J. Embryol. exp. Morph. 82, 97-117 (1984)
Printed in Great Britain © The Company of Biologists Limited 1984
97
The distribution of cytoplasmic actin in mouse
8-cell blastomeres
ByM. H. JOHNSON 1 AND B. MARO1-2
1
Department of Anatomy, Downing Street, Cambridge, CB2 3DY, U.K.
2
C.N.R.S.,
France.
SUMMARY
Three non-homogeneous patterns of cytoplasmic actin distribution have been demonstrated
in pairs of 8-cell blastomeres. Newly formed blastomeres showed an actin distribution
associated with the remnant of the previous mitotic spindle. Subsequently blastomeres showed
a zonal clearing of actin from regions of intercellular contact, the extent of the clearing
increasing with the extent of contact. A polarized distribution of actin was evident from the
early to mid 8-cell stage and coincided with the movement of nuclei towards the point of
intercellular contact. The detection of polar actin preceded by 2-4 h the detection of a surface
polarity as assessed by the FITC-Con A binding pattern and the distribution of cortical
microvillous actin. However, once a surface pole of microvilli had formed, it persisted under
conditions which led to loss of polar cytoplasmic actin. Incubation in cytochalasin D (CCD)
resulted in a dispersed homogeneous pattern of actin distribution but did not prevent the
formation of surface poles as assessed both by the Con A binding pattern and detection of
polar microvilli. However, the poles formed were less clearly defined and the density and
length of microvilli within them was variable. Moreover, when CCD was added early during
the 8-cell stage the position of the poles was frequently not on an axis perpendicular to the
point of intercellular contact. Cytochalasin D also affected the movement of the nucleus that
occurs during the process of polarization. On the basis of these experiments, we conclude that
actin is likely to be involved in the events of polarization, but that its precise role remains to
be determined.
INTRODUCTION
The 8-cell stage of mouse development is characterized by a number of
morphological changes that are described under the general heading of compaction. The changes include flattening of cells upon each other to maximize cell
contact and obscure intercellular boundaries (Lehtonen, 1980), the establishment of intercellular junctions (Ducibella, Albertini, Anderson & Biggers, 1975;
Magnuson, Demsey & Stackpole, 1977; Lo & Gilula, 1979; Goodall & Johnson,
1984), and the polarization of individual blastomeres (Ducibella, Ukena, Karnovsky & Anderson, 1977; Handyside, 1980; Reeve & Ziomek, 1981). It seems
likely that the changes in cell shape and organization that occur during compaction involve changes in the cytoskeleton but there is little direct evidence to
support this view (Ducibella etal. 1975; Lehtonen & Badley, 1980; Sobel, 1983).
In this paper we apply a technique devised for examining mounts of fixed,
98
M. H. JOHNSON AND B. MARO
detergent-extracted whole cells (Maro, Johnson, Pickering & Flach, 1984) to
examine the patterns of cytoplasmic actin distribution during the 8-cell stage. We
find three non-homogeneous patterns of cytoplasmic actin (i) a focal pattern,
that relates to the immediate aftermath of cleavage, (ii) a zonal pattern, that
relates to intercellular flattening, and (iii) a polar pattern, that relates to
polarization of intracellular and surface features. Furthermore, we present
evidence to suggest that whereas the turnover of intracellular actin may not be
required for the process of polarization and the stability of the polarized state,
it may have some role in correctly orienting the axis of polarity perpendicular to
the point of intercellular contact.
MATERIALS AND METHODS
1. Recovery of embryos
MF1 female mice (3-5 weeks; Olac) were superovulated by injections of 5 i.u.
of pregnant mare's serum gonadotrophin (PMSG; Intervet) and human chorionic
gonadotrophin (hCG; Intervet) 48 h apart. The females were paired overnight
with HC-CFLP males (Hacking & Churchill) and inspected for vaginal plugs the
next day. Unfertilized and fertilized eggs were recovered from females at
14-16 h post-hCG; 2-cell and 4-cell embryos were recovered at 46-50 h posthCG; 8-cell embryos were derived by overnight culture of 2- to 4-cell embryos;
early 16-cell embryos were recovered at 65-70 h post-hCG.
2. Preparation and handling of single cells
Two-cell embryos were recovered at 18.00 h (48 h post-hCG) and cultured in
Medium 16 containing 4 mg/ml BSA (M16+BSA; Whittingham & Wales, 1969)
under oil for 13 h at 37 °C in 5 % CO2 in air. All 4-cell embryos were exposed
briefly to acid Tyrode's solution (Nicolson, Yanagimachi & Yanagimachi, 1975)
to remove the zona pellucida, rinsed in Medium 2+BSA (Fulton & Whittingham, 1978), and placed in Ca2+-free M2+6 mg/ml BSA for 5-45min, during
which time they were disaggregated to single 4-cell blastomeres (1/4 cells) using
a flame-polished micropipette. Cells were cultured on Sterilin tissue-culture
dishes in drops of M16+BSA under oil at 37 °C in 5 % CO2 in air. Each hour the
cultures were inspected for evidence of division to 2/8 pairs. Each couplet was
removed and designated Oh old. Couplets were either cultured in M16+BSA as
natural 2/8 pairs, or were disaggregated by means of a flame-polished
micropipette in Ca2+-free M2+BSA to yield 1/8 blastomeres. These were then
aggregated in a 1/20 dilution of Gibco stock phytohaemagglutinin (PHA) either
with a second 1/8 blastomere to yield (1/8+1/8) reaggregated pairs or with a
fertilized or unfertilized egg. Frequently the remnant of the cleavage furrow, the
midbody, is evident in isolated 1/8 cells, and allows reaggregation of the cells in
various defined spatial arrays in relation to the original point of continuity with
the sibling cell. In some experiments, reaggregated cells were later disaggregated
Actin in 8-cell blastomeres
99
and reaggregated a second time in a new orientation, as described in the results
section.
In one experiment, late 2-cell embryos (48 h post-hCG) were isolated, freed
from the zona pellucida, disaggregated to 1/2 blastomeres, and 2/4 pairs were
collected at hourly intervals, classified as Oh-old 2/4 and then placed in culture
in M16+BSA.
In one series of experiments, whole 8-cell embryos were freed from their
zonae, disaggregated to single or paired blastomeres in Ca2+-free medium and
the blastomeres analysed immediately.
In some experiments cytochalasin D (CCD, Sigma: stock solution of 1 mg/ml
in dimethylsulphoxide (DMSO) stored at - 2 0 °C) was added to M16+BSA at a
final concentration of 0-5 /Ug/ml. As a control, DMSO was added to the various
media at a concentration of 0-5 /il/ml.
