/ . Embryol. exp. Morph. Vol. 58, pp. 231-249,1980
Printed in Great Britain © Company of Biologists Limited 1980
231
Changes in cell dimensions and
intercellular contacts during cleavage-stage cell
cycles in mouse embryonic cells
By E. LEHTONEN 1
From the Zoology Department, University of Oxford
SUMMARY
The cleavage behaviour of cells isolated from 1- to 8-cell-stage mouse embryos was studied
with time-lapse video equipment; changes in cellular dimensions and their timing were
recorded. The division of an isolated cell results in the formation of a twin-cell pair. The
divisions of these two cells were always asynchronous. In each division the volume of a
daughter cell was approximately half of that of the parental cell but its apparent surface area
was 59-65 % of that of the parental cell. Consequently, the ratio of apparent surface area to
volume increased in each division by 25-30%. The most noticeable changes were observed
in the relationship between the two daughter cells of each division. After cytokinesis, the
intercellular contact area gradually increased during the following cell cycle in the 2/8- and
2/16-cell pairs, whereas it hardly changed in the 2/2- and 2/4-cell pairs. The comparison of
the behaviour of the daughter cells on different substrates suggested that the zona pellucida
and the mid body might have a role in the contact development at the early stages. Scanning
electron microscopy was used for studying changes in the density of cell surface microvilli
in an attempt to explain how the cells regulate their intercellular contacts.
INTRODUCTION
The major factor influencing cell fate during preimplantation mouse development is the relative position of the cell in the embryonic structure (Tarkowski &
Wroblewska, 1967; Hillman, Sherman & Graham, 1972; Kelly, 1977). The
movements of the cells to their different relative positions principally occur
during cell division (Graham & Deussen, 1978; Graham & Lehtonen, 1979) and
the orientation of these divisions apparently depends on continuous cell interactions during cleavage-stage development (Graham & Lehtonen, 1979).
Furthermore, the cleavage divisions are asynchronous, and cell allocation to the
inner cell mass (ICM) and to the trophectoderm of the blastocyst is related to
the division order during the cleavage stages (Kelly, Mulnard & Graham, 1978).
The present study attempts to analyse some of the cell interactions involved
in the formation of the cleavage-stage structure. The movement of the cleavagestage cells at any moment during development depends on their contacts with
1
Author's present address: Department of Pathology, University of Helsinki, Haartmaninkatu 3, 00290 Helsinki 29, Finland.
232
E. LEHTONEN
other cells a moment ago (Graham & Deussen, 1978; Graham & Lehtonen,
1979). I have therefore analysed the development of cell contacts in isolated
cleavage-stage cells in vitro. During cell division the spherical cell body narrows
and elongates. This change of form frequently results in the formation of new
intercellular contacts as cells extend in division and it probably alters relative
cell position because dividing cells tend to maintain their previous intercellular
contacts during this shape change and move in association with the mitotic cell
(Graham & Deussen, 1978; Graham & Lehtonen, 1979). After division, the
daughter cells flatten against each other and slight relative cell movements
appear to be the consequence of this process. In order to understand the range
and efficiency of these possible morphogenetic factors I have quantitated the
changes in cell dimensions and intercellular contacts in cleavage-stage cells
developing in vitro.
MATERIALS AND METHODS
The supply of embryos
The embryos were from natural matings and were (C57BL6 x CBA)F2. They
were dissected into pre-warmed and pre-equilibrated Whitten's medium (1971).
The embryos were removed from the reproductive tract between 2 and 16 h
before the division stage which was to be studied.
Dissociation of embryos
The zona pellucida was partially dissolved by incubating each embryo in
pronase (Calbiochem. Co., U.K., technique of Mintz, 1967) at +37 °C for
5-5-6 min and then it was pipetted off 1-cell and 2-cell embryos after 30 min
incubation in the medium.
For the dissociation of 4-cell and 8-cell stages, each embryo was cultured
inside the pronase-thinned zona for 15 min in Whitten's medium in which the
calcium concentration had been lowered to 0-02 mM. The zona was then pipetted
off the embryo and thereafter the cells were blown apart by pipetting through a
flame-polished micro-pipette with a mouthpiece. The cells were replaced in
normal Whitten's medium immediately after dissociation. The low calcium
treatment was not needed for 1-cell- and 2-cell-stage embryos.
Culture
The isolated blastomeres were cultured in Whitten's medium in microdrops
under paraffin oil (selected for absence of toxicity to cultured embryos, Boots
Pure Drug Co., U.K.) in tissue culture flasks (Falcon Plastics, Oxnard, California) pregassed with a gas mixture of 5 % CO2, 5 % O2 and 90 % N 2 . The
microdrops were placed on the concave surface of a siliconized (Repelcote,
BDH, U.K.) capillary tube fragment.
