/. Embryol. exp. Morph. Vol. 48, pp. 215-223, 1978
Printed in Great Britain © Company of Biologists Limited 1978
215
Scanning electron microscopy of cells isolated from
amphibian early embryos
By MARTIN STANISSTREET 1 AND JAMES L. SMITH 1
From the Department of Zoology, University of Liverpool
SUMMARY
Cells have been dissociated from Xenopus and Ambystoma late blastulae, allowed to
adhere to glass coverslips, and studied by scanning electron microscopy. Xenopus ectoderm
cells initially show filopodia; later larger single pseudopodia are formed. Ambystoma ectoderm cells show fewer filopodia than Xenopus ectoderm, but later form pseudopodia.
Ectoderm cells of both Xenopus and Ambystoma show links between adjacent cells. Xenopus
endoderm cells do not show filopodia initially, but later show large pseudopodia.
INTRODUCTION
When amphibian early embryos are dissociated into single cells, the cells
become motile and can reaggregate (Holtfreter, 1946, 1947). Recently, the
properties of cells isolated from amphibian embryos have been studied, since
they might provide information on the mechanisms involved in normal morphogenesis. These studies have shown that the types of cell movement shown
in vitro and the proportion of cells exhibiting movement vary with the stage
of the embryo from which the cells are taken and are different for different
germ layers (Satoh, Kageyama & Sirakami, 1976; Johnson, 1976). This might
be expected if such movements are significant to normal morphogenesis. In
addition, hybrid embryos which do not gastrulate, or gastrulate abnormally,
show altered motility in vitro (Johnson, 1969, 1970, 1972), suggesting that
gastrulation is effected by the motile properties of individual cells.
The scanning electron microscope has revealed much about the cellular
changes which accompany early embryogenesis. Cells of the different germ
layers have different surface structures (Smith, Osborn & Stanisstreet, 1976)
and during gastrulation the invaginating cells change shape (Tarin, 1971) and
form microvilli (Monroy, Baccetti & Denis-Donini, 1976). However, less is
known about possible changes in the surface architecture of isolated amphibian
cells and so we have used the scanning electron microscope to study cells
dissociated from blastulae of Xenopus laevis and Ambystoma mexicanum. An
abstract of this work is being published in J. Anat.
1
Authors' address: Department of Zoology, University of Liverpool, Brownlow St.,
PO Box 147, Liverpool L69 3BX. U.K.
216
M. STANISSTREET AND J. L. SMITH
MATERIALS AND METHODS
Xenopus embryos were obtained by injecting pairs of adults with chorionic
gonadotrophin ('Chorulon', Intervet Ltd.) and were staged according to
Nieuwkoop & Faber (1956). The jelly coats were removed chemically by a
modification of the method of Dawid (1965): embryos were placed in 2 %
cysteine hydrochloride in 10 % Steinberg saline (Steinberg, 1957) brought to
pH 7-8 with 2 M NaOH. The embryos were washed and. subsequently cultured
in 10 % Steinberg saline, pH 7-3. When they had reached the late blastula stage
(stage 8x), the vitelline membranes were removed manually and the embryos
were dissected using iris scalpels to obtain ectoderm cells from the animal pole
and endoderm cells from the floor of the blastocoel. Ambystoma embryos
were obtained from natural matings of laboratory-maintained adults. When
the embryos had reached the blastula stage the jelly and vitelline membranes
were removed manually, and the embryos were dissected in the same way as
the Xenopus embryos.
The tissues were transferred to dissociation medium (Landesman & Gross,
1968) with 4 mM EDTA over an agar substrate for 1 h. At the end of this
period the cells were dispersed by gentle micropipetting. They were then
transferred to 10 mm glass coverslips in plastic Petri dishes containing Stearns'
medium (Stearns & Kostellow, 1958), allowed to settle, and were cultured for
various times (routinely \ h, 1 h and 3 h). The cells were then fixed by replacing
the medium with fixative (2-5 % glutaraldehyde plus 2 % paraformaldehyde
with 2-5 mM calcium chloride in 0-1 M cacodylate buffer, pH 7-2, modified from
Karnovsky, 1965). Following fixation overnight, the coverslips were washed
in changes of cacodylate buffer containing 2-5 mM calcium chloride over 2 h.
