J. Embryo!, exp. Morph. Vol. 69, pp. 169-182, 1982
Printed in Great Britain © Company of Biologists Limited 1982
Organization of actin filaments in
early chick embryo ectoderm: an ultrastructural
and immunocytochemical study
By J. WAKELY1 AND R. A. BADLEY2
From the Anatomy Department, University of Leicester and
Biosciences Division, Unilever Research
SUMMARY
By combining transmission electron microscopy (TEM), scanning electron microscopy
(SEM) and immunocytochemistry at both light and TEM levels the organization of actin
in early chick ectoderm has been demonstrated. Every cell is encircled by a ring of actin
filaments which are associated with apically situated intercellular junctions. In this way
the actin ring of each individual cell is connected to that of surrounding cells and a continuous
network is formed. As clearly shown by immunofluorescence and SEM studies on intact
and Triton X-100 treated whole-mount preparations this network extends throughout the
entire layer, interrupted only at the blastoderm margins and the primitive streak. A itiore
diffuse arrangement was sometimes seen at cell bases. It is suggested that the network
extending throughout the ectoderm gives it stability and cohesion and is important in gross
morphogenetic movements involving the entire layer. The less-organized arrangement at cell
bases is associated more with movements of individual cells.
INTRODUCTION
Epithelial cells in vitro show two major behaviour patterns which are not
generally seen in fibroblasts. They form extensive and longstanding lateral
adhesions with their neighbours and the continuous sheets thus created move
as a unit with only the marginal cells undergoing active locomotion (Di Pasqyale,
1975). Such united movements of whole sheets of cells implies the existence of
some co-ordinating mechanism. Albrecht-Buehler (1979) has demonstrated
that when PtKI cells form cohesive epithelial sheets in vitro, actin microfilament bundles appear to continue in direction across cellular boundaries
thus providing a 'tissue' level of cytomuscular organisation and the possibility
of a mechanical form of communication between cells. Meittenen, Virtanen &
Linder (1978) describe a similar situation for sheets of rat hepatocytes in vitro.
The ectoderm of the early chick embryo is a single relatively flat layer of
1
Author's address: Anatomy Department, University of Leicester, University Road,
Leicester, England. LEI 7RH, U.K.
2
Author's address: Biosciences Division, Unilever Research, Colworth Laboratory,
Colworth House, Sharnbrook, Bedford, M44ILQ, U.K.
170
J. WAKELY AND R. A. BADLEY
Fig. 1. Diagram to show general features of the early (approx. stage 4) chick
embryo, viewed from the ectodermal side. AO = areaopaca;AP = areapellucida;
s = primitive streak.
epithelial cells which ultimately participates in a wide range of morphogenetic
movements, and whose general ultrastructure is relatively well known (Trelstad,
Hay & Revel (1967), Bancroft & Bellairs (1975). The emphasis of the present
study was on the arrangement of actin filaments both within individual cells
and throughout the layer as a whole.
This was of interest partly from the point of view of using the ectoderm as
a generalized and easily obtained isolated example of an epithelium in vivo,
and partly because the contraction of actomyosin filaments has been widely
implicated in many developmental processes in which epithelial deformation
occurs (Wessels et ah 1971; Karfunkel, 1972; Burnside, 1971; Burgess &
Schroeder, 1979). Normal embryos were examined by scanning (SEM) and
transmission (TEM) electron microscopy and immunocytochemistry at light
(LM) and EM levels. Triton X-100 extracted preparations, as frequently used
for cultured cells, (Brown, Levinson & Spudich, 1976; Small, Isenberg & Celis,
1978; Bell, Miller, Carraway & Revel, 1978; Ip & Fischman, 1979) were made
of whole embryos and also examined by SEM and TEM. It was hoped that
in this way any physical continuity of the cytoskeleton from cell to cell might
be more clearly visible, unobscured by other cellular components.
MATERIALS AND METHODS
White Leghorn eggs were incubated at 37 °C to produce embryos at stages
3-8 (Hamburger & Hamilton, 1951).
