Unit: Stems Cells and Development - CNRS URA 2578
Director: Tajbakhsh Shahragim
Our principal interest is centered on how skeletal muscle stem cells are maintained, and how their fates become restricted during prenatal and postnatal development. We are using genetically modified mice to resolve the epistatic relationships between Pax3, Pax7, Myf5, Mrf4 and MyoD. These key transcription factors act on cells which divide with presumed symmetric or asymmetric kinetics to establish and maintain the skeletal muscle programme in the embryo and postnatally. We propose a cellular and developmental lineage model in which these events occur.
Genetic and cellular regulation of skeletal muscle identity :
Lina Kassar-Duchossoy*, Ellen Giacone*, Philippos Mourikis, Barbara Gayraud-Morel, Ramkumar Sambasivan, Danielle Gomès*, Gérard Dumas (*left the laboratory in 2005)
Skeletal muscle development serves as a excellent paradigm for the investigation of the acquisition of cell fate, the study of stem to differentiated cell transitions, as well as self-renewal of stem cells and tissue homeostasis. Key regulatory genes of the paired-box/homeodomain family, Pax3 and Pax7, the bHLH myogenic regulatory factors Myf5, MyoD, Mrf4 and Myogenin have been identified. The genetic regulation of skeletal muscle has proven to be complex. For example, regional differences such as head and body proper, are mirrored by distinct epistatic pathways since Pax3 and Myf5 act upstream of MyoD in the embryo (Tajbakhsh et al. 1997). These findings established a genetic hierarchy in somite derived muscles in the embryo. More recently, we identified Mrf4 as a third determination gene in this lineage in the embryo (Kassar-Duchossoy et al. 2004; in collaboration with D. Rocancourt, M. Buckingham, Pasteur Institute). We now believe that the genetic regulation of skeletal muscle is governed by anatomical location, and that it likely correlates with embryological patterning. In fact, new mutations at the Myf5 locus combined with MyoD mutant mice (Kassar-Duchossoy et al. 2004) showed that myogenesis can be uncoupled in space and time (Figure 1). We propose that stem cells are born genetically distinct in different regions of the embryo, and that the epistatic relationship between Pax3, Myf5, Mrf4 and MyoD can be modulated in different regions of the developing embryo (Tajbakhsh 2005). Our current work attempts to fine tune this epistasis and correlate the genetic regulation of myogenesis with the skeletal muscle lineage. A hierarchical order which defines cell populations in skeletal muscle was proposed (Tajbakhsh, 2003, 2005), and we have some genetic evidence supporting this view (Kassar-Duchossoy et. al. 2005). Whereas Pax3 plays a predominant role in prenatal myogenesis, Pax7 plays a key role in post-natal satellite cells. We are investigating the role of this gene, in combination with Myf5, in the embryo and in adult satellite cells. We have created new alleles at these loci to address questions concerning cell identity and self-renewal, as well as the role that these genes play in prenatal and postnatal stem and progenitor cells. Notably, prenatal muscle stem/progenitor cells require skeletal muscle as well as continued Pax7 expression for their survival.
Previously we showed that in the absence of Myf5, muscle progenitor cells remained multipotent and were capable of changing their fate in the embryo when located ectopically. We therefore created various knock-in alleles at the Myf5 locus to investigate these events in detail. In addition to investigating the genetic networks regulating skeletal myogenesis, these mice are designed to allow the molecular and cellular characterisation of skeletal muscle stem/progenitor cells. For example, Myf5GFP knock-in mice have permitted us to isolate by FACS post-natal satellite cells for engraftment studies in immuno-compromised recipient mice. Highly efficient engraftments are obtained from several hundred to several thousand satellite cells (in collaboration with A. Cumano, J. Di Santo, Pasteur Institute, and V. Mouly, Institut de Myologie, Paris).
From a developmental and embryological perspective, we are interested also in how cell lineages segregate in the somite. In particular, we are investigating the curious distal rib perturbations which are characteristic of some Myf5 null mice. The issue of the origins of distal rib progenitors - sclerotome vs. dermomyotome is disputed. Given that Myf5 is not expressed in the sclerotome, we have been investigating the causes of distal rib perturbations in these mutants. We have examined a number of alleles at the Myf5 locus and it appears that the regulation of this locus plays an intimate part in the outcome of progenitor cell segregation in the somite.
Identification and characterisation of skeletal muscle stem cells:
Vasily Shinin, Barbara Gayraud-Morel, Ramkumar Sambasivan, Fabrice Chrétien
Recently, we defined cell states characterising myogenic progression from stem to differentiated cells (Figure 2; Kassar-Duchossoy et al. 2005; Tajbakhsh, 2005). Using this model as a guideline, we are examining what events distinguish these cells states. The myogenic regulatory factors (MRFs) act within progenitor cells to produce myoblasts. Pax3 and Pax7 act in the stem cells and give rise to progenitors, at least in the embryo. The molecular and cellular events which guide this myogenic progession to a differentiated cell state are under investigation. During adult muscle regeneration, satellite cells leave the quiescent state, become activated and divide to generate precursor myoblasts which can fuse with regenerating muscle fibres. This model can be mimiced in culture where satellite cell activation and self-renewal can be studied. One strategy which we have used to mark muscle stem cells in the adult involves pulse-chase experiments with BrdU; this approach has previously identified other stem cells and their niche. We find that a subpopulation of satellite cells are label retaining thereby pointing to a heterogeneity in the satellite cell population (Figure 3). Interestingly, some BrdU label-retaining cells segregate the label to only one daughter cell during cell division, in vivo and in vitro. This observation is consistent with the model of old template or "immortal" DNA segregation to the stem cell proposed by Cairns (Nature, 1975).
One mode of stem cell self-renewal is by asymmetric cell division. Our studies using the cell fate determinant Numb reveal that the asymmetry apparatus is operational in satellite cell derived myoblasts (Figure 3), and we have direct time-lapse images of asymmetric cell divisions taking place in this lineage. We plan to extend these studies to the organism and use a transgenic approach to manipulate stem cell fate with a goal to modify cell identity decisions and identify the key regulators involved in this critical decision. Taken together these studies, as well as those indicated above, lead us to propose that some satellite cells have stem cell properties.
Where are skeletal muscle stem cells located, and how do they self-renew ?
Figure 1. Global genetic regulation in skeletal muscle of Pax and MRF genes. Skeletal muscle formation can be spatially and temporally uncoupled using various combinations of genetic mutants.
Figure 2. Identification of Pax3+/Pax7+ myogenic stem cells. By blocking cell fate progression and the formation of foetal muscles in Myf5GFP-P/GFP-P:MyoD-/- mutants, myogenic stem cells, marked here by Pax7 (A), are unveiled. These cells (red) are tightly associated with Desmin-positive differentiated embryonic fibres (green). Blue marks all nuclei (Hoechst staining). These findings lead us to propose a cell fate progression scheme for skeletal muscle (B).
Figure 3. Asymmetric divisions in primary adult muscle cells. A) asymmetric segregation of endogenous Numb protein during mitosis. B) antibody stainings of cell pairs reveal symmetric (top) and asymmetric (bottom) segregation of template DNA strands to daughter cells after pulse-chase labelling with BrdU.
Keywords: stem cells, asymmetric cellular division, skeletal muscle, somites, satellite cells, Myf5, MyoD, Mrf4, Pax3, Pax7, Numb