The effects of high hydrostatic pressure on microfilaments and microtubules in Xenopus laevis

During neurulation the neuroepithelium invaginates thereby transforming the flat neural plate into a neural tube. Two major modifications in the cell shape are observed during the process. Firstly, the cells elongate; it has been said that they become asymmetric. Secondly, they acquire a slender apical portion and a broad base; this has been referred to as an apico-basal dissymmetry (Messier, 1976a).

More than 10 years ago Waddington & Perry (1966) suggested that microtubules might be involved in neurulation in that they could act as shape generating structures capable of inducing cell elongation. Since then many reports, covering an impressive variety of cell systems, have shown the existence of a correlation between the presence of microtubules and the production and/or maintenance of cell asymmetry. Such a correlation has been tested experimentally and established in the cells making up the neuroepithelium of Xenopus (Karfunkel, 1971) and chick embryos (Auclair & Messier, 1974). Concerning the second feature, the acquisition of an apico-basal dissymmetry,it was proposed by Baker & Schroeder (1967) and Schroeder (1970) that once the cells acquire their columnar shape (asymmetry), an apical ring of microfilaments would contract thus reducing the surface area of the cellular apices. This production of an epithelium, made up of dissymmetric cells each endowed with a slender apex and a broad base, would create part of the forces that lead the neural plate into invagination, a prerequisite for neurulation.

To this concept were added a number of findings showing that the setting-in of an apico-basal dissymmetry also involves the phenomenon of interkinetic nuclear migration which induces the enlargement of the base of the cell by placing the nucleus in the basal region of the cell for the longer part of the cell cycle (Messier & Auclair, 1973, 1975; Messier, 1976a).

The role of microfilaments in constricting the cell apices was investigated by Karfunkel (1972). The author treated neurulating chick embryos with colchicine and cytochalasin B. Both chemicals, it was shown, blocked neurulation. Karfunkel reports that colchicine breaks down microtubules thus inducing a loss of asymmetry and that cytochalasin B selectively disrupts the ring of microfilaments thereby causing the loss of the apical construction and the resulting inhibition in neurulation. However, cytochalasin B, it is now most apparent, is a substance whose effects and cell target(s) are quite controversial (Burnside & Manasek, 1972; Holzer & Sanger, 1972; Auclair & Messier, a review, 1977). For instance, it is known now that cytochalasin B inhibits interkinetic nuclear migration in the chick embryo; thereby jeopardizing the apico-basal dissymmetry (Messier & Auclair, 1974). Also, it has been shown in the same tissue that the chemical greatly alters the morphology of the cell surface (Messier, 1976 b). Finally, cytochalasin B allows intercellular leakage of electron opaque tracers, thus suggesting that it may weaken the epithelial cohesiveness necessary for neurulation to proceed normally (Messier & Auclair, 1977).

Because of the uncertainties linked to experimentations with cytochalasin B, it was decided to reinvestigate the role of microfilaments and microtubules in neurulation firstly, via a different avenue and secondly, using a different species. Hence, we used high hydrostatic pressure as a means and the neurulating embryo of Xenopus laevis as a model. High hydrostatic pressures were selected because they are known to depolymerize microtubules and microfilaments (Zimmerman, 1971, 1970a).

Ovulation in Xenopus laevis was induced by injections of chorionic gonadotropin according to Gurdon’s method (1967). The gastrulae were chemically dejellied by treating them 10–15 min with a mercaptoacetic acid solution (0·7 % in 50 % Steinberg solution) brought to pH 8·5 with NaOH (Karfunkel, 1971). They were then kept at 23 °C in a 10 % Steinberg solution until they had reached stages 14–20 (according to Nieuwkoop & Faber, 1967), whence they were used for experimentation. A pressure-fixation chamber identical with that designed by Landau & Thibodeau (1962) was employed for pressurizing and fixation of specimens under hydrostatic pressure. The embryos were subjected to pressures of 500, 1500, 2000, 3000, 4000, 4500, 5000, 6000 and 10000 psi^* for intervals of 5, 15, 60, 90, 120, 180, 240 and 330 min. For each of the pressure/time combinations used a minimum of six embryos were analysed microscopically. However, for combinations (psi/min) of 4000/180, 4000/330, 4500/180 and 5000/90 at least 15 embryos were analysed.