Five normal embryos each for scanning (SEM) or transmission electron
microscopy (TEM) were dissected from the eggs into Pannet & Compton
Ectodermal actin organization
171
saline (1924) washed free of adherent yolk then fixed overnight in Karnovsky's
(1965) fixative. Fixed embryos were washed in 0-2 M-sodium cacodylate buffer
pH 7-3, post-fixed in 1 % cacodylate-buffered osmium tetroxide and dehydrated
in a graded ethanol series. For TEM embryos were transferred from absolute
alcohol to propylene oxide then embedded in Araldite. For SEM embryos
were transferred to acetone then critical-point dried after replacing acetotie
with liquid CO2 in a Polaron apparatus. Dried specimens were mounted on
aluminium stubs with double-sided Sellotape and coated with 20 nm of gold
in an Edwards SI50 sputter coating unit.
Ten embryos for Triton X-100 extraction were taken from the eggs and
washed in Hank's solution, then transferred to 0-5 % Triton X-100 in Hank's
solution at room temperature for 10-15 min. They were then fixed and prepared
for either SEM (5 embryos) or TEM (5 embryos) as above.
Antisera were prepared against chicken gizzard actin in New Zealand white
rabbits previously screened for low pre-immune activity against acetoneextracted fibroblasts. Serum collected from these rabbits after five injections
was found to show antiactin activity. Gamma globulin fractions were collected
from this serum.
Antibodies were further purified by affinity chromatography on columns of
sepharose 4B to which actin had been coupled (1-10 mg/g wet weight) using
cyanogen bromide.
The immunoglobulin was then eluted with 4 MMg Cl2 at pH6 dialysed
against phosphate-buffered saline. 200/tg/ml solutions were stored at 4°C in
the presence of 0-02 % sodium azide.
Antibody specificity was checked by comparing the staining activity of the
prepared antibody with preimmune antibody preparations and with absorbed
antibody, using rat embryo fibroblasts and chicken gizzard myofibrils as test
samples.
Embryos for immunofluorescence were fixed in 3 % formaldehyde in phosphate-buffered saline (PBS), for 15-30 min. They were washed overnight in
PBS before staining and as much as possible of the endoderm and mesoderm
dissected off. Staining was carried out on 30 embryos, processed on slides with
the embryos lying ectoderm-side uppermost. Of these embryos 8 were control
incubations performed either with absorbed antiserum (3) or without antiactin
IgG (5). Details of the staining procedure and the control incubations performed
were as described by Badley, Lloyd & Woods (1978). Immunoperoxidase
staining was performed using the method of Couchman et al. (1979). Treated
embryos and controls were examined as whole-mounts (3 treated, 3 control),
after alcohol dehydration and xylene clearing or, after Araldite embedding, by
LM and TEM in sections (4 treated, 4 control). The same pattern of
staining was obtained each time the procedure was carried out. The only
variations obtained were regional, as described in the results.
Sections were cut on a Porter Blum MTI Ultramicrotome. For LM 1 /im
172
J. WAKELY AND R. A. BADLEY
Ectodermal actin organization
173
sections were stained with 1 % Toluidine blue in 1 % borax. For T.E.M. some
cytochemical and all normal and Triton X-100 treated specimens were stained
after sectioning with uranyl acetate and lead citrate. Some immunocytochemical
material was examined unstained.
S.E.M. images were recorded on a ISI 60 Scanning Electron Microscope;
T.E.M. material was examined on a JEOL 100S and fluorescent and other
light microscope studies were made on a Zeiss Universal Microscope with an
epifluorescence u.v. illumination system, photographed on Kodak Ektachrome
400 or Ilford HP5 film.
RESULTS
Light, scanning and transmission electron microscopy - untreated material
The ectoderm cells varied in shape from a flat squamous epithelium in the
peripheral (area opaca) parts of the embryo to a low columnar form in the
central (area pellucida) regions (Figs. 1-3).
In all areas ectoderm sectioned in the plane of the epithelium formed a
tightly packed epithelial sheet (Fig. 4). Each cell was attached to surrounding
cells by junctions situated a short distance below the cell apex. Most of the$e
junctions were simple in structure, consisting of a plaque-like thickening of
the apposed cell membranes, with little narrowing of the intercellular gap.
