SEARCH

SEARCH BY CITATION

Keywords:

  • myostatin;
  • myogenesis;
  • chick;
  • embryo

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We have previously shown that Myostatin, a member of the transforming growth factor beta (TFG-β) family of signalling molecules, is expressed in developing muscle, and that treatment with recombinant Myostatin inhibited the expression of key myogenic transcription factors during chick embryogenesis. In this study, we followed the fate of muscle precursors after exposure to Myostatin. We report that in contrast to the down-regulation in expression of Pax-3, Myf-5, MyoD, and Myogenin, expression of Pax-7 was maintained. However, Myostatin completely inhibited cell division in the Pax-7-expressing cells. The inhibitory effect of Myostatin was reversible, as upon withdrawal myogenic cells re-initiated cell proliferation as well as expression of Pax-3 and MyoD. These results led us to investigate the temporal and spatial distribution of quiescent muscle precursors during development. To this end, we analysed distribution and mitotic behaviour of Pax-7-expressing cells during muscle development. Our studies revealed two populations of Pax-7-expressing cells, one that proliferated and incorporated BrdU, whilst the other did not. At early developmental stages, a high proportion of Pax-7-expressing cells proliferated, but there was a significant number of non-dividing Pax-7-expressing cells intermingled with differentiated muscle. Proliferating precursors became less frequent as development proceeded and at late fetal stages all Pax-7-expressing cells were mitotically quiescent. We suggest that Myostatin is an important signalling molecule responsible for imposing quiescence upon myogenic precursors during embryonic and foetal development. © 2006 Wiley-Liss, Inc. Developmental Dynamics 235:672–680, 2006. © 2006 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

During myogenesis, muscle precursors proliferate to expand the cell population before withdrawing from the cell cycle in order to differentiate. Inhibition of muscle precursor proliferation can result in an arrest of muscle growth and a reduced capacity to regenerate (Amthor et al., 1998; Quinlan et al., 1995, 1997). However, the generation of mitotically quiescent muscle precursors is equally important because it establishes a reserve cell population in adult muscle, called satellite cells, for future size adaptation or regeneration of skeletal muscle (Oustanina et al., 2004; Seale et al., 2000). Recently a number of groups have provided data demonstrating the generation of inactive muscle precursors during development (Gros et al., 2005; Relaix et al., 2005; Kassar-Duchossoy et al., 2005), greatly raising the significance of identifying key molecules that control the entry of muscle precursors into a mitotically quiescent state.

We have previously reported that Myostatin inhibited Pax-3, Myf-5, and MyoD expression during chick development and concluded this inhibition was responsible for the growth delay of muscle in the presence of Myostatin (Amthor et al., 2002, 2004). This deduction was in keeping with the phenotype of the Myostatin null mouse that displayed a two- to threefold increase in skeletal muscle mass resulting from an increase in fibre number as well as an increase in fibre size (McPherron et al., 1997). In this study, we investigated the fate of avian embryonic muscle precursors following their exposure to Myostatin. Myostatin has previously been implicated in imposing quiescence onto adult satellite cells (McCroskery et al., 2003). Here we asked whether Myostatin acts similarly during development as it was unclear from our previous studies whether Myostatin caused a loss of precursors, committed precursors towards another cell lineage, or whether it locked muscle precursors into a quiescent state.

We focused our study on the ability of muscle precursors to express Pax-3 and Pax-7, two genes of the paired box family of transcription factors, as these genes are also expressed in quiescent satellite cells (Buckingham et al., 2003). Previous studies have shown that Pax-3 and Pax-7 have distinct properties. Over-expression of Pax-7 led to the down-regulation of MyoD and Myogenin as well as withdrawing cells from the cell cycle (Olguin and Olwin, 2004), whereas Pax-3 seemed to promote proliferation of muscle precursors (Relaix et al., 2004).

In this study, we investigated the effect of recombinant Myostatin on the transcription of genes that regulate proliferation and differentiation of muscle precursors during limb development of chick embryos. Our data demonstrate that Myostatin differentially regulates myogenic gene expression by down-regulating expression of MyoD, Myogenin, Myf-5, and Pax-3, whilst maintaining expression of Pax-7. In addition, we also show that Myostatin inhibits proliferation of Pax-7-expressing muscle progenitors. Finally, we studied the distribution of proliferating and non-proliferating Pax-7 muscle precursors during chick development. We provide evidence that mitotically inactive Pax-7-expressing precursors predominantly accumulated in regions that express high levels of Myostatin, whereas mitotically active Pax-7-expressing precursors were predominantly present in regions outside of Myostatin expression.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Effect of Myostatin on MyoD, Myf-5, Myogenin, and Pax-3 Expression

We have previously shown that Myostatin inhibits the expression of genes that are involved in the differentiation cascade of embryonic muscle (Amthor et al., 2002).