The cells were dehydrated in an acetone series and the absolute acetone was
F I G U R E S 1-8
Fig. 1. Ectoderm cells isolated from Xenopus blastula and cultured on glass for \ h.
Cells show filopodia and 'cytoplasmic skirts', x 825.
Fig. 2. Ectoderm cell isolated from Xenopus blastula and cultured on glass for •} h.
Cell shows filopodia attached to substrate, x 1550.
Fig. 3. Ectoderm cells isolated from Xenopus blastula and cultured on glass for
i h. Adjacent cells are linked by filopodia. x 1865.
Fig. 4. Ectoderm cells isolated from Xenopus blastula and cultured on glass for 1 h.
Cells have become polarized and have formed single pseudopodia. x 870.
Fig. 5. Ectoderm cell isolated from Xenopus blastula and cultured on glass for 3 h.
Cell has a single pseudopodium. x 1640.
Fig. 6. Ectoderm cell isolated from Xenopus blastula and cultured on glass for
3 min. Cell shows many fine filopodia attached to substratum, x 1590.
Fig. 7. Ectoderm cell isolated from Ambystoma blastula and cultured on glass
for -| h. Cell has fewer, shorter filopodia than Xenopus cell, x 380.
Fig. 8. Ectoderm cells isolated from Ambystoma blastula and cultured on glass for
I h. Cell surface is smooth, x 415.
S.E.M. of amphibian embryonic cells
217
218
M. STANISSTREET AND J. L. SMITH
replaced with liquid C0 2 . The cells were then dried using the critical point
method, and the coverslips affixed to microscope stubs. Finally, the cells were
coated with gold-palladium and observed and photographed with a Cambridge
'Stereoscan' scanning electron microscope.
RESULTS
Xenopus ectoderm
Xenopus ectoderm cells fixed after \ h in Stearns' medium appeared slightly
flattened and attached to the glass by cytoplasmic 'skirts' (Fig. 1). Some cells
were covered with filopodia attached to the coverslip (Figs. 1 and 2) and, in
some cases, to adjacent cells (Fig. 3). After 1 h in Stearns' medium, some cells
appeared similar to those in the \ h sample, but some cells had become polarized
and were attached to the glass coverslip by single large pseudopodia (Fig. 4).
After 3 h, most of the cells were polarized and attached by single pseudopodia
(Fig. 5), and in some cases these pseudopodia formed a 'stalk' which raised
the cell from the substrate. Thus Xenopus ectoderm cells appeared to change
their method of attachment to glass over the 3 h period studied.
With these cells, an additional experiment was performed in an attempt to
determine the time of appearance of the filopodia. Cells were fixed after 3 min
in Stearns' medium. It was found that the cells were attached to the glass
after this time, although scanning electron microscopy showed that they were
less flattened than the cells of the \ h samples, and the cells possessed many
fine filopodia (Fig. 6). The upper surface of the cells showed collapsed filopodia,
and thus it is probable that the cells had developed filopodia while in the
dissociation medium.
F I G U R E S 9-16
Fig. 9. Ectoderm cell isolated from Ambystoma blastula and cultured on glass for
\ h. Cell surface is covered with projections, x 730.
Fig. 10. Ectoderm cell isolated from Ambystoma blastula and cultured on glass
for 1 h. Like Xenopus cells, cell has become polarized and has formed a single
pseudopodium. x 670.
Fig. 11. Ectoderm cell isolated from Ambystoma blastula and cultured on glass for
3 h. Cell shows single pseudopodium. x 780.
Fig. 12. Ectoderm cells isolated from Ambystoma blastula and cultured on glass
for 3 h. Links are seen between adjacent cells, x 670.
Fig. 13. Endoderm cells isolated from Xenopus blastula and cultured on glass for
{h. Cells have a 'raspberry-like' appearance and, unlike ectoderm, show no
filopodia. x38O.
