Final Report Summary - EPAXIALMYF5KO (Single-enhancer knockout analysis of Myf5 function in the developing mouse embryo)
The Myf5 gene is a member of the B-HLH superfamily of transcription factors and it has been implicated in the early specification of skeletal muscle formation during embryonic development.
For the past twenty years, most of the research carried out on this gene has concentrated on how it was regulated from a transcriptional point of view: which enhancer elements operated at what times and anatomical locations.
Classical knock-out experiments have given us a wide overview of the essential function carried out by Myf5 in the process of skeletal muscle development and adult muscle regeneration.
Despite the wealth of information, little is still known about the function of the gene in particular groups of muscle progenitors. Two main questions remain:
I/ Is the function of Myf5 equivalent between the different progenitors that give rise to the large variety of muscles in the adult body?
II/ Which other genes are downstream of Myf5 contributing to the process of muscle formation?
We planned to answer these complex questions by generating a new knock-out mouse strain in which a single enhancer element (the early epaxial enhancer, or EEE), which controls the activation of Myf5 in the dorsal-most portion of the somites, was removed without affecting the expression of the gene in the rest of the developing embryo. This was the starting point for the EU-funded project.
RESULTS DERIVED FROM THE MARIE CURIE INTEGRATION GRANT
1/ Analysis of the expression pattern of Myf5 in the new KO strain: Based on previous BAC-based research, we expected that only the dorsal-most aspect of the expression of Myf5 would be affected. In situ hybridisation of the new strain show this to be the case. The remaining pattern of expression remains unaltered by the deletion of this enhancer.
2/ Identification of genes downregulated in the EEE-KO: To determine if the elimination of Myf5 expression in the dorsal most portion of the somite does have an effect on the expression levels of putative Myf5 downstream targets, we isolated the first six somites from wild-type and EEE-KO embryos. Total RNA was extracted and RNA-seq was performed in duplicate. Dozens of genes were shown to be downregulated in the KO strain compared to the wild-type. In order to reduce the numbers of genes to be analysed, we made use of the g:Ghost Tool (part of the g:Profiler package; http://biit.cs.ut.ee/gprofiler/index.cgi) which allows to interrogate the hand- and in silico-curated gene networks based on functional relationships. From this analysis, we identified three networks that showed variation of expression levels in at least two members of the network.
3/ Networks controlled by Myf5 in the dorsal-most somitic region: From the analysis described above, we identified one network involved in broad transcriptional regulation, one network only associated with myogenesis and a third one related to somitogenesis, myogenesis and chondrogenesis. In total, seven genes belonging to these three networks showed significant differences in expression levels between wild-type and EEE-KO embryos.
4/ Validation of newly identified genes: We decided to use a double validation approach for the identified targets based on quantitative reverse-transcription PCR (qRT-PCR) and in situ hybridisation using mouse embryos between the stages of 8.5 days post coitum (dpc) and 10.5dpc. Of the seven genes identified, six showed significant variation by the qRT-PCR method and four displayed a clearly altered pattern of expression in early EEE-KO embryos (9.5dpc).
5/ The EEE-KO results in the misexpression of sclerotomal markers in the myotome: The myotome is the source of most of the skeletal muscles in the vertebrate embryo. It is located underneath the dermomyotome (which gives rise to dermis and further myotomal cells by delamination) and the sclerotome (which is the origin of vertebrae and ribs). A fourth compartment, known as the syndetome (origin of some tendons), has been proposed to reside between the myotome and the sclerotome. Myogenic markers are normally restricted to dermomyotome and myotome layers while chondrogenic markers are restricted to the sclerotome. In our EEE-KO, we can clearly identify the presence of cells expressing sclerotomal markers within the myotome, indicating that these cells have either adopted a new fate or misinterpreted migrational cues, thus ending in the wrong somitic compartment.
