Discussion

The processes regulating the development and maturation of the somites are being slowly unravelled. Whereas the segmentation of the paraxial mesoderm is driven by intrinsic properties (Cooke and Zeeman 1976; Pourquie 2000), the organisation of the same tissue and subsequent differentiation is regulated by extrinsic factors originating from adjacent tissues (Brand-Saberi et al. 1996; Rong et al. 1992). There is a clear distinction between segmentation and somitogenesis. Segmentation represents formation of repetitive portions of the paraxial mesoderm whereas somitogenesis is the epithelialisation of the segmentally organised paraxial mesoderm. Deletion of numerous genes has shown that somitogenesis is not a prerequisite for the differentiation of somitic derivatives (Burgess et al. 1996; Hrabe de Angelis et al. 1997; Johnson et al. 2001). However, although the Paraxis –/– mouse (which fails to form somites) still generates muscle and axial skeleton these tissues are disorganised (fused vertebrae, fused ribs and patterning defects of axial musculature; Burgess et al. 1996). The process of somitogenesis seems therefore to regulate the precise organisation and growth of tissues, allowing them to eventually differentiate in an orderly manner.

This study has identified Wnt6 originating from the dorsal ectoderm as a possible candidate that can regulate the process of somite formation. The role of ectodermal Wnt6 identified in this study seems to be similar to the role played by Paraxis, i.e. they both promote epithelialisation. We showed that removal of ectoderm leads to the loss of Paraxis and that Wnt6 rescues both Paraxis expression and epithelialisation of segmental plate. Paraxis expression is initiated in the segmental plate. Therefore, we suggest that Wnt6 signalling could either initiate or stabilise the expression of Paraxis, facilitating epithelialisation. The above experiments have shown that Wnt6 can influence the epithelialisation of the paraxial mesoderm. Since the application of the Wnt antagonist Sfrp2 interferes with somite epithelialisation, this demonstrates that Wnts play an active role in this process. In the cranial segmental plate, a region marked by the expression of Paraxis, SFRP2 did not totally prevent epithelialisation but led to the formation of small somites. Application of SFRP2 to the caudal, non- Paraxis expressing region of the segmental plate, however, completely inhibited epithelialisation. The size of the Pax3 expression domain was reduced, but levels were normal in the small Sfrp2-induced somites. When somite formation was totally inhibited, Pax3 expression levels were still found similar to the expression of this gene following genetic deletion of Paraxis (Burgess et al. 1995). Furthermore, we show that Wnt6 specifically regulates the expression of genes involved in the epithelialisation process. The role of promoting epithelialisation performed by Wnt6 seems to be unique amongst this family. We have previously investigated the role of Wnt1, Wnt3a and Wnt4 and found that they were involved in regulating myogenesis (Wagner et al. 2000).