3. Immunocytochemistry
Surface polarity was assessed by incubation of cells or embryos in 50 jug/ml
tetramethylrhodamine-labelled succinyl Concanavalin A (TMRTC-S-ConA,
Poly sciences) for 5min at room temperature, followed by two to three washes
in M2+BSA. Labelled cells were then placed in specially designed chambers
exactly as described in Maro et al. (1984) for fixation with 3-7 % formaldehyde
followed by extraction with 0-25 % Triton X-100. After washing, cells were
labelled either with NBD-phallicidin (5/ig/ml; Molecular Probes Inc) or with
affinity-purified rabbit anti-actin antibodies (Gounon & Karsenti, 1981) and
fluorescein-labelled anti-rabbit immunoglobulin antibodies. The detailed
characteristics of the reagents and procedures are reported in Maro et al. (1984).
Samples were mounted in 'mounting medium' (City University, London) and
viewed on a Zeiss Photomicroscope with filter set 48 77 09 for FITC-labelled
reagents and NBD-phallicidin and 48 77 15 for TMRTC-labelled reagents.
Photography was on Tri-X film with exposures of 3-45 s. The mounting medium
reduces fading of fluorescein labels but not of NBC-phallicidin.
Specimens were viewed with x40 and x63 objectives and examined for actin
localized in both surface microvilli and within the cells by focussing up and down
through the sample. The cortical microvillous actin distribution corresponded to
the pattern of Con A-binding receptors (e.g. seen clearly in Figs 5e,g & 6a,c),
confirming our previous results (Ziomek & Johnson, 1980; Johnson & Ziomek,
1981a). Throughout this paper, we are concerned exclusively with the
distribution within the cells of non-microvillous, cytoplasmic actin.
4. Scanning electron microscopy (SEM)
Whole embryos, freed from the zona pellucida, or pairs of 2/8 cells were
prepared for SEM analysis exactly as described in Johnson & Ziomek (1982).
Samples were viewed in a JSM-35CF Jeol microscope under 20 kV. Numbers of
samples examined were as follows: (A) Natural pairs of blastomeres: 9h in
100
M. H. JOHNSON AND B. MARO
DMSO = 36 pairs; 9h in CCD = 38 pairs; 6h in DMSO and 3h in CCD = 21
pairs. (B) Whole embryos: 9h in DMSO = 25; 9h in CCD = 14; 2h in DMSO
and 7 h in CCD = 34; 6 h in DMSO and 3 h in CCD = 26; 8-5 h in DMSO and 0-5 h
in CCD = 48.
RESULTS
1. Distribution of actin in whole embryos
We first examined whole embryos for evidence of actin localization. Fig. 1
shows examples of various stages of zona-free preimplantation embryos stained
with anti-actin antibodies. A localization of subcortical actin was observed in the
unfertilized egg (Fig. la,b) at the site of the second metaphase spindle (see Maro
etal. 1984 for discussion of this point and of the changes in the first few hours after
fertilization). At later cleavage stages, there is an impression of an apical actin
localization at the late 8-cell stage (Fig. li,j) at a time when polarization of
blastomeres might be expected to have occurred, as has also been reported by
Lehtonen & Badley (1980). In addition there is a general impression of less actin
staining adjacent to sites of intercellular contact at 2-, 4- and early 8-cell stages
(Fig. lc-h) and also within the inside cell population of the 16-cell stage (Fig.
lk,l). However, the density of stain in whole embryos makes the subcellular
identification of actin localizations difficult, in particular the distinction between
actin within the microvilli (designated here cortical microvillous actin) and
within the cell itself (designated here cytoplasmic actin). In order to gain a
clearer resolution of the actin distribution around the time of compaction, we
examined isolated pairs of 8-cell blastomeres for their actin-staining pattern.
2. Distribution of cytoplasmic actin in natural 2/8 pairs in relation to development of surface polarity
Isolated 4-cell blastomeres were cultured in vitro and at hourly intervals all 2/8
couplets formed were collected. Couplets of known age postdivision were then
labelled with TMRTC-Succinyl Con A, fixed, extracted and labelled with antiactin antibodies that were visualized with FITC-anti rabbit IgG (Fig. 2a-d,f,h),
or with NBD-phallicidin (Fig. 2j,l,n).
The patterns of cytoplasmic actin distribution were examined by focusing up
and down through the couplet, and cells were assigned to one of four categories
Fig. 1. Actin distribution in whole embryos, as revealed by use of anti-actin antibody
and FITC-labelled anti Ig, for embryonic stages; (a,b) unfertilized egg, note actin
localization associated with second metaphase spindle, (c,d) 2-cell embryo, (e,f) 4cell embryo, note relative absence of actin in areas of contact, (g,h) early 8-cell
embryo, note relative absence of actin in areas of contact, (i,j) late compacting 8-cell
embryo, note concentration of actin at outward-facing apical end of cell, (k,l) 16-cell
embryo, note strong apical staining and relatively weaker staining internally. (Mag.
X200).
_
I
c
k
r
i
1
H —P
F
10
\
H
g
F
•KK
.1
H
m
n
Actin in 8-cell blastomeres
101
Table 1. Incidence of cells showing distinct patterns of actin distribution in natural
pairs of 2/8 blastomeres at different times after their formation
No. (%) of cells showing distribution patterns
Age (in h post No. of cells
division)
scored
0
1
2
3
4
6
8
10
54
40
54
48
48
40
48
38
P*
32(59)
28(70)
3(5)
2(4)
0
0
0
0
2(4)
4(10)
1(2)
1(2)
4(8)
3(7)
0
0
20(37)
8(20)
29(54)
21(44)
21(44)
13(33)
10(21)
6(15)
0
0
21(39)
24(50)
23(48)
24(60)
38(79)
32(84)
* Focal, homogeneous, zonular and polar: see text and Fig. 2 for details.