6/ Muscle phenotype of the EEE-KO strain: We did not expect this KO to display major phenotypic features as none of the full Myf5-KO strains have obvious morphological deficiencies. In order to reveal the underlying effect of eliminating Myf5 expression from the dorsal-most somitic compartment, we crossed this new strain with the full MyoD KO, as it is thought that expression of MyoD is responsible for the lack of an overall phenotype in the Myf5 KO strains. Double EEE-KO/MyoD-KO animals died at postnatal day 0, probably by defects on diaphragm function, although a compete analysis of the diaphragm muscle is still underway. We thus decided to concentrate on earlier foetal stages to identify muscles compromised in the double KO. We performed serial sections on 13.5dpc-17.5dpc mouse embryos and could not identify any muscle group absent in the double KO animals indicating that other muscle progenitors, not from the dorsal somitic region, contribute to epaxial musculature. Interestingly, we noted that body muscles (with the exception of facial and limb muscles) had a clearly reduced mass. This indicates that the EEE regulatory element not only contributes to the activation of Myf5 in the dorsal-most portion of the somite but that it must have a secondary function (see point 8 below).
7/ Skeletal phenotype of the EEE-KO strain: Surprisingly, we identified several skeletal malformations in the double MyoD/EEE KO embryos. At 14.5dpc the curvature of the spine shows marked kyphosis (although we believe this to be the result of weak epaxial musculature development), rib cage malformations (wide variation on the number of ribs attached to the sternum, the number of sternum attachment points) and vertebral fusions. As these have been previously associated with Mrf4 KO strains, we extended the analysis to include double KO animals for different combinations of Mrf4, MyoD, EEE-KOs. We show that the kyphosis is exclusive of the EEE/MyoD double KO, as it is the case for vertebral fusions. If we put this together with the data on misexpression of sclerotomal markers in the myotome, we can conclude that the previously reported skeletal phenotypes are probably associated with the lack of Myf5 expression in the somites.
8/ The double function of the EEE element: The most puzzling result for us was the normal development of all muscle groups in the absence of MyoD and Myf5 in the dorsal progenitors coupled with a generalised reduction of muscle mass. We used transgenic analyses to show that the element previously described as the EEE has a dual function; one portion is solely responsible for the activation of Myf5 in the dorsal somitic region, while the second is essential to maintain Myf5 expression levels in all somites at later stages.
For the past twenty years, most of the research carried out on this gene has concentrated on how it was regulated from a transcriptional point of view: which enhancer elements operated at what times and anatomical locations.
Classical knock-out experiments have given us a wide overview of the essential function carried out by Myf5 in the process of skeletal muscle development and adult muscle regeneration.
Despite the wealth of information, little is still known about the function of the gene in particular groups of muscle progenitors. Two main questions remain:
I/ Is the function of Myf5 equivalent between the different progenitors that give rise to the large variety of muscles in the adult body?
II/ Which other genes are downstream of Myf5 contributing to the process of muscle formation?
We planned to answer these complex questions by generating a new knock-out mouse strain in which a single enhancer element (the early epaxial enhancer, or EEE), which controls the activation of Myf5 in the dorsal-most portion of the somites, was removed without affecting the expression of the gene in the rest of the developing embryo. This was the starting point for the EU-funded project.
RESULTS DERIVED FROM THE MARIE CURIE INTEGRATION GRANT
1/ Analysis of the expression pattern of Myf5 in the new KO strain: Based on previous BAC-based research, we expected that only the dorsal-most aspect of the expression of Myf5 would be affected. In situ hybridisation of the new strain show this to be the case. The remaining pattern of expression remains unaltered by the deletion of this enhancer.
2/ Identification of genes downregulated in the EEE-KO: To determine if the elimination of Myf5 expression in the dorsal most portion of the somite does have an effect on the expression levels of putative Myf5 downstream targets, we isolated the first six somites from wild-type and EEE-KO embryos. Total RNA was extracted and RNA-seq was performed in duplicate. Dozens of genes were shown to be downregulated in the KO strain compared to the wild-type. In order to reduce the numbers of genes to be analysed, we made use of the g:Ghost Tool (part of the g:Profiler package; http://biit.cs.ut.ee/gprofiler/index.cgi) which allows to interrogate the hand- and in silico-curated gene networks based on functional relationships. From this analysis, we identified three networks that showed variation of expression levels in at least two members of the network.
