Characterization and early embryonic expression of a neural specific transcription factor xSOX3 in Xenopus laevis.

Using the powerful RDA-PCR-technique we could identify a novel Xenopus specific Sox-gene (xSox3) a transcription factor closely related to the sox sub-group B, which contains a HMG box. In normogenesis the xSox3 gene is expressed in the presumptive central nervous system. Furthermore a maternal component is also found in oocytes and in early cleavage stages in the animal hemisphere only. By whole-mount in situ hybridization the first zygotic transcription activities can be detected in the late blastula in the dorsal ectoderm and the dorsal and lateral part of the marginal zone. The expression reaches the highest level atthe late gastrula till the late neurula and fades after stage 30. The expression is restricted from gastrulation onwards to the presumptive brain area and the lens epithelium. Furthermore we could show that the gene is expressed in isolated Spemann organizer with adjacent neuroectoderm. The signal can be suppressed by suramin treatment, which inhibits neural development and causes a shift of dorsal to ventral mesoderm. The treatment of whole embryos with LiCl and UV results in an overexpression or an inhibition of the expression, respectively. In exogastrulae (pseudo-exogastrulae) the gene is expressed in the close vicinity to the endomesoderm only, but not in the distal most part of the ectoderm. This result indicates that it is unlikely that the gene can be activated by planar signals. The gene can also be activated in dissociated gastrula ectoderm without mesodermal or neural inducers. That means that the gene can be expressed in ectodermal cells in a cell autonomous manner.

Partially purified factor from embryonic chick brain can provoke neuralization of Rana temporaria and Triturus alpestris but not Xenopus laevis early gastrula ectoderm.

A high neuralizing activity has been determined in forebrain of 7.5-day old chick embryos using Rana temporaria early gastrula ectoderm as reacting tissue (Mikhailov and Gorgolyuk, Soviet Scientific Reviews, Section of Physiology and General Biology, Vol. 1: 267-306, 1987). The corresponding protease-sensitive agent was extracted, partially purified by chromatography on DEAE-Toyopearl and Heparin-Ultragel columns, and its neuralizing activity was tested in vitro on ectoderm isolated from early gastrulae of R. temporaria, Triturus alpestris, and Xenopus laevis at different concentrations and for different periods of time (animal cap assay). Induction of neural structures was found in R. temporaria and T. alpestris explants (up to 100 and 60%, respectively), but not in cultures of X. laevis ectoderm. Under our experimental conditions, so-called "autoneuralization" of the ectoderm explants can safely be excluded. The results are discussed in relation to the neural competence of amphibian ectoderm and the mechanisms of neuralizing actions of different factors which might be involved in neural induction and patterning.

Close juxtaposition between inducing chordamesoderm and reacting neuroectoderm is a prerequisite for neural induction in Xenopus laevis.

The results of this study indicate that the induction of the central nervous system in Xenopus laevis depends on the close juxtaposition of inducing chordamesoderm and reacting ectoderm, which is necessary for the short distance migration of neural inducing factors. The examination of the neuroectoderm-chordamesoderm interface at intervals of 1 h up to 5 h showed that the onset of neural induction is correlated to the degree of contact formation between ectodermal and mesodermal cells. In the ectoderm cells the number of coated pits, a feature of receptor-mediated endocytosis, is increased. Furthermore there exist telophase bridges between some ectoderm cells, which are possibly correlated to secondary cell interactions.

Neural differentiation of Xenopus laevis ectoderm takes place after disaggregation and delayed reaggregation without inducer.

When Xenopus blastula or early gastrula ectoderm is disaggregated and cells are kept dispersed for up to 5 h prior to reaggregation, the resulting spheres will differentiate into large neural structures. In contrast, dissociated and immediately reaggregated ectoderm will only differentiate into ciliated epidermis (so-called 'atypical epidermis'). Ectoderm treated with mesoderm-inducing XTC-conditioned medium during the period of reaggregation immediately after disaggregation will only form one- or two-cell types (notochord and somites) only. Ectoderm treated with XTC-factor prior to disaggregation will differentiate into a large variety of cell types.

Extracellular matrix components prevent neural differentiation of disaggregated Xenopus ectoderm cells.

Neuralization (archencephalic brain formation) takes place after dissociation and delayed reaggregation of animal caps of early gastrula without inducer (Grunz, H. and L. Tacke: Cell Differ. Dev. 28, 211-218 (1989)). This autoneuralization can be prevented by the cell supernatant from dissociated ectoderm of Xenopus laevis, which contains extracellular matrix components. After phenol extraction of the supernatant, the aqueous phase does no longer show inhibitory activity. It can be concluded from these results that glycoconjugates responsible for the prevention of neuralization represent glycoproteins or proteoglycans which are loosely attached to integral plasma membrane components. Single early gastrula ectoderm cells mixed with non-competent late gastrula ectoderm or endoderm, which primarily form common aggregates, do not differentiate into neural derivatives. In these reaggregates the ectoderm cells remain separated from each other by heterologous cells (non-competent ectoderm or endodermal cells) during the period of competence. These data indicate that the quick recovery of extracellular matrix components together with the restoration of the former organization of the plasma membrane is responsible for the prevention of neuralization.