- Email: [email protected]
Factors involved in the formation and stabilization of cell aggregates obtained from amphibian embryonic explants
Cell Differentiation, 23 (1988) 69-76 Elsevier Scientific Publishers Ireland, Ltd.
69
CDF 00481
Factors involved in the formation and stabilization of cell aggregates obtained from amphibian embryonic explants Mats-Olof Mattsson, Huguette Lovtrup-Rein and Soren Lovtrup University of Ume&,Department of Zoophysiology, S-90I 87 Umed, Sweden (Accepted 29 September 1987)
The effect of factors influencing the formation and stability of animal and vegetal aggregates from
Xenopus iaevis and Ambystoma mexicanum was examined in the light and scanning electron microscopes. At extreme values of pH the surface coat covering the vegetal aggregates is dissolved and dissociation may take place. Animal aggregates are more resistant. At high tonicities vegetal aggregates may be dissociated, and in the animal aggregates the epidermal differentiation is suppressed. In the absence of Ca~+ the vegetal aggregates are dissociated, but the animal aggregates are not affected. The results obtained with the inhibitor selenate and from incorporation experiments indicate that sulfated glycosaminoglycans are involved in the formation of aggregates in both species. Corresponding observations with tunicamycin suggest that even glycoproteins may play a role in aggregate formation, particularly in the vegetal aggregates. Aggregate formation; Amphibian; Surface coat; Glycoprotein; Sulfated glycosaminoglycans
Introduction The existence of surface layers of varying thickness and composition in protozoans and in various marine animals has been known or at least surmised for a long time (cf. Martinez-Palomo, 1970; Luft, 1976). Holtfreter (1943, p. 266), claimed that in early amphibian embryos the cells are kept together by such a surface layer, a surface coat, which 'amalgamates the peripheral cells into a plastic super-cellular unit'. With the rise of electron microscopy and improved methods of fixation and staining, it has been possible to demonstrate the existence of a Correspondence address: M.-O. Mattsson, University of Ume~, Department of Zoophysiology, S-901 87 Ume~t, Sweden.
cellular surface coat as an almost ubiquitous phenomenon. Various lines of evidence suggest that this surface coat always contains carbohydrate. In fact, the name 'glycocalyx' was long ago proposed for this element of the cell cortex (Bennet, 1963). Aggregate formation by dispersed cells involves two aspects in which the participation of microfilaments may be envisaged: (1) the agglomeration of the individual cells and (2) the stabilization of the aggregates thus formed. It may be presumed that the former event is accomplished by microfilamentous filopodia. Microfilaments may also be involved in the stabilization or maintenance of aggregates, as suggested by the fact that cytochalasin may cause dissociation of aggregates of embryonic cells (Perry, 1975). Apart from these microfilaments, certain 'extracelhilar' factors, no-
0045-6039/88/$03.50 © 1988 Elsevier Scientific Publishers Ireland, Ltd.
70 tably glycoproteins and glycosaminoglycans (GAG), have been suspected to be involved in the formation of cell aggregates (see for instance Luft, 1976; Letourneau et al., 1980). When small explants are isolated in vitro from the blastula of either Ambystoma mexicanum or Xenopus laevis, the formation of two different kinds of cell aggregates may be observed (Lavtrup, 1983). Endodermal cells from the proximity of the vegetal pole or mesodermal cells from the interior marginal zone form more or less irregular aggregates of 'undifferentiated' cells (from here on called 'vegetal aggregates'). These aggregates, when observed in the scanning electron microscope, are seen to be covered by a surface coat. In contrast, ectodermal cells, excised from the animal hemisphere (' animal aggregates'), differentiate into vesicular aggregates made up of ciliated epidermis. In these there is no visible surface coat, but instead ridges are seen to border adjacent cells. In the present study we have attempted to establish possible differences in the nature of the extracellular substances which participate in the formation and maintenance of the two different kinds of aggregates in the amphibian species mentioned above. For this purpose we have investigated the effect of factors which may be expected to dissolve or otherwise interfere with the substances supposed to maintain the aggregates. We have further studied the effect of substances known to inhibit the synthesis of glycoproteins and sulfated G A G . In order to interpret the results thus obtained, it was necessary to investigate the synthesis of these compounds by means of isotope incorporation.