3/ Networks controlled by Myf5 in the dorsal-most somitic region: From the analysis described above, we identified one network involved in broad transcriptional regulation, one network only associated with myogenesis and a third one related to somitogenesis, myogenesis and chondrogenesis. In total, seven genes belonging to these three networks showed significant differences in expression levels between wild-type and EEE-KO embryos.
4/ Validation of newly identified genes: We decided to use a double validation approach for the identified targets based on quantitative reverse-transcription PCR (qRT-PCR) and in situ hybridisation using mouse embryos between the stages of 8.5 days post coitum (dpc) and 10.5dpc. Of the seven genes identified, six showed significant variation by the qRT-PCR method and four displayed a clearly altered pattern of expression in early EEE-KO embryos (9.5dpc).
5/ The EEE-KO results in the misexpression of sclerotomal markers in the myotome: The myotome is the source of most of the skeletal muscles in the vertebrate embryo. It is located underneath the dermomyotome (which gives rise to dermis and further myotomal cells by delamination) and the sclerotome (which is the origin of vertebrae and ribs). A fourth compartment, known as the syndetome (origin of some tendons), has been proposed to reside between the myotome and the sclerotome. Myogenic markers are normally restricted to dermomyotome and myotome layers while chondrogenic markers are restricted to the sclerotome. In our EEE-KO, we can clearly identify the presence of cells expressing sclerotomal markers within the myotome, indicating that these cells have either adopted a new fate or misinterpreted migrational cues, thus ending in the wrong somitic compartment.
6/ Muscle phenotype of the EEE-KO strain: We did not expect this KO to display major phenotypic features as none of the full Myf5-KO strains have obvious morphological deficiencies. In order to reveal the underlying effect of eliminating Myf5 expression from the dorsal-most somitic compartment, we crossed this new strain with the full MyoD KO, as it is thought that expression of MyoD is responsible for the lack of an overall phenotype in the Myf5 KO strains. Double EEE-KO/MyoD-KO animals died at postnatal day 0, probably by defects on diaphragm function, although a compete analysis of the diaphragm muscle is still underway. We thus decided to concentrate on earlier foetal stages to identify muscles compromised in the double KO. We performed serial sections on 13.5dpc-17.5dpc mouse embryos and could not identify any muscle group absent in the double KO animals indicating that other muscle progenitors, not from the dorsal somitic region, contribute to epaxial musculature. Interestingly, we noted that body muscles (with the exception of facial and limb muscles) had a clearly reduced mass. This indicates that the EEE regulatory element not only contributes to the activation of Myf5 in the dorsal-most portion of the somite but that it must have a secondary function (see point 8 below).
7/ Skeletal phenotype of the EEE-KO strain: Surprisingly, we identified several skeletal malformations in the double MyoD/EEE KO embryos. At 14.5dpc the curvature of the spine shows marked kyphosis (although we believe this to be the result of weak epaxial musculature development), rib cage malformations (wide variation on the number of ribs attached to the sternum, the number of sternum attachment points) and vertebral fusions. As these have been previously associated with Mrf4 KO strains, we extended the analysis to include double KO animals for different combinations of Mrf4, MyoD, EEE-KOs. We show that the kyphosis is exclusive of the EEE/MyoD double KO, as it is the case for vertebral fusions. If we put this together with the data on misexpression of sclerotomal markers in the myotome, we can conclude that the previously reported skeletal phenotypes are probably associated with the lack of Myf5 expression in the somites.
8/ The double function of the EEE element: The most puzzling result for us was the normal development of all muscle groups in the absence of MyoD and Myf5 in the dorsal progenitors coupled with a generalised reduction of muscle mass. We used transgenic analyses to show that the element previously described as the EEE has a dual function; one portion is solely responsible for the activation of Myf5 in the dorsal somitic region, while the second is essential to maintain Myf5 expression levels in all somites at later stages.