Fig. 14. Endoderm cell isolated from Xenopus blastula and cultured on glass for
3 h. Cell has formed a large pseudopodium. x 775.
Fig. 15. Endoderm cell isolated from Ambystoma blastula and cultured on glass
for \ h- Cell is relatively featureless, x 370.
Fig. 16. Endoderm cells isolated from Ambystoma blastula and cultured on glass
for 1 h. Cells at 3 h are similar in appearance, x 175.
S.E.M. of amphibian embryonic cells
219
220
M. STANISSTREET AND J. L. SMITH
Ambystoma ectoderm
Ambystoma ectoderm cells fixed after \ h in Stearns' medium showed fewer
filopodia than the corresponding Xenopus cells (Fig. 7). Two types of cells
were present; one with a smooth surface (Fig. 8) and the other with a surface
covered with short protrusions (Fig. 9). Since the blastula ectoderm is more
than one cell thick at the stage used, it is possible that these two cell types
correspond to inner and outer ectoderm, and further experiments will be
undertaken to attempt to clarify this problem. After 1 and 3 h, some cells had
become polarized and showed single pseudopodia (Figs. 10 and 11 respectively).
Thus Ambystoma ectoderm cells, like those of Xenopus, change their appearance
over the 3 h period studied. As was observed in Xenopus cultures, some adjacent
cells were linked by fine processes although these appeared somewhat different
from the links between Xenopus cells (Fig. 12).
Xenopus endoderm
In general, isolated Xenopus endoderm cells appeared similar to the endoderm of intact embryos (Smith et al. 1976) - the cell membrane had collapsed
onto the underlying yolk platelets to give a 'raspberry-like' appearance. Unlike
ectoderm, isolated endoderm cells did not show filopodia or obvious cytoplasmic 'skirts'. Cells fixed after -^h and 1 h in Stearns' medium appeared
similar and were slightly flattened onto the glass (Fig. 13). Endoderm cells
fixed after 3 h in Stearns' medium presented a different appearance - like the
3 h ectoderm cells they had formed pseudopodia (Fig. 14).
Ambystoma endoderm
The isolated Ambystoma endoderm cells were relatively featureless and
showed neither filopodia, 'skirts' nor pseudopodia (Fig. 15). They were fixed
after \ h, 1 h and 3 h in Stearns' medium, and did not alter their appearance
over this time (Fig. 16).
Mixed Xenopus ectoderm and Ambystoma ectoderm
In an attempt to determine whether the links between adjacent ectoderm
cells were the mechanism by which cells were rejoining or the result of incomplete separation of cells during dissociation, mixed cultures of Xenopus
and Ambystoma were examined, in the hope that size differences between
Xenopus and Ambystoma would act as a marker (see Discussion). In practice,
size was not an absolutely critical marker, and so other markers are being
sought.
S.E.M. of amphibian embryonic cells
221
DISCUSSION
The present results show that the ultrastructural appearance of cells isolated
from different regions of amphibian early embryos is different, and confirms
early light microscopical observations that when such cells are cultured in vitro
their appearance changes over the first few hours (Holtfreter, 1946, 1947).
Holtfreter (1947) described isolated neurula ectoderm cells as forming 'tapering
pseudopodia' which become 'web-like membranes', and these could correspond
to the filopodia and cytoplasmic 'skirts' reported here on ectoderm cells. More
recently, filopodia have been demonstrated by both transmission and scanning
electron microscopy on mesoderm cells undergoing migration in amphibian
gastrulae (Nakatsuji, 1975, 1976; Monroy et al. 1976) and are probably one
of the mechanisms by which these cells invaginate (Nakatsuji, 1976). However,
in a recent comprehensive S.E.M. study of Xenopus embryos, Keller & Schoenwolf (1977) showed that blastula ectoderm cells in situ do not have filopodia,
whereas migrating mesoderm cells of gastrulae do. Our finding that filopodia
are already present after only 3 min in Stearns' medium suggests therefore
that they are formed during dissociation in EDTA. Two possibilities are that
cells in EDTA become motile and therefore form filopodia, or that during
the time required for dissociation, cells age to become gastrula cells, and that
those cells with filopodia are those that would have become mesoderm cells.