Material and Methods
Naturally spawning Ambystoma mexicanum provided the urodele embryos. In Xenopus laevis spawning was induced by injecting chorion gonadotropin (Pregnyl, N.V. Organon, Oss, The Netherlands): 500 U per female, 250 U per male. The explants were taken from the animal or vegetal regions of late blastulae. The jelly surrounding the embryos was removed either by mechanical dissection (Ambystoma) or, in the case
of Xenopus, by short exposure to thioglycolic acid (1%) in 10% modified Stearns' solution (MSS) (Nakatsuji and Johnson, 1982), p H 7.8-8.0. The explants were obtained and cultured according to the methods described by Lovtrup and Perris (1983), with the exceptions that we used undiluted MSS, p H 7.4-7.5, to which benzylpenicillin (100 U / m l ) and streptomycin (100 /~g/ml) were added, while serum albumin was omitted. The effects of p H were studied in the interval of 2.4-11.9 during 10 rain to 1 h. The pH was adjusted with HC1 or N a O H to the desired value. For testing the importance of the tonicity of the medium we used MSS corresponding to a 0.5. 0.75, 1.5, 2.0 and 2.5-fold tonicity of the standard solution. The control medium had an osmolarity of 177 mOsm, with a NaCI concentration of 75 mM (150 mOsm). The other tonicities were obtained by adjusting the NaC1 concentration to 30 m M (87 m O s m total osmolarity), 52.5 mM (132 mOsm), 118 m M (263 mOsm), 163 mM (353 mOsm) and 208 m M (443 mOsm). The aggregates were incubated in the respective media until the animal controls became ciliated (24 h for Xenopus, 3 days for Arnbystoma). The effect of the Ca 2+ concentration was studied with MSS free of Ca 2+ or containing 5 /~M, 50 #M, 0.2 mM, 0.4 mM, 0.8 mM and 1 mM Ca 2+, respectively. After one day (Xenopus) or 3 days (Ambystoma) in control medium with 2 mM Ca 2+, the aggregates were exposed to these different media from 15 min up to 2 days. The antibiotic tunicamycin (Sigma) was added to the standard medium at a concentration of 5 /~g/ml. Na-selenate was added to the cultures in concentrations of 10 and 30 raM. Both substances were added from the beginning of the cultures. The experiments were observed and documented in a scanning electron microscope (SEM) or in an inverted microscope with camera attachment. For electron microscopy the specimens were fixed overnight at 4 ° C in 2.5% glutaraldehyde in 0.l M phosphate buffer or 0.1 M cacodylate buffer. After rinsing in buffer the specimens went through a graded series of ethanol and transferred to a critical point drying apparatus for further drying. The specimens were coated with a gold layer
71
(900-1000 A) and studied in a Jeol JSM-P15 electron microscope with a beam voltage of 15 kV. For isotope experiments, explants were cultured as described above in the presence of 0.50 /~Ci/ml L-[6-3H]fucose, with or without 5 / ~ g / m l of tunicamycin; or 5 /~Ci/ml of 35SO42+; or 0.15 t t C i / m l D-[1-14C]galactosamine hydrochloride plus 0.15 / t C i / m l D-[1-14C]glucosamine hydrochloride. The samples were sonicated, digested with pronase, and loaded on a Sephadex G-50 column after delipidation with methanolchloroform. Radioactivity in the fractions containing glycoproteins was determined in a scintillation counter. The GAG, emerging in the excluded volume, were further purified by ion exchange chromatography on a Whatman DE-52 column. Fractions containing G A G were used partly for direct determination of total incorporation, partly for determination of chondroitin sulfate and heparan sulfate c o n t e n t by enzymatic d e g r a d a t i o n (Lovtrup-Rein, in preparation).
Results
The effect of pH Holtfreter (1943) found that elevation of pH to above 10 will cause dissolution of the surface coat and disaggregation of the peripheral cells in the embryo. Working with very small aggregates it is necessary to employ much more extreme p H values, acid as well as alkaline, before any effects on the state of aggregation are observable. When effects were observed at the p H values listed in Table I, further attempts to dissociate the aggregates were given up. In Ambystoma the exposure of vegetal aggregates to changes of p H in both directions leads to dissolution of the surface coat and to partial dissociation. Elevation of the p H is much more efficient than lowering, a finding which corroborates earlier observations by Holtfreter. The vegetal aggregates from Xenopus are more resistant. The animal aggregates from both species are more susceptible to low values of pH, but are very resistant to alkaline media (Fig. l a - e ) .
The effect of tonicity Holtfreter observed a tendency towards dissociation when embryos were reared at high tonicities. When the tonicity of our media was raised 2-2.5 fold, the formation of the surface coat was suppressed in Ambystoma vegetal aggregates, whereas in the animal aggregates no epidermal differentiation (ciliation) took place (Fig. 2a). The Xenopus aggregates were more susceptible to changes in tonicity; thus, in both types of aggregate, dissociation or lack of aggregate formation was observed when the tonicity was twice the normal or higher (Fig. 2b-e). In both species it was found that when the tonicity of the medium was half of that in the controls, epidermal differentiation was favoured, manifested by a higher percentage of ciliated vesicles. No effect was observed on the vegetal aggregates. The results of incorporation of aminosugars and sulfate at high and low tonicities are given in Table II. In both species, the synthesis of G A G was reduced as compared with the controls, more extensively in Ambystoma than in Xenopus. However, among the two G A G that were synthesized by the aggregates, namely chondroitin sulfate and heparan sulfate, the production of the latter was almost completely suppressed in aggregates from Xenopus at high tonicities. It is interesting, and perhaps of significance, to observe that in Am-
TABLE I The effect of pH changes
A mbystoma Animal pH 3.0 1 h aggregates Partial dissociation Cells cytolyzed pHll.41h No visible effect Vegetal pH 2.4 1 h aggregates Partial dissociation pH 11.4 20 rain Dissolution of surface coat Partial dissociation
Xenopus pH 3.0 1 h Partial dissociation Cells cytolyzed pHll.91 h No visible effect pH 3.0 1 h Partial dissociation Cells cytolyzed pH 11.5 1 h No visible effect
Deviations from neutrality less than those reported here did not have any visible effect.
72
Fig. 1. (a) Animal control aggregate from Xenopus. Note the ciliation and epidermal ridges (ri). A corresponding Amb),stoma control is similar to this aggregate. Bar, 50/xm. X 280 in a f. (b) The corresponding vegetal control. A surface coat obscures the cell borders. (c) Animal aggregate from Ambvstoma exposed to pH 2.4 for I h. Cells are cytolyzed and the epidermal characteristics are missing. (d) Vegetal aggregate from Amt~vstorna exposed to pH 2.4 for 1 h. Surface coat is partly absent. (e) Ambystoma vegetal aggregate exposed to pH 11.4 for 20 min. Surface coat is totally absent and the aggregate starts to dissociate. (f) Xenopus vegetal aggregate in Ca2+-free medium for 15 min. Surface coat is absent and the aggregate is starting to dissociate, a process which is completed within 1 h.
73
bystoma the degree of sulfation was n o t affected by tonicity, b u t in Xenopus the G A G were u n d e r sulfated at high a n d oversulfated at low tonicities.
