, a clear decrease in the proliferation rate of satellite cell-derived myoblasts was apparent in satellite cell-derived cultures. The reduction of the proliferation rate of Myf5 mutant myoblasts was also reflected by a delayed transition from proliferation to differentiation, resulting in a reduced number of myotube nuclei after 6 and 7 days of culture. We reason that Myf5 supports efficient skeletal muscle regeneration by enabling transient myoblast amplification.

Disclosure of potential conflicts of interest is found at the end of this article.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

Skeletal muscle is a highly regenerative tissue that is capable of regeneration even after extensive damage. Since skeletal muscle myotubes are postmitotic, repair and regeneration rely on the existence of satellite cells, which reside between the myofiber basal lamina and plasma membrane [1]. In healthy adult muscle, satellite cells are quiescent [2]. To accomplish their roles of myonuclear turnover, myofiber hypertrophy, or muscle repair, satellite cells have to be activated from the quiescent state. Directly after muscle damage, myofibers hyalinize, vacuolate, and lyse, leading to a marked inflammatory response, which is characterized by infiltrating lymphocytes and macrophages that phagocytose necrotic myofibers [3]. Subsequent to their activation, satellite cells start to proliferate, generating myogenic precursor cells and myoblasts, which serve as transient amplifying cells. Proliferation, which peaks between day 2 and day 3 after injury, is followed by a phase of differentiation where myoblasts withdraw from the cell cycle and form small centronucleated myotubes. The skeletal muscle tissue architecture is eventually restored within 2–4 weeks by fusion of myoblasts and further growth of the newly regenerated centronucleated myofibers [4]. Muscle regeneration is normally an extremely efficient process, which enables muscles to retain effective regenerative abilities even after repeated injury cycles (up to 50) [5].

It has long been claimed that muscle regeneration from satellite cells merely recapitulates the process of embryonic myogenesis, although no systematic comparisons of pathways have been performed. Instead, the claim was mainly based on the finding that proliferating and differentiating satellite cells re-express myogenic regulatory factors (namely MyoD, Myf5, Myogenin, and Myf6) [6, 7]. Given the fact that regenerating muscle lacks positional cues, which are present during embryonic myogenesis in somites, and that muscle regeneration takes place within the basal lamina of necrotic myofibers, it seems reasonable to assume that significant differences exist between development and regeneration, although some basic mechanisms might be shared. The hypothesis that satellite cells and embryonic myogenic cells are rather different cell populations is illustrated by distinct function of Pax3/Pax7 during embryonic myogenesis and adult muscle regeneration. Pax3 mutant mice lack limb muscles due to a defect of limb muscle precursor cells [8]. Pax3 and Myf5 compound mutant mice miss body wall and limb muscles, suggesting a regulatory network where Pax3 is genetically upstream of the MyoD family members (i.e., MyoD) [9, 10]. Pax3 appears not to be expressed during muscle regeneration in adult life [11, 12] (E. Vorobyov, S. Ustanina, T. Braun, unpublished observations) and hence seems dispensable for satellite cell viability. In contrast, the lack of Pax7 does not affect normal muscle development during embryogenesis [13] but severely impairs maintenance of satellite cells and muscle regeneration [14, 15], indicating that the fate of satellite cells and embryonic myogenic cells are controlled by different sets of genes.

The role of myogenic regulatory genes of the MyoD family, which is the core pathway of skeletal muscle formation, has been addressed extensively during embryonic myogenesis but only incompletely for adult skeletal muscle regeneration and homeostasis. Myf5 homozygous mice display a severe myotomal deficiency prior to E10.5, but this deficiency is overcome when MyoD expression is initiated so that newborn mice have a grossly normal muscle phenotype at birth [16, 17]. MyoD knockout mice have a virtually normal musculature during embryonic development but exhibit a muscle regeneration phenotype in adult life, apparently due to a reduced differentiation capacity of muscle satellite cells [18, 19]. MyoD/Myf5 double knockout mice fail to generate skeletal muscle precursor cells [20]. So far, the perinatal lethality of the original Myf5 and Myogenin knockout strains and alterations of the transcription of the Myf5 gene by insertions in the neighboring Myf6 gene prevented a thorough analysis in adult musculature [21, 22].

Detailed transcriptome analysis following cardiotoxin-induced muscle injury [12] revealed an induction of MyoD and Myf6 within 2–6 hours of injury, suggesting a role in activation of the satellite cell population [6, 19] [23], whereas the paired box transcription factor Pax7 is already expressed in quiescent satellite cells [14]. Activated satellite cells initially seem to coexpress Pax7 and MyoD. Then, most activated satellite cells proliferate, downregulate Pax7, and differentiate, while a small satellite cell subpopulation maintains Pax7 but loses MyoD and withdraws from proliferation and differentiation to convert to a reserve cell-like state [24]. Myf5 is delayed in expression compared with MyoD and peaks within 5 days of injury similarly to myogenin, although the Myf5 locus seems to be active in resting satellite cells based on the activity of LacZ reporter knockin mouse strains [24, 25]. Here, we investigate the function of Myf5 for maintenance and regeneration of adult skeletal muscle using a Myf5 mutant strain, which lacks the rib phenotype of the original mutant that caused perinatal lethality.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

Origin of Mouse Mutants, Induction of Muscle Regeneration, and Immunohistochemical Analysis

The generation of Myf5loxPΔ mutant mice has been published [26]. Mutant mice were maintained on a mixed 129Sv/C57/BL6 background and genotyped by Southern blot analysis. For control purposes, mice of the same genetic background were used. C57/BL6 mice were used as an additional wild-type control in some experiments. We did not detect a significant difference in the regenerative capacity of 129Sv/C57/BL6 and C57/BL6 mice in our assays. The BAC195APZ reporter strain has been described before [27].