The fact that animal pole ectoderm was used in our experiments favours the
former of these possibilities, and the observations that embryonic neural retina
cells produce filopodia after dissociation, and that cell surface structure is
partly dependent upon the method of dissociation, show that the morphology
of cells can alter upon isolation (Ben-Shaul & Moscona, 1975).
In the present experiments Xenopus endoderm and ectoderm and Ambystoma
ectoderm formed pseudopodia by which they adhered to the glass. Similar
pseudopodia have been observed with light microscopy on isolated cells
(Johnson, 1976). They are also found on migrating cells in situ, and have
been ascribed a functional role in cell migration (Nakatsuji, 1974).
Spiegel & Spiegel (1977) have recently reported the results of S.E.M. observations of cells dissociated from sea-urchin early embryos, and they described
structures very similar to those which we show here. Sea-urchin cells are
initially covered with microvilli, but after isolation become polarized and form
pseudopodia and cytoplasmic 'skirts'. The pseudopodia may form stalks which
raise the cell above the substratum, and we have observed similar forms in
isolated amphibian cells.
The conditions under which the cells were cultured in our experiments are
the same as those which allow reaggregation (Stanisstreet & Smith, 1978). In
some cases, filopodia were observed linking adjacent cells in both Xenopus and
Ambystoma ectoderm. Similar features have been seen in reaggregating seaurchin cells (Spiegel & Spiegel, 1977), and in reaggregating embryonic chick
15
EMB 48
222
M. STANISSTREET AND J. L. SMITH
heart cells (Shimada, Moscona & Fischman, 1974), and it is tempting to
suggest that cells make initial contact in this way. However, with amphibian
ectoderm, particularly Xenopus ectoderm, it is difficult to ensure complete
separation of cells, since the outer layer of pigmented ectoderm is quite resistant
to EDTA dissociation; and it could be that such links are the result of incomplete separation of blastomeres following dissociation or cell division.
For this reason, mixed cultures of Ambystoma and Xenopus ectoderm were
made, in the hope that the size difference of the cells could be used as a marker.
In long term cultures mixed aggregates did form, but size difference was not
an absolutely critical marker for S.E.M. Under the conditions used, amphibian
cells would not only reaggregate, but would also sort out (Townes & Holtfreter,
1955; Stanisstreet & Smith, 1978), and it is of interest that endoderm cells did
not form filopodia. It is not known what controls the sorting of amphibian
cells, but it might be significant that cells of different tissues show different
surface features when isolated.
It is a pleasure to thank C. J. Veltkamp for his expert help with the scanning electron
microscopy, and R. Wall who provided the Ambystoma embryos and suggested improvement
to the manuscript.
REFERENCES
Y. & MOSCONA, A. A. (1975). Scanning electron microscopy of embryonic
neural retina cell surfaces. Devi Biol. 44, 386-393.
DAWID, I. B. (J965). Deoxyribonucleic acid in amphibian eggs. /. molec. Biol. 12, 581-599.
HOLTFRETER, J. (1946). Structure, mobility and locomotion in isolated embryonic amphibian
cells. /. Morph. 79, 27-62.
HOLTFRETER, J. (1947). Observations on the migration, aggregation and phagocytosis of
embryonic cells. /. Morph. 80, 25-55.
JOHNSON, K. E. (1969). Altered contact behaviour of presumptive mesodermal cells from
hybrid amphibian embryos arrested at gastrulation. / . exp. Zool. 170, 325-332.
JOHNSON, K. E. (1970). The role of changes in cell contact behaviour in amphibian gastrulation. /. exp. Zool. 175, 391-428.
JOHNSON, K. E. (1972). The extent of cell contact and the relative frequency of small and
large gaps between presumptive mesodermal cells in normal gastrulae of Rana pipiens
and the arrested gastrulae of Rana pipiens $ x Rana catesbeiana $ hybrid. / . exp. Zool.