All specimens were fixed with 4 % glutaraldehyde buffered with 0·1 M phosphate (pH 7·2). Fixation was carried out for 15 min at each of the pressures used. Upon decompression, specimens and fixative were transferred to Petri dishes where, to enhance fixation, the upper halves of the eggs were dissected with iris scissors. Fixation was pursued on the isolated embryos for an additional 3–6 h. Other embryos, used as controls, were grown at atmospheric pressure (14,7 psi) and fixed at intervals where they had reached stages 14, 16, 18 and 20. In order to control the effect of a possible oxygen depletion (see Tilney & Gibbins, 1969) 20 embryos were inserted in the pressure bomb and kept there under atmospheric pressure for 330 min. These controls neurulated exactly as those grown under atmospheric pressure. All the controls were fixed as above.

Following fixation, the specimens were washed overnight in the phosphate buffer and post-fixed for 90 min in 0·1 M phosphate buffered osmium tetroxide. Specimens were dehydrated and embedded in Epon. Using an LKB ultrotome, sections covering an average length of 50 μm were cut transversely to the cephalocaudal axis in the mid-trunk region. For light microscopy, the sections were 1 μm thick and stained with a 1 % aqueous solution of toluidine blue; thin sections for electron microscopy were stained with uranyl acetate and lead citrate (Reynolds, 1963) and examined with a Siemens 1A microscope.

The response exhibited by microfilaments and microtubules was always observed in more than 95 % of the neuroepithelial cells of each individual embryo. Furthermore, in most cases, at each of the pressure/time combinations used, 100 % of the embryos reacted in the same way; cases making exception are mentioned in the Results.

The histology and cell ultrastructure of the neuroepithelium of early Xenopus laevis embryos have already been described (Schroeder, 1970; Tarin, 1972). Moreover, morphometric data related to this tissue is also available (Mathieu & Messier, 1977). Therefore, we will only recall here that which is pertinent to the present title.

In Xenopus laevis the neuroepithelium is made up of two layers of epithelial cells. They are termed the superficial and deep layers. The cells whose ultrastructure we will be concerned with are those touching the floor of the neural groove. They belong to the superficial layer and occupy the topographical zone termed median by Schroeder (1970) or those zones labelled proximal superficial and suprachordal superficial by Mathieu & Messier (1977). The cells making up these zones always exhibit a narrow apex and a broad base (dissymmetry) and show asymmetry in that they are either slightly elongated (suprachordal superficial) or highly elongated (proximal superficial) (Fig. 1I). In the literature, these cells have been referred to as ‘bottle cells’, ‘flask cells’ and ‘wedgeshaped cells’. They contain an apically situated ring of microfilaments (Fig. 2A) and numerous microtubules oriented parallel to the cell’s longer axis (Fig. 2B).

The fate of the neuroepithelium, as it progresses from the neural plate to the neural tube, is illustrated in Figs. 1A-D. This series of micrographs shows the degree of development attained under atmospheric pressure, starting from stage 14 or stage 16, which were the ones selected as starting points in our experimentation. The series helps in evaluating how much a given pressure, applied for a given length of time, disturbed neural organogenesis. For instance, pressurizing stage-14 embryos (Fig. 1A) for 180 min at 4000 psi impeded the invagination of the neuroepithelium and gave rise in 75 % of the cases, to specimens whose neural groove was wide open (Fig. 1G). By comparison, an embryo developed to approximately stage 14 and grown for 180 min under atmospheric pressure reached stage 18; at which time its neural groove was deep and narrow (Fig. 1C). In any of the pressure/time combinations mentioned later, such comparisons could be made provided that the treatment is not so deleterious as to render comparison unnecessary. The effects that the various combinations used had on microfilaments and microtubules are summarized in Table 1.