Immunochemistry, Electron Microscopy, LacZ Staining, and Muscle Regeneration Assay

Tissues were prepared for paraffin sectioning and subsequent hematoxylin and eosin staining using established techniques [28]. Identification of satellite cells in musculi (Mm.) interossei by electron microscopy and LacZ staining has been outlined before [15]. Muscle regeneration was induced as described previously [15]. We used both male and female mice matched to each other in the regeneration assays without noticing a significant difference.

Quantitative Reverse Transcription-Polymerase Chain Reaction and Northern Blot Analysis

Total RNA was isolated using TRIzol reagent according to the instruction manual of the manufacturer (Invitrogen, Carlsbad, CA, http://www.invitrogen.com). Northern blot analysis of 20–50 μg of total RNA from pooled hind limb muscles was done by using established procedures [15]. Quantitative real-time polymerase chain reaction (PCR) was done as described previously [29]. Relative quantification of expression was carried out by using the comparative cycle threshold method [29].

Primary Cultures of Myogenic Cells and Preparation of Isolated Myotubes

Primary myoblasts were prepared using muscles of both hind limbs of a mouse according to Bischoff [30]. Cultures of isolated single myotubes were prepared according to Rosenblatt et al. [31]. Single myotube suspension was attached to glass cover slides using Matrigel basement membrane matrix (354234; BD Biosciences, San Diego, http://www.bdbiosciences.com) and fixed for further staining.

Statistical Analysis

A minimum of three and up to five replicates was done for each experiment shown if not indicated otherwise. Representative fields were chosen randomly. The person who performed the counting was unaware of the genotype of the sample. Data are presented as means and standard errors of the mean. Comparisons between groups were done using Kruskal-Wallis comparison and a Dunn's multiple comparison post hoc test; p < .05 was considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

Skeletal Muscles of Myf5-Deficient Mice Show Normal Tissue Architecture but a Slight Decrease in Muscle Mass and Myotube Diameter

To gain insight into possible functions of Myf5 in adult skeletal muscles, we generated the Myf5ΔloxP mouse strain, in which a 3-kilobase (kb) genomic fragment of the Myf5 locus was removed by Cre-mediated deletion. The deleted fragment includes exon I, which contains most of the Myf5 coding region with the basic helix-loop-helix (bHLH) DNA binding region, and 2 kb of upstream sequences [26]. Macroscopical inspection of adult Myf5-deficient mice disclosed a virtually normal musculature and no signs of muscle loss or degeneration (data not shown). Examination of hematoxylin and eosin stained sections of the musculus (M.) tibialis anterior, M. soleus, the diaphragm, the M. quadriceps, and the M. latissimus dorsi from Myf5-deficient mice did not reveal any major structural alterations, although some muscles showed a slight decrease in muscle mass and myotube diameter (Fig. 1A–1D). We did not find any evidence for an increased incidence of degeneration/regeneration events since the number of myofibers with centrally located nuclei that are characteristic of regenerating muscles was equal between mutant and wild-type animals (data not shown). Likewise, we did not find an increase in collagen VI and laminin deposition in the M. tibialis anterior of Myf5 mutant mice (supplemental online Fig. S1).

thumbnail image

Figure Figure 1.. Myf5-deficient mice display regular muscle architecture but slightly reduced muscle mass and myotube diameters. Analysis of H&E-stained M. tibialis anterior (A, B) and diaphragm (C, D) revealed a normal muscle morphology of Myf5-deficient (B, D) in comparison with wild-type (A, C) mice. Myf5 mutant muscle showed reduced muscle diameters in some muscles (compare [C] and [D]) (200-μm scale bars). (E–G): Normal number, length, and general structure of muscle spindles in Myf5-deficient muscles. (E): Transversal section through a muscle spindle (H&E staining, 200-μm scale bar). (F): Number of muscle spindles per single M. soleus. (G): Average length of muscle spindles in μm; p < .01. Abbreviation: M., musculus.

Download figure to PowerPoint

We also investigated the effect of the loss of Myf5 on muscle spindles, since Myf5 is not only expressed in satellite cells but also in the nuclei of both nuclear bag and nuclear chain fibers of muscle spindles [24]. Muscle spindles are stretch sensitive mechanoreceptors sensitive to muscle movement. They are composed of distinct small diameter muscle fiber types that lie in parallel with the myofibers and are innervated by both sensory and motor axons [32]. Muscle spindles of soleus muscles of Myf5-deficient and wild-type mice were quantified, and the muscle spindle number and length were compared. Neither the number of muscle spindles nor their length and general structure were significantly affected by the loss of Myf5 (Fig. 1E–1G).