Here we modified our previous experimental procedure to ensure a more homogenous exposure of muscle cells to Myostatin through the use of evenly spaced affinity beads. In addition, we monitored the effect of Myostatin at different intervals after bead implantation, which enabled us to study temporary changes in gene expression.

To this end, we exposed developing limb muscle to recombinant Myostatin and monitored the effect on gene expression using in situ hybridisation. We have previously noted a concentration-dependent response of myogenic cells to Myostatin and, therefore, designed our experiments to yield a maximal response. We, therefore, soaked Affigel blue beads in 1 mg/ml recombinant Myostatin or in PBS for controls and inserted multiple evenly spaced beads into the dorsal mesenchyme of right wing buds covering the extent of the dorsal premuscle mass. We reincubated operated embryos between 1 and 24 hr. The non-operated left wing buds served as internal controls for successful in situ hybridisation.

Control experiments in which we inserted beads soaked in PBS into the dorsal mesenchyme of HH stage 21–25 limbs had no influence on gene expression or limb outgrowth when examined 6 hr to one day after bead implantation (Fig. 1).

thumbnail image

Figure 1. Effect of recombinant Myostatin on MyoD, Myf-5, Myogenin, and Pax-3 expression in developing chick wing buds revealed by in situ hybridisation. A,D,G,J: Dorsal view at non-operated control wing buds. B,E,H,K: Dorsal view at wing buds after implantation of PBS-beads C,F,I,L: Dorsal view at wing buds after implantation of Myostatin beads. A: Expression profile of MyoD at day 4 of development in the dorsal premuscle mass of a non-operated control wing bud. B: Implantation of control beads exerted no effect on the expression profile of MyoD in the dorsal premuscle mass 6 hr after bead implantation. C: Implantation of Myostatin beads led to complete loss of MyoD expression in the dorsal premuscle mass within 4 hr. D: Expression profile of Myf-5 at day 4 of development in the dorsal premuscle mass of a non operated control wing bud. E: Implantation of control beads exerted no effect on the expression profile of Myf-5 in the dorsal premuscle mass 6 hr after bead implantation. F: Expression of Myf-5 was down-regulated in the dorsal premuscle mass 6 hr after Myostatin bead implantation. G: Expression profile of Myogenin at day 5 of development in the dorsal premuscle mass of a non-operated control wing bud. H: Implantation of control beads exerted no effect on the expression profile of Myogenin of the dorsal premuscle mass 8 hr after bead implantation. I: Myostatin beads caused severe loss in Myogenin expression in the dorsal premuscle mass 6 hr after bead implantation. J: Expression profile of Pax-3 at day 4 of development in the dorsal premuscle mass of a non-operated control wing bud. K: Implantation of control beads exerted no effect on the expression profile of Pax-3 of the dorsal premuscle mass 6 hr after bead implantation. L: Complete loss in Pax-3 expression of the dorsal premuscle mass 7 hr after Myostatin bead implantation.

Download figure to PowerPoint

Myostatin and MyoD.

We first tested the effect of Myostatin on MyoD expression. We inserted Myostatin beads into the limb bud mesenchyme after the onset of MyoD expression in the dorsal muscle mass, which begins at HH-stage 21. Operations were performed on HH-stage 21–25. In all cases, we found a severe down-regulation of MyoD expression in the dorsal and ventral muscle masses 3 to 8 hr after the operation (n = 15; Fig. 1A–C). Total down-regulation of MyoD expression in the dorsal muscle mass was observed in nearly all cases (13/15) and also in the ventral muscle masses in 4/15 cases. One day after Myostatin bead insertion, the proximal region of the limb bud appeared smaller in size in some cases. Remarkably, MyoD expression did not extend distally compared to the non-operated left wing bud or to control limb buds (n = 8; data not shown).

Myostatin and Myf-5.

A down-regulation of Myf-5 expression was detected 4–7 hr after Myostatin bead implantation, but in contrast to MyoD the expression of the gene was never completely extinguished (n = 18; Fig. 1D–F). Twenty-four hours after Myostatin bead insertion, we found a significant reduction in Myf-5 expression with only a remnant of expression remaining in the proximal part of the limb bud that failed to extend distally (7/8; data not shown).

Myostatin and Myogenin.

Operations were performed on HH-stage 25–26 embryos, immediately after the onset of Myogenin expression in the wing muscle. We found that 6 hr after Myostatin bead exposure, the extent of the expression domain remained the same compared to the PBS-bead treated wings. However, the intensity of Myogenin expression was severely down-regulated in every case (n = 9; Fig. 1G–I).

