Search

Embryogenesis requires cell-cell communication in which individual cells send and receive signals that determine their final fate. The transforming growth factor-β (TGF-β) superfamily of molecules encodes secreted factors which are capable of regulating cellular growth and differentiation. Recently, members of the TGF-β superfamily have emerged as candidates for many postulated embryonic inductive interactions. As a step towards understanding such interactions, this thesis investigates the hypothesis that TGF-β superfamily molecules may act as embryonic inducing factors in the mouse. The embryonic expression of a number of TGF-β superfamily members and a putative inhibitor of one of these, follistatin, were investigated. While activin (a mesoderm inducing factor in Xenopus animal caps) was not expressed in the mouse embryo, follistatin, BMP-7, BMP-2 and Vgr-2 were expressed during gastrulation and early organogenesis. Strategies were designed to ectopically express two of these molecules, BMP-7 and Vgr-2 in embryogenesis. Firstly, embryonic stem (ES) cells were used as a vehicle to secrete Vgr-2 throughout the embryo. No effect on embryonic morphology was detected, although conditioned medium from the ES cells is capable of inducing mesoderm in Xenopus. Secondly, COS cells were used to deliver a localised source of BMP-7 to the cranial neural plate in cultured embryos. Such ectopic expression resulted in the induction of ectopic dorsal neural marker expression (Msx1 and AP-2), the diminished expression of a ventral marker (Shh), and a localised expansion of neurectodermal tissue. The localisation of endogenous BMP-7 mRNA during formation and patterning of the neural axis, together with the activities shown in this assay, suggest that BMP-7 may direct one or more aspects of early central nervous system formation in the mouse.

The neural crest is an embryonic population of migratory, multipotent cells that are formed at the boundary of the neural and non-neural ectoderm around the time of neural tube closure. After migrating to locations throughout the embryo the neural crest cells differentiate into many different cell types including osteoblasts and chondrocytes of the cranio-facial skeleton and neurones and glia of the peripheral nervous system. Gain of function experiments demonstrate that many genes, including the Zic gene family, are involved in neural crest formation. In order to determine whether they have a role in endogenous neural crest development, loss of function studies are required. The ability to perform genetic analysis in the mouse makes it an ideal organism in which to do this. When this study commenced Zic1, Zic2 and Zic3 were assigned as members of the mammalian Zic gene family. A comparison of their expression patterns during mouse neural crest development reveals Zic2 and Zic3 to be the most likely to have roles in neural crest formation. Analysis of neural crest induction and migration in loss of function alleles of Zic2 and Zic3 reveals that loss of either gene function alone causes both a delay in the onset of trunk neural crest production and a reduction in the number of neural crest progenitor cells. Additionally, loss of Zic2 gene function results in a heightened response to BMP signalling by the neural tube. A more severe neural crest depletion occurs in embryos lacking Zic3 and heterozygous for a mutation in Zic2. These results indicate that Zic2 and Zic3 cooperate in the formation of the neural crest by mediating the competence of the dorsal neural tube to respond to inductive signals. This work provides the first genetic evidence that Zic2 and Zic3 are involved in neural crest development and that they function together during mouse development.