The effect of Ca 2 + O b s e r v a t i o n s have shown that Ca 2+ is often necessary for the f o r m a t i o n a n d stabilization of cell aggregates, a n d Holtfreter's o b s e r v a t i o n s
(1943) suggest that C a 2+ may even be necessary for the integrity, of the a m p h i b i a n surface coat. In the present study the vegetal aggregates from b o t h species u n d e r w e n t partial dissociation when kept in a Ca2+-free m e d i u m for 1 h. S c a n n i n g electron m i c r o g r a p h s showed that the surface coat had d i s a p p e a r e d (Fig. lf). N o dissociation was observed when Ca 2+ was present in the m e d i u m ,
Fig, 2. (a) Animal aggregate from Ambystoma exposed to high tonicity, corresponding to 1.5 times the normal. The ciliation is suppressed, but otherwise the aggregate is intact. Bar, 20 #m in both a and b. × 290. (b) Animal aggregate from Xenopus exposed to a tonicity corresponding to 2.0 times the normal. The explant is partly disaggregated, many ceils are cytolyzed and no 'normal' features are discernible, x 450. (c-e) A series of vegetal Xenopus aggregates exposed to tonicities corresponding to 1.0, 1.5 and 2.0 times the normal, respectively. A progressive disaggregation is observable ending up with a total dissociation in e. y = yolk platelets. Bar, 50 p.m. x150.
74 TABLE II Rates of incorporation of radioisotopes into the GAG of animal and vegetal explants at different tonicities, compared with controls at normal tonicities 14C-labelled amino sugars
[35S]sulfate
Animal
Vegetal
Animal
Vegetal
15% 50%
15% 50%
17% 8%
17% 8%
19% 57%
20% 57%
22% 82%
22% 82%
High tonicity
Ambystoma Xenopus Low tonicity
Ambystoma Xenopus
not even at the lowest concentrations. No effect was observed when the animal aggregates were treated similarly.
The effect of tunicamycin If glycoproteins are involved either in the formation of the surface coat or as cell adhesives, we should expect tunicamycin to affect aggregate formation of the embryonic cells. We found that at a concentration of 5 /~g/ml, tunicamycin had no effect on vegetal aggregates from Ambystoma. In the corresponding Xenopus aggregates no surface coat was formed. The formation of animal aggregates was suppressed in both species. Synthesis of glycoproteins, as evidenced by the incorporation of fucose (Table III), took place in the vegetal aggregates from both species and in TABLE III Rates of incorporation of [3H]fucose into glycoproteins of animal and vegetal explants (cpm//~g protein, _+standard error of mean)
Ambystoma
Xenopus
Animal Vegetal Animal Vegetal Controls
4430 3 715 910 5 395 (_+310) (_+390) (_+105) (_+305)
Tunicamycin-treated
565 (_+60)
305 685 480 (_+38) (_+67) (_+42)
Number of experiments 5
5
4
4
Ratio control vs. treated 7.84
12.18
1.33
11.24
the animal aggregates from Ambystoma, and was completely suppressed by tunicamycin. Little incorporation of fucose could be observed in animal aggregates from Xenopus.
The effect of selenate If Na-selenate was added to the culture medium, no obvious effect was observable on the vegetal aggregates from Ambystoma, but in 10 mM selenate this type of aggregates from Xenopus remained completely dissociated. The animal aggregates from both species were also affected, and in much the same way; at 10 m M they remained partly dissociated, and at 30 m M they never aggregated.
Discussion Since a surface coat is found only in vegetal aggregates, a comparison between our results and Holtfreter's must concern primarily this type of aggregate. We have been able to confirm the observations made by Holtfreter as far as the effects of pH, tonicity and C a 2+ a r e concerned, even though our aggregates are more resistant to extreme p H values than the whole embryos studied by Holtfreter (1943). It is also consistent with Holtfreter's results that high tonicities are detrimental to the surface coat of these aggregates. With regard to Ca 2+ sensitivity, two different cell-cell adhesion systems may be distinguished, viz. a Ca2+-dependent system (CDS) and a Ca 2+independent one (CIDS) ( N o m u r a et al., 1986). From ours, as well as from other studies (Suzuki et al., 1986; N o m u r a et al., 1986), it may be inferred that both these systems are present in amphibian embryos. Ca 2+ seems to be indispensable for the preservation of the surface coat and for the maintenance of vegetal aggregates. The necessary concentration must be very low, since reducing the amount of Ca 2+ to 5 /zM does not affect these aggregates. The vegetal aggregates, comprising undifferentiated cells, may be considered more 'primitive' than the epidermal vesicles. Our finding of a CDS in the former agrees with the notion advocated by N o m u r a et al. (1986), who pos-
75 tulated that CDS precedes CIDS during development. They did not find CIDS at all in Xenopus embryos up to the gastrula stage. This is corroborated by our finding that such a system of cell adhesion is found in the animal aggregates, which consist of ciliated, epidermal cells, corresponding to a developmental stage occurring after gastrulation in vivo. The animal aggregates, consisting of ciliated epidermis, are sensitive to low p H and high tonicities. In Xenopus the explants fail to aggregate properly when the tonicity is twice the normal. Under the same conditions, ciliation, but not aggregate formation, is suppressed in Ambystorna. The differentiation into ciliated epidermis is thus repressed at high tonicities. When embryos were reared under these conditions, Holffreter (1943) observed that cells from the interior leave the embryo passing through the ectoderm. This finding, indicating the failure of epidermis formation, is thus corroborated by our results. On the contrary, when tonicity is half the normal, the epidermal differentiation is promoted, as observed previously by Lovtrup and Perris (1983). Tunicamycin is a compound which inhibits glycosylation of asparaginyl residues in glycoproteins (Takatsuki and Tamura, 1975; Olden et al., 1978; Ekblom et al., 1979). The results obtained with this substance are ambiguous. In animal aggregates from Ambystoma the suppression of glycoprotein synthesis is parallelled by dissociation. In vegetal aggregates from the same species, no morphological effect of tunicamycin is observed; yet the synthesis of glycoproteins is completely suppressed. This observation may imply either that although glycoproteins under some circumstances are involved in aggregate formation, they are not an indispensable component, or that the glycoproteins necessary for aggregation are present at the outset. The circumstance that a surface coat is formed even when tunicamycin is added seems to support the latter alternative. In vegetal aggregates from Xenopus tunicamycin prevents the formation of the surface coat, suggesting that the latter contains glycoproteins. Although almost no synthesis of glycoprotein occurs in animal explants from Xenopus, aggregate formation is still affected by tunicamycin. This
