Concanavalin A binding to amphibian embryo and effect on morphogenesis

Amphibian gastrulation is a complex morphogenic process in which all the embryonic cells are involved (Vogt, 1929; Holtfreter, 1939). During this period, cellular adhesiveness is markedly modified (Townes & Holtfreter, 1955; Curtis, 1961; Jones & Elsdale, 1963; Patricolo, 1967; Johnson, 1970) as are intercellular contact and cell morphology (Baker, 1965; Tarin, 1971; Monroy, Baccetti & Denis-Donini, 1976; Smith, Osborn & Stanisstreet, 1976). Recent studies on biological membranes have demonstrated the important role played by carbohydrate-containing cell surface macromolecules in cellular function (Moscona, 1971; Kleinschuster & Moscona, 1972; Collins, 1974; Marchase, Vosbeck & Roth, 1976). Lectins, which interact with carbohydrate-containing cell surface macromolecules (Sharon & Lis, 1972; Burger, 1973; Nicolson, 1974; Bourrillon, 1975), may prove to be useful tools for the investigation of cell surface modifications with respect to morphogenesis.

Concanavalin A (Con A) interacts with the methyl-D-mannopyranosyl residues located in cell surface binding sites (Goldstein, Holleman & Smith, 1965; Sharon & Lis, 1972). These binding sites may play a role in embryo-cell migration (Moscona, 1974; Krach, Green, Nicolson & Oppenheimer, 1974; Moran, 1974a). Con A interacts preferentially with the chordomesoderm migratory cells in the amphibian gastrula (O’Dell, Tencer, Monroy & Brachet, 1974), and inhibits the development of Ambystoma maculatum gastrulae (Moran, 1974 b).

Following our earlier studies (Boucaut, 1974; Aubery & Bourrillon, 1976), the present report describes the effects of Con A on the morphogenic cellular movements of Pleurodeles waltlii embryos, in relation with the lectin binding to cell surface.

[l–¹⁴C]Acetic anhydride (27·6 mCi/mmol) was obtained from the Radiochemical Centre (Amersham, England), α-methyl-D-mannopyranoside (α-MM) was purchased from Sigma (St. Louis, Missouri).

Con A (Grade III) was purchased from Sigma. Fluorescein isothiocyanate Concanavalin A (FITC-Con A) was obtained from LB.F. (Clichy, France).

Labelled Con A was prepared according to Miller & Great (1972), using [¹⁴C]acetic anhydride. Lectin was purified by gel filtration on Biogel P 6, the eluting buffer being 0·01 NaHCO₃ (pH 7·5) containing 0·15 M-NaCl. The specific activity of the lectin was 10⁴ dpm/μg. The labelled lectin behaved exactly like native lectin in the erythroagglutination test (Bourrillon & Font, 1968). The lectin solution was prepared at a concentration of 500 μg/ml of eluting buffer.

Quantitative binding studies were performed in centrifuge tubes by incubating four embryos with various concentrations of labelled lectin (from 10 to 200 μg/ml) in 1 ml of Steinberg’s solution. After incubation for various times at 4 °C or 18 °C, the embryos were decanted and washed three times with Steinberg’s solution (15 ml total). They were then dissolved in 0·5 ml H₂O₂ 110 vol for 24 h at 37 °C. The resulting solution was added to 10 ml of A-C-S scintillation fluid (Aqueous Counting Scintillation, Amersham, Searle, U.S.A.) and counted in a liquid scintillation spectrometer (SL 300, Inter technique).

Specific binding of Con A was expressed in terms of the difference between the quantity of ;lectin bound in the presence and in the absence of α-MM (0µ M) a specific binding was 10–15%.

Four embryos were incubated with 100 μg/ml of fluorescent Con A during a 24 h incubation at 18 °C in 1 ml of Steinberg’s solution. After incubation, the embryos were decanted and washed three times with Steinberg’s solution (15 ml total), and the fluorescence was observed using a Zeiss universal microscope with an excitation filter KP 500 and a chromatic beam splitter FL 500.

The specificity of FITC-Con A was determined in the presence of α-MM (0·05 M). The fluorescence was observed before and after removing the vitelline membrane.