Dimensional changes in cleavage-stage cells
233
Observations
For the measurement of cellular dimensions the cells were observed continuously in a +37 °C room with a video recording system fitted to a Wild
inverted microscope. The recording system consisted of a WV-1350 camera
(Matsushita Electric Trading Company, Osaka, Japan), a NV-8030 time-lapse
video-tape recorder (Matsushita Company) and a VM-172 AV video monitor
(Hitachi Denshi Limited, Lodge Road, London, U.K.). The records were made
with a recording interval of 1-44 sec. The illumination was kept to a minimum,
and the light passed through a 4 mm-thick glass heat filter and a 3 cm-thick
filter containing tissue-culture medium; under these conditions twenty 2-cell
embryos each formed a morphologically normal blastocyst on the fourth day of
development, and it is therefore unlikely that the illumination disturbed the
cellular events.
The cellular dimensions were measured continuously from the recording, and
the changes in dimensions were traced on transparent film. The cells were rarely
completely spherical. Therefore the largest and shortest diameters were measured
and the mean diameter was calculated. The development of an isolated cell was
followed through at least two consecutive divisions, and data was excluded if
the cell failed to produce four viable descendants (see Results, section 6).
Scanning electron microscopy
For the scanning electron microscopy (SEM) data all the isolated cells of an
embryo were cultured simultaneously in microdrops on a bacteriological grade
dish. The cells were fixed individually at varying periods of time after division.
In addition, some cells were fixed before division, and to ensure that these cells
were viable only cells with dividing sister cells were included in the data. The
cells were rinsed quickly in 0-1 M cacodylate buffer (pH 7-6) with 2 mM-CaCl2
at + 4 °C and fixed in 1-5% (v/v) glutaraldehyde in 0-09 M cacodylate buffer
(pH 7-6) with 2 mM-CaCl2 at + 4 °C for 60 min. After fixation the cells were
attached individually to poly-1-lysine-treated glass surface (Graham &Lehtonen,
1979) in such a way that both cells of a twin-cell pair lay on the glass substrate
and had similar contact to the glass. Thereafter the cells were processed for
SEM as described elsewhere (Graham & Lehtonen, 1979).
For quantitative data the surface microvilli were counted over an area of
about 30 jiim2 on the uppermost part of each cell directly opposite the area in
contact with the glass surface. In this area, the cell surface appears to be perpendicular to the electron beam and minimal distortion of the surface area of these
curved objects should occur in this region. The microvilli were counted from
micrographs with final magnifications of about x 15000. Micrographs of a
grating replica (2160 lines/mm) were used for calibration.
234
E. LEHTONEN
Control experiments
The following experiments were conducted to discover if the culture conditions
described above were likely to affect cell behaviour during and after cell division.
In addition to culture on the siliconized glass surfaces, cells were cultured in the
medium on an agarose surface (1 % agarose in water) which had previously been
equilibrated with the medium for 30min, or in microdrops on the plastic
surface of bacteriological grade dishes (Sterilin Ltd., Richmond, Surrey, U.K.).
In some experiments the plastic surface was scratched with a glass needle before
culture.
To test the effects of the dissociation methods on cleavage behaviour 1-celland 2-cell-stage embryos were subjected to 5-5-7 min pronase treatment and
then cultured undissociated either inside or outside the thinned zona pellucida.
In these preliminary experiments the observations were intermittent and made
with the video recording system or by photography. The data were excluded
from the quantitative results.
RESULTS
The results are presented in six sections. (1) There is data on cell cycle durations and division asynchrony of isolated cleavage-stage cells in culture. (2) The
physical dimensions of isolated cleavage-stage cells in culture are given. (3) and
(4) The changes in cellular dimensions during and after cell division in culture
are described. (5) There is data on cell cycle differences in the density of surface
microvilli of the cleavage-stage cells. (6) There are observations on the effect of
culture substratum on cleavage behaviour.
The quantitative study of cellular dimensions is based on continuous measurements made on isolated cleavage-stage cells which divided twice on a siliconized glass surface producing four descendants. Twenty-five video records
were made in which the behaviour of cells was clearly recorded. In three of these
experiments the cell failed to produce four viable descendants and these records
were discharged.
1. Cell cycle durations and division asynchrony
The cell cycle durations of cleavage-stage cells were calculated by measuring
the time from the division of an isolated cell to the divisions of the cell pair thus
formed. The mean cell cycles of the 2-, 4-, 8- and 16-cell stages were 24-2, 12-3,
13-3, and 12-0 h, respectively (Table 1). The cell cycle times varied considerably
within each stage studied. The differences between the shortest and the longest
2nd, 3rd, 4th and 5th cell cycle were, respectively, 13, 36, 16, and 16% of the
shortest cycle (Table 1).
The divisions of the two daughter cells of the isolated cell were always
asynchronous. The mean division interval (93 min) between the faster and
slower dividing cells of the 2/2-cell pairs represents 6-4 % of the mean cell cycle
1/1-cell
1/2-cell
1/4-cell
1/8-cell
2/2-4/4 (2nd)
2/4-4/8 (3rd)
2/8-4/16 (4th)
2/16-4/32 (5th)
No. of
observations
1405
704
786
707
Mean
1372-1432
618-801
737-837
673-750
Range
1498
771
810
735
Mean
1431-1556
679-843
. 763-853
709-780
Range
Cycle duration of the
slower cell (min)
93
68
24
29
Mean
59-146
27-148
16-34
10-36
Range
Interval from first to
second division (min)
The cell cycle durations were determined by observing an isolated zona-free cell to divide twice to produce four descendants. The time was counted
from the beginning of cytokinesis. Interval from first to second division refers to the asynchrony between the faster and slower dividing cell in
each individual experiment.