(Fig. 2a-c). Immediately after division, granular foci of actin were evident between the peripherally placed nucleus and the overlying plasma membrane and
also associated with the remnants of the mitotic spindle (Fig. 2a). The incidence
of cells scored as being of this type, designated F for focal, is shown in Table 1,
column 3; it falls off rapidly with increasing time after division to 2/8 cells. In a
few cells, the distribution of actin elements was spread throughout the cell and
although within this dispersed distribution there were occasional local concentrations such cells were designated homogeneous (H; Table 1, column 4; Fig. 2b,
lower cell). More commonly, particularly at earlier times after division, a zone
relatively deficient in cytoplasmic actin elements was detected adjacent to
the point of contact with the other cell in the couplet (designated Z; Table 1,
Fig. 2. Pairs of 8-cell blastomeres, double-labelled to reveal patterns of actin
distribution (as revealed by use of anti-actin antibodies a,b,c,d,f,h, or NBDphallicidin j,l,n) and surface Con A binding (e,g,i,k,m,o). Throughout large
arrowhead indicates position of nucleus, small arrowhead indicates limits of actin
staining; F= focal actin, H = homogeneous actin; Z = zonal actin and P — polar
actin. (a) Natural 2/8 pair 1 h old; both cells show focal actin, associated with spindle
remnant, (b) Natural 2/8 pair 2 h old; both cells scored as homogeneous although the
left cell shows a slight polarization of actin. (c) Natural 2/8 pair 7 h old, one cell zonal
and the other polar; small arrowheads define the limits of actin-positive areas. (d,e)
Natural 2/8 pair l h old, both cells show focal actin and apolar Con A receptor
distribution patterns. (f,g) Natural 2/8 pair, 9h old; both cells show apolar Con A
binding, upper right cell shows polar actin and lower left homogeneous actin changing to zonal. (h,i) Reaggregated (1/8+1/8) pair 11 h old; both cells polar for both
actin distribution and Con A binding. (j,k) Natural 2/8 pairs, l h old; focal actin
stained with phallicidin and associated with spindle remnants, and apolar Con A
binding. (l,m) Natural 2/8 pairs, 2-4h old; homogeneous actin stained with phallicidin, and apolar Con A binding. (n,o) Natural 2/8 pairs, 6-8 h old; polar actin
stained with phallicidin, and Con A binding apolar (left) and polar (right). Mag.
X330.
102
M. H. JOHNSON AND B. MARO
column 5; Fig. 2c, lower cell). Finally, from 2h after division onwards, a more
restricted polar distribution of cytoplasmic actin elements was observed (designated P; Table 1, column 6; Fig. 2c, upper cell), localized to the area lateral to
and above the nucleus. The same four types of pattern were identifiable using
NBD-phallicidin (Fig. 2j,l,n), the differences from pairs of cells stained with
anti-actin antibody being (i) absence of intranuclear staining, and (ii) lower
cytoplasmic background staining due to lack of staining of nonpolymerized actin
(see discussion Maro et al. 1984). Although it is probable that the four patterns
described form part of a continuum, most cells could be assigned unambiguously
to one of the four categories. The occasional cell did show a somewhat intermediate pattern (for example slight zonal clearing in a cell otherwise scored as
homogeneous e.g. lower cell Fig. 2f and left-hand cell Fig. 2b; in such cases the
cell was classified as being of the least restricted in phenotype e.g. H in Fig. 2b
and2f).
The surface binding of Concanavalin A in cells stained with anti-actin
antibodies (and with NBD-phallicidin) was examined in order to establish the
relationship between the cytoplasmic actin distribution and surface polarity. The
surfaces of all cells scored as homogeneous for actin (Fig. 2f,g, lower cell) and
of all except three cells scored as having a focal distribution of actin (Fig. 2d,e)
stained homogeneously with Con A and showed an even distribution of cortical
actin corresponding to the microvilli. The three exceptions showed a slight
increase in surface Con A labelling immediately over the actin focus. The incidence of surface polarity in cells showing actin distribution patterns type Z and
P is shown in Table 2. The incidence of surface polarity rises from 4-6 h onwards,
confirming previous observations. However, it is clear that the distribution of
cytoplasmic actin in both zonal and polar patterns preceeds by several hours the
Table 2. Incidence of surface polarity (as assessed by Con A binding pattern) in
natural 2/8 pairs* showing zonal (Z) or polar (P) actin distribution
% polar surface labelling patterns for cells showing
A
r
pattern
Z-type actin
A
P-type actin pattern
Age (h post
division)
Total
% Polar'
Total
% Polar^
0
1
2
3
4
6
8
10
20
8
29
21
21
13
10
6
0
0
10
0
4
15
30
50
0
0
21
24
23
24
38
32
—
9
0
21
50
55
91
* data is for cells shown in columns 5 and 6 of Table 1.
A
Actin in 8-cell blastomeres
100 n
103
Focal
"-Z
0
100
H
T
10 h
0
2
4
6
rm
r—n
i—i
i—i
%50
Homogeneous
Zonal
0J
1=1
Actin distribution patterns:
[H Focal+Homogeneous
Polar
Zonal LJ Polar
Con A binding pattern:
I Polar
Fig. 3. Schematic summary of distribution patterns of cytoplasmic actin (righthand side) and their changing incidence with time in natural 2/8 pairs (graph) in
relation to the development of surface polarity as assessed by Con A binding
(nomogram).
104
M. H. JOHNSON AND B. MARO
Table 3. Relative incidence of polar plus zonal cytoplasmic actin patterns in
relation to surf ace polarity of 1/8 blastomeres disaggregated from 8-cell embryos
1-11 h old
Age of 8-cell
embryos (h)
Total No. of
cells examined
% of cells that were polar
or zonular for actin but
not polar for Con A
binding
1
3
4
6
7
8
46
36
38
45
30
41
0
17
31
20
13
17
67
61
11
40
3
92
% of cells that were both
polar or zonular for actin
and were also polar for
Con A binding
0
17
11
33
detection of polarity at the surface (compare Fig. 2f,g with 2h,i). These results
are summarized in Fig. 3.
In order to determine whether the patterns of actin observed in pairs of 2/8
blastomeres were also observed in cells developing within intact embryos, 8-cell
embryos were cultured for varying times after their formation, disaggregated to
single cells or to couplets and analysed for cytoplasmic actin distribution. The
contents of Table 3 and Fig. 4a,b reveal that whereas focal patterns were not
observed, homogeneous, zonal and polar patterns were, and, moreover, some
cells in which no evidence of surface polarity was detected were nonetheless
scored as polar or zonal for actin. These results suggest that actin patterns
observed in pairs of cells correspond to the in situ condition and confirms the
impression that cytoplasmic actin may show a polar distribution before the
appearance of surface polarity.
Table 4. Incidence of cells showing distinct patterns of actin distribution in
reaggregated (1/8+1/8) pairs at different times after division from 1/4 cells
No. (%) cells showing distribution patterns
Age (in h post No. of cells
division)
scored
1
3
5
7
9
11
36
30
38
47
48
37
F
H
1(3)*
1(3)*
0
0
0
0
22(61)
6(20)
8(21)
7(15)
6(12)
0
13(36)
23(77)
17(45)
16(34)
3(6)
0
In each case the focus of actin was opposite the residual midbody.