Under the plasma membrane a concentration of fibrillar material was present,
coextensive with the membrane thickening. Sometimes a little electron-dense
material was seen in the intercellular gap at a junction but true desmosomes
were very rare. Occasional small junctions showing close membrane apposition
were also seen. Their structure was not investigated further to identify them
positively as gap or tight junctions. At the level of the junctions each cell was
ringed by a bundle of (8 nm) microfilaments, lying immediately beneath the
plasma membrane. The filament bundle appeared to be connected to the
fibrillar material associated with cell junctions (Fig. 5). No filament bundles
were seen other than in association with apical cell junctions. Intermediate
Fig. 2. 1 /*m Araldite section of anterior area pellucida ectoderm stained with
toluidine blue. The cells form a pseudostratified columnar epithelium. The basal
surface (arrowed) of the cells has artificially separated from the mesoderm. Some
basal cell processes (p) can be seen. Stage-4 embryo.
Fig. 3. 1 /tm Araldite section of the same embryo as Figure 2, showing the cells
of the area opaca ectoderm (ec) and the large yolky cells of the endoderm (en).
Fig. 4. Scanning electron micrograph of mid-area pellucida ectoderm of stage-5
embryo. The epithelial cells are closely packed. The visible apical surfaces show
long and short microvilli (wv); and beaded threads (0Fig. 5. Transmission electron micrograph of tangentially sectioned ectoderm of
stage-4 embryo, passing through the level of the junctions. Deep to the opposing
cell membranes is a zone containing many microfilaments (mf). At junctions a
more electron-dense fibrous mat may be seen (arrow).
174
J. WAKELY AND R. A. BADLEY
- . - * >
7 •"
.
-
•
•
:
>
Ectodermal actin organization
175
filaments (10 nm) were not evident. Microtubules were present, crossing the
cell apex in all directions, not obviously aggregated into bundles or preferentially associated with any other organelle.
Deeper into the cell, at the level of the nucleus, the peripheral ring of
microfilaments disappeared and no specialized cellular junctions were seen
(Fig. 6).
The base of the cells rested on a thin basement membrane. This was thicker
and more compact in appearance in samples taken from the area pellucida
than from the area opaca. It was also better developed in the older embryos
in the series. The plasma membrane of the cell was separated by a gap 10-30 nrn
wide from the basement membrane. No plaques or other specialized contacts
between the plasma and basement membranes were observed. Contacts between
adjacent cell bases were complex, although no membrane specializations were
seen. Many cells formed processes, which either extended for a short distance
under the flat base of the neighbouring cell, partly separating it from its
portion of the basement membrane (Fig. 7), or interlocked with a similar
process from a neighbouring cell. These specializations were particularly
common in the area pellucida near the primitive streak. Near such cell contacts
and in the processes fine bundles of microfilaments were sometimes seen in the
cytoplasm immediately next to the plasma membranes (Fig. 7). In many
places, however, the general cytoplasmic features of free ribosomes, intracellular yolk, mitochondria, and microtubules extended right down to the
membrane.
Scanning and transmission electron microscopy - Triton X-100treated material
Cell ultrastructure was grossly disrupted. Nuclei and large intracellular yolk
droplets were still recognisable but most other organelles including microtubules were seen by TEM to have dispersed during treatment. In genefal
plasma membranes had also disappeared, except for the junctional sites aftd
a small membrane area immediately surrounding the junction. Apical microfilament bundles were clearly seen. It was clear that whole microfilament
bundles did not insert end-on into the junctions but passed alongside them
and were linked by the fibrillar mat to the plasma membrane (Fig. 8). It was
not possible to ascertain whether individual filaments terminated at the
junctions. Sections below the cell apex did not show this organization.
In maximally extracted embryos viewed by SEM the entire sheet of ectoderm
Fig. 6. Stage-4 ectoderm sectioned at the level of the nuclei (n). No microfilament
bundles can be seen.