Step at which
culture initiated
Cell cycle measured
Cycle duration of the
faster cell (min)
Table 1. Cell cycle durations and asynchrony between divisions of twin cells
x
1
r
a'
I
ZJD
E. LEHTONEN
Table 2. The diameters and volumes of cleavage-stage cells in culture
Cell1 diameter
Stage
1-cell (a)
2-cell (b)
(a)
4-cell (b)
(a)
8-cell (b)
(a)
16-cell (b)
No. of
observations
5
10
6
12
5
10
6
12
Om)
Volume difference
Surface
area
Volume
i(x 103 /tm2) (xlO 3 /tm 3) Mean
Range
A
Mean
Range
68-8
530
54-4
43-2
43 0
34-5
34-7
26-9
67-9-69-5
49-6-54-7
53-3-55-7
40-8-45-0
42-2-43-5
33-6-361
33-6-35-8
25-8-28-2
14-87
8-82
9-30
5-86
5-80
3-75
3-77
2-28
170-3
77-9
84-3
All
41-6
21-6
21-8
10-2
—
—
116
50-21-9
51
1-6-8-2
4-3
0-11-8
61
1-2-10-5
Cell surface areas and volumes are calculated from the mean diameters of the cells. Volume
difference refers to unequal cell division in individual experiments. This difference between
the two daughter cells is expressed as percentage of the bigger volume.
(a) The mean diameters of isolated cells were measured immediately before they divided.
(b) The mean diameters were measured for both sister cells immediately after division of
the isolated parental cell. The (6)-values of each stage are from the same experiments as the
(a)-values of the preceding stage.
duration (1452 min). Correspondingly, the division asynchrony within 2/4-, 2/8and 2/16-cell pairs represents 9-2, 3-0 and 4-0%, respectively, of the mean cell
cycle time (Table 1). This data on asynchrony supplements the less accurate
data previously obtained by intermittent observations (Kelly, Mulnard &
Graham, 1978), and confirms the view that asynchronous divisions of daughter
cells is common in the development of these embryos.
2. The dimensions of isolated cleavage-stage cells
The mean diameters of 1- to 16-cell-stage cells were measured both immediately before an isolated cell divided in culture and immediately after the division.
These measurements were used for calculating cell volumes and apparent
surface areas (Table 2). At each cleavage stage, the post-division observations
('£'-values in Table 2) are exclusively on the daughter cells formed by cells
isolated at the previous stage ('a'-values in Table 2). Comparison of these predivision and post-division volumes suggests that a slight volume loss takes place
near the time of the first division of the zona-free zygote: the combined volume
of the two daughter cells is about 91 % of that of the late 1-cell stage. In the
following divisions the combined volume of the daughter cells is approximately
equal to the mother cell volume. Consequently the total cellular volume of the
8- to 16-cell-stage embryo is approximately that of the late 1-cell-stage embryo
(Table 2). The cleavage divisions are rarely completely equal, volume differences
of up to 21-9 % between viable sister cells were observed (Table 2).
Dimensional changes in cleavage-stage cells
237
Table 3. Changes in dimensions during cell division
Stage
1/11/21/41/8-
to
to
to
to
2/2-cell
2/4-cell
2/8-cell
2/16-cell
Width reduction
Elongation (/Am)
No. of
observations
Mean
5
6
5
6
37-5
33-4
260
181
Range
34-7-38-9
29-8-36-4
25-0-27-4
15-8-19-3
0/
/o
55
61
60
52
Mean
15-8
12-5
9-3
81
Range
150-171
11-4-14-0
80-11-5
7-0-8-6
Om)
/o
23
23
22
23
Elongation is the difference between the mean cell diameter before division and the
maximum length of the daughter cell pair at any moment during 150 min after the beginning
of division. Width reduction is the difference between the mean cell diameter before division
and the minimum mean width of the daughter cell pair at any moment during 150 min after
the beginning of division. The mean changes are expressed also as % of the mean diameter of
the parental cell. The material is the same as that in Table 2.
During the cleavage-stage development, the ratio of apparent cell surface
area to volume increases (the apparent surface area is the surface area calculated
from these light microscope observations and does not take into account microvillar projections from the cell surface). The average apparent surface area of a
daughter cell is between 59 and 65 % of the area of the parental cell (Table 2).
Consequently, the calculated total apparent surface area of the embryo increases,
and the total apparent area of all disaggregated cells of the 16-cell-stage embryo
is about 2-4 times that of the late 1-cell stage. Similarly, the ratio of apparent
surface area to volume increases in each division by between 25 and 30%; and
this ratio for 16-cell-stage cells (0-224) is about 2-6 times that of the late 1-cell
stage (0-087).