0
0
13(34)
24(51)
39(82)
37(100)
Actin in 8-cell blastomeres
105
Table 5. Incidence of surface polarity (as assessed by Con A binding patterns) in
reaggregated (1/8+1/8) pairs* showing zonal (Z) or polar (P) actin distribution
% polar surface labelling patterns for cells with
A
Z-type actin pattern
P-type actin pattern
A
A
Age (h post
division)
Total
% Polar
1
3
5
7
9
11
13
23
17
16
3
0
0
4
24
25
0
f
Total
% Polar
0
0
13
24
39
37
92
75
87
100
* data is for cells shown in columns 5 and 6 of Table 4.
3. Distribution of cytoplasmic actin in reaggregated 2/8 pairs in relation to
development of surface polarity
The characteristic changes in actin distribution patterns with time described in
section 2 could represent a sequel to the previous mitotic cleavage, a response
to cell contact or a combination of both these events. One way to examine this
question is to disaggregate newly formed Oh 2/8 pairs to single 1/8 cells and,
using the remnant of the midbody as a marker, to reaggregate them together
again oriented in a new contact relationship. The effect of such an experimental
approach on the actin distribution pattern and its relationship to development of
surface polarity is shown in Tables 4 and 5 (compare with Tables 1 and 2). Two
major differences from natural pairs were observed. First, only two examples of
Table 6. Position of nucleus in relation to the axis perpendicular to contact point
with companion cell
Reaggregated (1/8+1/8) pairs
Natural 2/8 pairs
A
Time
Total
0
1
2
3
4
5
6
7
8
9
10
11
54
40
54
48
48
—
40
—
48
—
37
—
A
On axis (%) Off axis (%)
53(98)
38(95)
46(85)
43(89)
43(89)
1(2)
2(5)
8(15)
5(11)
5(11)
32(80)
8(20)
40(83)
8(17)
29(78)
8(22)
Total
—
36
—
30
—
37
—
44
—
48
—
30
On axis (%) Off axis (%)
17(47)
19(53)
20(67)
10(33)
28(76)
9(24)
36(82)
8(18)
42(88)
6(12)
25(84)
5(16)
Total
50
40
54
48
48
—
40
—
48
—
37
Time
0
1
2
3
4
5
6
7
8
9
10
0
0
2
6
9
—
16
—
23
—
32
Near (%)
6
3
24
42
56
—
57
—
52
—
43
Midway (%)
Natural 2/8 pairs
94
97
74
52
35
—
27
—
25
—
25
Far (%)
r
36
—
30
—
37
—
44
—
48
—
Total
17
—
20
—
5
—
24
—
13
—
Near (%)
22
—
50
—
30
—
40
—
56
—
Midway (%)
Reaggregated (1/8+1/8) pairs
Table 7. Distance of nucleus from point of contact with companion cell
61
—
30
—
65
—
36
—
31
—
Far (%)
O
KA
a
o
o
Actin in 8-cell blastomeres
107
a focal-type actin distribution were scored, both at early times and in both cases
the focus of actin was not related to the point of contact with the companion cell
but rather was at the opposite end of the cell to the remnant of the preceding
cleavage furrow, the midbody. Second, a much larger number of cells with a
homogeneous staining pattern was observed than in natural pairs, especially at
the early time points. In contrast, the changes in zonular and polar actin patterns
with time were broadly similar in natural and reaggregated pairs, although the
reaggregated population tended to complete polarity development (both actin
and surface) earlier than the natural pairs. Moreover, and almost without exception, the zonular and polar patterns of cytoplasmic actin developed about an axis
perpendicular to the new point of intercellular contact as described previously,
and confirmed here, for surface polarity (Ziomek & Johnson, 1980).
Two other differences between reaggregated and natural 2/8 pairs were
noticed in this study. Nuclear position in each cell was scored (i) as being either
on an axis perpendicular to the contact point or off that axis (see Fig. 5g), and
(ii) as being either near to or distant from the point of cell contact (see Fig. 5c).
The results are summarized in Tables 6 & 7. Most nuclei in natural pairs were onaxis regardless of time of examination, whereas nuclei in newly reaggregated
cells were initially off-axis but became increasingly on-axis with time. In natural
pairs, nuclei were initially distant from the point of intercellular contact but
moved towards it with time, confirming the results of Reeve & Kelly (1983). In
newly reaggregated pairs, nuclei were more randomly positioned although a
move towards the new contact point with time was evident.
From these results we conclude that cell interactions can influence the orientation of zonal and polar actin-staining patterns, as well as the alignment and
position of nuclei, but that focal actin staining is likely to be related simply to the
orientation of the previous cleavage plane.
4. Does any type of cell interaction lead to development of zonal and polar actin
redistribution?
We undertook three types of experiment to determine whether the zonal and
polar redistribution of actin observed in 8-cell blastomeres is a general response
to cell interaction.
Table 8. Patterns of actin distribution in 2/4 natural pairs of various ages
Actin distribution No. (%)
Age (h post
division)
0
1
5
9
Total
r
F
H
z
P
10
50
54
40
7(70)
6(12)
6(11)
2(5)
2(20)
17(34)
4(7)
0
1(10)
27(54)
42(78)
38(95)
0
0
2(4)
0
108
M. H. JOHNSON AND B. MARO
First, we examined natural 2/4 blastomeres of various ages (Table 8 and Fig.
4c,d). All cells examined showed a homogeneous distribution of bound Con A.
Four patterns of cytoplasmic actin distribution, as described in 8-cell blastomere,
were also found in 4-cell blastomeres, and the data on their occurrence with time
are shown in Table 8. The only major differences from the 8-cell stage is the very
low incidence of cells scored as having a polar actin distribution (2 out of 154),
and the relatively high incidence of homogeneous cells in newly formed natural
pairs. This result suggested that polar actin distribution might be a characteristic
feature of the 8-cell response to cell interaction whereas a zonal actin distribution
might represent a more general response to cell contact.
In a second experiment, 22 newly formed and isolated 8-cell blastomeres were
each aggregated with an unfertilized or a newly fertilized egg, and the aggregate
was cultured for 8 h. In half of the aggregates the midbody on the 8-cell blastomers
was oriented away from the point of contact with the egg, and in the other half
the midbody on the 8-cell blastomere was oriented towards the egg. We have
shown previously (Johnson & Ziomek, 1981b) that, whilst the egg itself is a very
poor inducer of surface polarity in 8-cell blastomeres, when the midbody of the
8-cell blastomere is allowed to flatten on the egg, it appears to function as a 'mini-
8-cell' and induces a surface pole on the opposite side of the 8-cell blastomere.