For each set of experiments, the embryos were selected from a single batch. Fertilized eggs were obtained from a stock of Pleurodeles waltlii kept at the Laboratoire de Biologie Animale. They were reared at 18 °C and were staged according to the instructions of Gallien & Durocher (1957). The jelly was removed from the embryos with forceps, the embryos rinsed twice and allowed to develop in 10 % Steinberg’s solution, pH 7-8, until they reached the desired stage. Three stages of development were investigated : late blastula (stage 7), young gastrula (stage 8 a, small blastopore groove), and late gastrula (stage 12, small slitshaped blastopore).

When the embryos had reached the appropriate stage, they were transferred and reared for 6 days in 10% Steinberg’s solution containing 25, 50, 100, 150 or 200 μg/ml of Con A.

The effects of each Con A concentration were followed on twenty embryos from each of the developmental stages used (stage 7, stage 8, stage 12). The embryos were observed and photographed daily throughout the incubation period. These procedures and observations were carried out at 18 °C under sterile conditions.

Control experiments were performed in order to test the specific action of Con A and the viability of the treated embryos.

Inhibition assays were carried out with α-MM (0·05 M).

The solutions of lectin were pre-incubated for 30 min with α-MM (0·05 M). To test the viability of the embryos after Con A treatment, they were washed with 10% Steinberg’s solution and allowed to develop in this culture medium. For each experiment twenty embryos were placed in 10% Steinberg’s solution so as to study and stage normal development.

[¹⁴C] Con A was used in quantitative assays at 4 °C (to prevent endocytosis) in order to determine the amount of Con A bound to the eggs. The binding of Con A increased regularly from 0 to 100 μg/ml of lectin added and then reached a plateau. The embryos rapidly bound labelled Con A to plateau levels within 10 min at 4 °C, and all embryos were saturated with 100 μg/ml. No significant differences in specifically bound labelled Con A were obtained at concentrations of Con A ranging from 10 to 100 μg/ml after a 10 min incubation in regard with the age of the embryos. Thus, at saturation 0·18 ±0·06 μg, 0·13 ±0·01 μg and 0·24 ± 0-09 μg of labelled Con A were bound per embryo in the late blastula, young gastrula and late gastrula stages, respectively. These results were obtained from four separate and duplicate experiments. The specificity of lectin binding was investigated in the presence of α-MM (0-05 M) which is an inhibitor of Con A (Goldstein et al. 1965). The specific binding of the lectin reached 8590% of the total binding.

1. Con A was added at final concentrations ranging from 25 to 200 μg/ml. The development rate and the types of morphological variation were analysed in relation to the lectin dose and time of incubation. Embryos from a control medium without lectin developed normally (Fig. 1). Incubation with 25 μg/ml of Con A caused very few anomalies, even though the development time was significantly delayed, since the embryos reached the tail-bud stage without being malformed (Fig. 1). Following treatment with 50 μg/ml of lectin, development was retarded. The neural tube remained open indicating that a dose of 50 μg/ml was required to inhibit morphogenic development. However, all treated embryos reached the tail-bud stage.

The effect of Con A on morphogenesis was significantly marked as the dose was increased from 50 to 100 or 200 μg/ml. The most striking result was the inhibition of gastrulation and neurulation. In treated embryos developed from the late blastula or young gastrula stage, exogastrulation occurred over a period of 2 days. Subsequently the abnormalities observed were all non-closures of the neural tube, ranging from neural folds separated by the subsisting yolk plug to complete anencephaly. Finally, morphogenesis ceased during the fifth day and the treated embryos died. Similarly, the same lectin levels (100–200 μg/ml) applied on late gastrula embryos caused a significant delay in neurulation, after which development was arrested and necrosis followed.

1. Administration of Con A doses ranging from 25 to 200 μg/ml, associated with α-MM (0-05 M) produced no morphological variations. In addition, no indication of defective development was observed among the embryos incubated in α-MM (0-05 M) solution alone. Therefore, these results show that the Con A effect can be prevented by its carbohydrate inhibitor.