3. Changes in cellular dimensions during cell divisions in culture
The cellular dimensions of isolated 1- to 8-cell-stage cells were measured
continuously before and during cell division, and during subsequent development until the next cell division. In the conditions used in the video recordings,
the two cells produced by the division of an isolated 1-cell-stage cell did not
develop an intercellular membrane contact; in each experiment the mid body
was seen separating the cells during the 2-cell stage. One cell pair out of seven
produced by the division of an isolated 2-cell-stage cell did not develop an
intercellular membrane contact. In this case, the mid body was seen separating
the daughter cells during the following cell cycle and this record was not included
in the cell dimension data. In all other cases, the daughter cells formed a contiguous pair with contact between the membranes of the daughter cells.
In cell division the spherical cell body elongates and narrows to produce a
daughter cell pair. The relative magnitude of these dimension changes was
similar at all stages studied. On average, the maximum division associated
elongation of the cell body was 52-61 % of the mean diameter of the parental
16-2
238
E. LEHTONEN
90
-1
••••
•
•
a.
80
70
60
50
40
30
20
•*•-••••.•»%.
10
Q
o
I
i
i
3
4
i
7
8
9
I
8
9
i
i
i
i
11
12
13 14
I
I
1
I
10
11 12 13 14
10
-2
70
60
3
<D
50
"5
40
h or
..I
e
30
•3
20
10
n, 8
0
J
I
J
I
1
1
50
« 40
•2 30
I20
"So
J 10
8
0 1
i
i
i
2
3
R
i n i
i
i
4
5 6
7
8 9
Time after cytokinesis (h)
FIGURES
i
10
i
i
11 12
i
13
14
1-3
An example of changes in cellular dimensions during 1/2- to 2/4-cell (Fig. 1),
1/4- to 2/8-cell (Fig. 2) and 1/8- to 2/16-cell (Fig. 3) division. The upper diagram
of each figure represents the maximum length of the undivided parental cell or that
of the daughter cell pair after division. Each of the lower diagrams shows the
diameter of the cleavage furrow or that of the contact area between the daughter
cells in the same experiments. The beginning of cleavage furrow formation was
chosen for time point 0. The diagram pairs are from individual experiments; the
data from these diagrams are included in the Tables.
Dimensional changes in cleavage-stage cells
239
Table 4. The plateau diameters of contact areas between
two sister cells
Stage
No. of
observations
2/4-cell
2/8-cell
2/16-cell
6
5
6
Contact diameter (jim)
0/
/o
of width
Mean
Range
Mean
Range
% of cell
surface
14-5
12-9-161
32-5-38-9
26-8-27-6
35
95
98
31-38
90-98
95-100
2-8
26-5
25-5
35-6
27-2
For the plateau diameter of the contact area, the peak value of the apparent contact
plateau was averaged with values 60 min on either side of that point. In two experiments the
2/16-cell structure turned so that the cells were lying on top of each other. This appeared to
happen soon after the plateau contact was reached. In these occasions the plateau contact
diameter was calculated by averaging the last measurable contact value with the value 30 min
before that. The plateau contact diameters are expressed also as % of the simultaneous
mean widths of the cell pairs. The percentage of the contact area compared to the total cell
surface area ('b '-values from Table 2) is calculated assuming that the contact area is spherical
and that the total cell surface area does not change during the cell cycle. The material is the
same as that in Table 2.
cell (Figs 1-3, 9; Table 3). Correspondingly, the division-associated width
reduction of the cell body was 22-23 % of the mean diameter of the parental
cell (Figs 1-3, 9; Table 3).
4. Changes in cellular dimensions after cell division in culture
The extent of the contact between the two daughter cells changes during the
following cell cycle. Immediately after cytokinesis the contact area is relatively
small. Subsequently it increases, reaches an apparent plateau and then decreases
to some extent just before the next cell divisions (Figs 1-3). The maximum
contact diameters for 2/4-, 2/8- and 2/16-cell pairs during interphase were,
respectively, 35, 95 and 98 % of the simultaneous mean widths of the cell pairs
(Table 4). Assuming that the contact area is spherical, these contact diameters
represent 2-8, 26-5 and 25-5 %, respectively, of the corresponding total cell
surface areas (lb '-values in Table 2).
The changes in the intercellular contact area are closely connected with
changes in the total length of the cell pair: increase in the contact diameter is
accompanied by simultaneous decrease in the length of the cell pair (Figs 1-3,
Table 5). The mean post-division length reductions in 2/4-, 2/8- and 2/16-cell
pairs were, respectively, 5-4, 30-5 and 24-0% of the mean peak lengths of the
cell pairs (Table 5).
The rate of changes in the above dimensions is not constant, and the kinetics
of the changes appears to differ from stage to stage (Figs 1-3). In order to
quantify this impression, the time which elapsed from the beginning of cytokinesis to the period when the diameter of cell contact hardly increased was
240
E. LEHTONEN
Table 5. Post-division length reduction of contiguous sister cell pairs
in culture
Stage
No. of
observations
2/4-cell
2/8-cell
2/16-cell
6
5
6
Length reduction (/tm)
A
,
N
Mean
Range
4-8
21-1
12-5
1-6- 6-8
18-9-22-5
11-4-14-3
,
% of peak length
*
Mean
Range
5-4
30-5
240
1-9- 7-9
27-1-33-6
21-8-26-9
The peak length of the newly divided sister cell pair was compared with the bottom length
of the pair. The length reduction is expressed also as % of the peak length. The material is
the same as that in Table 2.
determined. To calculate the onset of the latter period, the rate of increase of the
contact diameter was studied: the midpoint of the last 30 min period during
which the intercellular contact diameter still increased by at least 5 % of the
total increase was used. In the case of 2/4-cell pairs the dimension changes were
so small that reliable calculations were not possible although it is obvious that
cell contact diameter does not increase markedly at times later than 60 min after
division (Fig. 1). In the case of 2/8-cell pairs this time period ranged from 453
to 525 min, and the mean was 482 min. In the case of 2/16-cell pairs this time
period ranged from 129 to 177 min, the mean being 155 min.