observation must remain unexplained for the time being. The presence of sulfated substances in early amphibian embryos is well documented (see for instance Kosher and Searls, 1973; Ht~glund and Lovtrup, 1976). A rise in the incorporation of sulfate was found to coincide with gastrulation, confined particularly to the invaginating cells in the anurans Xenopus and Rana pipiens (Tarin, 1971; Kosher and Searls, 1973). Normal aggregate formation is suppressed at high tonicities in both species and in both types of aggregates. At low tonicities the formation of epidermis is promoted, while no effect was observed regarding the vegetal aggregates. The synthesis of G A G is decreased at high, as well as low tonicities, but is not completely suppressed. Thus the participation of G A G in aggregate formation cannot be excluded. This conclusion is fairly well supported by the observations on the effect of selenate, since this substance gives rise to dissociation in all cases except in vegetal Ambystoma aggregates, and this finding may perhaps be accounted for on the assumption made above, namely, that the glycoproteins needed for the formation of a surface coat are present in the cells at the time our cultures are started. The effect of extracellular substances and cell surface components on cell adhesion and morphogenesis has been discussed repeatedly during the last decade (cf. Toole, 1981; Edelman, 1983, 1984; Thiery et al., 1985). Altogether, our results suggest that both glycoproteins and G A G are involved in the formation of aggregates by embryonic cells. The former substances seem to be of particular significance for the formation of vegetal aggregates, presumably because the surface coat is made up of glycoproteins. But as pointed out in the introduction, there are other factors involved in aggregate formation; thus the ridges which interconnect the cells in animal aggregates are probably formed by cortical microfilaments.
Acknowledgements We are indebted to Mrs. Mabel Jonsson for typing the manuscript, and to Mrs. Astrid Rehn-
76
holm and Mrs. Birgitta Grahn for technical assistance.
References Bennet, H.S.: Morphological aspects of extracellular polysaccharides. J. Histochem. Cytochem. 11, 14-23 (1963). Edelman, G.M.: Cell adhesion molecules. Science 219, 450-457 (1983). Edelman, G.M.: Cell adhesion and morphogenesis. The regulator hypothesis. Proc. Natl. Acad. Sci. USA 81, 1460-1464 (1984). Ekblom, S., L. Nordling, L. Saxrn, M.-L, Rasilo and O. Renkonen: Cell interactions leading to kidney tubule determination are tunicamycin sensitive. Cell Differ. 8, 347-352 (1979). HSglund, L. and S. Lovtrup: Changes in acid mucopolysaccharides during the development of the frog Rana temporaria. Acta Embryol. Exp. 63-79 (1976). Holtfreter, J.: Properties and functions of the surface coat in amphibian embryos. J. Exp. Zool. 93, 251-323 (1943). Kosher, R.A. and R.L. Searls: Sulfated mucopolysaccharide syntheses during the development of Rana pipiens. Dev. Biol. 32, 50-68 (1973), Letourneau, P.C., P.N. Ray and M.R. Bernfield: The regulation of cell behaviour by cell adhesion. In: Biological Regulation and Development, vol. III, ed. R.F, Goldberger (Plenum Press, New York), pp. 339-376 (1980). Lovtrup, S.: Epigenetic mechanisms in the early amphibian embryo: Cell differentiation and morphogenetic elements. Biol. Rev. 58, 91-130 (1983). Lovtrup, S. and R. Perris: Instructive induction or permissive activation? Differentiation of ectodermal cells isolated from the axolotl blastula. Cell Differ. 12, 171-176 (1983). Luft, J.H.: The structure and properties of the cell surface coat. Int. Rev. Cytol. 45, 291-382 (1976).
Martinez-Palomo, A.: The surface coat of animal cells. Int. Rev. Cytol. 29, 29-75 (1970). Nakatsuji, N. and K.E. Johnson: Cell locomotion in vitro by Xenopus laevis gastrula mesodermal cells. Cell Motil. 2, 149-161 (1982). Nomura, K., M. Uchida, H. Kageura, K. Shiogawa and K. Yamana: Cell to cell adhesion systems in Xenopus laevis, the South African clawed frog I. Detection of Ca 2+ dependent and independent adhesion systems in adult and embryonic cells. Dev. Growth Differ. 28, 311-319 (1986). Olden, K., R.M. Pratt and K.M. Yamada: Role of carbohydrates in protein secretion and turnover: Effects of tunicamycin on the major cell surface glycoprotein of chick embryo fibroblasts. Cell 13, 461-473 (1978). Perry, M.M.: Microfilament in the external surface layer of the early amphibian embryo. J. Embryol. Exp. Morphol. 33. 127-146 (1975). Suzuki, A.S., T. Ueno and T. Matsusaka: Alteration of cell adhesion system in amphibian ectoderm cells during primary embryonic induction: changes in reaggregation pattern of induced neuroectoderm cells and ultrastructural features of the reaggregate. Roux's Arch. Dev. Biol. 195, 85-91 (1986). Takatsuki, A. and G. Tamura: Effect of tunicamycin on the synthesis of macromolecules in cultures of chick embryo fibroblasts infected with Newcastle disease virus. J. Antibiot. 24, 785-794 (1975). Tarin, D.: Histological features of neural induction in Xenopus laevis. J. Embryol. Exp. Morphol. 26, 543-570 (1971). Thiery, J.P., J.L. Duband and G.C. Tucker: Cell migration in the vertebrate embryo: Role of cell adhesion and tissue environment in pattern formation. Annu. Rev. Cell Biol. 1, 91-113 (1985). Toole, B.P.: Glycosaminoglycans in morphogenesis. In: Cell Biology of Extracellular Matrix, ed, E.D. Hay (Plenum Press, New York) pp. 259-294 (1981).
