Isolated two dorsal animal blastomeres of 8-cell-stage Xenopus embryos differentiated in about 24% of the cases into mesodermal structures, while the two ventral animal ones formed exclusively atypical epidermis. Of special interest is the fact that most of the dorsal animal blastomeres without mesodermal structures, differentiated into atypical epidermis with large parts of cement gland. Cement glands could not be detected in the derivatives of the ventral animal blastomeres. In earlier concepts it has been suggested that the animal hemisphere of the amphibian egg is an uncommitted area, which receives its instructions for further determination and differentiation from the vegetal part of the embryo. However, the results, shown in this paper, support the view that the developmental determinants are distributed in distinct gradients already in the early cleavage stages. Although the highest concentrations of these putative determinants are located in the dorsal vegetal area (Melton, 1995) and the dorsal marginal zone, lower concentrations of these substances are thought to be present in the animal part, especially in the dorsal animal hemisphere of the egg and the early embryo. Our results indicate that dorsal animal blastomeres of early cleavage stage embryos possess the capacity to form mesodermal derivatives after their separation from the vegetal hemisphere.
INTRODUCTION
Fate maps of the early embryo are very important for the study and understanding of the mechanisms at the cellular and molecular level during early embryonic development. Many informations are now available about the formation of the animal/vegetal and dorsal/ventral polarity during early embryogenesis resulting in the establishment of a rather complex body axis in the larval stage (Grunz, 1993a; Dawid and Taira, 1994; Lamaire and Gurdon, 1994). Different methods were used to study these mechanisms and processes, i.e. dissection, cell lineage techniques with different tracers or UV-irradiation, etc. (Ancel and Vintemberger, 1948; Nieuwkoop, 1969a,b; Gimlich and Gerhart, 1984; Gimlich, 1986). The classical vital dye staining method of Vogt (1929) could recently be extended by powerful microdisection and cell lineage techniques (Nakamura and Kishiyama, 1971; Hirose and Jacobson, 1979; Jacobson and Hirose, 1981; Dale and Slack, 1987; Moody, 1987; Moody and Kline, 1990; Keller, 1991). It could be shown that rRNA is synthesized in differnt rates in cells derived from the dorsal and ventral blastomeres of 4-cell stage-embryos (Shiokawa and Yamana, 1979). Circumstantial evidence suggested that determinants for endoderm and mesoderm are distributed in animal-vegetal and dorsal-ventral gradients (Nakamura et al., 1971; Tiedemann, 1975; Nakamura, 1978; Grunz and Tacke, 1986; Takasaki, 1987; Tannahill and Melton, 1989; Gallagher et al., 1991; Köster et al., 1991; Hainsky and Moody, 1992; Tiedemann et al., 1993; Grunz, 1992, 1993a, b; Rebagliati and Dawid, 1993; Dohrmann et al., 1993).
In earlier papers it could be shown that the animal blastomeres of eight cell stage embryos will differentiate in high percentages into dorsal mesodermal structures (Grunz, 1977, 1994). The results indicate that the early amphibian embryo contains maternal factors which are distributed in form of gradients rather than in form of distinct centers. Although there is no doubt that certain factors are mainly located in the vegetal half of the egg, it cannot be excluded that the same factors are also present in low concentrations in the marginal zone and the animal half. Recently it could be shown that Vg1 biological active processed from the inactive precursor into the mature protein. The mature molecule is presumably located in a restricted area only, i.e. the dorsal vegetal zone (Thomsen and Melton, 1993; Melton, 1995). Until now it is not clear in detail how the gradients are formed, how certain genes are expressed in a specific spatial/temporal manner and how the complicated body axis develops from the relatively simple organized egg. Two main mechanisms are most important: 1. the presence of maternal factors including transcription factors and growth factors (determination factors, embryonic inducers) and 2. complicated cell-to-cell interactions including embryonic induction during cleavage, gastrulation and neurulation. There exists a distinct animal/vegetal polarity of the egg before insemination. After fertilization correlated with cortical rotation certain factors will be shifted preferentially into the dorsal zone opposite to the sperm entry point suggesting that mesodermal determinants will also be included in the animal part of the egg (Vincent and Gerhart, 1987; Fujisue et al., 1993). By this event a dorsal/ventral polarity will be established in addition to the animal/vegetal polarity, in some species indicated by the so-called grey crescent. The animal cap of later stages (after midblastula transition) differentiates into ciliated (so called atypical epidermis) only. However, there exist observations that the dorsal ectoderm is more susceptible to inducing signals than the ventral ectoderm (Sokol and Melton, 1991). This suggests that the dorsal ectoderm contains factors, which are not present in the ventral ectoderm. Recent experiments with Rana indicate that there exist no strict boundaries between the presumptive germ layers ectoderm, mesoderm and endoderm in the 32-cell stage. In contrast there exist overlapping zones in the area of the animal hemisphere and the marginal zone (Saint-Jeannet and Dawid, 1994).