Deletion of Myf5 Does Not Affect the Expression of Myf6 and Other Muscle-Specific Genes in Adult Myf5-Deficient Mice

We and others have described that manipulation of the Myf5 locus does affect the expression of the neighboring Myf6 gene in cis [21, 22]. To rule out position effects of the Myf5 deletion on Myf6 expression in adult muscle, the expression of Myf6 was analyzed by Northern blot analysis using RNA isolated from intact hind limb muscles of Myf5-deficient and wild-type mice. As shown in Figure 2A, no significant change in the expression of Myf6 was evident. Likewise, we did not observe major alterations of the expression of several muscle-specific genes such as embryonic myosin heavy chain (MyHC), postnatal MyHC, fast adult MyHC, Troponin fast, and myosin light chain 1 in Myf5 mutants, although the expression of MyHC was slightly decreased. To analyze the expression of MyoD and Myogenin, we used real-time quantitative reverse transcription-PCR, since the expression level of these transcription factors is below the detection level of regular Northern blots in intact muscles. The absence of Myf5 did not cause a compensatory increase (or decrease) of the expression of MyoD (data not shown), which is in contrast to the upregulation of Myf5 in MyoD mutant mice during embryogenesis and might indicate that both genes fulfill different functions during adult myogenesis. Similarly, we did not observe a change in the expression level of Myogenin, suggesting that ongoing differentiation processes proceed regularly (data not shown).

thumbnail image

Figure Figure 2.. Loss of Myf5 leads to increased dystrophic changes in the musculature of mdx mice despite a relatively normal expression of Myf6 and other muscle-specific genes in Myf5-deficient mice. (A): Northern blot analysis of muscle-specific mRNAs isolated from hind limb muscles of Myf5-deficient(−/−) and wild-type mice. (B): The number of centrally located nuclei in mdx/Myf5−/− muscles was reduced (four times) in comparison with mdx muscles. (C–F): H&E-stained paraffin sections of M. quadriceps (C, D) and diaphragm (E, F) muscles from mdx(C, E) and mdx/Myf5−/− (D, F) mice. Muscles of mdx/Myf5−/− mice display more severe dystrophic changes: more extensive necrotic areas (arrows), more mononuclear cells, and smaller myofiber calibers are seen. Note the presence of centrally located nuclei that are characteristic for regenerating myotubes. Scale bars, 200 μm (** = p < .01). Abbreviations: embMHC, embryonic myosin heavy chain; faMHC, fast adult myosin heavy chain; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; M., musculus; MHCβ, myosin heavy chain β; MLC1, myosin light chain 1; pnMHC, postnatal myosin heavy chain; Tn, troponin; wt, wild-type.

Download figure to PowerPoint

Loss of Myf5 Increases Dystrophic Changes in the Musculature of mdx Mice

To determine whether Myf5 might play a role in skeletal muscle regeneration, we interbred Myf5−/− mutants to mdx mice. The mdx strain, a mouse model of Duchenne muscular dystrophy, carries a loss-of-function mutation in dystrophin, which is a component of the membrane-associated dystrophin-glycoprotein complex. Muscle fibers of mdx mice undergo extensive necrosis by 2 weeks of age, which is repaired by newly formed fibers. Later during life, this cycle of degeneration and regeneration continues, albeit at a slower pace, generating an excellent model of constant muscle regeneration [33].

Histological examination of the diaphragm and limb muscles (M. tibialis anterior and M. soleus) from 5-month-old animals revealed no significant differences between mdx and mdx/Myf5(−/−) mice (data not shown). This situation changed later during life: several muscles of 12-month-old mdx/Myf5(−/−) mice showed enhanced dystrophic changes (Fig. 2D, 2F) in comparison with mdx mutants (Fig. 2C, 2E). Diaphragm muscles of double mutants contained considerably more mononuclear cells and more small-sized myofibers compared with mdx mice that contained myotubes of different sizes caused by asynchronous regeneration and hypertrophy (Fig. 3F). We also detected a massive increase in the number of necrotic calcified fibers in double mutant mice as indicated by alizarin red staining and an increased deposition of extracellular matrix as revealed by trichrome staining. We did not find calcified fibers in M. quadriceps and M. tibialis (supplemental online Fig. S2). Furthermore, areas with dystrophic changes were observed in several muscles of double mutant mice along with large numbers of infiltrating mononuclear cells (Fig. 2D, 2F, arrows). In addition, the number of centrally located nuclei in mdx/Myf5(−/−) muscles was dramatically reduced in comparison with mdx muscles, where most of the nuclei were located in the center of the myotubes (Fig. 2B–2D), indicating an impairment of the continuous and efficient regeneration of myofibers that occurs in mdx mice. The increased presence of necrotic myotubes, the large number of mononuclear cells, and the severely reduced number of centrally located nuclei all indicate a significantly reduced regeneration capacity of skeletal muscles in mdx/Myf5−/− mice.