Myostatin and Pax-3.

Operations were performed on HH-stage 21–24. The first detectable down-regulation was found as early as 2 hr after Myostatin exposure. In the 3–7-hr time window after Myostatin bead insertion, Pax-3 expression in the dorsal premuscle mass was completely lost in 12/18 cases and in the remaining cases only a few expressing cells were detectable (Fig. 1J–L). Pax-3 expression in the ventral premuscle mass was also severely compromised. After 1 day, all operated wing buds showed some Pax-3 expression, though the expression domain was much smaller compared to controls and confined to the proximal limb region (data not shown).

In summary, expression of MyoD, Myf-5, Myogenin, and Pax-3 was down-regulated following exposure to Myostatin within 2–8 hr. However, the differing genes showed individual susceptibilities to Myostatin with Pax-3 and MyoD being completely down-regulated in contrast to the partial loss of Myf-5 and Myogenin expression. The effect seemed transient, in that some cells reinitiated MyoD and Pax-3 expression after an extended period. However, the expression domain was never completely restored to normal levels.

The Reversible Effect of Myostatin on MyoD and Pax-3 Expression

In the next set of experiments, we determined whether expression of Pax-3 and MyoD could be reinitiated after inhibition. We inserted 8–12 Myostatin beads into the dorsal mesenchyme of wing buds at HH-stages 21–23 and re-incubated the embryos for 6 hr. Our previous experiments demonstrated a severe inhibiting effect of Myostatin on MyoD and Pax-3 expression at this time point. We then removed the beads and re-incubated for another 18–24 hr. In all cases, we found an almost complete restoration of MyoD expression (n = 6) as well as Pax-3 expression (n = 8; Fig. 2). Thus, in spite of the fact that Myostatin down-regulated a number of key myogenic transcription factors, these cells could execute the myogenic programme once the inhibitory influence was removed.

thumbnail image

Figure 2. The reversible nature of the inhibitory effect of recombinant Myostatin on Pax-3 and MyoD expression in developing chick wing buds revealed by in situ hybridisation. A,B: Five-day-old non-operated control wing buds. F–J: Five-day-old operated wing buds. Implantation of Myostatin beads at day 4 of development and bead removal 6 hr thereafter, which was then followed by an 18-hr re-incubation. A: Expression profile of Pax-3 at day 5 of development in the dorsal premuscle mass of a non operated control wing bud. B: Expression profile of MyoD at day 5 of development in the dorsal premuscle mass of a non-operated control wing bud. C: Almost complete restoration of Pax-3 expression 18 hr after Myostatin bead removal. Note that not only is the expression of nearly the same intensity as the control, but also the size of the expression domain. D: Almost complete restoration of MyoD expression 18 hr after Myostatin bead removal. Note that the expression intensity is nearly the same as the control. However, the size of the expression domain is slightly smaller.

Download figure to PowerPoint

Effect of Myostatin on Pax-7 Expression

Thus far, all the myogenic genes examined had been down-regulated to some extent following exposure to Myostatin. Work on adult muscle has suggested that Myostatin may induce quiescence of satellite cells (McCroskery et al., 2003). Therefore, we next tested the effect of Myostatin beads on Pax-7 expression, a gene transcribed in non-dividing muscle cells (Buckingham et al., 2003). Operations were performed at HH-stage 21–23. In contrast to the complete down-regulation of MyoD and Pax-3 expression, we found no discernible down-regulation of Pax-7 expression 6–7 hr after Myostatin bead implantation (n = 10; Fig. 3A–E). After one day, we found a remnant of expression at the proximal part of the limb bud that failed to extend distally (8/8; Fig. 3F–J).

thumbnail image

Figure 3. Effect of recombinant Myostatin on Pax-7 expression in developing chick wing buds revealed by in situ hybridisation. A–E: Four-day-old wing buds. Implantation of Myostatin and control beads into dorsal wing bud mesenchyme of 4-day-old chick embryos and re-incubation for up to 6 hr. F–J: Five-day-old wing buds. Implantation of Myostatin and control beads into dorsal wing bud mesenchyme of 4-day-old chick embryos and re-incubation for 1 day. A: Expression profile of Pax-7 at day 4 of development in the dorsal premuscle mass of a non-operated control wing bud (arrow). B: Implantation of control beads exerted no effect on the expression profile of Pax-7 of the dorsal premuscle mass 6 hr after bead implantation. C: Expression of Pax-7 of the dorsal premuscle mass is largely maintained 6 hr after Myostatin bead implantation. D: Expression profile of Pax-7 at day 4 of development in the ventral premuscle mass of a non-operated control wing bud (arrow). E: Expression of Pax-7 is largely maintained in the ventral premuscle mass 6 hr after Myostatin bead implantation. F: Expression profile of Pax-7 at day 5 of development in the dorsal premuscle mass of a non-operated control wing bud (arrow). G: Implantation of control beads exerted no effect on the expression profile of Pax-7 of the dorsal premuscle mass one day after bead implantation. H: Truncation of the Pax-7 expression domain of the dorsal premuscle mass. Expression remained mainly in the proximal part of the wing 1 day after Myostatin bead implantation. I: Expression profile of Pax-7 at day 5 of development in the ventral premuscle mass of a non-operated control wing bud (arrow). J: Expression of Pax-7 was largely maintained in the ventral premuscle mass 1 day after Myostatin bead implantation.

