The formation of skeletal muscle [Ted Chang, Philippe Daubas, Juliette Hadchouel, Jacqueline Perreau, Didier Rocancourt, Ralf Spörle, Shahragim Tajbakhsh collaborations with the laboratories of G. Cossu, P. Maire, T. Partridge, M. Polimeni, P. Rigby]
In 1997 we showed that Myf5 and Pax3 intervene upstream of MyoD in the genetic hierarchy that regulates the onset of myogenesis. Our current research focuses on the regulation, function and manipulation, with the introduction of other coding sequences, of these genes.
We have analysed Myf5 regulation with YAC transgenes which carry large fragments of DNA from around the locus. Several regions within 96 kbp upstream of the gene have been identified which are necessary for the complete spatio-temporal pattern of expression in the mouse. These include several enhancers containing sequences which specifically target different sites of myogenesis. The fine analysis of the expression of Myf5 and other marker genes has led to the identification of subdomains in the somite which are also of interest in evolutionary comparisons with other species. Another unexpected aspect of the expression of Myf5 is its transcription in certain domains of the central nervous system where the protein is not detectable and there is no apparent mutant phenotype. Surprisingly a specific Myf5 regulatory sequence targets these sites.
The first targetted mutation in Myf5, with introduction of the nlacZ reporter, permitted us to analyse the mutant phenotype in the embryo, but not posnatally because of a rib defect which is lethal to respiration at birth. This defect probably results from an effect in cis on another gene in the Myf5 locus. A mutant allele with a different configuration now avoids this problem, giving viable homozygotes. The analysis of adult muscle fibres reveals a role for Myf5 in the proliferation of the satellite cells which are associated with fibres and essential for muscle regeneration. The fact that Myf5, visualized by the expression of an Myf5-nlacZ allele, marks these cells, led to the identification, in conjunction with other markers, of a subpopulation which may have stem cell properties.
Given the complexity of the Myf5 locus which contains the MRF4 myogenic factor gene, as well as other possible coding sequences, we are now making precisely targeted mutations in MRF4 and Myf5 which should avoid secondary effects which may complicate interpretation of the mutant phenotype. We have also introduced the GFP reporter into an allele of Myf5 in order to follow the movement of muscle precursor cells in vivo. Other genetic manipulations include the introduction of the MyoD coding sequence in place of Myf5 with the aim of comparing the function of these two determination factors in the absence of differences in their expression, in the embryo and adult.
The Pax3 gene is also regulated by a complex combination of DNA sequences which target its expression to non-muscle as well as muscle precursors. The introduction of nlacZ and alkaline phosphatase reporter sequences in Pax3, permits us to follow the cells which express the gene in normal and mutant mice. We observe, unexpectedly, that cells in adult muscles are positive, suggesting that Pax3 like its orthologue Pax7, may play a key role in myogenesis in the adult as it does in the embryo.
In the regulatory cascade that leads from Pax3 to MyoD and the activation of myogenesis, other transcription factors and co-factors have recently emerged as potential players, such as the Six family of homeoproteins. We have targetted an allele of Pax3 with a dominant negative Six (Six4) sequence in order to investigate a role for these proteins prior to the onset of skeletal muscle formation.
We are interested in the signalling pathways which control the initiation of myogenesis. The sequences whcih regulate the Myf5 gene are being studies in this context. In order to determine whether Notch signalling is important in the specification of muscle cells we have targetted an allele of Pax3 and Myf5 with a constitutively active intracellular form of Notch. Perturbations in myogenesis are particularly evident with the Pax3 allele which also affects neural crest derivatives.
Cardiogenesis [Robert Kelly, Marguerite Lemonnier, Sigolène Meilhac, Andrew Munk collaborations with the laboratories of N. Brown and J.-F. Nicolas]
The regionalized expression of certain transgenes which are expressed in cardiac as well as skeletal muscle, provided us with an interesting tool with which to follow different subdomains of the myocardium. Thus in explants of the early cardiac tube we were able to identify a time frame when left/right atrial identity is still flexible whereas that of the ventricule is already acquired. We have constructed adenovirus vectors expressing different signalling molecules and factors which enable us to manipulate this acquisition of identity on the left/right and potentially also on the anterior/posteriorl axis.
In most cases regulatory sequences of the transgenes are responsible for their regionalized transcription in the heart. We are trying to define these sequences. In vitro manipulation of cardiocytes prepared from different regions of the myocardium, shows that the expression of the transgene becomes fixed early during development, probably because of chromatin conformation and/or DNA modification.
One transgenic line, where the expression is due to an integration site effect, has provided insight into the unexpected contribution of anterior mesodermal cells to the myocardium of the outflow tract. DiI labelling experiments confirm that cells in the anterior heart field are recruited into this part of the developing heart. The transgene has integrated upstream of the Fgf10 gene and its expression suggests that Fgf10 is implicated into this process. The identification of an Fgf10 enhancer element which may be regulating the transgene, together with the analysis of heart formation in the absence of Fgf signalling are in progress.
The question of the origin of the cell populations which colonise different parts of the developing heart is also being addressed by a retrospective clonal analysis. We have targetted the cardiac actin gene with an nlaacZ reporter sequence in which a rare combination event will lead to a functional nlacZ and consequent labelling of a cell expressing the cardiac actin allele with b-galactosidase. The analysis of clones in the embryonic heart visualised in this way already shows that the pool of founder cells is large. The comparison with skeletal muscle, also analysed with this transgene, shows interesting differences.