By our earlier papers (Grunz, 1977, 1994) it could be shown that the four animal blastomeres of the 8-cell-stage are able to differentiate into mesodermal structures. However, it was not clear if the dorsal or the ventral animal blastomeres contributed preferentially to these mesodermal structures. Therefore in this article we present data, which show that the dorsal animal blastomeres are more potent to differentiate into mesodermal structures than the ventral ones.
MATERIALS AND METHODS
Embryos of Xenopus laevis were obtained by artificial insemination after injection of female frogs with 1000 IU human chorion gonatropin (HCG, Schering AG, Berlin). They were dejellied with 3.5% cysteinium chloride (pH 7.4) at the 8-cell stage, and carefully washed several times in Holtfreter solution. Strong variations in the cleavage pattern exist in different spawnings of Xenopus. The frequencies of “regular” cleavage (described in the textbooks) is sometimes very low. The results reported here are from those series of experiments in which the frequencies of the regular cleavage pattern was about 50% of the total number of embryos at 8-cell stage. Only embryos with stereotypic radial (regular) cleavage pattern were used for the experiments (see Grunz, 1977, 1994). The two-dorsal-animal-half (2DA) blastomeres of these embryos were easily distinguished from the two-ventral-animal-half (2VA) blastomeres by the smaller size and lighter pigmentation. These dejellied, “regular” embryos were transferred into 50% Leibovitz-Medium (L-15; pH 7.4) supplemented with 10% fetal calf serum (FCS) on petri dishes coated with 1% agar. In earlier reports it could be shown that FCS has no inducing capacity under our experimental conditions (Grunz, 1977, 1994). The vitelline membrane was manually removed with sharpened watch-makers forceps. First the four-animal blastomeres were isolated from the four-vegetal ones as described elsewhere (Grunz, 1977, 1994) and Figure 1a, b. Then the 2DA-blastomeres were separated from the 2VA-blastomeres (Fig. 1c,d) with fine glass needles with ball-like melted tips. These 2DA and 2VA-blastomeres were transferred into another agar-coated dish containing 50% L-15 medium (plus 10% FCS) and cultured at 20°C. The possible debris were discarded timely all along the incubation process. The L-15 was replaced stepwise by Holtfreter solution (pH 7.4) after 4 hours incubation when the blastomeres divided several times and formed round spheres. Explants were cultured for 3 days at 20°C prior to the fixation in Bouin's fixative, when normal embryos have reached stage 42-46 (Nieuwkoop and Faber, 1967). After gross staining in borax-carmine, the tissue was dehydrated and embedded in paraffin. Section (6 μm) were stained with aniline-blue-orange G.
RESULTS
During the first hours of culture, the 2 DA- and 2 VA-blastomeres divided quickly. After 3 days by macroscopic observation over 20% of the cases of the 2DA differed significantly from the 2VA explants. The 2DA explants had formed vesicles and protrusions, while all 2VA explants have differentiated into wrinkled atypical epidermis only.
Differentiation of the two ventral animal blastomeres (VA)
From a larger number (about 70) of isolated 2VA blastomeres 44 survived in good conditions for 3 days at 20°C. Most of them (93%) developed exclusively into atypical epidermis (Fig. 2, Table 1). No dorsal mesoderm were found in the 2VA blastomeres derived explants. Only one case (3%) differentiated into ventral mesoderm (coelomic epithelium), mesenchyme and normal epidermis. In 5% the explants formed yolk-rich derivatives without any differentiations. Of special significance is the fact that in contrast to the DA-series no cement glands were found in the explants of this series.
Differentiation of the two dorsal animal blastomeres (DA)
Sixteen explants (33% of the cases) differentiated exclusively into atypical epidermis (Table 1). Twenty two (46%) differentiated into atypical epidermis with large cement glands. A typical representative is shown in Fig. 3a. In 17% of the cases the explants contained normal epidermis. Mesenchyme and coelomic epithelium (ventral mesoderm) was found in 17% of the cases. Four of these cases have also formed heart-like structures (Fig. 3b). Dorsal mesoderm (notochord) differentiated in 10%, muscle (somites) in 8% of the cases (Table 1, Fig. 4a). Neural tissue was found in 14% of the cases (eye [tapetum] 6%, brain 10%; Table 1, Fig. 4b). it must be pointed out that the cement glands were relatively large (Figs. 3a, 4a, b).
Table 1.