Download figure to PowerPoint

These experiments show that Pax-7 expression was initially unaffected by the presence of Myostatin, whereas MyoD, Myf-5, Myogenin, and Pax-3 were considerably down-regulated. However, the expression domain of Pax-7 failed to expand in the presence of Myostatin during further development.

Myostatin Reversibly Inhibited the Proliferation of Pax-7-Expressing Muscle Precursors

We had noted that even though the expression of Pax-7 was relatively unaffected by Myostatin, the expression domain did not expand in its normal manner. One explanation for these results could be changes in the rate of cell division in Pax-7-expressing cells. To investigate this possibility, we inserted Myostatin beads into limb buds at HH-stages 21–25 as previously described. Limb buds were treated with BrdU 30 min prior to fixation. In non-operated left wing buds, a substantial number of mesenchymal cells were labelled with BrdU (Fig. 4A,B). The positions of the premuscle masses were easily identified by the presence of Pax-7-positive cells. Approximately 30–50% of Pax-7-expressing cells co-localised with BrdU-positive nuclei. In contrast, 7–10 hr after Myostatin exposure, there was a severe decrease in the number of BrdU-positive cells in all cases (n = 8; Fig. 4C,D). Pax-7-positive cells could still be found; however, they rarely (<5%) co-localised with BrdU-positive cells.

thumbnail image

Figure 4. The reversible nature of the inhibitory effect of recombinant Myostatin on proliferating Pax-7-expressing muscle precursors in developing chick wing buds. A,C,E: Transverse section through the proximal region of wing buds at day 4 of development (A,C) and day 5 of development (E). Position of premuscle masses in green. Inlays are shown as close-ups in B,D,F. Imaging by laser scanning microscopy after immunhistochemistry; Pax-7-expressing cells appear in green, BrdU in red (30 min BrdU pulse before fixation); cells that incorporated BrdU and expressed Pax-7 appear in yellow. A,B: Non-operated control. About half the number of Pax-7-expressing cells incorporated BrdU and appear in yellow and thus are proliferating. C,D: Severe reduction in cells that incorporated BrdU 7 hr after exposure to Myostatin. Pax-7-expressing cells are still plentiful; however, they were not positive for BrdU, hence, the proliferation of Pax-7-expressing cells decreased compared to the control depicted in A,B. E,F: Myostatin beads were removed after an initial treatment for 7 hr and embryos allowed to develop for another 18 hr. Following bead removal, a substantial number of Pax-7-expressing cells were also positive for BrdU indicating that Pax-7 cells re-initiated proliferation.

Download figure to PowerPoint

To clarify whether the inhibiting effect of Myostatin on the proliferation of muscle precursors was also of a reversible nature, we removed the Myostatin beads after an initial exposure of 7 hr and re-incubated the embryos for an additional 18 hr. Whereas 7 hr after Myostatin exposure the proliferation of Pax-7-expressing cells was severely inhibited (see above), removal of Myostatin beads resulted in a re-initiation of cell proliferation and in all cases we found a high proportion of Pax-7-positive nuclei that co-localised with BrdU-positive nuclei (30–40%) (n = 5; Fig. 4E,F).

These results demonstrate that myogenic precursors continually expressed Pax-7 even when Myostatin has mitotically silenced them. Furthermore, the Pax-7 cells initiated cell division when freed from the influence of Myostatin.

Distribution of Mitotically Active and Mitotically Inactive Pax-7 Muscle Precursors During Normal Development

The above results show that muscle precursors continue to express Pax-7 in the presence of Myostatin, but cease to proliferate. We, therefore, aimed to investigate the emergence of non-proliferating Pax-7-expressing cells during development in order to gain insight on their potential in vivo relevance.