Of the 11 blastomeres with the midbody oriented away from the egg, 10
developed a zonal actin distribution and were uniformly stained with Con A (Fig.
4e,f) and 1 developed a polar distribution of both actin and surface Con A.
Of the 11 blastomeres with the midbody oriented between the egg and the
Fig. 4. Panels a, c & e show Con A binding patterns; panels b, d & f show actin
distribution. (a,b) 8-cell blastomeres after disaggregation from 4h-old intact 8-cell
embryos and examination for actin distribution with anti-actin antibody and for
surface organization as assessed by Con A binding. All cells are polar for actin but
only one cell clearly polar for Con A. (c,d) Natural 2/4 pair l h old, both cells
showing apolar surface organization and zonular actin distribution. (e,f) 1/8 blastomere aggregated with a fertilized egg, cultured for 8h and then stained with antiactin antibody (f) or Con A (e); note zonal actin staining and apolar Con A staining
in 1/8 cell. Mag. x200 for a, b and x330 for c-f.
Fig. 5. Reaggregated (1/8+1/8) pairs that were incubated for 2 h (a,b) or 6h (c-h)
before disaggregation and reaggregation and a further period in culture until the cells
were aged a total of 9h. Arrowheads indicate off-axis poles of Con A binding and
the corresponding pole of cortical microvillous actin. (a,b) Both cells are on-axis and
polar for actin, Con A binding and nuclear position - note midbody remains on upper
cell. (c,d) Both cells are polar for Con A binding, the left-hand one being on-axis and
having an on-axis zonal actin pattern, the other being off-axis and having an off-axis
polar actin pattern. (e,f) Both cells are polar and off-axis for Con A binding, the
upper one being zonal and on-axis for subcortical actin (although note microvillous
actin pole corresponding to Con A pole) whereas the lower one is homogeneous for
actin - note nuclei are off-axis. (g,h) Upper cell is polar and on-axis for both Con A
binding and actin, lower one being off-axis for Con A binding and nuclear position,
and zonal and on-axis for cytoplasmic actin binding; again note pole of cortical
microvillous actin corresponding to Con A binding pole. Mag. x330.
\
A
•P
^c-
V^-r-**
H
r
\
d
g
Actin in 8-cell blastomeres
109
blastomere, all developed a polar Con A binding pattern and all except two
developed a polar cytoplasmic actin distribution opposite to the position of the
midbody. The two exceptions developed a zonal actin-staining pattern. These
results reinforced the conclusion that zonal actin staining might represent a
general response to cell contact, whereas a polar actin distribution was related
to the polarization of blastomeres observed at the 8-cell stage.
In a third experiment, newly formed natural 2/8 pairs of blastomeres were
cultured for 8-9 h either in control medium or in an antiserum to embryonal
carcinoma cells that prevents extensive cell flattening and adhesion but not cell
polarization. Those cells showing zonal actin staining at harvesting were compared, and the extent of zonal staining was seen to be related to the extent of cell
flattening.
5. Stability of zonal and polar patterns of cytoplasmic actin staining in 8-cell
blastomeres
We have shown previously that a stable axis of polarity perpendicular to the
point of intercellular contact is formed between 2 and 5 h after aggregation; thus
when cells were disaggregated and then reaggregated in a new contact relationship after this period, they 'remembered' the position of their old contact point
and developed surface poles opposite to it. In contrast, pairs that were
disaggregated and reaggregated in a new contact relationship prior to this period,
developed surface poles opposite to the new contact point (Johnson & Ziomek,
19816). We attempted to determine whether the polar actin distribution behaved
in a similar way to the surface poles.
Newly formed, isolated 1/8 cells were reaggregated in pairs (1/8+1/8), cultured for 2 or 6 h and then treated in one of two ways. Pairs in one group at each
Fig. 6. Effect of treatment of 8-cell blastomeres with CCD for varying periods on
orientation of polarity as assessed by (a & c) FITC-Con A binding, (b & d) actin
distribution and (e-1) distribution of microvilli as assessed by scanning electron
microscopy. Throughout, arrows indicate poles that are off-axis with respect to
point(s) of contact, (a-d; Mag. x330) 2/8 pairs incubated for (a,b) 9h in CCD or
(c,d) 2 h in DMSO and 7 h in CCD; note (a,c) off-axis poles of FITC-Con A binding
and (b,d) the dispersed granular distribution of cytoplasmic actin contrasted with the
cortical, microvillous actin which is distributed in positions corresponding to the
poles of FITC-Con A binding. Note also off-axis nuclei (d). (e-g; Mag. x350) 2/8
pairs incubated (e) for 9 h in DMSO, both cells polar and on-axis, (f) for 6 h in DMSO
and 3 h in CCD, left cell polar and on-axis but microvilli are more variable in length,
right cell not obviously polarized and note absence of intercellular flattening, (g) for
9h in CCD, both cells polar and off-axis, note absence of flattening, (h-1; Mag.
x 350) Whole embryos incubated in (h) DMSO for 9 h and decompacted in Ca++-free
medium, note clearly defined on-axis poles, (i) DMSO for 8-5 h and CCD for 0-5 h,
note slightly more ragged but on-axis poles, (j) DMSO for 6 h and CCD for 3 h, note
longer microvilli, poles less well defined but on-axis, (k) CCD for 9 h and (1) DMSO
for 2 h and CCD for 7 h - note poles are present but less well organized and some
clearly off-axis.
(2)
9
9
6
7
8
9
60
96
42
42
40
145
No. of cells
scored
(3)
18
22
36
33
30
27
R*
(4)
82
68
7
19
20
23
"on
(5)
0
10
57
48
50
50
Poff
(6)
18
23
47
22
33
40
H
(7)
64
67
53
78
67
55
Z
(8)
18
10
0
0
0
5
P
(9)
0
1
37
18
43
13
H
(10)
15
5
22
43
21
34
zt
(11)
A
85
94
41
39
36
53
P*
(12)
% of polar-staining Con A
cells, with actin pattern
* R = ring stain - apolar; P = polar stain that is on (POn) or off (Poff) axis with respect to contact point with companion cell,
t Where Con A was polar and off-axis and actin was zonal, the actin staining was on-axis in all cases.
t Where both Con A and actin were polar, the poles coincided in all cases except one.
2
6
6
6
6
Controls
Time of rotation Time of analysis
(h post
(h post
aggregation)
aggregation)
(1)
A.