Differentiations of the two dorsal or two ventral animal blastomeres of eight-cell-stage embryos
DISCUSSION
In this paper we could show that the two dorsal animal blastomeres of the 8-cell-stage differentiated in over 24% into mesodermal structures. On the other hand all explants, derivatives of the two ventral blastomeres, differentiated into atypical epidermis only. Traditionally the animal hemisphere of the amphibian egg has been considered as an uncommit-ted zone, which exclusively receives its signals for differen-tiation from the vegetal part of the egg. This view was sup- ported by the fact that animal caps isolated from middle blastula or early gastrula stages differentiated into so called atypical (ciliated) epidermis only (Grunz et al., 1975). There-fore the animal caps were used as an excellent test system for inducing factors, for example the vegetalizing factor, identical with activin (Grunz, 1983; Tiedemann, 1986; Knöchel et al., 1987). However, it could be shown that in the blastula stage the animal cap is not an homogeneous entity (Sokol and Melton, 1991). The dorsal ectoderm reacts much more sensitive to inducing factors than the ventral ectoderm. In our earlier studies we could show that the animal blas-tomeres of the 8-cell-stage cannot be considered as an ex-clusively permissive part of the egg, receiving all its informa-tion from the vegetal part of the egg (Grunz, 1977,1994; Kageura and Yamana, 1983). In a large number of cases the four isolated animal blastomeres formed mesodermal struc-tures instead of atypical epidermis. Recently it could be shown that in the 32-cell-stage of Rana pipiens there exist no deterministic boundaries between the blastomeres. Most of the blastomeres give rise to multiple tissues (Saint-Jeanett and Dawid, 1994). The data above show that at least the dorsal animal blastomeres are able to differentiate into mesodermal structures, although the whole animal hemi-sphere formerly has been considered as an uncommitted area, which will form atypical epidermis only without instruc-tions from the marginal part of the egg. This indicates that dorsal blastomeres in contrast to their ventral counterparts contain after the third cleavage (8-cell-stage) components, which allow their differentiation into mesodermal structures. On the other hand Kinoshita et al. (1993) could show that single dorsal animal or single vegetal blastomeres do not au-tonomously differentiate into mesodermal structures. The absence of mesodermal structures (Kinoshita et al., 1993) and the lower percentage of mesodermal derivatives than in the earlier papers (Grunz, 1977, 1994) can be explained by the smaller initial cell mass. in the earlier experiments (isola-tion and cultivation of all animal blastomeres of the 8-cell-stage) the two ventral animal blastomeres (VA) could sup-port the mesodermal differentiation of the two dorsal animal blastomeres (DA). The community effect and the larger initial cell mass could be favourable for the mesodermal differen-tiation (Grunz, 1979; Gurdon, 1988). In the two isolated VA-blastomeres alone or single DA- or DV-biastomeres the threshold concentrations are apparently too low to favour a mesodermal pathway of differentiation. This view is sup-ported by the fact that all isolated two ventral animal blas-tomeres differentiated into atypical epidermis. Under in vivo conditions (normogenesis) mesodermalization of the animal part of the egg may be prevented by inhibitory molecules located in the ectoderm [presumptive epidermis and neural tissue] (Kato and Gurdon, 1994). Follistatin-like proteins could be good candidates for such mechanisms (Hemmati-Brivanlou et al., 1994). After separation of the animal blas-tomeres from the rest of the embryo the activation of mesodermalizing factors may be enhanced in several cases resulting in the differentiation of mesodermal derivatives. On the other hand the dissociation of animal pole cells at blastula and gastrula stages causes the neuralization of the cells (Grunz and Tacke, 1989). The reason could be the dilu-tion or elimination of mesodermalizing factors like BMPs from the single cells leading to the neuralization of the re-aggregated ectoderm material (Grunz and Tacke, 1990; Wil-son and Hemmati-Brivanlou, 1995). These results indicate that mesodermalizing factors may be present in very low concentration or as unprocessed precursors, which prevent the ectoderm from its differentiation into neural tissue, the postulated default state of ectoderm. In normogenesis neuralizing factors may act like competitive inhibitors by blocking BMP-receptors (Xu et al., 1995) and signal transfer of BMP-like morphogens resulting in the shift of the pre-sumptive epidermal state of ectoderm into the neural path-way of differentiation.
Acknowledgments
We thank Sabine Effenberger for skilful assistance. This investi-gation was supported by the Deutsche Forschungsgemeinschaft and in part by the Forschungspool of the University GH Essen. The stay of Li Shuhong, Institute of Developmental Biology, Chinese Academy of Sciences, Beijing, China in our lab was supported by Akademisches Auslandsamt with a grant from the Ministerium fÅúur Wissenschaft und Forschung des Landes Nordrhein-Westfalen.