Fig. 7. Vertically sectioned ectoderm cells, showing the base of the cell and the
basement membrane {bm). The cell on the left sends a process under the cell on
the right (p). The process and the cell which it underlaps contain microfilaments
176
J. WAKELY AND R. A. BADLEY
0-25 //m
Fig. 8. Transmission electron micrograph of Triton X-100-treated ectoderm. The
cytoplasm has largely disappeared, leaving yolk particles (y) and other debris
(d). Surviving portions of the plasma membrane (m) show junctions (j) associated
with a fibrillar mat (/) and bundles of microfilaments (m/).
Fig. 9. Scanning electron micrograph of Triton X-100-treated ectoderm. The
epithelial organisation is still apparent, as a basket like framework of Tritonresistant cytoskeletal material (arrows) outlining former cell boundaries, in which
nuclei («) and yolk (y) are suspended.
Ectodermal actin organization
111
cells was represented by a basket-like meshwork of linked filaments. Deep to
the meshwork regularly spaced nuclei were seen, sometimes suspended by
sparse fine filaments. Some spherical yolk droplets were also identified. Examination of specimens showing varying degrees of disruption by the detergent
showed that the Triton X-100-resistant filaments followed ceil perimeters so
that each unit in the meshwork represented an encircling band of filaments
associated with one cell. At high magnifications the filaments of adjacent cells
appeared to be separated by a narrow space representing the position of the
cells' membranes (Fig. 9). Apart from variation in size, the area pellucida cells
being smaller in diameter than those of the area opaca, there were no
regional or age differences.
Immunochemistry
Because of the thickness of whole-mount specimens background fluorescence
was invariably high in immunofluorescent-stained embryos although endoderm
removal helped to some extent to reduce this problem. Against the background
a patterned distribution of specific fluorescence was seen. This was not present
in control embryos processed in the absence of antiactin IgG, or with the use
of absorbed antisera.
Fluorescence associated with the presence of actin outlined the perimeter
of every cell throughout the area pellucida and area opaca of the embryos.
Thus a continuous fluorescent network was formed extending throughout the
ectoderm almost without interruption (Fig. 10). The pattern changed at the
blastoderm margins where the regular network was replaced by irregular
fluorescent strands and at the primitive streak where increased thickness of
the specimen made it impossible to resolve cell outlines clearly but the general
impression was that the pattern was destroyed. Differential focussing showed
that the fluorescent network was situated at the cell apices. Focussing deeper
into the cells usually failed to demonstrate discrete patterns of fluorescence
at other levels in the cytoplasm. However a more diffuse fluorescence could
be detected nearer to the base of the cells and occasionally short fluorescent
strands were seen more clearly in the flattest cells from the area opaca, where
the fluorescent image was less obscured by an overlying thickness of cytoplasm
than in the area pellucida (Fig. 11).
Immunoperoxidase staining examined by light microscopy in whole-mount
embryos demonstrated a pattern very similar to that seen by immunofluorescence.
A band of dark reaction product outlined the apex of every cell, giving the
epithelium as a whole a net-like appearance. The pattern was not seen in
control embryos from which the specific antiactin antibody alone, or both that
and the peroxidase-labelled goat anti-rabbit serum were omitted from the
staining procedure. By T.E.M. electron-dense deposits were seen associated
with bands of filaments passing alongside apical cell junctions. They were not
seen on other cellular organelles or in control embryos. (Figs. 12, 13).
178
J. WAKELY AND R. A. BADLEY
HL.
f1
• *
•
0-25 /im
0-25 /im
i
12
13
i
Ectodermal actin organization
179
DISCUSSION
The combination of techniques used in this investigation has demonstrated
the presence of a band of actin filaments encircling every cell in the ectoderrfl
of gastrulating chick embryos and attached tangentially to the junctions linking
each cell to its neighbours. Thus, as clearly demonstrated by S.E.M. of
detergent-extracted material, the peripheral ring of actin in each individual
cell is linked to those of the surrounding cells. In effect there is a coherent
network of actin forming a 'skeleton' to the entire epithelium, broken only
where the epithelial organization of the ectoderm is itself destroyed; at the
primitive streak and at the margins. The arrangement of actin and junctions
in chick ectoderm is similar to that described by Perry (1975) for amphibian
embryonic ectoderm and quite different from that Seen by Albrecht-Buehl^r
(1979) & Meittenen et al. (1978) in cultured epithelial sheets where bundles
of filaments stretch across the cell apex from junction to junction. Retinal
pigment epithelium in vitro shows an intermediate situation with filaments
both encircling and crossing the cell apex (Crawford, 1979). The intercellular
junctions are also of a type common in developing epithelia (Perry, 1975;
Sanders, 1973; Mobbs & MacMillan, 1979). The nearest adult equivalent is
the zonula adherens Farquhar & Palade, 1963) associated with actin in cultured
cells (Miettenen et al. 1978).