It was found that a pressure of 3000 psi did not inhibit the invagination of the neural plate. Stage-14 embryos subjected to such a pressure for as long as 180 min developed to stage 18 (Fig. 1E) just as controls did (Fig. 1C). However, although neurulation always occurred, some deformities in the neuroepithelium were frequently noted (Fig. IF). Most often these deformities took the aspect of an abnormal mediad convergence of the epidermis which caused a crowding in of the neurocoele. Raising the pressure to the 4000 psi range disturbed the invagination. For instance stage-14 embryos which showed a flat neural plate (Fig. 1 A) exhibited, in 75 % of the cases, a wide open neural groove following a 180 min exposure at 4000 psi (Fig. 1G). Raising the pressure still more (4500 psi) while shortening the pressurizing period (120 min) induced less of an accentuated widening of the groove (Fig. 1H). At a pressure of 4000 psi, maintained for 180 min, the so-caHed bottle-shaped cells (Fig. II) lost their dissymmetry; their apices widened and assumed a width equal to that of the base (Fig. 1 J).

Ultrastructurally, these cells lost the apical ring of microfilaments (Fig. 3 A) which is normally encountered in untreated embryos (Fig. 2 A). These effects were reversible. Indeed, in all stage-14 embryos treated at 4000 psi for 180 min, the invagination resumed upon decompression. Filaments reappeared in the cells, dissymmetry was achieved anew and neurulation proceeded though it led to the appearance of deformities in the neural tubes. One typical example of such deformities was the production of an abnormally large neurocoele (Fig. 4F). When a pressure of 4000 psi was applied to stage-14 embryos for as long as 330 min, 75 % of the pressurized specimens showed rounded neuroepithelial cells in a completely disrupted epithelium (Fig. 3B). Similar results were obtained in 88 % of the stage-16 embryos kept at 4500 psi for 180 min (Fig. 3C). At pressure/time combinations of 4000 psi/330 min and 4500 psi/180 min, microtubules were no longer found in the neuroepithelial cells, which, we emphasize, became round. This situation was not reversible upon decompression. However, the general aspect of the epithelium could resist higher hydrostatic pressures provided the duration of treatment was made shorter. For instance, a pressure of 5000 psi, held for 90 min, induced in 10 % of the embryos a widening of the neural groove (Fig. 3D). In such specimens, the cells, though devoid of microfilaments, still contained microtubules (Fig. 3 F). in the remaining 90% of the embryos exposed to 5000 psi/90 min, both microfilaments and microtubules were lost and the neuroepithelium was unrecognizable (Fig. 3E).

A pressure of 6000 psi, maintained for 15 min, disturbed only slightly the general aspect of the neuroepithelium (Fig. 4A). Here, the neuroepithelial cells lost their filaments but they still contained some microtubules as small portions of these could occasionally be found (Fig. 4C). Following a 45-min exposure to this same pressure, the neuroepithelium became unrecognizable, all cells assumed a round shape (Fig. 4B) and they lost all of their microtubules. Finally, a pressure of 10000 psi applied for 5 min, barely modified the epithelialnature of the neuroepithelium (Fig. 4D). The treatment did not rid the cells of all of their microtubules. In most cells, very short segments of microtubules, though admittedly quite scarce, could still be found. Prolonging the treatment to 15 min induced the cells to become rounded (Fig. 4E) and a total loss of microtubules.

In cases where the pressure used destroyed the apical bundle of microfilaments, yolk platelets, mitochondria and pigment granules were found closer to the apical surface of the cell. The apical surface became smooth and devoid of its characteristic microvilli.