We have previously shown that the expression of Myostatin during muscle development is restricted to specific regions of both the trunk and limb (Amthor et al., 2002, 2004). In the following set of experiments, we determined the distribution of Pax-7-expressing cells in the somites and correlated this to cell division and the previously documented expression of Myostatin during normal development. We carried out a detailed examination of the somites of 4-day-old chick embryos. At this stage, Myostatin is expressed at high levels in the intermediate portion of the dermomyotome but not in the dorsomedial and ventrolateral dermomyotomal lips (Amthor et al., 2002, 2004). We found a high proportion of Pax-7+ cells that were also BrdU+ at the dermomyotomal lips, whereas there were less frequently Pax-7+/BrdU+ expressing cells in the intermediate portion of the dermomyotome (6 hr of BrdU incubation; Fig. 5M–P). Interestingly, Pax-7-expressing muscle precursors were not confined to the dermomyotome but also located in between differentiated muscle cells of the myotome that expressed Myosin Heavy Chain (Fig. 5Q–S) confirming recent reports from Gros et al. (2005), Relaix et al. (2005), and Kassar-Duchossoy et al. (2005).

thumbnail image

Figure 5. Characterisation of Pax-7-expressing cells during chick prenatal development of limb and somite muscle revealed by immunostaining against Pax-7, BrdU, and Myosin Heavy Chain and laser scanning microscopy. Double immunostaining against Pax-7 (green) and BrdU (red) on sections from 5-day-old wing buds (A–D), from 10-day-old wing buds at zeugopod level (E–H), from 19-day-old calf muscle (I–L), and from 4-day-old thorax region (M–P) of chick embryos. Double-expressing cells appear yellow in overlay (D,H,K and M–P). Double immunostaining against Pax-7 (green) and Myosin Heavy Chain (MHC; red) on sections from 4-day-old thorax region (Q–S) and from 10-day-old wing muscle at zeugopod region (T) of chick embryos. A: Overlay of Pax-7 expression (green) in dorsal and ventral premuscle masses and phase image (grey) of a 5-day-old chick wing. Inlay is shown as close-ups in B,C,D. B: Close-up of inlay in A shows densely packed Pax-7 expressing cell nuclei in dorsal premuscle mass. C: Close-up of inlay in A and same field of view as in B shows cell nuclei that incorporated BrdU (red). D: Overlay of B and C revealed that many cell nuclei were positive for Pax-7 and BrdU and appeared yellow. However, many cell nuclei were only positive for Pax-7. E: Overlay of Pax-7 expression (green) in zeugopod muscles and phase image (grey) from 10-day-old chick embryos. Inlay is shown as close-ups in F–H. F: Close-up of inlay in E shows that wing muscle from 10-day-old embryos was less densely packed with Pax-7-expressing cells compared to day 5 (compare with B). G: Close-up of inlay in E and same field of view as in F show cells that incorporated BrdU (red). H: Overlay of F and G revealed that only a few cells that were positive for Pax-7 were also positive for BrdU (cell nuclei in yellow), whereas the majority were positive for Pax-7 only (cell nuclei in green). I: Overlay of Pax-7 expression (green) and phase image (grey) revealed that only a few Pax-7-expressing cell nuclei located in between muscle fibres of calf muscle from 19-day-old chick foetus. J: Same field of view as in I shows few BrdU positive cell nuclei (red) located in between muscle fibres. K: Overlay of staining against Pax-7 (green) and BrdU (red) revealed no Pax-7-expressing cell nuclei that were also positive for BrdU. L: Nuclear stain of same field of view as in I–K. M: Double staining against Pax-7 (green) and BrdU (red) at thorax level from 4-day-old chick embryos at low magnification. Inlays mark different regions of the dermomyotome and are shown in N–P. Pax-7 expression to the right of the letter N is located in the dorsal part of the neural tube. N: Close-up of inlay N in M shows densely packed Pax-7-expressing cells at the dorsomedial portion of the dermomyotome that were partly positive for Pax-7 and BrdU (yellow) or only positive for Pax-7 (green). O: Close-up of inlay O in M shows Pax-7-expressing cells at the intermediate portion of the dermomyotome. Less Pax-7-expressing cells were also positive for BrdU (yellow) when compared to N or P. P: Close-up of inlay P in M shows Pax-7-expressing cells at the ventrolateral portion of the dermomyotome. Many Pax-7-expressing cells were also positive for BrdU (yellow). Q: Overlay of staining against Pax-7 (green) and phase image (grey) at thorax level from 4-day-old chick embryos at low magnification. Inlay marks intermediate region of the dermomyotome/myotome that is shown in S. R: Overlay MHC expression in muscle of the myotome (red) and phase image (grey) of same field of view as in Q. Inlay marks same region shown in Q and S. S: Close-up and overlay of inlays in Q and R revealed that Pax-7-expressing cells (green) were not only confined to the dermomyotome but also located in between MHC-expressing muscle cells of the myotome (red). T: Overlay of staining against Pax-7 (green) and MHC (red) in zeugopod muscle of 10-day-old chick wings. Pax-7-expressing cells were intercalated in between MHC-expressing cells.