69
CDF 00481
Factors involved in the formation and stabilization of cell aggregates obtained from amphibian embryonic explants Mats-Olof Mattsson, Huguette Lovtrup-Rein and Soren Lovtrup University of Ume&,Department of Zoophysiology, S-90I 87 Umed, Sweden (Accepted 29 September 1987)
The effect of factors influencing the formation and stability of animal and vegetal aggregates from
Xenopus iaevis and Ambystoma mexicanum was examined in the light and scanning electron microscopes. At extreme values of pH the surface coat covering the vegetal aggregates is dissolved and dissociation may take place. Animal aggregates are more resistant. At high tonicities vegetal aggregates may be dissociated, and in the animal aggregates the epidermal differentiation is suppressed. In the absence of Ca~+ the vegetal aggregates are dissociated, but the animal aggregates are not affected. The results obtained with the inhibitor selenate and from incorporation experiments indicate that sulfated glycosaminoglycans are involved in the formation of aggregates in both species. Corresponding observations with tunicamycin suggest that even glycoproteins may play a role in aggregate formation, particularly in the vegetal aggregates. Aggregate formation; Amphibian; Surface coat; Glycoprotein; Sulfated glycosaminoglycans
Introduction The existence of surface layers of varying thickness and composition in protozoans and in various marine animals has been known or at least surmised for a long time (cf. Martinez-Palomo, 1970; Luft, 1976). Holtfreter (1943, p. 266), claimed that in early amphibian embryos the cells are kept together by such a surface layer, a surface coat, which 'amalgamates the peripheral cells into a plastic super-cellular unit'. With the rise of electron microscopy and improved methods of fixation and staining, it has been possible to demonstrate the existence of a Correspondence address: M.-O. Mattsson, University of Ume~, Department of Zoophysiology, S-901 87 Ume~t, Sweden.
cellular surface coat as an almost ubiquitous phenomenon. Various lines of evidence suggest that this surface coat always contains carbohydrate. In fact, the name 'glycocalyx' was long ago proposed for this element of the cell cortex (Bennet, 1963). Aggregate formation by dispersed cells involves two aspects in which the participation of microfilaments may be envisaged: (1) the agglomeration of the individual cells and (2) the stabilization of the aggregates thus formed. It may be presumed that the former event is accomplished by microfilamentous filopodia. Microfilaments may also be involved in the stabilization or maintenance of aggregates, as suggested by the fact that cytochalasin may cause dissociation of aggregates of embryonic cells (Perry, 1975). Apart from these microfilaments, certain 'extracelhilar' factors, no-
0045-6039/88/$03.50 © 1988 Elsevier Scientific Publishers Ireland, Ltd.
70 tably glycoproteins and glycosaminoglycans (GAG), have been suspected to be involved in the formation of cell aggregates (see for instance Luft, 1976; Letourneau et al., 1980). When small explants are isolated in vitro from the blastula of either Ambystoma mexicanum or Xenopus laevis, the formation of two different kinds of cell aggregates may be observed (Lavtrup, 1983). Endodermal cells from the proximity of the vegetal pole or mesodermal cells from the interior marginal zone form more or less irregular aggregates of 'undifferentiated' cells (from here on called 'vegetal aggregates'). These aggregates, when observed in the scanning electron microscope, are seen to be covered by a surface coat. In contrast, ectodermal cells, excised from the animal hemisphere (' animal aggregates'), differentiate into vesicular aggregates made up of ciliated epidermis. In these there is no visible surface coat, but instead ridges are seen to border adjacent cells. In the present study we have attempted to establish possible differences in the nature of the extracellular substances which participate in the formation and maintenance of the two different kinds of aggregates in the amphibian species mentioned above. For this purpose we have investigated the effect of factors which may be expected to dissolve or otherwise interfere with the substances supposed to maintain the aggregates. We have further studied the effect of substances known to inhibit the synthesis of glycoproteins and sulfated G A G . In order to interpret the results thus obtained, it was necessary to investigate the synthesis of these compounds by means of isotope incorporation.
Material and Methods
Naturally spawning Ambystoma mexicanum provided the urodele embryos. In Xenopus laevis spawning was induced by injecting chorion gonadotropin (Pregnyl, N.V. Organon, Oss, The Netherlands): 500 U per female, 250 U per male. The explants were taken from the animal or vegetal regions of late blastulae. The jelly surrounding the embryos was removed either by mechanical dissection (Ambystoma) or, in the case
of Xenopus, by short exposure to thioglycolic acid (1%) in 10% modified Stearns' solution (MSS) (Nakatsuji and Johnson, 1982), p H 7.8-8.0. The explants were obtained and cultured according to the methods described by Lovtrup and Perris (1983), with the exceptions that we used undiluted MSS, p H 7.4-7.5, to which benzylpenicillin (100 U / m l ) and streptomycin (100 /~g/ml) were added, while serum albumin was omitted. The effects of p H were studied in the interval of 2.4-11.9 during 10 rain to 1 h. The pH was adjusted with HC1 or N a O H to the desired value. For testing the importance of the tonicity of the medium we used MSS corresponding to a 0.5. 