179, 227-238.
JOHNSON, K. E. (1976). Circus movements and blebbing locomotion in dissociated embryonic
cells of an amphibian, Xenopus laevis. J. Cell Sci. 22, 575-583.
KARNOVSKY, M. J. (1965). A formaldehyde/glutaraldehyde fixative of high osmolarity for
use in electron microscopy. / . Cell Biol. 27, 137A-138A.
KELLER, R. A. & SCHOENWOLF, G. C. (1977). An S.E.M. study of cellular morphology,
contact and arrangement, as related to gastrulation in Xenopus laevis. Roux's Archives of
Developmental Biology 182, 165-186.
LANDESMAN, R. & GROSS, P. R. (1968). Patterns of macromolecular synthesis during
development of Xenopus laevis. 1. Incorporation of radioactive precursors into dissociated
embryonic cells. Devi Biol. 18, 571-589.
MONROY, A., BACCETTI, B. & DENIS-DONINI, S. (1976). Morphological changes in the
surface of the egg of Xenopus laevis in the course of development. 111. Scanning electron
microscopy of gastrulation. Devi Biol. 49, 250-259.
NAKATSUJI, N. (1974). Studies on the gastrulation of amphibian embryos: pseudopodia in
the gastrula of Bufo bufo japonicus and their significance to gastrulation. /. Embryol. exp.
Morph. 32, 795-804.
BEN-SHAUL,
S.E.M. of amphibian embryonic cells
223
N. (1975). Studies on the gastrulation of amphibian embryos: light and electron
microscopic observation of a urodele Cynops pyrrhogaster. J. Embryol. exp. Morph. 34,
669-685.
NAKATSUJI, N. (1976). Studies on the gastrulation of amphibian embryos: ultrastructure of
the migrating cells of anurans. Roux's Archives of Developmental Biology 180, 229-240.
NIEUWKOOP, P. D. & FABER, J. (1956). Normal Table o/Xenopus laevis Daud. Amsterdam:
North Holland Publishing Co. Ltd.
SATOH, N., KAGEYAMA, T. & SIRAKAMI, K-T. (1976). Motility of dissociated embryonic cells
in Xenopus laevis: its significance to morphogenetic movements. Development, Growth and
Differentiation 18, 55-67.
SHIMADA, Y., MOSCONA, A. A. & FJSCHMAN, D. A. (1974). Scanning electron microscopy
of cell aggregation: cardiac and mixed retina-cardiac cell suspensions. Devi Biol. 36,
428-446.
SMITH, J. L., OSBORN, J. C. & STANJSSTREET, M. (1976). Scanning electron microscopy of
lithium-induced exogastrulae of Xenopus laevis. J. Embryol. exp. Morph. 36, 513-522.
SPIEGEL, E. & SPIEGEL, M. (1977). A scanning electron microscope study of early sea urchin
reaggregation. Expl Cell Res. 108, 413-420.
STANISSTREET, M. & SMITH, J. L. (1978). Scanning electron microscopy of aggregates of
cells from normal embryos and lithium-induced exogastrulae of Xenopus laevis. Acta
Embryol. Exper. issue 1 (in the Press).
STEARNS, R. N. & KOSTELLOW, A. B. (1958). Enzyme induction in dissociated embryonic
cells in The Chemical Basis of Development (ed. W. D. McElroy & B. Glass), pp. 448-457.
Baltimore: Johns Hopkins Press.
STEINBERG, M. (1957). Carnegie Jnst. Wash. Yearbook 56, 347 (report by J. D. Ebert).
TARIN, D. (1971). Scanning electron microscopical studies of the embryonic surface during
gastrulation in Xenopus laevis. J. Anat. 109, 535-548.
TOWNES, P. L. & HOLTFRETER, J. (1955). Directed movements and selective adhesion of
embryonic amphibian cells. /. exp. Zool. 128, 53-120.
NAKATSUJF,
{Received 3 June 1978, revised 27 July 1978)
15-2