Con A binding pattern (%)
% of ring-staining Con A
cells, with actin pattern
Table 9. Type and orientation of various actin staining and Con A binding patterns in pairs of reaggregated (1/8+1/8)
blastomeres, the relative orientation within which was changed during their 9 h incubation
w
\J
•—I
>
O
(A
E
o
<—i
E
&,
Actin in 8-cell blastomeres
111
time point were disaggregated to single cells and rotated prior to reaggregation
in a new contact orientation; pairs in the second group were left as pairs but were
exposed to Ca2+-free medium and PHA as a control. Pairs in both the 2 h-rotated
group, and its control, were cultured for a further 7 h before analysis. Pairs in the
6 h-rotated group were cultured for 0, 1, 2 or 3 h; cells in the 6h control group
were cultured for 3 h. All pairs were then examined for the distribution pattern
and orientation of their cytoplasmic actin and surface Con A receptors (Table 9).
Cells in control pairs showed a high incidence of Con A poles (82 %; Table 9,
line 1, columns 5 & 6) all of which were on-axis and almost all of which were
associated with a polar actin distribution oriented on the same axis (Table 9, line
1, columns 10-12). Pairs, in which the constituent cells had been rotated at 2h,
differed only in the inclusion of a few cells with off-axis poles for both Con A
binding and actin (Table 9, line 2, columns 5, 6 & 12; Fig. 5a,b); this result
reflects the fact that in most cells the axis of polarity is not stabilized until later
than 2h (but before 5h) into the fourth cell cycle (Johnson & Ziomek, 19816).
However, when cells were rotated at 6h, after stabilization of the axis of cell
polarity, the proportion of cells that were off-axis for Con A binding rose,
regardless of the time of analysis (Table 9, lines 3-5, columns 5 & 6). Moreover,
of the cells that were clearly polarized as judged by their Con A binding pattern,
many fewer (about half) also had a polar cytoplasmic actin distribution compared
to the controls or to 2 h-rotated groups (column 12 of Table 9; compare Fig. 5a,b
with 5e,f). Where present, the orientation of the actin pole always corresponded
to that of the surface pole (see Fig. 5c,d and 5g,h). There is a suggestion (line 5,
column 12) that the proportion of cells with a polar actin distribution rises with
time. Amongst those cells that did not show a polarized cytoplasmic actin
distribution, many showed a zonal actin distribution that was always on-axis with
respect to the point of contact (Table 9; Fig. 5c,e,g).
This result leads us to conclude that once the axis of polarity is stabilized, the
cell may retain its surface polar phenotype whilst losing its polar distribution of
cytoplasmic actin. It also supports the conclusion that zonal staining is related to
cell contact but not to cell polarization.
6. Effect of cytochalasin D (CCD) on incidence and orientation of actin and Con
A polarity
Cytochalasin D (CCD) acts to prevent microfilament growth. Thus, microfilaments that are turning over within a cell will disassemble in the presence of the
drug, whereas stable microfilaments, such as are found in striated muscle or in
the microvilli of brush borders will be relatively resistant to the drug (discussed
in Maro et al. 1984; Mak, Trier, Serfilippi & Donaldson, 1974). Cortical microvilli of mouse blastomeres have been shown to contain relatively resistant microfilaments (Pratt, Chakraborty & Surani, 1981; Sutherland & Calarco-Gillam,
1983). We used CCD to investigate the effect of disrupting the more dynamic
microfilaments on the development and stabilization of polarity in blastomeres.
112
M. H. JOHNSON AND B. MARO
Table 10. Surface polarity and relative nuclear position in natural 2/8 pairs
cultured in CCD
% of cells with nuclei
A
conHitirjrK
N o rpll<i
(h)
analysed
R
Ion
Poff
36
42
22
29
67
4
11
67
81
43
19
57
48
42
12
46
38
62
48
42
42
16
79
21
DMSO (9h)
CCD (9h)
M16(2h) +
CCD(7h)
M16(6h) +
CCD(3h)
A.
to cell contact
Newly formed natural 2/8 pairs were treated in one of four ways. One group
was placed in DMSO (the solvent for CCD) for 9 h as a control. Two other groups
were placed in CCD for 9 h or in DMSO for 2h followed by CCD for 7 h; these
groups should reveal the effect of CCD on the generation of polarity. A fourth
group was placed in DMSO and was transferred to CCD after 6 h and cultured
for a further 3h; this group should reveal the effect of CCD on the stability of
the axis of polarity. In all pairs in CCD, the cells had not flattened on each other
by the time of harvesting. Cells were analysed for their distribution of cytoplasmic actin and surface Con A receptors and for the position of their nuclei in
relation to the point of cell contact. The results are summarized in Table 10.
Cells present for the whole, or the terminal 3 or 7 h, of the incubation in CCD
showed, in most cases, a cytoplasmic actin distribution that was homogeneous
or not clearly classifiable (Fig. 6b,d); occasional zonal patterns were seen in cells
that had been placed in CCD for the last 3h, and in many cells surface
microvillous actin was not homogeneous (Fig. 6b,d & below). The incidence of
surface polarity as assessed by Con A binding was similar in cells incubated in
DMSO and CCD throughout, although somewhat lower in cells transferred to
CCD at 2 or 6 h (Table 10). However, amongst the cells with polar Con A
binding, the proportion with poles off-axis with respect to the point of intercellular contact was higher when CCD was added at or before 2 h, than it was in
the DMSO controls or after addition of CCD at 6 h, a time at which a stable axis
of polarity has formed in most cells (Fig. 6a,c; arrowed). Similarly, most nuclei
were on-axis with respect to the point of cell contact in DMSO controls or cells
transferred to CCD at 6h, but when CCD was added at earlier times, many
nuclei were off-axis (Table 10, Fig. 6d).
In order to confirm that the abnormally oriented and less discretely organized
polar patterns of Con A binding observed in CCD-treated cells did indeed reflect
cortical polarity, 2/8 pairs of cells were placed in CCD at 0 or 6 h and whole
embryos were placed in CCD at 0, 2, 6 or 8-5 h, and all samples were incubated
Actin in 8-cell blastomeres
113
for a total of 9h and examined on the scanning electron microscope (SEM).
Control embryos and pairs contained cells with a polar distribution of microvilli
as described previously (Fig. 6e,h, and Ziomek & Johnson, 1980; Reeve &
Ziomek, 1981). Many cells from embryos and pairs placed in CCD within the
first 2h after their formation also showed a polar distribution of microvilli.
However, the patterns differed from controls in two ways. First, the poles were
less clearly defined and the microvilli within them less evenly distributed and
variable in length (Fig. 6f, left cell). Second, many poles were not on-axis to the
point(s) of intercellular contact (Fig. 6g,k,l). Cells from embryos or pairs placed
in CCD for the terminal 0-5 or 3 h only of their incubation also possessed poles
that were on-axis (thereby confirming the results of Handyside, 1980) but these
poles were also not as well defined or evenly composed as in controls, especially
when CCD was present for the terminal 3h (Fig. 6i,j).