0.75, 1.5, 2.0 and 2.5-fold tonicity of the standard solution. The control medium had an osmolarity of 177 mOsm, with a NaCI concentration of 75 mM (150 mOsm). The other tonicities were obtained by adjusting the NaC1 concentration to 30 m M (87 m O s m total osmolarity), 52.5 mM (132 mOsm), 118 m M (263 mOsm), 163 mM (353 mOsm) and 208 m M (443 mOsm). The aggregates were incubated in the respective media until the animal controls became ciliated (24 h for Xenopus, 3 days for Arnbystoma). The effect of the Ca 2+ concentration was studied with MSS free of Ca 2+ or containing 5 /~M, 50 #M, 0.2 mM, 0.4 mM, 0.8 mM and 1 mM Ca 2+, respectively. After one day (Xenopus) or 3 days (Ambystoma) in control medium with 2 mM Ca 2+, the aggregates were exposed to these different media from 15 min up to 2 days. The antibiotic tunicamycin (Sigma) was added to the standard medium at a concentration of 5 /~g/ml. Na-selenate was added to the cultures in concentrations of 10 and 30 raM. Both substances were added from the beginning of the cultures. The experiments were observed and documented in a scanning electron microscope (SEM) or in an inverted microscope with camera attachment. For electron microscopy the specimens were fixed overnight at 4 ° C in 2.5% glutaraldehyde in 0.l M phosphate buffer or 0.1 M cacodylate buffer. After rinsing in buffer the specimens went through a graded series of ethanol and transferred to a critical point drying apparatus for further drying. The specimens were coated with a gold layer
71
(900-1000 A) and studied in a Jeol JSM-P15 electron microscope with a beam voltage of 15 kV. For isotope experiments, explants were cultured as described above in the presence of 0.50 /~Ci/ml L-[6-3H]fucose, with or without 5 / ~ g / m l of tunicamycin; or 5 /~Ci/ml of 35SO42+; or 0.15 t t C i / m l D-[1-14C]galactosamine hydrochloride plus 0.15 / t C i / m l D-[1-14C]glucosamine hydrochloride. The samples were sonicated, digested with pronase, and loaded on a Sephadex G-50 column after delipidation with methanolchloroform. Radioactivity in the fractions containing glycoproteins was determined in a scintillation counter. The GAG, emerging in the excluded volume, were further purified by ion exchange chromatography on a Whatman DE-52 column. Fractions containing G A G were used partly for direct determination of total incorporation, partly for determination of chondroitin sulfate and heparan sulfate c o n t e n t by enzymatic d e g r a d a t i o n (Lovtrup-Rein, in preparation).
Results
The effect of pH Holtfreter (1943) found that elevation of pH to above 10 will cause dissolution of the surface coat and disaggregation of the peripheral cells in the embryo. Working with very small aggregates it is necessary to employ much more extreme p H values, acid as well as alkaline, before any effects on the state of aggregation are observable. When effects were observed at the p H values listed in Table I, further attempts to dissociate the aggregates were given up. In Ambystoma the exposure of vegetal aggregates to changes of p H in both directions leads to dissolution of the surface coat and to partial dissociation. Elevation of the p H is much more efficient than lowering, a finding which corroborates earlier observations by Holtfreter. The vegetal aggregates from Xenopus are more resistant. The animal aggregates from both species are more susceptible to low values of pH, but are very resistant to alkaline media (Fig. l a - e ) .
The effect of tonicity Holtfreter observed a tendency towards dissociation when embryos were reared at high tonicities. When the tonicity of our media was raised 2-2.5 fold, the formation of the surface coat was suppressed in Ambystoma vegetal aggregates, whereas in the animal aggregates no epidermal differentiation (ciliation) took place (Fig. 2a). The Xenopus aggregates were more susceptible to changes in tonicity; thus, in both types of aggregate, dissociation or lack of aggregate formation was observed when the tonicity was twice the normal or higher (Fig. 2b-e). In both species it was found that when the tonicity of the medium was half of that in the controls, epidermal differentiation was favoured, manifested by a higher percentage of ciliated vesicles. No effect was observed on the vegetal aggregates. The results of incorporation of aminosugars and sulfate at high and low tonicities are given in Table II. In both species, the synthesis of G A G was reduced as compared with the controls, more extensively in Ambystoma than in Xenopus. However, among the two G A G that were synthesized by the aggregates, namely chondroitin sulfate and heparan sulfate, the production of the latter was almost completely suppressed in aggregates from Xenopus at high tonicities. It is interesting, and perhaps of significance, to observe that in Am-
TABLE I The effect of pH changes
A mbystoma Animal pH 3.0 1 h aggregates Partial dissociation Cells cytolyzed pHll.41h No visible effect Vegetal pH 2.4 1 h aggregates Partial dissociation pH 11.4 20 rain Dissolution of surface coat Partial dissociation
Xenopus pH 3.0 1 h Partial dissociation Cells cytolyzed pHll.91 h No visible effect pH 3.0 1 h Partial dissociation Cells cytolyzed pH 11.5 1 h No visible effect
Deviations from neutrality less than those reported here did not have any visible effect.
72
Fig. 1. (a) Animal control aggregate from Xenopus. Note the ciliation and epidermal ridges (ri). A corresponding Amb),stoma control is similar to this aggregate. Bar, 50/xm. X 280 in a f. (b) The corresponding vegetal control. A surface coat obscures the cell borders. (c) Animal aggregate from Ambvstoma exposed to pH 2.4 for I h. Cells are cytolyzed and the epidermal characteristics are missing. (d) Vegetal aggregate from Amt~vstorna exposed to pH 2.4 for 1 h. Surface coat is partly absent. (e) Ambystoma vegetal aggregate exposed to pH 11.4 for 20 min. Surface coat is totally absent and the aggregate starts to dissociate. (f) Xenopus vegetal aggregate in Ca2+-free medium for 15 min. Surface coat is absent and the aggregate is starting to dissociate, a process which is completed within 1 h.
73
bystoma the degree of sulfation was n o t affected by tonicity, b u t in Xenopus the G A G were u n d e r sulfated at high a n d oversulfated at low tonicities.