Download figure to PowerPoint

Interestingly, we found that limb premuscle masses at day 5 of development were densely populated with Pax-7+ cells, of which a high proportion were also BrdU+ (30 min BrdU incubation; Fig. 5A–D). At day 10 of development, Pax-7+ cells were less densely populated in limb muscle, and also less frequently appeared BrdU+ than at day 5 of development (30 min BrdU incubation; Fig. 5E–H). Interestingly, already at this stage, Pax-7+ cells mixed in between differentiated limb muscle that expressed Myosin Heavy chain (Fig. 5T). At day 19 of development, Pax-7+ cells appeared sparse in limb muscle and they did not proliferate even when we incubated embryos up to 7 hr with BrdU (Fig. 5I–L).

This demonstrates that Pax-7-expressing muscle precursors intermingled with differentiated muscle cells at early stages of muscle development, and a fraction of these muscle precursors proliferated and incorporated BrdU while another fraction did not. In somites, non-proliferating Pax-7-expressing cells accumulated at positions that we previously described as expressing Myostatin.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Our experiments demonstrate that non-proliferating Pax-7-expressing muscle precursors are generated from early developmental stages onwards, both during somite and limb muscle development. We have demonstrated that Myostatin is a signalling molecule that inhibits proliferation of Pax-7-expressing muscle precursors, and furthermore that Myostatin down-regulated transcription factors that control proliferation and differentiation of embryonic muscle precursors, such as Pax-3, MyoD, Myf-5, and Myogenin, which confirmed previously published data (Amthor et al., 2002, 2004). However, in addition to our previously published data, we now show that myogenic gene response to Myostatin is subject to significant molecular variation. MyoD and Pax-3 most strongly responded to Myostatin and expression of both genes was completely lost, whereas expression of Myf-5 was inhibited, but not completely lost, and expression of Pax-7 was not affected. This shows that Myostatin induced a dormant muscle precursor cell state whereby cells maintained expression of Pax-7 and some remnant Myf-5 expression, but ceased to proliferate and differentiate. Importantly, the quiescence of muscle precursors was of a transient nature and they re-initiated proliferation and differentiation upon release from the influence of Myostatin.

In this study, we have implanted multiple beads soaked in Myostatin to attain uniform levels of protein into the experimental tissue. Although the carrier beads were soaked in non-physiological levels of Myostatin, the actual amount delivered to the tissue is a fraction of the stock solution; plus the quantity of Myostatin used may not reflect activity. We are simply using a concentration of recombinant protein that consistently produces a biological effect. We believe that the effects reported in this study are specific to the action of Myostatin as we do not observe cellular features induced by other members of the TGF as apoptosis or ectopic bone development.

We have previously shown that Myostatin is expressed in both the limbs and somites and yet it has the capacity to inhibit gene transcription and muscle precursor proliferation. However, muscle still develops at both these sites. We suggest that both the timing and the spatial distribution of Myostatin expression, together with its antagonists (such as Follistatin), allow myogenesis to proceed and for Myostatin to regulate its development. For example, we have shown that the expression of Myostatin is relatively late in comparison to the onset of myogenesis (especially in the somites), as robust expression of Myostatin is not detected until after stage 19 (Amthor et al., 2002). Furthermore, even when transcription has been initiated, Myostatin tends to be expressed in a subset of the myogenic anlagen.

We have shown that expression of Pax-7 was not sufficient to maintain proliferation of muscle precursors after Pax-3 expression was down-regulated in response to Myostatin. This situation is similar to the behaviour of Pax-7-expressing muscle precursors in a genetically modified mouse model, in which cells expressed Pax-7 in place of Pax-3 (Relaix et al., 2004). Replacement of Pax-3 by Pax-7 caused a reduced migration of muscle precursors into the limb bud and reduced their rate of proliferation. Our results strongly support the notion that Pax-3 and Pax-7 have diverse functions during muscle development, and we suggest that signalling molecules such as Myostatin modulate cell function by differentially regulating the two genes. Our proliferation assays revealed two populations of Pax-7-expressing cells, one that proliferated and another that did not. We hypothesise that proliferating Pax-7-expressing cells co-express Pax-3, but not quiescent Pax-7-expressing cells. Our data support the idea that there is a concentration-dependent antagonism of Pax genes (Relaix et al., 2004), and signalling molecules such as Myostatin or BMPs will inhibit or promote proliferation by regulating Pax-3 expression (Amthor et al., 1999). We have tried to prove that proliferating Pax-7-expressing cells also express Pax-3 at the single cell level. However, in our hands none of the Pax-3 antibodies available specifically mark Pax-3, but mark a subpopulation of Pax-7-expressing cells instead.