From these results we conclude that a surface pole or clump of microvilli can
develop without turnover of CCD-sensitive polymerized actin, but that CCD
does disturb the organization of this surface pole and the location of both the
surface pole and the cell nucleus such that they do not lie on an axis normal to
the point of intercellular contact.
DISCUSSION
We have examined 8-cell blastomeres of various ages, and under various
conditions, for evidence of their actin distribution in relation to the presence and
position of a surface pole (summarized in Fig. 3). The use of fixed, extracted
couplets of cells has permitted resolution of details not observable clearly in
intact embryos. Actin was localized cortically in microvilli and its distribution
there corresponded to that of Con A binding. In this paper we have been concerned exclusively with the distribution of cytoplasmic actin, for which three nonhomogeneous patterns were recognized in both couplets and in cells isolated
from whole embryos; each pattern seemed to be associated with a different
aspect of blastomere function.
The focal distribution of actin was related to the immediately preceding mitotic
division. The evidence for this conclusion is seen in the time of maximum incidence of the pattern in both early 4- and early 8-cell blastomeres, the obvious
physical association with spindle remnants, and an orientation that related to the
residual midbody in those few disaggregated blastomeres in which the focal
pattern was detected. An association between actin and the spindle remnant has
been detected in many other systems including the second meiotic spindle of the
mouse oocyte (Aubin, Weber & Osborn, 1979; Maro etal. 1984).
The subcortical zone adjacent to areas of intercellular contact and relatively
clear of actin (zonal pattern) appeared to be associated with the cell apposition
that occurs during early cleavage and is particularly evident during compaction
at the 8-cell stage. Thus, apposed areas of both 4- and 8-cell blastomeres were
114
M. H. JOHNSON AND B. MARO
deficient in actin, and where apposition was absent or reduced as a result of
disaggregation of cells or incubation in antiserum to EC cells, zonal staining was
also absent or reduced in extent. Moreover, incubation of embryos or pairs of
cells in cytochalasin D prevents or reverses cell flattening implying a role for
microfilaments in this process (Table 10; Handyside, 1980; Pratt et at. 1981;
Pratt, Ziomek, Reeve & Johnson, 1982; Sutherland & Calarco-Gillam, 1983).
Independent confirmation that polymerized actin is cleared from contact zones
has been obtained by decorating microfilaments in detergent-extracted pairs of
blastomeres with myosin Si subunits and their examination on the high-voltage
microscope (unpublished work with J. Van Blerkom). Moreover, endogenous
myosin seems to demonstrate a similar contact-dependent zonal clearing (Sobel,
1983).
There appears to be no obvious association between zonal clearing of actin
from contact areas and the development of polarity in blastomeres since (i) zonal
clearing was observed in 4-cell blastomeres that did not polarize, (ii) oocytes or
eggs that did not induce polarity could induce zonal clearing, (iii) antiserum to
EC cells did not prevent or reduce development of polarity but did reduce the
extent of both cell flattening and zonal clearing, (iv) disaggregation and
reaggregation of some pairs of cells, in which the axis of surface polarity had
already stabilized, led to dissociation of the orientation of the axis of the surface
polar pattern from that of the actin zonal pattern in the same cell.
The polar pattern of cytoplasmic actin was detected earlier than surface polarity
as assessed by both Con A binding and cortical microvillous actin, and was first
observed at about the time that the axis of polarity is becoming stabilized (Johnson & Ziomek, 1981b). However, the subsequent appearance of polar Con A
binding always overlay the polar actin. These observations could mean that
polarization of actin forms part of a causal sequence that results ultimately in
surface polarity. Alternatively, both actin and surface polarization could
represent independent manifestations of some underlying controlling process.
The results of two types of experiment do not clearly distinguish between these
alternatives. If cells, in which a stable axis of polarity had been laid down, were
disturbed by disaggregation and reaggregation in a new orientation, then many
cells that manifested the stable surface polarity lacked polarity of cytoplasmic
actin distribution. Since all undisturbed cells that are polar at their surface also
show polar cytoplasmic actin patterns, this result implies a greater stability at
disaggregation of surface polarity than of cytoplasmic actin polarity. Stability of
surface polarity during mitotic division has been described previously, at least
one polar cell always being among the progeny (Johnson & Ziomek, 1981a,
1983). Thus, if the polar organization of actin does prove to have a causal role
in the sequence leading to surface polarity, then that role would have to be
transitory. Cytoplasmic actin poles do not appear to be required for stability of
the surface pole once formed.
The results obtained with use of the drug cytochalasin D provide some
Actin in 8-cell blastomeres
115
evidence for cytoplasmic actin having a role to play in the sequence of events leading to surface polarity, but do not allow precise identification of that role. Thus,
continuous incubation in the drug did not affect the incidence of poles, although
the tight organization of the pole of microvilli was affected. This is not surprising
in light of the microvillous content of cortical filamentous actin (e.g. Figs 5e,g &
6a,c) which appears to be more CCD-resistant presumably because of a much
lower turnover rate (Pratt et al. 1981). However, although surface poles were
evident in CCD-treated cells, they did not always lie on the expected axis of polarity relative to cell contacts. When CCD was added prior to the stabilization of the
axis of polarity, the relationship between the cell-contact point(s) and the position
of the surface pole observed in controls was lacking. When CCD was added after
the axis of polarity had been stabilized, the position of the surface pole in most cells
was similar to that observed in controls. The disorienting effect of CCD applied
early in the 8-cell stage could result from an effect of CCD on the setting up of a
primary axis of polarity thereby leading to misalignment of all polarized features
of the cell. Alternatively the primary axis might be oriented correctly in the
presence of the drug, which could act by interfering with the secondary arrangement of cell organelles and structures in relation to that axis. The latter alternative
seems more likely by analogy with the disruptive effects of CCD on oriented responses in other systems. Usually aspects of cell movement are affected and not the
orientation process itself (Malech, Root & Gallin, 1977; Gottlieb, Subramanian &
Kalnins, 1983). At the 8-cell stage, movement and/or reorganization occur not
only of micro villi but also of the nucleus (Tables 6 & 7 and Reeve & Kelly, 1983),
of the endocytotic uptake (Reeve, 1981) and processing systems of the cell (T.
Fleming & S. Pickering, in preparation), and of intracellular clathrin (B. Maro, S.
Pickering, M. Johnson & D. Louvard, in preparation). We are presently examining these and other features of cell polarization to determine whether in every case
their occurrence and/or alignment is sensitive to CCD. With this approach, it may
be possible to discriminate between the alternative mechanisms outlined above.