thumbnail image

Figure Figure 3.. Lack of Myf5 results in a delay of skeletal muscle regeneration. H&E staining of cardiotoxin injured M. tibialis anterior of wild-type (A, C, E, G) and Myf5-deficient mice (B, D, F, H). Extensive degeneration and cellular infiltration is seen at day 3 after injury ([A, B]: 200-μm scale bars). Newly formed myotubes show centrally located nuclei (white arrowheads in [E, F]). Degenerating damaged myofibers are pale after H&E staining (white arrows in [A, B, D]). Undamaged myotubes stain more intensively. Some peripherally located nuclei can also be found in injured muscles (black arrows in [A–D]). Seven days after injury ([C, D]: 200-μm scale bars; [E, F]: 100-μm scale bars), muscles of Myf5−/− mice display more fibrotic tissue, more mononuclear cells, more degenerating damaged myofibers, smaller myotube diameters, and a lower number of newly formed myofibers, which were more heterogenous in diameter compared with wild-type myofibers. At day 15 after injury ([G, H]: 200-μm scale bars), newly formed myofibers of Myf5 knockout mice were smaller and more heterogeneous in diameter.

Download figure to PowerPoint

The Lack of Myf5 Delays Skeletal Muscle Regeneration

The enhanced dystrophic changes in the musculature of mdx/Myf5−/− double mutant mice indicated that Myf5 is required for efficient skeletal muscle regeneration. We therefore directly investigated the timing of skeletal muscle regeneration of Myf5-deficient mice. Although numerous injury models have been proposed to examine skeletal muscle regenerative mechanisms including crush and freeze injuries, the damage protocol, which is based on chemical agents including cardiotoxin, notexin, or barium chloride, is probably the most reproducible and reliable [4, 34]. These chemicals initiate a well known chain of events that causes widespread degeneration followed by an extensive regeneration of muscle fibers. One day after injection of cardiotoxin into the M. tibialis anterior of Myf5-mutant and wild-type mice, degeneration of muscle fibers was just starting, a process that was virtually identical in both strains (supplemental online Fig. S3). Three days after injury, extensive infiltration of cells including myoblasts, leukocytes, and macrophages was found both in wild-type and knockout mice (Fig. 3A, 3B). However, some morphological differences between knockout and wild-type mice, such as an enhanced presence of degenerated myotubes, became apparent at this stage. By day 7, damaged muscles of wild-type mice showed no or only few remaining degenerated muscle fibers, whereas muscles of Myf5−/− mice still contained numerous dystrophic, damaged myofibers and hyaline material, indicating a decreased time course of muscle regeneration (Fig. 3C, 3D). Moreover, we found an enhanced presence of fibrous tissue and more mononuclear cells in regenerating muscles of Myf5-deficient mice (see also supplemental online Fig. S4 for a trichrome staining). We did not find calcified fibers in regenerating muscles of Myf5 mutant mice as indicated by alizarin red staining (data not shown). Newly regenerated myofibers were present at lower numbers, in general had smaller diameters, and were characterized by increased diameter heterogeneity. Quantitative analysis revealed that Myf5 mutant mice had significantly larger numbers of smaller myofibers and lower numbers of larger myofibers than wild-type mice both 7 and 15 days after cardiotoxin injection (supplemental online Fig. S5). However, it needs to be stressed that, both in wild-type and knockout mice, numerous regenerating myofibers were found (Fig. 3E, 3F). At day 15 after injury, differences between both strains were still visible, but the regeneration gap had narrowed. In particular, we again observed smaller myofibers in mutant mice, which were more heterogeneous in diameter (Fig. 3G, 3H; supplemental online Fig. S5). Two weeks later and 4 weeks after injury, these differences were not present anymore: both wild-type and knockout mice showed a complete restoration of skeletal muscle architecture (supplemental online Fig. S3). We can also rule out that Myf5 mice suffer from a slower inflammatory response to damage or have a reduced capacity to attract inflammatory cells, since we did not find any significant difference between wild-type and Myf5 mutant mice regarding the infiltration of damaged muscle by macrophages and hematopoietic cells using antibodies against F4/80 (a macrophage marker) and CD45 (a marker of hematopoietic cells) (supplemental online Figs. S6, S7). Taken together, our data indicate a slower pace of skeletal muscle regeneration of Myf5-deficient mice, which could be compensated over time.