The effect of Ca 2 + O b s e r v a t i o n s have shown that Ca 2+ is often necessary for the f o r m a t i o n a n d stabilization of cell aggregates, a n d Holtfreter's o b s e r v a t i o n s
(1943) suggest that C a 2+ may even be necessary for the integrity, of the a m p h i b i a n surface coat. In the present study the vegetal aggregates from b o t h species u n d e r w e n t partial dissociation when kept in a Ca2+-free m e d i u m for 1 h. S c a n n i n g electron m i c r o g r a p h s showed that the surface coat had d i s a p p e a r e d (Fig. lf). N o dissociation was observed when Ca 2+ was present in the m e d i u m ,
Fig, 2. (a) Animal aggregate from Ambystoma exposed to high tonicity, corresponding to 1.5 times the normal. The ciliation is suppressed, but otherwise the aggregate is intact. Bar, 20 #m in both a and b. × 290. (b) Animal aggregate from Xenopus exposed to a tonicity corresponding to 2.0 times the normal. The explant is partly disaggregated, many ceils are cytolyzed and no 'normal' features are discernible, x 450. (c-e) A series of vegetal Xenopus aggregates exposed to tonicities corresponding to 1.0, 1.5 and 2.0 times the normal, respectively. A progressive disaggregation is observable ending up with a total dissociation in e. y = yolk platelets. Bar, 50 p.m. x150.
74 TABLE II Rates of incorporation of radioisotopes into the GAG of animal and vegetal explants at different tonicities, compared with controls at normal tonicities 14C-labelled amino sugars
[35S]sulfate
Animal
Vegetal
Animal
Vegetal
15% 50%
15% 50%
17% 8%
17% 8%
19% 57%
20% 57%
22% 82%
22% 82%
High tonicity
Ambystoma Xenopus Low tonicity
Ambystoma Xenopus
not even at the lowest concentrations. No effect was observed when the animal aggregates were treated similarly.
The effect of tunicamycin If glycoproteins are involved either in the formation of the surface coat or as cell adhesives, we should expect tunicamycin to affect aggregate formation of the embryonic cells. We found that at a concentration of 5 /~g/ml, tunicamycin had no effect on vegetal aggregates from Ambystoma. In the corresponding Xenopus aggregates no surface coat was formed. The formation of animal aggregates was suppressed in both species. Synthesis of glycoproteins, as evidenced by the incorporation of fucose (Table III), took place in the vegetal aggregates from both species and in TABLE III Rates of incorporation of [3H]fucose into glycoproteins of animal and vegetal explants (cpm//~g protein, _+standard error of mean)
Ambystoma
Xenopus
Animal Vegetal Animal Vegetal Controls
4430 3 715 910 5 395 (_+310) (_+390) (_+105) (_+305)
Tunicamycin-treated
565 (_+60)
305 685 480 (_+38) (_+67) (_+42)
Number of experiments 5
5
4
4
Ratio control vs. treated 7.84
12.18
1.33
11.24
the animal aggregates from Ambystoma, and was completely suppressed by tunicamycin. Little incorporation of fucose could be observed in animal aggregates from Xenopus.
The effect of selenate If Na-selenate was added to the culture medium, no obvious effect was observable on the vegetal aggregates from Ambystoma, but in 10 mM selenate this type of aggregates from Xenopus remained completely dissociated. The animal aggregates from both species were also affected, and in much the same way; at 10 m M they remained partly dissociated, and at 30 m M they never aggregated.
Discussion Since a surface coat is found only in vegetal aggregates, a comparison between our results and Holtfreter's must concern primarily this type of aggregate. We have been able to confirm the observations made by Holtfreter as far as the effects of pH, tonicity and C a 2+ a r e concerned, even though our aggregates are more resistant to extreme p H values than the whole embryos studied by Holtfreter (1943). It is also consistent with Holtfreter's results that high tonicities are detrimental to the surface coat of these aggregates. With regard to Ca 2+ sensitivity, two different cell-cell adhesion systems may be distinguished, viz. a Ca2+-dependent system (CDS) and a Ca 2+independent one (CIDS) ( N o m u r a et al., 1986). From ours, as well as from other studies (Suzuki et al., 1986; N o m u r a et al., 1986), it may be inferred that both these systems are present in amphibian embryos. Ca 2+ seems to be indispensable for the preservation of the surface coat and for the maintenance of vegetal aggregates. The necessary concentration must be very low, since reducing the amount of Ca 2+ to 5 /zM does not affect these aggregates. The vegetal aggregates, comprising undifferentiated cells, may be considered more 'primitive' than the epidermal vesicles. Our finding of a CDS in the former agrees with the notion advocated by N o m u r a et al. (1986), who pos-
75 tulated that CDS precedes CIDS during development. They did not find CIDS at all in Xenopus embryos up to the gastrula stage. This is corroborated by our finding that such a system of cell adhesion is found in the animal aggregates, which consist of ciliated, epidermal cells, corresponding to a developmental stage occurring after gastrulation in vivo. The animal aggregates, consisting of ciliated epidermis, are sensitive to low p H and high tonicities. In Xenopus the explants fail to aggregate properly when the tonicity is twice the normal. Under the same conditions, ciliation, but not aggregate formation, is suppressed in Ambystorna. The differentiation into ciliated epidermis is thus repressed at high tonicities. When embryos were reared under these conditions, Holffreter (1943) observed that cells from the interior leave the embryo passing through the ectoderm. This finding, indicating the failure of epidermis formation, is thus corroborated by our results. On the contrary, when tonicity is half the normal, the epidermal differentiation is promoted, as observed previously by Lovtrup and Perris (1983). Tunicamycin is a compound which inhibits glycosylation of asparaginyl residues in glycoproteins (Takatsuki and Tamura, 1975; Olden et al., 1978; Ekblom et al., 1979). The results obtained with this substance are ambiguous. In animal aggregates from Ambystoma the suppression of glycoprotein synthesis is parallelled by dissociation. In vegetal aggregates from the same species, no morphological effect of tunicamycin is observed; yet the synthesis of glycoproteins is completely suppressed. This observation may imply either that although glycoproteins under some circumstances are involved in aggregate formation, they are not an indispensable component, or that the glycoproteins necessary for aggregation are present at the outset. The circumstance that a surface coat is formed even when tunicamycin is added seems to support the latter alternative. In vegetal aggregates from Xenopus tunicamycin prevents the formation of the surface coat, suggesting that the latter contains glycoproteins. Although almost no synthesis of glycoprotein occurs in animal explants from Xenopus, aggregate formation is still affected by tunicamycin. This