Treatments which induced a rounding of the cells also induced cytolysis in some cells. The degree of cytolysis observed was related to the pressure/time combination employed. The cells from the superficial layer were more prone to cytolysis just as they were to rounding and dissociation. Generally, rounded cells displayed denser cytoplasm and round nucleus; their usual perinuclear clumps and islets of heterochromatin were virtually absent. Under high pressure (e.g. 4000 psi/180 min) a swelling of the perinuclear space and of the endoplasmic reticulum was regularly observed (Fig. 4G, H).

Although most of the work done on the effects of high pressures dealt with protozoa, isolated cells and marine eggs, some data are available concerning metazoa. Along that line, Ebbecke (1944) indicated that pressures of 15000–20000 psi are required for the glandular cells of the frog’s pharynx to become round. Also, Tilney & Cardell (1970) showed that a pressure of 6500 psi, applied for 30 min, causes the disruption of the microfilaments contained in the microvilli of the cells forming the small intestine of the salamander. Finally, O’Connor, Houston & Samson (1974) did not succeed in depolymerizing neuronal microtubules in adult frogs even after a 30 min exposure to 10000 psi.

High hydrostatic pressures have been reported to depolymerize cytoplasmic microtubules and to produce shape changes in several cell types (Zimmerman, 1971 ; a review). The pressure/time combinations reported to be needed to destroy these organelles vary somewhat depending on the experimental model used. Depolymerization is achieved at 4000 psi/10 min in Actinosphaerium (Tilney, Hiramoto & Marsland, 1966) while it occurs at 10000 psi/10 min in HeLa cells (Salmon, Goode, Maugel & Bonar, 1976). It is also known that high pressure affects microfilaments. For instance, in the eggs of Arbacia, microfilaments are disrupted and furrowing is inhibited by pressure/time combinations ranging from 5000 psi/4 min (Marsland, 1956) to 6500 psi/ 0·1 min (Marsland, 1970).

We found that the microfilaments which are usually assembled in the form of a ring at the apex of the neuroepithelial cells in Xenopus, were more sensitive to high pressure than were microtubules. Microfilaments were always lost before microtubules disappeared. For a pressure treatment lasting a given period of time (e.g. 180 min), a small increment of 12 % over 4000 psi (bringing the pressure to 4500 psi) induced the disappearance of microtubules (Table 1). Yet, at this pressure of 4500 psi, 120 min did not suffice to rid the cells of their microtubules; whereas an increment of 10% over that 4500 psi achieved it, even if the duration of treatment was reduced by 25 %. Clearly, microtubules became increasingly sensitive as the pressure exerted was increased. The same applied to microfilaments.

The pressure-induced disappearance of microfilaments was always accompanied by a release of the apical constriction, that is to say by a loss of cell dissymmetry. Furthermore, in the reversibility tests, we have always observed the reappearance of microfilaments together with a return to dissymmetry. In this respect, our observations concur with Karfunkel’s view (1972) derived from work done on chicken embryos exposed to cytochalasin B, that there exists a relationship between microfilaments and apical constrictions. Although our approach avoided the pitfalls of cytochalasin B, it gave no indications as to the cellular processes that might have been disturbed by the high pressures used. On that account, the effects of pressure on a variety of cell functions have been studied extensively (Zimmerman, 1970a, 1971, reviews). However, such a study has not been found for an amphibian. In essence, the study of ‘primitive systems’ has shown that pressures less than 6000 psi usually have a negligible effect on the rate of cell metabolism. Higher pressures are required to delay significantly DNA, RNA and protein synthesis. Moreover, cellular permeability in Arhacia is not affected at 4000 psi (Zimmerman, 1970b); while it seems that it is affected in the onion Allium cepa at a pressure of 7500 psi (Murakami, 1963). Our work confirmed the existence of a relationship between the elongated state of neuroepithelial cells and the presence of microtubules. Indeed, conditions that rid the cells of their microtubules induced a loss of cell asymmetry. Additionally, we showed that the pressure-induced disintegration of microtubules was, in most cases, not reversible. This is of interest, for in the neuroepithelial cells of the chick embryo the depolymerization of microtubules under cold exposure was shown to be reversible (Auclair & Messier, 1974). Here, it appears as though once dissymmetry is gone, the additional loss of asymmetry will leave the cells incapable of reverting to their original shape. The difficulty is not one of geometry only, for upon releasing those pressures that destroy microtubules neither they, nor microfilaments, will reappear. Our results suggest that there may exist a factor responsible for the polymerization of microfilaments and microtubules which is irreversibly affected beyond a given pressure/time combination. A similar view is held by O’Connor et al. (1974) who have suggested that pressure could cause depolymerization of microtubules indirectly either by a pressure stimulus response or by a pressure effect on some stabilizing or control factor. Forer & Zimmerman’s (1976) recent finding agrees with this suggestion.