Tn single non-muscle cells cytoplasmic actin forms part of an actomyosin
contractile system, which is involved both in the generation of movement and
the stabilization of cell shape (Fleischer & Wohlfarth-Bottermann, 1975;
Pollard, 1976; Isenberg et al. 1976; Adelstein, Scordilis & Trotter, 1979);
Epithelia in vitro characteristically behave as sheets rather than as individual
cells (Di Pasquale, 1975). Perry (1975) suggested that in amphibian embryos
a similar 'tissue level' organization of actin filaments to that which we have
seen in the chick might act as a strengthening framework for the ectoderm
during gastrulation. In the gastrulating chick embryo the ectoderm appears to
Fig. 10. Immunofluorescence micrograph of part of the inner area opaca ectoderm
of a stage-5 embryo, (antiactin staining). A peripheral band of fluorescence surrounds every cell.
Fig. 11. Immunofluorescence micrograph of outer area opaca ectoderm of stage-5
embryo. As well as the peripheral localization of actin small strands are faintly
discernible nearer to the centre of the cell (arrows).
Fig. 12. Transmission electron micrograph of junctional region of Triton X-100
extracted ectoderm cells studied by the immunoperoxidase method to demonstrate
actin. Bundles of filaments (arrowed) associated with the junctions (j) show an
overall electron density and larger electron dense deposits. These features are
absent from the control specimen. Fig. 13. From which the specific antiactin
antibody was omitted. In neither case was any heavy metal staining applied to the
sections.
180
J. WAKELY AND R. A. BADLEY
be in a state of tension. Ectodermal wounds immediately gape into round
holes (Stanisstreet, Wakely & England, 1980). Cells round the wound margins
bulge as though released from tension. Ectodermal tension is at least partly
caused by blastoderm expansion (New, 1959; Bellairs, Bromham & Wylie,
1967) and also by an opposite flow of cells towards and into the primitive
streak (Nicolet, 1971; Rosenquist, 1966). Cytoskeletal continuity could give
the epithelium necessary cohesion to resist being disrupted by this internal
tension. The actin net may itself contribute to the tension by participating in
an isometric contractile process (Isenberg et ah 1976). Also forces generated
by cell-shape changes in one part of the embryo may be transmitted beyond
the actively deforming cells so producing morphogenetic changes in whole
areas of tissue.
Electron microscopy reveals some sparse filaments at cell bases, some forming
a sheet and some forming more organised fibre bundles often associated with
cellular processes. The arrangement parallels the general background and short
strands seen at cell bases by immunofluorescence. It is reminiscent of that
seen by Sanders & Prasad (1979) in the endodermal cells of chick embryos of
a comparable age in vitro and Di Pasquale (1975) and Crawford (1979) in
cultured epithelial cells. The association of organized microfilament bundles
with cell-substrate contacts is also a consistent feature of fibroblasts growing
on substrates in culture (Izzard & Lockner, 1976; Badley, Woods, Smith &
Rees, 1980; Heath & Dunn, 1978). Similarities such as these suggest an
association with cell locomotion. The basal processes and their associated
microfilament bundles in ectoderm cells are much smaller and less highly
organized than the sole plate of a fibroblast and there are no focal contacts,
as judged by the lack of cytoplasmic dense plaques with associated actin
filaments. The similarity does, however, suggest a similar functional association
between the cells and their substrate (the basement membrane) and that therefore during morphogenetic movements, for example, the migration of ectoderm
towards the primitive streak, the cells could migrate over their own basement
membrane, until they reach the primitive streak, where the membrane is
broken down and epithelial arrangement is lost as the cells join the mesoderm.
The morphology of cell bases did not vary regionally within the ectoderm,
suggesting that all cells were capable of some movement of this nature.