Our results also show that there is a concentration-dependent antagonism of the Myogenic Regulatory Factors. Whereas MyoD expression is completely down-regulated by Myostatin expression, Myf5 expression was maintained. The inhibitory effect of Myostatin on Pax-3 and MyoD expression are significant as it shows that Myostatin can control muscle developing not only at the myogenic precursor proliferation stage as previously mentioned but also at the differentiation stage. Previous work has shown that the activation of the Myostatin signalling pathway leads to phosphorylated Smad3 binding to MyoD and preventing it from carrying out its transcription activity (Langley et al., 2002). Myostatin would lead to a down-regulation of MyoD as this transcription factor auto-regulates its expression. We suggest although differentiating muscle cells can survive in the absence of MyoD, it is essential that they maintain Myf5 expression as recent work has shown that deletion of both these factors leads to myogenic apoptosis (Kassar-Duchossoy et al., 2005). In addition, we suggest that although Myostatin can down-regulate MyoD transcription, this may not be occurring in early development as we have seen the two molecules often co-expressed. Instead, we suggest that the primary function of Myostatin is to control proliferation of precursor cells and this is achieved by threshold levels being too low to inhibit MyoD expression. We are currently determining whether Myostatin has morphogenic properties.

We investigated the generation of quiescent Pax-7-expressing muscle precursors during development. In our previous studies, we showed that the highest rates of cell division within the dermomyotome were localised at the dorsal medial and ventral lateral lips (Amthor et al., 1999). These results have been confirmed by Gros et al., 2005 (Fig. 1N in Gros et al., 2005). Pax-7-expressing cells are located along the entire length of the dermomyotome. We found that during somite development, proliferating Pax-7-expressing cells predominantly located at the dorsomedial and ventrolateral third of the dermomyotome, whereas the intermediate portion was predominantly populated with Pax-7-expressing cells that proliferated less. We previously demonstrated that Myostatin expression was initially confined to the intermediate portion of the dermomyotome (Amthor et al., 2002, 2004). Thus, there is a correlation in the spatio-temporal expression of Myostatin and non-proliferating Pax-7-expressing cells. This is further emphasised at a slightly later stage in development once the cells have left the dermomyotome and entered the myotome. At this stage, Gros et al. (2005) showed actively dividing Pax-7 cells in the Myotome (Fig. 3E and F in Gros et al., 2005). However it has to be remembered that the expression of Myostatin is firstly localised to the central portion of the dermomyotome and then at a later time spreads to the myotome. Therefore, Pax-7-expressing cells in the central portion of the dermomyotome were initially able to divide and only stopped following the local upregulation of Myostatin. Once they translocated out of this region and into the myotome, they could reinitiate proliferation. We suggest that as the muscle developed, there is an up-regulation of Myostatin, which again would inhibit proliferation activity of Pax-7-expressing cells. This further supports our hypothesis that Myostatin induces reversible quiescence of muscle precursors during development.

Interestingly, we revealed that during development, proliferating Pax-7-expressing cells became less frequent, and by the end of fetal stages we found only non-proliferating Pax-7 cells. We suggest that quiescent satellite-like cells are formed early during development, and the early cell cycle arrest is an important mechanism to prevent excessive muscle growth. In fact, lack of Myostatin in the constitutive Myostatin knockout mouse leads to a twofold increase in the number of muscle fibres, pointing to an excessive proliferation and differentiation of muscle precursors during development (McPherron et al., 1997).

We have shown that Myostatin is able to induce mitotic quiescence of Pax-7-expressing muscle precursors at embryonic stages in our experimental setup. Additionally, we have shown that mitotically quiescent Pax-7-expressing cells are more likely to accumulate in regions of high Myostatin expression. This observation led us to ask whether Myostatin is required for the formation of quiescent satellite cells in adult muscle, and, interestingly, analysis of the Myostatin knockout mouse revealed a higher percentage of satellite cells in cell cycle (McCroskery et al., 2003).

The situation in embryonic muscle, therefore, is reminiscent of that in adult muscle, in which Myostatin has been implicated in imposing quiescence on satellite cells. Furthermore, like embryonic precursors, most satellite cells express Pax-7 and Myf-5, and only upon activation do they initiate proliferation, express MyoD, and differentiate (Beauchamp et al., 2000; Zammit et al., 2004).