As part of our experiments, relatively young embryos (stage 15–16) were subjected for 15 min to a pressure of 6000 psi. In these, the neuroepithelial cells maintained their elongated shape and microtubules were present. Older embryos (stage 20) submitted to the same conditions also showed elongated cells containing microtubules. It follows that, in the domain of the stages analysed, microtubules were not less sensitive from one stage to the other. However, when stage-15–16 embryos were pressurized for 90 min at 5000 psi their cells became round and they lost their microtubules; whereas stage-20 embryos treated in the same way had slightly elongated cells showing microtubules. Thus, it appears that microtubules in older specimens were more resistant than those found in younger embryos. More work is needed to determine whether the increased resistance derived from the fact that (a) the subunits microtubules in older specimens possess a greater degree of macromolecular bonding and therefore are better protected or (b) more differentiated cells host stabilizing factors capable of counteracting the effect of pressure or (c) the inclusion of a cell in a more differentiated (more cohesive, tissue endows it with better protection.

Our results underline, once again, the role of microfilaments and microtubules in neurulation. The loss of one of the organelles is sufficient to jeopardize neurogenesis. In previous studies we have said much on the possibility that the phenomenon of interkinetic nuclear migration might also have something to do with neurulation. In that sense, it would have been interesting to learn how the nuclear movements are affected by high pressures. However, regrettably, our present experimentation offers no arguments concerning interkinetic nuclear migration. These nuclear movements cannot be followed with precision in the neuroepithelial cells of Xenopus. A similar study is presently being carried out on the better suited neural epithelium of Triturus viridescens.

Des embryons de Xénopus laevis, ayant atteint les stades 14 à 20, ont été soumis, pour des périodes variant de 5 à 330 min, à des pressions hydrostatiques allant de 500 à 10000 psi. Les spécimens ont été fixés à des pressions correspondantes et leur neuroépithélium a été étudié en microscopie photonique et électronique. La pression de 3000 psi, maintenue pour aussi longtemps que 180 min, n’a pas empêché la neurulation de se produire bien qu’elle ait induit des malformations mineures du neuroépithélium. La pression de 4000 psi, appliquée pendant 180 min, a détruit l’anneau de microfilaments normalement observé à l’apex des cellules et a inhibé aussi la neurulation. Les cellules ont perdu leur dissymétrie. L’effet était réversible. Le fait d’augmenter la durée du traitement jusqu’à 330 min a entraîné la perte des microtubules et provoqué l’arrondissement des cellules. Cet état n’était pas réversible. Nos résultats indiquent que les microfilaments sont plus sensibles aux pressions que ne le sont les microtubules, que les deux organites deviennent de plus en plus labiles au fur et à mesure qu’augmente la pression et que, finalement, les microtubules des embryons les plus âgés résistent mieux aux pressions. Enfin, on a montré une corrélation entre la présence des microfilaments et l’état de constriction des apex cellulaires de même qu’une corrélation entre la présence des microtubules et la forme allongée des cellules.

This work was supported by the Medical Research Council of Canada. We are grateful to Mme Geneviève Anglade and to Mlle Lucie Héroux for their technical assistance and to Mrs Barbara Lafrenière for her help with the English presentation of this text.