In summary, then, actin filaments in chick ectoderm appear to be organized
in two patterns. The majority of the protein forms an apical band surrounding
every cell and linked to that of neighbouring cells via cell junctions. A second
smaller system is situated near the basement membrane and associated with
basal cytoplasmic processes and less closely linked to that of neighbouring
cells. The two structural arrangements reflect two different functions performed
by actin in the ectoderm. The basal concentrations of actin are perhaps more
directly associated with active locomotion, with the basement membrane as
substrate, than the association of actin and cell junctions at the cell apices
Ectodermal actin organization
181
which gives the epithelium as a whole cohesion and stability to enable it to
perform complex morphogenetic movements involving large numbers of cells,
while at the same time retaining its epithelial integrity.
We are grateful for the skilled technical assistance of Stephanie Bulman and Alison Cole,
to Mr G. L. C. McTurk for operation of the IS1 60 Scanning Electron Microscope in the
Leicester University S.E.M. Unit, and to Diane Dolman for typing the manuscript.
REFERENCES
R. S., SCORDILIS, S. P. & TROTTER, J. A. (1979). The cytoskeleton and cejl
movement: general considerations. Meth. Achiev. exp. Path. 8, 1-41.
ALBRECHT-BUEHLER, G. (1979). Group locomotion of PtKl cells. Expl Cell Res. Ill, 402<407.
BADLEY, R. A., LLOYD, C. W. & WOODS, A. (1978). Mechanisms of cellular adhesion. III.
Preparation and preliminary characterisation of adhesions. Expl Cell. Res. Ill, 231-244.
BADLEY, R. A., WOODS, A., SMITH, C. G. & REES, D. A. (1980). Actomyosin relationships
with cell surface features in fibroblast adhesion. Expl Cell Res. 126, 263-272.
BANCROFT, M. & BELLAIRS, R. (1975). The onset of differentiation in the epiblast of the
chick blastoderm (SEM & TEM). Cell Tiss. Res. 155, 399-418.
BELL, P. B., MILLER, M. M., CARRAWAY, K. L. & REVEL, J. P. (1978). SEM revealed changes
in the distribution of the Triton-insoluble cytoskeleton of Chinese hamster ovary cells
induced by dibutyryl cyclic AMP. In Scanning Electron Microscopy 1978, vol. 2 (edR. P. Becker & O. Johari), pp. 899-906. Scanning Electron Microscopy lnc, Illinois
BELLAIRS, R., BROMHAM, D. R. & WYLIE, C. C. (1967). The influence of the area opaca on
the development of the young chick embryo. /. Embryol. exp. Morph. 17, 195-212.
BROWN, S., LEVINSON, W. & SPUDICH, J. A. (1976). Cytoskeletal elements of chick embryo
fibroblasts revealed by detergent extraction. /. Supramole. Struct. 5, 119-130.
BURGESS, D. R. & SCHROEDER, T. E. (1979). The cytoskeleton and cytomusculature in
embryogenesis - an overview. Meth. Achiev. exp. Path. 8, 171-189.
BURNSIDE, B. (1971). Microtubules and microfilaments in newt neurulation. Devi Biol. 16,
416-441.
ADELSTEIN,
COUCHMAN, J. R., GIBSON, W. T., THOM, D., WEAVER, A. C , REES, D. A. & PARISH, W. E.
(1979). Fibronectin distribution in epithelial and associated tissues of the rat. Archs
Dermatol. Res. 266, 295-310.
CRAWFORD, B. (1979). Cloned retinal pigment epithelium. The role of microfilaments in
the differentiation of cell shape. /. Cell Biol. 81, 301-315.
DI-PASQUALE, A. (1975). Locomotory activity of epithelial cells in culture. Expl Cell Res.
94, 191-215.
FARQUHAR, M. G. & PALADE, G. E. (1963). Junctional complexes in various epithejia.
J.Cell Biol. 17, 375-412.
FLEISCHER, M. & WOHLFARTH-BOTTERMAN, K. E. (1975). Correlation between tension force
generation, fibrillogenesis and ultrastructure of cytoplasmic actomyosin during isometric
and isotonic contraction of protoplasmic strands. Cytobiol. 10, 339-365.