Enumeration of Quiescent Satellite Cells in Myf5-Deficient Mice

Muscle satellite cells are the key players in skeletal muscle growth and regeneration. We reasoned that the compromised regeneration of Myf5 deficient muscle might be either due to a reduced number or to an impaired proliferation/differentiation defect of satellite cells. To address these possibilities, we first determined the number of satellite cells in regenerating and resting muscles of Myf5−/− mice using several methods. Serial ultrathin sections of Mm. interossei were analyzed using transmission electron microscopy, which still remains the most reliable technique to identify quiescent muscle satellite cells. Random fields were chosen for the examination. We counted 500–1,000 nuclei per muscle, and the number of satellite cells per 100 myotube nuclei was calculated. We decided to analyze 1-year-old, wild-type, and Myf5−/− mice to address mice that had already been exposed for a longer period to an absence of Myf5 and hence might show more severe pathological changes. However, we did not find differences in the ultrastructure of myofibers of wild-type and Myf5−/− (Fig. 4A, 4B) mice, nor did we detect a major reduction of the number of satellite cells. Myf5 mutant satellite cells carried nuclei with the characteristic heterochromatic appearance and very small cytoplasm. In Myf5-deficient mice, 2.9% satellite cells were counted compared with 3.5% in wild-type sections (Fig. 4C), which is within the normal range, although the differences between both strains were statistically significant (p < .05). We also stained isolated myofibers with an antibody against CD34, a marker of quiescent satellite cells (Fig. 4D), and counted the number of CD34+ cells per 100 nuclei (2,000–5,000 nuclei were counted per individual experiment [n = 5] to acquire a significant number of CD34+ cells). Again, we found only a minor reduction of the number of satellite cells in Myf5 mutants (2.8%) compared with wild-type control animals (3.5%), although the differences were statistically significant (* = p < .05) (Fig. 4C).

thumbnail image

Figure Figure 4.. The number of satellite cells in resting muscles of Myf5-deficient mice is within the normal range of variation. Transmission electron microscopy of musculi interossei from wild-type (A) and Myf5−/− (B) mice revealed structurally normal quiescent satellite cells in Myf5−/− muscles. The plasma membrane separating satellite cells from adjacent myofibers, the basal lamina that surrounds satellite cells, and myofibers, the condensed heterochromatin in the nuclei of satellite cells and small cytoplasm, are clearly visible (1-μm scale bars). (C): Satellite cell quantification in resting musculus interosseus from wild-type (black) and Myf5−/− (gray) mice by electron microscopy and counting of CD34+ cells on myotubes. The number of satellite cells per 100 myotube nuclei was calculated using both methods. A slight reduction of the satellite cell number in Myf5-deficient mice was found using either quantification method (* = p < .05). (D): Myotubes isolated from interossei muscles were stained with an anti-CD34 antibody. The arrows in (D) indicate brown CD34-positive cells (50-μm scale bar). Abbreviation: EM, electron microscopy.

Download figure to PowerPoint

To analyze the number of activated, proliferating satellite cells in regenerating muscles of Myf5-deficient and wild-type mice, we used an antibody against MyoD (Fig. 5A–5D) and counted the number of MyoD positive cells 3 and 8 days after cardiotoxin injection. We also monitored the number of satellite cell-derived myoblasts, which activated the Myf5 locus in the absence of Myf5 protein using the BAC195APZ strain. This strain carries a Myf5 reporter construct, which recapitulates the endogenous Myf5 expression [27] and thus allows tracking of the Myf5 expression by β-galactosidase (LacZ) or anti-LacZ antibody staining. No apparent changes in the number of activated satellite cells were seen by either method in regenerating muscles of Myf5-deficient mice (Fig. 5B, 5D, 5F, 5H) in comparison with wild-type (Fig. 5A, 5C, 5E, 5G).

thumbnail image

Figure Figure 5.. Regenerating muscles of Myf5-deficient mice have virtually normal numbers of proliferating satellite cells. Musculus (M.) tibialis anterior from wild-type (A, C) and Myf5−/− mice (B, D) was injured using cardiotoxin injections. Cryosections from damaged muscles were analyzed using an anti-MyoD antibody 3 days (A, B) and 8 days (C, D) after injury. Brown staining (arrows) indicates activated muscle precursor cells, which are positive for MyoD (200-μm scale bars). The numbers of MyoD-positive cells were virtually identical between wild-type and mutant mice. (E–H): No visible reduction of proliferating Myf5-LacZ satellite cells in Myf5-deficient mice. M. tibialis anterior from the Myf5 reporter strain BAC195APZ(E, G) and BAC195APZ/Myf5−/− mice (F, H) was injured using cardiotoxin injections. Cryosections from damaged muscles were analyzed using LacZ staining 3 days (E, F) and 7 days (G, H) after injury to identify Myf5-LacZ muscle precursor cells (200-μm scale bars). The numbers of LacZ-positive cells were virtually identical between wild-type and mutant mice.