In addition to the cell cycle changes in dimensions, the cells regularly showed
other motile features. During the first 60-90 min after division cell membrane
blebbing of varying degree was frequently observed at all stages studied.
Simultaneously the cells often wobbled with the intercellular contact as an axis;
this continued during the interphase and occasionally resulted in rolling of the
cells on the glass surface. Moreover, the cells showed contraction and expansion movements of varying amplitudes and frequencies. These rhythmical
pumping movements were most pronounced in the 8- and 16-cell-stage cells
where their amplitude was often about 10% of one dimension. In the 8-cellstage cells the pumping movements usually started about 2 h after division and
continued at a frequency of about 1 cycle of movements in 1 min through most
of the interphase. The result is based on two experiments, in each of which about
10 cycles were followed. About 30-60 min before next cell division the movements slowed down and stopped about 10 min before division. The above
movements were less striking in the 4-cell-stage cells and usually very mild or
negligible in 1- and 2-cell-stage cells. Comparable rhythmical contraction and
expansion have been reported in rabbit ova (Ogawa, Satoh & Hashimoto, 1971)
and in mouse blastocyst (Kuhl & Friedrich-Freksa, 1936). After the second round
of cell divisions the above described cell movements resulted in the establishment
of new intercellular contacts in several experiments.
Dimensional changes in cleavage-stage cells
241
FIGURES 4-7
Scanning electron micrographs of a 2/8-cell pair about 20 min (Figs 4-5) and
about 5 h (Figs 6-7) after the division of the isolated 4-cell-stage cell. The distribution
of the microvilli is relatively even, although there seems to be some concentration
in the intercellular contact area and in the polar regions. The microvillar density
appears lower 20 min after cytokinesis than it does 5 h after cytokinesis. MB, mid
body.
242
E. LEHTONEN
4-cell
stage
8-cell
stage
12
10
•
*
•
•
{
Q
O
6 4 -
'
:
•
2 -
i
0
1
i
2
i
i
i
i
3
4
5
6
7
Time after cytokinesis (h)
Fig. 8. Microvillar densities of isolated late 4-cell-stage cells and of 8-cell-stage
cells derived from the division of isolated 4-cell-stage cells in culture. The diagram
is based on scanning electron micrographs.
5. Surface microvilli during cell cycle
In order to explain the cell behaviour described above, the surface characteristics of 4- to 8-cell-stage cells were studied with SEM. It was supposed that
the differences in contact area between cells at different stages of development
might correlate with the number of microvilli on the blastomeres; these surface
organelles are known to make contact between adjoining cells.
In isolated late 4-cell-stage cells the distribution of microvilli was relatively
even. During cytokinesis concentration of microvilli to the cleavage furrow area
was regularly observed, but already about 20 min after cytokinesis the distribution of the microvilli appeared quite even (Figs 4 and 5). Later during the
8-cell-stage cell cycle some polar microvillus concentration was seen occasionally (Figs 6 and 7), but usually the distribution appeared relatively even. The
microvillar densities were determined from an area about 1/4 of a cell circumference away from the intercellular contact point both because this is the area
which would probably touch other cells in the intact embryo and because it was
impossible to obtain accurate counts at the junction between the two daughter
cells. At the late 4-cell stage and the late 8-cell stage the density of surface microvilli is similar (Fig. 8). The microvillar density drops for 60 to 90 min after 4to 8-cell division, but it then again reaches the level scored for late 4-cell-stage
cells. The microvillar densities which were measured within 1 h after cytokinesis
are less than those measured later during the cell cycle (P < 0-001) (Fig. 8).
The microvillar densities of isolated late 8-cell-stage cells were similar to those
of the late 2/8-cell pairs.
Dimensional changes in cleavage-stage cells
243
Table 6. The effect of culture substratumon cleavage behaviour
Contiguous cell pairs/number of
experiments
Division observed
Siliconized glass
Bacteriological dish
Scratched bacteriological dish
Agarose
1/1-to
2/2-cell
1/8-to
2/16-cell
0/8
5/11
8/8
10/11
11/11
8/8
8/8
12/12
Isolated 1-cell-stage and 8-cell-stage cells were cultured on different substrata. The possible
contiguity of the resulting daughter cell pairs was observed about 5-6 h after division.