HAMBURGER, V. & HAMILTON, H. L. (1951). A series of normal stages in the development
of the chick embryo. /. Morph. 88, 49-92.
HEATH, J. P. & DUNN, G. A. (1978). Cell to substratum contacts of chick fibroblasts
and their relation to the microfilament system. A correlated interference reflexion and high
voltage electron microscope study. /. Cell Sci. 29, 197-212.
IP, W. & FISCHMAN, D. A. (1979). High resolution scanning electron microscopy of isolated
and in situ cytoskeletal elements. /. Cell Biol. 83, 249-254.
1SENBERG, G., ROTHKE, P. C , HULSMANN, N . , FRANKE, W . W. & WOHLFARTH-BOTTERMlAN,
K. E. (1976). Cytoplasmic actomyosin fibrils in tissue culture cells. Direct proof of contractility by visualisation of ATP- induced contraction in fibrils isolated by laser
beam dissection. Cell Tiss. Res. 166, 427-433.
182
J. WAKELY AND R. A. BADLEY
G. S. & LOCHNER, L. (1976). Cell to substrate contact in living fibroblasts; an
interface reflexion study with evaluation of the technique. /. Cell Sci. 21, 129-159.
KARFUNKEL, P. (1972). The activity of microtubules and microfilaments in neurulation in
the chick. /. exp. Zool. 181, 289-302.
KARNOVSKY, M. J. (1965). A formaldehyde-glutaraldehyde fixative of high osmolality for
use in electron microscopy. /. Cell Biol. 21, 137A.
MIETTENEN, A., VIRTANEN, I. & LINDER, E. (1978). Cellular actin and junction formation
during reaggregation of adult rat hepatocytes into epithelial cell sheets. /. Cell Sci. 31,
341-353.
MOBBS, I. G. & MCMILLAN, D. B. (1979). Structure of the endodermal epithelium of the
chick yolk sac during early stages of development. Am. J. Anat. 155, 287-310.
NEW, D. A. T. (1959). The adhesive properties and expansion of the chick blastoderm.
/. Embryol. exp. Morph. 7, 145-154.
NICOLET, G. (1971). Avian gastrulation. Advances in Morphogenesis 9, 231-262.
PANNET, G. A. & COMPTON, A. (1924). The cultivation of tissues in saline embryonic juice.
Lancet 102, 381-384.
PERRY, M. M. (1975). Microfilaments in the external surface layer of the early amphibian
embryo. /. Embryol. exp. Morph. 33, 127-146.
POLLARD, T. D. (1976). Cytoskeletal functions of cytoplasmic contractile proteins. /. Supramolec. Struct. 5, 317-334.
ROSENQUIST, G. C. (1966). A radioautographic study of labelled grafts in the chick blastoderm. Development from primitive streak stages to stage 12. Contr. Embryol. Carnegie
Institn. 38, 71-110.
SANDERS, E. G. (1973). Intercellular contact in the chick embryo. Z. Zellforsch mikrosk.
Anat. 141, 459-468.
SANDERS, E. J. & PRASAD, S. (1979). Observations of cultured embryonic epithelial cells in
side view. / . Cell Sci. 38, 305-314.
SMALL, J. V., ISENBERG, G. & CELIS, J. E. (1978). Polarity of actin at the leading edge of
cultured cells. Nature 272, 638-639.
STANISSTREET, M., WAKELY, J. & ENGLAND, M. A. (1980). Scanning electron microscopy
of wound healing in xenopus and chicken embryos. J. Embryol. exp. Morph. 59, 341-353.
TRELSTAD, R. L., HAY, E. D. & REVEL, J. P. (1967). Cell contact during early morphogenesis
in the chick embryo. Devi Biol. 16, 78-106.
IZZARD,
WESSELS, N. K., SPOONER, B. S., ASH, J. F., BRADLEY, M. D., LUDUENA, M. A., TAYLOR,
E. L., WRENN, J. T. & YAMADA, K. M. (1971). Microfilaments in cellular and developmental processes. Science 171, 135-143.
(Received 10 August 1981, revised 16 December 1981)