We wish to acknowledge the technical assistance of Gin Tampkins, Sue Pickering, Ken
Thurley, Tim Crane, Roger Liles, Raith Overhill and Ian Edgar. We thank Dr E. Karsenti and
P. Gounon for the generous gift of anti-actin antibodies. The work was supported by a grant
from the Medical Research Council of Great Britain. B.M. is an EMBO fellow.
REFERENCES
J. E., WEBER, K. & OSBORN, M. (1979). Analysis of actin and microfilamentassociated proteins in the mitotic spindle and cleavage furrow of Pt K2 cells by immunofluorescence microscopy. Expl Cell Res. 124, 93-109.
DUCIBELLA, T., ALBERTINI, D. F., ANDERSON, E. & BIGGERS, J. (1975). The preimplantation
mammalian embryo: characterization of intercellular junctions and their appearance during
development. Devi Biol. 45, 231-250.
DUCIBELLA, T., UKENA, T., KARNOVSKY, M. & ANDERSON, E. (1977). Changes in cell surface
and cortical cytoplasmic organization during early embryogenesis in the preimplantation
mouse embryo. /. Cell Biol. 74, 153-167.
AUBIN,
116
M. H. JOHNSON AND B. MARO
B. P. & WHITTINGHAM, D. G. (1978). Activation of mammalian oocytes by
intracellular injection of calcium. Nature 273, 149-151.
GOODALL, H. & JOHNSON, M. H. (1984). The nature of intercellular coupling within the
preimplantation mouse embryo. /. Embryol. exp. Morph. 19, 53-76.
GOTTLIEB, A. I., SUBRAMANIAN, L. & KALNINS, V. I. (1983). MTOC and cell migration: effect
of inhibition of migration and microtubule disruption in endothelial cells. /. Cell Biol. 96,
1266-1272.
GOUNON, P. & KARSENTI, E. (1981). Involvement of contractile proteins in the changes in
consistency of oocyte nucleoplasm of the newt Pleurodeles waltlii. J. Cell Biol. 88,410-421.
HANDYSIDE, A. H. (1980). Distribution of antibody- and lectin-binding sites on dissociated
blastomeres from mouse morulae: evidence for polarization at compaction. /. Embryol.
exp. Morph. 60, 99-116.
JOHNSON, M. H. & ZIOMEK, C. A. (1981a). The foundation of two distinct cell lineages within
the mouse morula. Cell 24, 71-80.
JOHNSON, M. H. & ZIOMEK, C. A. (1981ft). Induction of polarity in mouse 8-cell blastomeres:
specificity, geometry and stability. J. Cell Biol. 91, 303-308.
JOHNSON, M. H. & ZIOMEK, C. A. (1982). Cell subpopulations in the late morula and early
blastocyst of the mouse. Devi Biol. 91, 431-439.
JOHNSON, M. H. & ZIOMEK, C. A. (1983). Cell interactions influence the fate of mouse blastomeres undergoing the transition from the 16- to the 32-cell stage. Devi Biol. 95, 211-218.
LEHTONEN, E. (1980). Changes in cell dimensions and intercellular contacts during cleavagestage cell cycles in mouse embryonic cells. /. Embryol. exp. Morph. 58, 231-249.
LEHTONEN, E. & BADLEY, R. A. (1980). Localization of cytoskeletal proteins in preimplantation mouse embryo. /. Embryol. exp. Morph. 55, 211-225.
Lo, C. W. & GILULA, N. B. (1979). Gap junctional communication in the preimplantation
mouse embryo. Cell 18, 399-409.
MAGNUSON, T., DEMSEY, A. & STACKPOLE, C. W. (1977). Characterization of intercellular
junctions in the preimplantation mouse embryo by freeze-fracture and thin-section electron
microscopy. Devi Biol. 61, 252-261.
MAK, K. M., TRIER, J. S., SERFILIPPI, D. & DONALDSON, JR., R. M. (1974). Resistance of adult
mammalian intestinal mucosa to cytochalasin B. Expl Cell Res. 86, 325-332.
MALECH, H. L., ROOT, R. K. & GALLIN, J. I. (1977). Structural analysis of human neutrophil
migration: centriole, microtubule and microfilament orientation and function during
chemotaxis. J. Cell Biol. 75, 666-693.
MARO, B., JOHNSON, M. H., PICKERING, S. & FLACH, G. (1984). Changes in actin distribution
during fertilization of the mouse egg. J. Embryol. exp. Morph. 81, 211-237.
NICOLSON, G. L., YANAGIMACHI, R. & YANAGIMACHI, H. (1975). Ultrastructural localization
of lectin binding sites on the zonae pellucidae and plasma membranes of mammalian eggs.
/. Cell Biol. 66, 263-274.
PRATT, H. P. M., CHAKRABORTY, J. & SURANI, M. A. H. (1981). The effect of cytochalasin D
on the differentiation of the mouse blastocyst. Cell 26, 279-292.
PRATT, H. P. M., ZIOMEK, C. A., REEVE, W. J. D. & JOHNSON, M. H. (1982). Compaction of
the mouse embryo: an analysis of its components. /. Embryol. exp. Morph. 70, 113-132.
REEVE, W. J. D. (1981). Cytoplasmic polarity develops at compaction in rat and mouse
embryos. /. Embryol. exp. Morph. 62, 351-367.
REEVE, W. J. D. & ZIOMEK, C. A. (1981). Distribution of microvilli on dissociated blastomeres
from mouse embryos: evidence for surface polarization at compaction. /. Embryol. exp.
Morph. 62, 339-350.
REEVE, W. J. D. & KELLY, F. P. (1983). Nuclear position in the cells of the mouse early
embryo. /. Embryol. exp. Morph. 75, 117-139.
SOBEL, J. S. (1983). Cell-cell contact modulation of myosin organisation in the early mouse
embryo. Devi Biol. 100, 207-213.
SUTHERLAND, A. E. & CALARCO-GILLAM, P. G. (1983). Analysis of compaction in the preimplantation mouse embryo. Devi Biol. 100, 328-338.
WHITTINGHAM, D. G. & WALES, R. G. (1969). Storage of two-cell mouse embryos in vitro.
Austr. J. biol. Sci. 22, 1065-1068.
FULTON,
Actin in 8-cell blastomeres
111
C. A. & JOHNSON, M. H. (1980). Cell surface interaction induces polarization of
mouse 8-cell blastomeres at compaction. Cell 21, 935-942.
ZIOMEK,
(Accepted 21 March 1984)