6. The effect of culture conditions on cleavage behaviour
In order to test the possible effect of culture conditions on cleavage behaviour
four different culture substrata were used, viz. siliconized glass, bacteriological
grade plastic dish, bacteriological dish scratched with a glass needle, and agarose
(Table 6). The division of an isolated, zona-free 1-cell-stage cell never produced
a contiguous cell pair with an intercellular membrane contact on siliconized
glass surface (none out of eight experiments); in each experiment the mid body
was seen separating the cells during the 2-cell stage. On bacteriological grade
plastic surface five experiments out of eleven resulted in a contiguous cell pair
with an intercellular membrane contact; on scratched plastic surface or on
agarose surface the resulting 2/2-cell pairs were nearly always contiguous
(Table 6). During further cultivation both cells of the 2/2-cell pairs divided. The
resulting four 4-cell-stage cells always established at least three intercellular
membrane contacts even if the previous 2/2-cell pair appeared to be separated
by the mid body. During further cultivation most of these embryos formed
blastocysts. The ability to develop to this stage was not dependent on culture
substratum or membrane contact at the 2/2-cell stage. The division of an isolated
8-cell-stage cell always produced a contiguous cell pair regardless of the culture
substratum (Table 6). During further culture these experiments regularly
resulted in the formation of blastocyst vesicles. Clearly, the substratum can
influence the division behaviour of blastomeres from early stages of development
(see Discussion).
The following experiments were performed in order to test the effects of the
dissociation methods on cleavage behaviour. (1) 2-cell-stage embryos were
pronased for 5-5-7 min, the zona pellucida were removed, and the embryos were
cultured on bacteriological dish undissociated. In this experiment fourteen out of
fourteen 2/2-cell pairs remained contiguous; and in a typical experiment five out
244
E. LEHTONEN
Length reduction
I
1
Width reduction
' —
U — Plateau --*•!
contact diameter
Fig. 9. Cellular dimensions changing during cleavage-stage cell cycles.
five embryos developed to the blastocyst stage. (2) 1-cell-stage embryos were of
pronased for 5-5-7 min and cultured inside the thinned zona pellucida. About
10-12 h after division the zona pellucida were mechanically removed. In subculture on bacteriological dishes eight out of eight 2/2-cell pairs remained
contiguous; and in a typical experiment three out of four embryos developed to
the blastocyst stage. In neither of these experiments was the capacity to develop
to the blastocyst stage dependent on the length of the pronase treatment.
Throughout the present study no difference could be observed in the behaviour
of cells which were either a long (8-9 h) or short (30 min) time in culture before
division. Therefore the recovery period after dissociation is not critical for the
development of these dimensions and any possible effects of the dissociation
procedures must affect all the cells to a similar extent.
DISCUSSION
The preimplantation mouse cell lineage has regularities (Wilson, Bolton &
Cuttler, 1972; Graham & Deussen, 1978; Kelly et ah 1978) that apparently
depend on continuous interaction between the cells of the embryo (Graham &
Lehtonen, 1979). The mechanism of these interactions is unknown, but it seems
to be connected with basic cellular relationships such as cell contacts and timing
of cell divisions (Graham & Lehtonen, 1979; Kelly et al. 1978). The present
study gives quantitative data on these possible morphogenetic factors.
1. Cellular dimensions during the cleavage stages
The present results on cell dimensions and contact diameters are summarized
in Fig. 9 and Table 7. The main observations of this paper are from isolated
zona-free cleavage-stage cells developing in vitro. There are three main reasons
Dimensional changes in cleavage-stage cells
245
Table 7. Summary of dimensional changes during
cleavage-stage cell cycles (from Tables 2-5)
Cell cycle measured
A.
Dimension (jim)
Diameter
Elongation
Width reduction
Plateau contact
diameter
Length reduction
1/1-2/2
(1st)
1/2-2/4
(2nd)
1/4-2/8
(3rd)
1/8-2/16
(4th)
34-5
68-8
530
43-2
37-5(55%) 33-4(61%) 26-0(60%) 18-1 (52%)
15-8(23%) 12-5(23%) 9-3(22%) 8-1(23%)
—
14-5(35%) 35-6(95%) 27-2(98%)
—
4-8(5%)
21-1 (30%) 12-5(24%)
1/16-2/32
(5th)
26-9
—
—
—
—
for the use of isolated cells in this study. First, in the zona-contained eggs the
extent of active intercellular contact is difficult to distinguish from passive
membrane apposition possibly resulting from constraints imposed by the zona
pellucida. Second, there are variations in the cell patterns of the cleavage-stage
embryos (Lewis & Wright, 1935; Graham & Deussen, 1978). Therefore the
possible effects of other cells on a particular cell-to-cell contact must be
excluded. Third, the measurement of cellular relationships from undissociated
embryos is practically impossible after compaction at the 8-cell stage.
Cell size. The cleavage-stage cell dimensions measured here for isolated
blastomeres developing in culture are similar to the values previously measured
for blastomeres in undissociated embryos which were in culture medium for
short periods (Lewis & Wright, 1935; Abramczuk & Sawicki, 1974). It is
therefore improbable that these dimensions were greatly influenced by the
culture conditions used in this study. It is well known that the cell size of daughter
cells may vary considerably in cleavage-stage embryos. In the present material
the volume differences between newly divided twin cells ranged up to around
10 % at each cleavage division and up to 21-9 % at the first division (Table 2).
