Experimental investigation of the control and function of the muscle segment homeobox genes during vertebrate embryological development
Collinson, Jon Martin
The control and function of the msh-class (muscle segment homeobox) genes, Msxl and Msx2, was investigated during vertebrate embryonic development. These genes are widely expressed in the embryo, often in structures which are elaborations on the presumed ancestral chordate body-plan. It is thought that duplication events of a single ancestral msh gene, and subsequent divergence early in the vertebrate line, may have given the genes a special role in development of vertebrate-specific structures.The limb was chosen as a model developmental system in which to analyse patterns of Msx gene expression by whole -mount in situ hybridisation in mouse and chicken embryos, to relate these patterns to developmental processes known to be occurring and hence to elucidate potential roles for the genes. Expression of Msx genes in a number of areas of cell death, proliferation and differentiation suggested that they are involved in a wide range of developmental decisions in several tissues, although their molecular role is still uncertain. The potential for functional redundancy between the genes was addressed. There are distinct differences between the expression patterns of Msxl and Msx2 during limb development; these differences are largely quantitative, however, and the broad similarities between the genes suggest that a large degree of functional redundancy may be possible.At all stages of limb development the expression of the Msx genes and the reported expression of bone morphogenetic protein -4 (BMP4) coincide. By implanting beads, soaked in BMP4, into the chick wing in ovo, I showed that BMP4 is an upstream activator of Msxl and Msx2 in the limb. This made it possible to suggest elements of the genetic pathways which may be involved in epithelial-mesenchymal interactions at disparate sites around the body, and to suggest a scenario for the evolution of Msx gene expression patterns.The control of Msxl was further investigated using two reporter genes which express 13- Gal/Msxl chimaeric protein under control of Msxl regulatory sequences. Both reporters transcribe lacZ from 4.7kb of Msxl 5' promoter, but one also contains 7kb of .Msxl sequence 3' of lacZ, including the intron, 3'UTR and 3' genomic sequence.An assay was developed, by means of which the reporter constructs could be introduced into cells in culture by microinjection, then grafted into the developing chicken limb in ovo, using X -Gal staining to visualise expression of the gene. The primary aim was to find sequences in the promoter of Msxl which are responsible for induction of Msxl expression in response to the signal from the apical ectodermal ridge (AER) of the limb.1OT' /2 cells were injected with these constructs; the shorter construct, pH7lacA3', was expressed constitutively, whereas the longer one, pH7lacT, was infrequently expressed, showing that there are elements 3' of the transcription start site which had a negative regulatory effect on the expression of lacZ. The experiments demonstrated that the 4.7kb is almost certainly not the whole of the Msxl promoter; in grafted cells, expression of pH7lacT was not induced in response to the AER. The failure of the AER response could have been due to deficiencies within the promoter or to a failure of the signal transduction pathway within the cells. Radioactive in situ hybridisation showed endogenous Msxl and Msx2 were not induced in 10TV, cells grafted underneath the AER. However, inability to express endogenous Msxl need not preclude expression of the introduced promoter, so further work was necessary to determine whether the reporter is capable of responding to the AER.The behaviour of the 4.7kb Msxl /lacZ reporter gene was therefore further investigated in embryonic tissue from several lines of transgenic mouse which contained the gene. Analysing expression of the transgene in these mice confirmed that the 4.7kb was not the whole of the promoter, although it could reproducibly drive an expression pattern similar to elements of expression of Msxl. Grafting transgenic limb mesenchyme under the AER of chicken limbs did not induce lacZ, providing further evidence that the AER -response element was not in the 4.7kb.The transgene was strongly expressed in the cells of the ventrolateral dermamyotome of the somites which are fated to form limb muscles. Expression of Pax3 was previously the only known molecular marker for these cells. Expression of the transgene is retained in a subset of the developing muscles in the anterior of the limb. Hence the transgene provides a molecular marker for these differentiating myoblasts which can be used in experimental situations, and which I used to show that mouse myoblasts can contribute to chicken muscle.BMP4 could neither induce nor maintain expression of the transgene in cultured limb mesenchyme. The missing AER/BMP4 response element(s) may be one reason why expression of the transgene differs from that of the endogenous Msxl, but it is not the only reason. Although transgenic mouse limb tissue is initially responsive to position -specific signals within the chicken which can maintain its expression, it later becomes refractory to induction by these same signals (at stages when the endogenous gene is still responsive), and non -expressing transgenic mesenchyme does not reinitiate expression when grafted into appropriate areas of the chicken or the mouse.This has the characteristics of an epigenetic silencing effect; it raises the possibility that chromosomal location is important even for regulation of non -clustered homeobox- containing genes, but again suggests that sequences missing from the 4.7kb allow the promoter to fall under the control of repressor proteins which do not affect the endogenous gene.