Download figure to PowerPoint

Delayed Proliferation of Myf5-Deficient Satellite Cells

The number of proliferating satellite cells in vivo might not only depend on the number of quiescent satellite cells and the cell autonomous proliferation potential but also on external signals that stimulate satellite cell proliferation. For example, a decreased ability of Myf5 mutant satellite cells to proliferate might be compensated by additional stimuli that are released when the regeneration process does not meet physiological requirements. To avoid such potential compensatory interference, we carefully investigated the proliferation and differentiation of Myf5-deficient satellite cells in primary cultures. Primary satellite cell-derived myogenic cultures were established from hind limb muscles of control Myf5 reporter mice (BAC195APZ) and Myf5-deficient mice (BAC195APZ/Myf5−/−), plated at the same density and cultured for 4, 6, and 7 days in high serum growth medium (Fig. 6A–6F; supplemental online Fig. S8). Interestingly, we found a significant reduction of the number of Myf5-LacZ and MyoD-positive (data not shown and Fig. 7) satellite cell progeny in samples from Myf5 mutants in comparison with controls at day 6 and day 7 after plating (Fig. 6C–6F). Quantitative assessment of the results revealed that the initial proliferation rate of satellite cell-derived cells between day 4 and day 6 was reduced by 50% in Myf5 mutants (Fig. 6G, 6H). At later stages (between day 6 and day 7), this difference vanished, and the proliferation rate was comparable with the control (Fig. 6G, 6H). Similar results were obtained by counting satellite cell-derived myoblasts that had been pulse-labeled for 2 hours with 5-bromo-2′-deoxyuridine (BrdU) at different days of culture. Using this technique, we detected a reduction of the number of proliferating cells from Myf5 mutants by 50% at culture day 5. At day 6 and day 7, this difference was not longer detectable (supplemental online Fig. S9). To rule out that Myf5 mutant satellite cell-derived myoblasts show an increased rate of apoptosis, which might affect the proliferation rate and the number of BrdU positive cells, we performed terminal deoxynucleotidyl transferase dUTP nick-end labeling assays. No significant increase of the number of apoptotic cells was monitored in cultures of Myf5 mutant myoblasts at days 4, 5, 6, and 7 when compared with wild-type mice (supplemental online Fig. S10).

thumbnail image

Figure Figure 6.. The proliferation of Myf5 deficient satellite cells is initially delayed. Myogenic cultures isolated from pooled hind limb muscles of BAC195APZ(A, C, E) and BAC195APZ/Myf5−/− mice (B, D, F) at day 4 (A, B), day 6 (C, D), and day 7 (E, F) after plating. Cultures were plated at the same densities and maintained in growth medium. Cultures were stained for Myf5-LacZ expression and counterstained with 4,6-diamidino-2-phenylindole (data not shown) to identify all nuclei. At least 13 randomly chosen fields were microscopically analyzed for each genotype (200-μm scale bars). Cultures from Myf5-deficient mice contained less LacZ-positive cells than control cultures at days 6 and 7 (G, H). Satellite cells stained for LacZ expression in primary cell cultures derived from three Myf5-deficient (BAC/Myf5-1, -2, -3) and three control (BAC1, -2, -3) animals and cultured for 4, 6, and 7 days. The numbers of LacZ-positive cells were counted in 13 representative fields and calculated for each individual sample.(G): Total numbers of LacZ-positive cells per 13 microscopic fields at days 4, 6, and 7. (H): Average fold change of the total numbers of LacZ-positive cells between day 6 and day 4 and day 7 and day 6. At the initial growth phase (between day 6 and day 4), myogenic LacZ-positive cells from Myf5 mutants expanded two times slower than control cells. This difference in growth rate disappeared at later stages (between day 7 and day 6) (* = p < .05). Abbreviations: BAC, bacterial artificial chromosome; d, day.

Download figure to PowerPoint

thumbnail image

Figure Figure 7.. Normal distribution of individual satellite cell subpopulations in Myf5-deficient mice. Immunofluorescent images of myogenic primary cultures prepared from hind limb muscles are shown (A–E). Cultures were prepared in parallel from control BAC195APZ and Myf5-deficient mice (BAC195APZ/Myf5−/−). Cultures were fixed at day 1, day 2, day 3, day 4, day 6, and day 7 after plating and stained using antibodies against MyoD ([A], green), Pax7 ([B], red), or Myf5-LacZ (red, data not shown), myosin heavy chain ([E], blue), and nuclear dye Draq5 ([C], white) (100-μm scale bars). (D): Merged Pax7 and MyoD pictures. Total numbers of cells and purity of cultures were quantified based on nuclear staining and immunolabeling. (F–I): Kinetics of satellite cell growth and differentiation. The numbers of cells expressing MyoD and Myf5-LacZ ([H] and left table), cells expressing MyoD and Pax7 ([F, G, I] and right table), and nuclei in myotubes were counted, and the numbers of cells in different subpopulations (MyoD+/Myf5+ and MyoD+/Pax7+ shown in blue; MyoD+/Myf5− and MyoD+/Pax7− in yellow; MyoD−/Pax7+ in green;myotube nuclei in red) were averaged for 14 microscopic fields and are presented as stack bars. Diagrams (F) and (G) represent early stages after plating (day 1 to day 4, high cell density); diagrams (H) and (I) represent later stages after plating (day 4 to day 7, low cell density). The average number of cells represents counting results from three parallel cultures per time point. At each time point, the following antibody combinations were used for triple immunolabeling of parallel cultures: MyoD/LacZ/MF20 and MyoD/Pax7/MF20. No statistically significant differences of satellite cell subpopulations were observed between wild-type and Myf5(−/−) mice. Abbreviation: d, day.