Thus the diversity in cell size observed in early embryos may arise at any
cleavage division. This contrasts with the situation in many vertebrate and
invertebrate embryos where unequal cell divisions characteristically occur at
particular cleavage stages (e.g. Biggelaar & Guerrier, 1979; Hara, 1977).
Cell contact. There are no previous data on cell cycle changes in cellular
dimensions or contacts during cleavage stages. In zona, the cells of the cleavagestage embryo always appear to be in contact with each other. In vitro, however,
the contiguity of the cleavage-stage cells depends on culture environment. For
the main observations in this paper, culture conditions were used in which it
appeared that there was little interaction between the cells and the substratum
(see Results, section 6). The comparison of the behaviour of daughter cells on
different substrates suggests that interaction with substrate may promote cell
adhesion, and further that the zona pellucida may provide such a substratum
246
E. LEHTONEN
in vivo. Thus 2/2-cell structure derived from the division of a zona-free 1-cellstage cell falls apart or remains contiguous depending on the culture substratum. On the other hand, division inside an intact or a pronase-thinned zona
results in 2/2-cell pairs that always remain contiguous even after the removal of
the zona. These observations suggest that the initiation of contact formation is
an active cellular function for which the zona pellucida provides a necessary
substratum. After the 2-cell stage, it appeared that cell behaviour was qualitatively similar on a variety of substrata, and the zona pellucida may not play a
role in blastomere adhesion after this stage.
The mid body may also have a role in the development of cell contacts as it
was always seen between the noncontiguous twin cells. Furthermore, the
development of contiguous cell pairs was frequently seen to involve gradual
tilting of the two cells round the mid body to reach a membrane contact. This
suggests that the mid body may have some influence on the orientation and
positioning of the cell contacts. Presumably, after the 2-cell stage also the other
cells of the embryo may influence the development of a particular cell contact.
The data we have obtained describes the properties of individual cells and cell
pairs which, when acting in concert with the other cells, may account for
morphogenesis of the blastocyst. Although this analysis is not yet complete, it is
already possible to suggest that one of the most obvious features of mouse
development, 'compaction', is a progressive event which is cell cycle dependent.
Towards the end of the 8-cell stage the blastomeres are known to undergo the
process of compaction whereby their intercellular contacts are maximized. At
earlier stages, however, the cells appear to lie in a looser arrangement without
clear cell cycle changes in the contact areas. In the present study the 2/8- and
2/16-cell pairs progressively increased their intercellular contacts up to around
8 and 2-5 h after division, respectively, whereas there was little change in the
2/2- and 2/4-cell pairs during the cell cycle. Thus the ability to compact characterizes the developmental stage of the cell and is not dependent on zona pellucida
or on the cell number.
2. Surface microvilli during cell cycle
The mechanism by which the cleavage-stage cells regulate their intercellular
contacts is unknown. The cells have microfilament containing surface microvilli,
and it has been suggested that the microvilli may approximate the cell membranes
in compaction (Ducibella, Ukena, Karnovsky & Anderson, 1977). The present
study confirms the impression (Graham & Lehtonen, 1979) that the density of
the microvilli in 8-cell-stage cells is reduced for a short period of time after
cytokinesis, but it then again reaches a steady level in about 60-90 min. In a
rather similar way, somatic cells early in the cell cycle have less microvilli than
cells late in the cycle (Porter, Prescott & Frye, 1973; Knutton, Sumner &
Pasternak, 1975; Knutton, 1976). This time period of increasing microvillar
density coincides with the beginning of the contact increase in 2/8-cell pairs.
Dimensional
changes in cleavage-stage cells
247
However, the amount of microvilli alone cannot account for the subsequent
development of cell contacts as the microvillar densities of the late 4-cell- and
late 8-cell-stage cells are similar but the contact diameter in 2/4-cell pairs
(14-5 iim) is 35% and in 2/8-cell pairs (35-6 fim) is 9 5 % of the cell diameter.
The failure to discover a correlation between microvillar density and extent of
cell contact may be due to the fact that the density of these organelles was
measured on that part of the cell surface away from the actual contact point.
3. Mouse cell lineage and the present observations
In the mouse cell lineage, the descendants of the first dividing cell of the 2-cellstage embryo tend to divide ahead of the descendants of the last dividing cell;
some of these fast dividers also tend to segregate into the interior of the embryo
and eventually into the ICM (Kelly et ah 1978; Graham & Deussen, 1978;
Graham & Lehtonen, 1979). The present quantitative data show that during the
cell cycle a twin-cell pair increases its inherited intercellular contact. Additional
cell contacts result from cytokinesis associated and cell cycle changes in cellular
dimensions (see for instance fig. 2 in Graham & Lehtonen, 1979); and our
preliminary observations suggest that these contacts behave in a similar way as
the inherited ones. This apparently implies that during the early phases of each
cleavage stage the descendants of the first dividers tug the other cells around
them and therefore tend to segregate into an internal position. However, if the
data on division asynchrony (this study; Kelly et al. 1978) are compared to the
timing of changes in cell contacts, it appears that even the last dividers have the
possibility to maximize their contacts and to reverse the morphogenetic effects
of the earlier contacts before next cell divisions. The implication is that changes
in stage-dependent and cell-cycle-dependent cell contacts observed in isolated
cell pairs cannot by themselves account for the internalization of fast dividing
cells.