observation must remain unexplained for the time being. The presence of sulfated substances in early amphibian embryos is well documented (see for instance Kosher and Searls, 1973; Ht~glund and Lovtrup, 1976). A rise in the incorporation of sulfate was found to coincide with gastrulation, confined particularly to the invaginating cells in the anurans Xenopus and Rana pipiens (Tarin, 1971; Kosher and Searls, 1973). Normal aggregate formation is suppressed at high tonicities in both species and in both types of aggregates. At low tonicities the formation of epidermis is promoted, while no effect was observed regarding the vegetal aggregates. The synthesis of G A G is decreased at high, as well as low tonicities, but is not completely suppressed. Thus the participation of G A G in aggregate formation cannot be excluded. This conclusion is fairly well supported by the observations on the effect of selenate, since this substance gives rise to dissociation in all cases except in vegetal Ambystoma aggregates, and this finding may perhaps be accounted for on the assumption made above, namely, that the glycoproteins needed for the formation of a surface coat are present in the cells at the time our cultures are started. The effect of extracellular substances and cell surface components on cell adhesion and morphogenesis has been discussed repeatedly during the last decade (cf. Toole, 1981; Edelman, 1983, 1984; Thiery et al., 1985). Altogether, our results suggest that both glycoproteins and G A G are involved in the formation of aggregates by embryonic cells. The former substances seem to be of particular significance for the formation of vegetal aggregates, presumably because the surface coat is made up of glycoproteins. But as pointed out in the introduction, there are other factors involved in aggregate formation; thus the ridges which interconnect the cells in animal aggregates are probably formed by cortical microfilaments.
Acknowledgements We are indebted to Mrs. Mabel Jonsson for typing the manuscript, and to Mrs. Astrid Rehn-
76
holm and Mrs. Birgitta Grahn for technical assistance.
References Bennet, H.S.: Morphological aspects of extracellular polysaccharides. J. Histochem. Cytochem. 11, 14-23 (1963). Edelman, G.M.: Cell adhesion molecules. Science 219, 450-457 (1983). Edelman, G.M.: Cell adhesion and morphogenesis. The regulator hypothesis. Proc. Natl. Acad. Sci. USA 81, 1460-1464 (1984). Ekblom, S., L. Nordling, L. Saxrn, M.-L, Rasilo and O. Renkonen: Cell interactions leading to kidney tubule determination are tunicamycin sensitive. Cell Differ. 8, 347-352 (1979). HSglund, L. and S. Lovtrup: Changes in acid mucopolysaccharides during the development of the frog Rana temporaria. Acta Embryol. Exp. 63-79 (1976). Holtfreter, J.: Properties and functions of the surface coat in amphibian embryos. J. Exp. Zool. 93, 251-323 (1943). Kosher, R.A. and R.L. Searls: Sulfated mucopolysaccharide syntheses during the development of Rana pipiens. Dev. Biol. 32, 50-68 (1973), Letourneau, P.C., P.N. Ray and M.R. Bernfield: The regulation of cell behaviour by cell adhesion. In: Biological Regulation and Development, vol. III, ed. R.F, Goldberger (Plenum Press, New York), pp. 339-376 (1980). Lovtrup, S.: Epigenetic mechanisms in the early amphibian embryo: Cell differentiation and morphogenetic elements. Biol. Rev. 58, 91-130 (1983). Lovtrup, S. and R. Perris: Instructive induction or permissive activation? Differentiation of ectodermal cells isolated from the axolotl blastula. Cell Differ. 12, 171-176 (1983). Luft, J.H.: The structure and properties of the cell surface coat. Int. Rev. Cytol. 45, 291-382 (1976).
Martinez-Palomo, A.: The surface coat of animal cells. Int. Rev. Cytol. 29, 29-75 (1970). Nakatsuji, N. and K.E. Johnson: Cell locomotion in vitro by Xenopus laevis gastrula mesodermal cells. Cell Motil. 2, 149-161 (1982). Nomura, K., M. Uchida, H. Kageura, K. Shiogawa and K. Yamana: Cell to cell adhesion systems in Xenopus laevis, the South African clawed frog I. Detection of Ca 2+ dependent and independent adhesion systems in adult and embryonic cells. Dev. Growth Differ. 28, 311-319 (1986). Olden, K., R.M. Pratt and K.M. Yamada: Role of carbohydrates in protein secretion and turnover: Effects of tunicamycin on the major cell surface glycoprotein of chick embryo fibroblasts. Cell 13, 461-473 (1978). Perry, M.M.: Microfilament in the external surface layer of the early amphibian embryo. J. Embryol. Exp. Morphol. 33. 127-146 (1975). Suzuki, A.S., T. Ueno and T. Matsusaka: Alteration of cell adhesion system in amphibian ectoderm cells during primary embryonic induction: changes in reaggregation pattern of induced neuroectoderm cells and ultrastructural features of the reaggregate. Roux's Arch. Dev. Biol. 195, 85-91 (1986). Takatsuki, A. and G. Tamura: Effect of tunicamycin on the synthesis of macromolecules in cultures of chick embryo fibroblasts infected with Newcastle disease virus. J. Antibiot. 24, 785-794 (1975). Tarin, D.: Histological features of neural induction in Xenopus laevis. J. Embryol. Exp. Morphol. 26, 543-570 (1971). Thiery, J.P., J.L. Duband and G.C. Tucker: Cell migration in the vertebrate embryo: Role of cell adhesion and tissue environment in pattern formation. Annu. Rev. Cell Biol. 1, 91-113 (1985). Toole, B.P.: Glycosaminoglycans in morphogenesis. In: Cell Biology of Extracellular Matrix, ed, E.D. Hay (Plenum Press, New York) pp. 259-294 (1981).