2. Since morphogenesis was delayed or arrested following treatment with Con A, the question arose whether the lectin effect was reversible. To test this, further experiments were carried out with doses of Con A ranging from 100 to 200 μg/ml. After incubation times of 24 h, 48 h and 72 h, the embryos were washed and transferred to normal Steinberg’s medium without Con A. Embryos incubated in a solution containing 200μg/ml of Con A supplemented with α-MM (0·05 M) served as controls. The experiments gave similar results for the stages of development choosen (late blastula, young or late gastrula). After an incubation time of 24 h, organogenesis began, and tissue differentiation could take place in the treated embryos. This showed that the effect of Con A can be reversed after a 24 h treatment. In contrast, the Con A effect was not reversible with Con A treatment for 48 and 72 h, since it was observed a total arrest of development and death. These results suggest that as from 48 h of incubation with Con A, the morphological modifications (exogastrulation) induced by lectin are such that normal development could not continue. For the rest, the development of the controls seemed absolutely normal.

At saturation (100 μg/ml), the amount of bound labelled Con A was evaluated at 18 °C after a 24 h incubation under the conditions described above. At 18 °C the binding of labelled Con A to eggs with a vitelline membrane was similar to those obtained at 4 °C (Table 1).

In the presence of Con A at a final concentration of 100 μg/ml, late blastula, young gastrula, and late gastrula embryos bound 0-·24 ±0·03 μg, 0·31 ±0-01 μg and 0·28 ± 0·01 μg, respectively. These results were obtained from four separate and duplicate experiments.

After a 24 h incubation, the vitelline membrane was removed with forceps and the amount of labelled Con A bound to the eggs (s.s.) was determined and compared with the total lectin bound to the eggs with a vitelline membrane. The results are given in Table 1. Of the total labelled Con A 7-9% remained bound to the eggs after the vitelline membrane had been removed. These results suggest that Con A migrates through the vitelline membrane and that there is a direct interaction between the lectin and embryo cell surface glycoproteins.

These results were confirmed by experiments using FITC-Con A (Fig. 2). The fluorescent lectin bound predominantly to the vitelline membrane. However, after the vitelline membrane had been removed, FITC-Con A remained bound to the cell surfaces and particularly to intercellular junctions. Possibly, Con A was evenly distributed over the blastomere surface and thus it would appear to be more concentrated in the troughs between the cells. The interaction of FITC-Con A with cell surfaces was specific since α-MM (0-05 M) prevented the binding of Con A (Fig. 2).

The data obtained clearly demonstrate that Con A altered the development of Pleurodeles waltlii embryos since both exogastrulation and partial neurulation were observed. The modifications depended on the dose as well as the stage of the treated embryo.

Our results have to be compared to those obtained by Lallier (1972), Moran (1974a, b) and Lee (1976) who worked on sea urchin, amphibian (Ambystoma maculatuni) and chick embryo, respectively. The mor ph ogenic alterations induced by Con A could be due to the toxicity of the lectin (Inbar, Ben-Bassat & Sachs, 1972; Nicolson, 1974). However, in our experiments this possibility was excluded as the washing experiments clearly showed that the effect of the lectin could be reversed, allowing tissue differentiation to take place.

These findings are in agreement with those of Lamon & Duprat (1976). The effects of Con A could be the result of the binding of the lectin itself (Nicolson 1974). Indeed, a rapid and specific binding of labelled Con A to the eggs occurred at 4 °C.

Slight differences in lectin binding were observed depending on the stage of embryo development. These differences may be explained in terms of random differences in the topography of vitelline membrane glycoproteins since the total volume remained rather constant during the stages studied.

Although Con A was predominantly bound to the vitelline membrane, after 24 h under rearing conditions, a significant percentage (7–10 %) remained bound to the eggs after the membrane had been removed. Using fluorescent Con A we were able to detect it near the ‘interblastomeric’ junctions, in agreement with the results reported by O’Dell et al. (1974) on Xenopus eggs with the vitelline membrane removed.

All these results suggest a direct interaction between Con A and cell surface components, The well known characteristics of Con A (Goldstein et al. 1965; Sharon & Lis, 1972) and the specificity of the binding and the effect of the lectin as demonstrated by the use of α-MM (0·05 M) mean that cell surface glycoproteins involved in morphogenic development are Con A binding sites. The alterations in morphogenic development induced by Con A could therefore be correlated with the binding of lectin to cell surface glycoproteins. Con A would inhibit the cell’s relative mobility by cell to cell cross-linking and/or mask the specific glycoproteins involved in cellular recognition and migratory activity.

We are indebted to Dr H. Denis for helpful advice and suggestions. The technical assistance of J. Desrosiers is acknowledged.

This work was supported in part by grants from INSERM, ATP No 43-76-75 and CRL No 77-1-087-3.