Download figure to PowerPoint

To analyze whether Myf5-null myoblasts were more dependent than normal on extracellular matrix factors, we cultured wild-type and Myf5 satellite cell-derived myoblasts with or without Matrigel. We did not find a significant difference between both culture conditions. Instead, we again observed a slower growth rate of Myf5 mutant myoblasts, which was most pronounced at day 5/day 6 (supplemental online Fig. S11).

Heterogeneity of Satellite Cells from Muscles of Myf5-Deficient Mice

Muscle satellite cells represent a heterogeneous cell population that expresses different myogenic markers at different phases of satellite cell development [7, 24, 25, 35]. Most satellite cells express MyoD after activation (Pax7+/MyoD+) and proliferate before they switch off Pax7 and differentiate (Pax7−/MyoD+). A minor population seems not to express Myf5 or MyoD but maintains only Pax7 expression. Such slowly dividing Pax7+/MyoD− cells are believed to replenish the satellite cell subpopulation that proliferates rapidly and eventually differentiates into myotubes. However, it is not clear yet whether Pax7+/MyoD+ or Pax7+/Myf5+ cells might revert to Pax7+/MyoD− cells.

To study the role of Myf5 in the generation of different satellite cell-derived subpopulations, we analyzed cultures from pooled hind limb muscles of young (3 months old) control (BAC195APZ) and Myf5-deficient mice (BAC195APZ/Myf5−/−). The cultures were analyzed for proliferation (total cell numbers), activation (Pax7+/MyoD+ cells), differentiation (Pax7−/MyoD+ cells and nuclei in MyHC+ myotubes), and generation of reserve cells (Pax7+/MyoD−). Two separate staining experiments (n = 3) were performed. Half of the cells were stained for MyoD, Myf5-LacZ (antibody staining), MyHC, and Draq5 nuclear dye (data not shown). The other half was stained in parallel for MyoD (Fig. 7A, green), Pax7 (Fig. 7B, red), Draq5 nuclear dye (Fig. 7C, white), and myosin heavy chain (Fig. 7E, blue). In each case, positive cells were counted in 14 representative fields, and the number of cells within the different subpopulations was calculated (Fig. 7F–7I). In general, a higher proliferation rate was observed in controls compared with Myf5-deficient cells, resulting in higher numbers of mature cells (nuclei in myotubes, MyoD+/Myf5−, and MyoD+/Pax7− subpopulations). No significant differences in the differentiation rate of Myf5-deficient cells were detected when the initially reduced proliferation rate was taken into account. In the initial experiments, which were done using rather low cell densities and analyzed relatively late (4–7 days after plating), we were unable to detect the slowly dividing Pax7+/MyoD− “reserve cells” that are believed to replenish the satellite cell pool (Fig. 7I). However, these differences disappeared when the staining was performed at earlier time points (1–4 days after plating) and at higher concentrations of cells (Fig. 7F, 7G). Most likely, the impaired ability of Myf5 mutant satellite cells for initial proliferation in vitro did also affect the relatively rare Pax7+/MyoD− reserve cells, resulting in lower numbers of these cells after the initial proliferation phase. Since Pax7+/MyoD− reserve cells tend to lose this phenotype outside their niche, Myf5 mutant reserve cells might be unable to compensate an initial delay of proliferation. In contrast, early after plating and at higher densities, the effects of a reduced proliferation rate on the expansion of satellite cell subpopulations might be less important. Although we assume that the absence of Myf5 did not affect the formation of different satellite cell subpopulations, we formally cannot exclude that we have missed effects of the Myf5 mutation on satellite cell subpopulations in vivo within the stem cell niche, since our analysis was confined to ex vivo studies in this respect.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

The current interest in regenerative processes in adult organisms has also brought renewed attention to differentiation of organ-specific precursor cells during embryogenesis, since it is often claimed that tissue regeneration recapitulates important aspects of embryonic development. The myogenic factor Myf5 defines the onset of myogenic development in mammals, and its decisive role in embryonic muscle development is well established [36]. Hence, it appeared logical to address the role of Myf5 in adult myogenesis during skeletal muscle regeneration, which became possible after bypassing the perinatal lethality of the original Myf5 knockout strains [26].

Although the kinetics of skeletal muscle formation were delayed in Myf5-deficient mice and the degree of dystrophic changes in Myf5/mdx compound mutant mice was moderately enhanced, we only detected a restrained reduction of the initial proliferation rate of muscle satellite cells. Furthermore, we were able to rule out that Myf5 plays a decisive role for the generation of the satellite cell lineage and for the determination of specific subpopulations of cells within the satellite cell pool.

These findings suggest that the role of Myf5 during adult myogenesis is quite distinct from its role during embryogenesis, where it determines the formation of the first wave of myogenic cells [17, 37]. Myf5 has also been implicated in the control of the expression of MyoD during development [10]. We did not find evidence for this hypothesis during regeneration of skeletal muscle since the expression of MyoD was clearly not affected by the loss of Myf5. Similarly, the reduction of the initial proliferation rate of muscle satellite cells during adult myogenesis is not reflected by a comparable proliferation deficit during embryogenesis. If one accepts that Myf5 might play different roles during embryonic and adult myogenesis, it is interesting to ask whether the redundancy among MyoD and Myf5, which we have demonstrated previously for embryonic muscle formation, is relevant during adult myogenesis. The delayed induction of Myf5 expression during muscle regeneration compared with MyoD argues for distinct roles of both genes. Moreover, an increased propensity for self-renewal and an impairment of myoblast differentiation have been noted in MyoD-deficient primary myoblasts [19, 38], which were not detected in Myf5 mutant mice. On the other hand, several similarities of the phenotypes caused by the absence of either MyoD or Myf5 are evident: (a) neither the absence of Myf5 nor the inactivation of MyoD leads to a significant reduction of the number of satellite cells, (b) the combined inactivation of MyoD/mdx and Myf5/mdx results in increased dystrophic changes in the musculature, and (c) the mutation of both MyoD and of Myf5 causes a reduced proliferation rate of myoblasts, although in the case of MyoD a paradoxical increase of the numbers of cells with a myogenic potential was observed, which was not present in Myf5 mutant cells.

The reduced initial proliferation rate of Myf5-deficient myoblasts seems inconsistent with the documented role of bHLH proteins to inhibit rather than to stimulate cell cycle progression. In particular, MyoD has been shown to cause growth arrest when expressed in different cell lines [39]. However, it should be stressed that these results were obtained by directed overexpression, whereas the endogenous expression of Myf5 and MyoD follows a different scheme. The expression of MyoD peaks in mid G1, whereas Myf5 showed maximal protein levels in G0 and G2 cells [40]. This means that cells display the highest expression of MyoD during G1 at a time when cells exit the cell cycle and enter differentiation, whereas the highest level of Myf5 is reached in G0 when cells fail to differentiate. Hence, any disruption of a specific MyoD/Myf5 ratio, which might determine whether a cell proliferates or differentiates, will cause changes in the proliferation characteristics as observed in MyoD and Myf5 deficient myoblasts. Compensatory changes in the regulatory network that governs satellite cell proliferation might also complicate the situation. In fact, we have observed a diminished proliferation rate in vitro but not a reduced number of activated satellite cells in vivo, which suggests that the decreased ability of Myf5 mutant satellite cells to proliferate is compensated in vivo by additional stimuli to match physiological requirements.

It was interesting to note that the absence of Myf5 did not affect activation of the Myf5 locus as indicated by the expression of the Myf5-LacZ reporter gene. In this situation, the autoregulatory loop that is known to contribute to the maintenance of Myf5 expression [41] is disrupted, and any signals that result in the activation of the Myf5 have to be derived either from the molecular “memory” of the satellite cell or from external cues [42]. Since the positional signals that initiate and maintain expression of Myf5 during somitic cell development are missing in regenerating muscles, the driving forces that stimulate expression of Myf5 differ probably between the embryonic and adult state. Most likely this is reflected by a different set of regulatory elements in the enhancer that drives expression of Myf5 in satellite cells as already suggested by Zammit et al. [43].

The relatively moderate phenotype of Myf5-deficient mice during adult myogenesis, which does not affect the viability of Myf5 mutant mice, and the compensation of the loss of Myf5 activity by MyoD during development [20] again raise the long-known question about a specific function of the Myf5 gene, which cannot be compensated by other family members. The Myf5 gene is evolutionarily well conserved from fish to mammals, and it seems hard to accept that the evolutionary pressure to achieve this conservation is solely based on a slight reduction of the regenerative capability. However, it cannot be excluded that we have not yet disclosed a function of Myf5 that becomes vital under specific (patho)physiological circumstances or requirements. In summary, the absence of Myf5 clearly reduced the initial expansion of satellite cell-derived transient amplifying cells and resulted in a shift of the ratio of satellite cell subpopulations but did not affect the specification and generation of specific subpopulations of satellite cell-derived cells such as reserve cells, amplifying cells, and differentiating mature myogenic cells.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

This work was supported by the Max Planck Society, the DFG, and the European Commission (MYORES, Network of Excellence). We thank Nicole Gensch for help with the counting of cells and myofibers and Fikru Belema Bedada for fluorescence-activated cell sorting analysis. The authors declare that they have no conflicting commercial interests related to this work.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
Supplemental_Figure_1.pdf177KSupplemental Figure 1
Supplemental_Figure_10.pdf137KSupplemental Figure 10
Supplemental_Figure_11.pdf76KSupplemental Figure 11
Supplemental_Figure_2.pdf313KSupplemental Figure 2
Supplemental_Figure_3.pdf488KSupplemental Figure 3
Supplemental_Figure_4.pdf571KSupplemental Figure 4
Supplemental_Figure_5.pdf129KSupplemental Figure 5
Supplemental_Figure_6.pdf247KSupplemental Figure 6
Supplemental_Figure_7.pdf118KSupplemental Figure 7
Supplemental_Figure_8.pdf153KSupplemental Figure 8
Supplemental_Figure_9.pdf131KSupplemental Figure 9

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.