From the results of the present study, we can conclude that the generation of quiescent muscle precursors is a continuous process that begins at early developmental stages and proceeds to the formation of satellite cells. Myostatin may be a key signalling factor that imposes quiescence and allows muscle precursors to be reactivated upon release from its inhibitory effect.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Preparation of Chick Embryos

Fertilised chick eggs were incubated at 38°C, and the embryos were staged according to Hamburger and Hamilton (1992). Experiments were performed on embryos at HH stages 22 to 24, re-incubated for periods stated in the text, sacrificed, and processed for whole-mount in situ hybridisation.

Myostatin Bead Preparation and Application to Limb Buds

Recombinant mouse Myostatin protein was purchased from R&D Systems, dissolved according to the manufacturer's protocols, and applied to Affigel beads (Sigma, St. Louis, MO). The protein was loaded onto beads as described by Cohn et al. (1995) in a concentration of 1 mg/ml. For control bead implantation, beads were soaked in PBS only. For bead implantation, the dorsal ectoderm and mesenchyme of the right wing were punctured with an electrolytically sharpened tungsten needle, and beads were inserted into the punctured mesenchyme using a blunt glass needle. Beads were implanted at HH-stages stated in the text. For bead removal experiments, beads were extracted using a tungsten needle.

Whole-Mount In Situ Hybridisation

All chick embryos were washed in PBS and then fixed overnight in 4% paraformaldehyde at 4°C. Anti-sense RNA probes were labelled with digoxygenin, and whole-mount in situ-hybridisation was performed as described by Nieto et al. (1996). The following probes were used in this study: MyoD, clone CMD9 1.5-kb full-length fragment (a gift from Professor Bruce Patterson); Pax-3, a 645-bp fragment corresponding to nucleotides 468–1,113 (a gift from Dr. Martin Goulding); Myf-5, a 1.1-kb fragment and Myogenin, a 1.2-kb fragment (both kind gifts from Professor Antony Graham); Pax-7 a 582-bp fragment (a kind gift from Dr. Susanne Dietrich).

Immunohistochemistry

Embryos that were subjected to immunohistochemistry were incubated with BrdU from 30 min to 7 hr prior to fixation. Three hundred microliters of a 16-mM BrdU solution (Sigma, B-5002) was added on top of each embryo in ovo. Embryos were fixed in 4% paraformaldehyde in PBS overnight at 4°C, embedded in 4% agar, and sectioned using a vibratome at a thickness of 100 μm. Alternatively, embryos were embedded in Tissue Tec (Jung Tissue Freezing Medium) and cryosectioned at 10 μm. For the preparation of the double staining against BrdU and Pax-7, sections were incubated in 100% methanol for 30 min, and then incubated in 2 N HCl for 30 min. Subsequently, the samples were neutralised in 0.1M sodium borite (Merck 102674E) for 15 min. Sections were subsequently incubated in rat monoclonal anti-BrdU antibody (1:100 in 20% goat serum; Abcam, ab6326) together with mouse monoclonal anti-Pax-7 antibody (1:4 in 20% goat serum; supernatant from DSHB Pax-7 cell line) overnight at 4°C, washed with PBS, incubated with biotinylated rabbit anti-rat IgG (1:200 in goat serum; DAKO, E 0468) for 1 hr at room temperature, washed in PBS, incubated in Alexa 594 streptavidin conjugate (1:200 in goat serum; Molecular Probes, Eugene, OR; S-11227) and Alexa 433 goat anti-mouse IgG (1:200 in goat serum; Molecular Probes, A-11029) for 1 h at room temperature, washed in PBS, and mounted. Some sections were additionally stained with TO-PRO-3 iodide (1:500 in PBS; Molecular Probes, T-3605) as a nuclear stain.

For double staining against Pax-7 and Myosin Heavy Chain (MHC), we preincubated sections in 20% goat serum for 20 min, followed by an incubation in mouse monoclonal anti-Pax-7 antibody (1:4 in 20% goat serum; supernatant from DSHB Pax-7 cell line) overnight at 4°C, washed with PBS, incubated in Alexa 433 goat anti-mouse IgG (1:200 in goat serum; Molecular Probes, A-11029) for 1 hr at room temperature, washed in PBS, incubated in biotinylated mouse anti-pan-MHC (1:50 in 20% goat serum; purified and biotinylated supernatant from hybridoma cell line, clone A4.1025; gift from Dr. Simon Hughes), washed in PBS, incubated in Alexa 594 streptavidin conjugate (1:200 in goat serum, Molecular Probes, S-11227) and mounted. Imaging was performed by confocal microscopy (Zeiss LSM 510).

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

This work was supported by the following grants: Deutsche Forschungsgemeinschaft (Am 151/2-1) to H.A. and Biotechnology and Biological Sciences Research Council (BBS/S/A/2004/11020) to A.O. and K.P. We thank Elaine Shervill for technical assistance.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES