Localization of microfilament bundles in the upper layer of the primitive-streak-stage chick blastoderm

In several investigations microfilament bundles have been observed just under the apical surface of the neural plate cells of amphibian embryos (Baker & Schroeder, 1967; Burnside, 1971 ; Löfberg, 1974). Their presence was correlated with the apical constriction of the cells, and with neural tube formation. Similar observations have been made in chick neurulae (Karfunkel, 1972; Bancroft & Bellairs, 1975; Nagele & Lee, 1980). During these studies it was suggested that microfilament bundles might be characteristic of all cells lining grooves in early morphogenesis.

In the present investigation, this hypothesis was tested in the primitive-streakstage chick blastoderm. During avian gastrulation, cells from the upper layer ingress along an anteroposterior axis, resulting in the formation of a groove. Examination of the primitive-streak region in the chick blastoderm revealed that as the upper layer cells approach the primitive streak, many of them elongate and have an irregular shape (Balinsky & Walther, 1961 ; Ruggeri, 1966 and 1967). Variations in apical cell morphology between cells destined to ingress in the groove and cells situated more laterally were observed by scanning electron microscopy (SEM) (Bancroft & Bellairs, 1974; Jacob, Christ, Jacob & Bijvank, 1974). SEM studies of the primitive streak revealed that the cells become progressively elongated and flask-shaped (Solursh & Revel, 1978). This technique did not allow presumptive mesoblast to be distinguished from the presumptive endoblast cells during ingression (Wakely & England, 1977). A basal lamina is present in close contact with the ventral surface of the upper layer (Low, 1967; Mestres & Hinrichsen, 1974; Sanders, 1979; Wakely & England, 1979). It is continuous in the lateral region, gradually becomes interrupted towards the primitive streak, and only sparse fragments of basal lamina can be recognized in the primitive streak (Vakaet, Vanroelen & Andries, 1980). Examination of the junctional complexes between the upper layer cells revealed the presence of gap and tight junctions in the region close to the dorsal surface (Bellairs, Breathnach & Gross, 1975).

The aim of the present investigation was to study the architecture of the microfilament bundles within the upper layer cells, and the distribution of the upper layer cells showing microfilament bundles. These results will allow us to test the hypothesis that microfilament bundles are restricted to cells lining grooves.

Eggs from commercial stock were incubated for 16 h at 38 °C. Blastoderms of stage 5 of Vakaet (1970) (corresponding to stage 3+ of Hamburger & Hamilton, 1951) were selected. The blastoderms were removed from the eggs and fixed as described by Sanders (1979). Summarizing, a freshly prepared solution of 1 % (w/v) glutaraldehyde and 1 % (w/v) tannic acid (mol. wt. 1701) in 0 ·1 M cacodylate buffer at pH 7 ·4 was used as a fixative, since previous observations suggest that this hypotonic fixative results in an improved visualization of microfilament bundles (Vanroelen & Vakaet, 1981). The embryos were fixed at room temperature for 2 h. After a thorough rinse, the embryos were post-fixed in a 1 % (w/v) osmium tetroxide solution in the same buffer for 1 h. After dehydration in a graded series of ethanol, the embryos were embedded in Epon 812. Ultrathin sections were stained in a 2 % (w/v) aqueous solution of uranyl acetate for 30 min and in lead citrate for 30 s (Reynolds, 1963). The sections were examined with a JEOL 100 B electron microscope. Low-magnification electron micrographs (× 1700) of each ultrathin section were mounted. A drawing was made from these mounted micrographs in order to obtain the exact localization of the details. The thickness of the microfilaments was estimated on electron micrographs at a final magnification factor of × 100000 and × 200000, using a Peak scale 10 × magnifier. Sections of 14 blastoderms were examined.

A drawing of a stage-5 (Vakaet, 1970) chick blastoderm is represented in Fig. 1. The embryos were studied in three regions of the area pellucida: (1) the primitive streak, (2) lateral to the primitive streak, and (3) the neural plate. An account on the fine structure of the upper layer cells in the different regions has been reported by others (see Introduction), and will therefore not be repeated in detail here.

Our technique is not adequate for the localization of individual microfilaments. Therefore, only the localization of microfilament bundles will be described.

Figure 2 is a drawing made from mounted low-magnification electron micrographs of a transverse section of the primitive streak. It demonstrates that the upper layer cells in the primitive streak do not have a regular shape. The cells are situated above one another, some of them are extremely elongated. In the centre of the streak, some of the upper layer cells appear narrower near the dorsal surface.

Micrographs at higher magnification show clearly delineated plasma membranes. Just under the apical surface, the lateral plasma membranes show a junctional complex, together with plasma-membrane-associated dense accumulations. In all the cells, microfilament bundles are present in close association with these dense accumulations. High-magnification electron micrographs reveal that the diameter of the microfilaments varies from 4 to 6 nm. In some cells, the microfilament bundles appear to be restricted to a 0 ·5 μm array from the junctional complex (Fig. 5B and 6 A). We have not seen transverse sections of microfilament bundles in the apical cytoplasm at some distance from the junctional complex. In other cells, the microfilament bundles extend from one junctional complex to the other, within the cell apex. This is especially apparent in the cells of the centre of the primitive streak. These cells are extremely elongated and often very narrow at their apical surface (Fig. 5 A).

Microfilament bundles extending from one junctional complex to the other are also observed in a few cases in presumptive columnar cells. In these cases the possibility exists that the sections are tangential to the lateral plasma membrane of the upper layer cell (Fig. 5C). Small microfilament bundles were also observed at the bases of apical surface projections and in microvilli.

Tangential sections of the apical surface of the primitive-streak cells illustrate the close association of the microfilament bundles with the junctional complexes (Fig. 7). As the apical surface is very irregular and shows a lot of projections, perfectly tangential sections are hard to obtain. This perhaps explains why we were not able to demonstrate circumferential rings of microfilament bundles in the cell apex, although their presence is inferred from the observations obtained from our transverse sections. Mitochondria, small vacuoles (0 ·1 μm), and microtubules are observed in the apical cytoplasm.

The region studied is represented in Fig. 3. In this region, an epithelium with cells situated above one another is observed beside an epithelium with columnar cells covering the entophyllic crescent. The presence of cells situated above one another in the section does not exclude the possibility that these cells, as well, are extending from the dorsal to the ventral surface of the upper layer. Indeed, it may be the result of an even small angle between the columnar cell axis and the knife during sectioning.

All the cells have apical microfilament bundles in close contact with the junctional complexes (Fig. 6B). When the upper layer cells are narrower at their apical surface, the microfilament bundles extend from one junctional complex to the other. Transverse sections of microfilament bundles could not be detected. However, some differences to the results from the primitive streak appear: (a) microfilament bundles are less obvious, (b) the junctional complex, if compared to the lateral plasma membranes, is not so highly contrasted, and (c) the surface projections seem to be different from these observed in the groove. The latter is confirmed by previous SEM observations (Bancroft & Bellairs, 1974; Vakaet, unpublished).

The region selected for ultrastructural observation is represented in Fig. 4. The cells in this layer are elongated, extending from the dorsal to the ventral surface, or they appear above one another. Rounded cells in mitosis can be recognized near the apical surface.

In this region, microfilament bundles are also ultrastructurally observed in close contact with the junctional complex (Fig. 6C). Cells, apparently narrower near the apical surface, show microfilament bundles extending from one junctional complex to the other.

Do all epithelial cells lining grooves during morphogenesis possess microfilament bundles at their apical ends? Microfilament bundles have been observed during neurulation in the chick (Karfunkel, 1972; Nagele & Lee, 1980) and in Amphibia (Baker & Schroeder, 1967; Burnside, 1971; Lof berg, 1974), just under the apical surface of the folding neural plate cells. Bancroft & Bellairs (1977) showed the presence of microfilament bundles in cells which will form the otic, nasal and lens placodes, but could not detect them in cells destined to form the head epidermis. Balsinky & Walther (1961) suspected the presence of microfilament bundles in the centre of the primitive-streak-stage chick blastoderm, and described them as ‘strips of denser cytoplasm arranged transversely to the long axis of the cell’. Our study demonstrates that microfilament bundles are present just under the apical surface of the cells in the centre of the primitive streak, where they can be observed as dense bands of microfilaments in TEM sections. Cells in the centre of the primitive streak are very narrow at their dorsal surface. They are wedge- or flask-shaped, as was shown by SEM studies of this array (Wakely & England, 1977; Solursh & Revel, 1978; Vakaet, unpublished). Moreover, the nucleus of these cells frequently was observed in the same plane of the section as the narrow apical end. The microfilament bundles seem to form a continuous meshwork beneath the apical surface of these cells.

Is the presence of microfilament bundles in epithelia during morphogenesis restricted to cells lining grooves? From the literature (see above), it appeared that the presence of dense bands of microfilaments is correlated with the presence of grooves. Outside these grooves microfilament bundles were not detected. In our study, microfilament bundles can be detected in cells outside the centre of the groove, in the lateral region and the neural plate region. Outside the centre of the groove, presumptive columnar cells are present beside presumptive wedge- or flask-shaped cells. In the latter cells, microfilament bundles can be seen as dense bands (probably a continuous meshwork). The organization of microfilament bundles in a circumferential ring was widely accepted, and recently doubted by Nagele & Lee (1980) who observed discrete microfilament bundles. Our experiments, however, do not allow us to reject or accept the hypothesis of circumferential organization of the microfilament bundles. Indeed, on the one hand, the presence of a circumferential ring of microfilament bundles in the cell apex is probable because they were always observed near the junctional complex. Moreover, transverse sections of microfilament bundles at some distance from the junctional complex could never be detected. Finally, the tangential section of the dorsal surface showed half a ring. On the other hand, microfilament bundles extending from one junctional complex to the other within the cell, could suggest the presence of discrete bundles in the cell apex. However, in those cases, the possibility that the section was tangential to the lateral plasma membrane cannot be eliminated. Although present over the whole area pellucida, the electron density of the microfilament bundles compared to that of the cytoplasm is higher in the primitive-streak region than outside this region.

Several observations suggest that microfilament bundles are contractile. We observed that microfilament bundles were always more obvious in cells which were narrower at their apical surface and therefore probably constricted at their apical end. Baker & Schroeder (1967) remarked that the apical cell membrane is always folded, while the microfilament bundles are not. More conclusive evidence was provided by Karfunkel (1972), who used cytochalasin B, a microfilament-modifying agent, in the chick during neurulation. The absence of microfilament bundles and flattening of the neural folds after treatment with this product was observed. Recently, Nagele & Lee (1980) demonstrated that, during neurulation in the chick, actin is a component of the microfilament bundles.

Summarizing, the presence of microfilament bundles in epithelia, during embryogenesis, is certainly not restricted to cells lining grooves. As it appears from our study, at least the cells beside the groove, contain microfilament bundles in the cell apex. We think that they form an interconnected meshwork through the junctional complexes over the complete area pellucida. They appear as a continuous meshwork in the cell apex of presumptive wedge- and flaskshaped cells. The architecture of microfilament bundles in the apices of columnar cells remains uncertain. Because they are ubiquitous in the area pellucida of the primitive-streak-stage chick blastoderm, they cannot be regarded as a marker of differentiation in this tissue. If we agree with the view that microfilament bundles are contractile, they may be regarded as cellular organelles with a function during every morphogenetic event in the upper layer. These not only include the obvious phenomena, such as the formation and shortening of the streak, and the folding of the neural groove, but also the movements in the plane of the upper layer. Therefore, microfilament bundles may have a role in the further determination and differentiation of the different anlage fields of the upper layer.

The authors would like to thank Mr Ch. De Schepper for technical assistance, Mr F. De Bruyn for artwork, Mr J. Van Ermengem for photographic processing and Mrs N. Van den hende-Bol for typing our manuscript.

This work was supported by grant n° 3.9001.81 of the Belgian Fonds voor Geneeskundig Wetenschappelijk Onderzoek.