There are at least two possible mechanisms for preventing this deshaping
effect of the last dividing cells and each of them would depend on interactions
between all the cells of the embryo. First, the number of intercellular contacts
tends to increase during and soon after cytokinesis. As a consequence, cell
contacts occupy a rapidly growing proportion of the surface of the fast dividers'
descendants. The slow dividers therefore tend to segregate into an outside
position, and as they divide their descendants have a reduced probability to
establish new contacts, especially as the earlier cell contacts tend to direct
cytokinesis towards the exterior of the embryonic structure (Graham & Lehtonen, 1979). The relatively low deformability of the spherical cells results in
simple physical limits for the area of a cell contact and therefore the slow
dividers' descendants with fewer contacts are likely to stay in an outside position
and to form the outer surface of the embryo. Second, after reaching the maximum
contact area, the cell contacts may change so that they resist the reversing forces
that result from the development of new membrane contacts. Such a stabilizing
248
E. LEHTONEN
factor could be related to the formation of gap and tight junctions (Ducibella &
Anderson, 1975; Ducibella, Albertini, Anderson & Biggers, 1975; Magnuson,
Demsey & Stackpole, 1977) or to the decrease in Con A-induced agglutinability (Rector & Granholm, 1978; Rowinski, Solter & Koprowski, 1976) in
8-cell-stage embryos.
The present and our previous results suggest that the mouse cell lineage is a
result of continuous cell interactions starting from the 2-cell stage. This epigenetic control of the lineage can be partly understood in terms of simple spatial
and temporal relationships of the cells and especially their intercellular contacts
during the cleavage stages. The molecular mechanism by which the development
of cell contacts is controlled is unknown, but changes in cell surface microvilli
(this study; Ducibella et al. 1977; Graham & Lehtonen, 1979) and the organization of cytoskeletal structures (Lehtonen & Badley, 1980) may be involved in
this process. Furthermore, as suggested by Lewis & Wright (1935), the internal
pressure of the cells and the surface tension or deformability of the cell cortex
may regulate the extent of intercellular contacts.
I would like to thank M.R.C. and the Finnish Culture Foundation for funding this work,
E. M.B.O. for providing a Fellowship for me, and Chris Graham for discussion. I am
grateful to Ms Margaret K. Arnold and Dr K. A. Harrap of the Natural Environment Unit
of Invertebrate Virology, Oxford, for the use of the Jeol 100 CX Temscan electron microscope.
REFERENCES
J. & SAWICKI, W. (1974). Variation in dry mass and volume of nonfertilized
oocytes and blastomeres of 1-, 2- and 4-celled mouse embryos. / . exp. Zool. 188, 25-34.
BIGGELAAR VAN DEN, J. A. M. & GUERRIER, P. (1979). Dorsoventral polarity and mesentoblast determination as concomitant results of cellular interactions in the mollusc Patella
vulgata. Devi Biol. 68, 462-471.
DUCIBELLA, T. & ANDERSON, E. (1975). Cell shape and membrane changes in the eight-cell
mouse embryo: prerequisites for morphogenesis of the blastocyst. Devi Biol. 47, 45-58.
DUCIBELLA, T., ALBERTINI, D. F., ANDERSON, E. & BIGGERS, J. D. (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.
GRAHAM, C. F. & DEUSSEN, Z. A. (1978). Features of cell lineage in preimplantation mouse
development. / . Embryol. exp. Morph. 48, 53-72.
GRAHAM, C. F. & LEHTONEN, E. (1979). Formation and consequences of cell patterns in
preimplantation mouse development. / . Embryol. exp. Morph. 49, 277-294.
HARA, K. (1977). The cleavage pattern of the axolotl egg studied by cinematography and
cell counting. Wilhelm Roux's Arch. Devi Biol. 181, 73-87.
HILLMAN, N., SHERMAN, M. I. & GRAHAM, C. (1972). The effect of spatial arrangement on cell
determination during mouse development. / . Embryol. exp. Morph. 28, 263-278.
KELLY, S. J. (1977). Studies on the developmental potential of 4- and 8-cell stage mouse
blastomeres. / . exp. Zool. 200, 365-376.
KELLY, S. J., MULNARD, J. G. & GRAHAM, C. F. (1978). Cell division and cell allocation in
early mouse development. / . Embryol. exp. Morph. 48, 37-51.
KNUTTON, S. (1976). Structural changes in the plasma membrane of synchronized P815Y
mastocytoma cells. Expl Cell Res. 102, 109-116.
ABRAMCZUK,
Dimensional changes in cleavage-stage cells
249
S., SUMNER, M. C. B. & PASTERNAK, C. A. (1975). Role of microvilli in surface
changes of synchronized P815Y mastocytoma cells. /. Cell Biol. 66, 568-576.
KUHL, W. & FRIEDRICH-FREKSA, H. (1936). Richtungskorperbildung und Furchung des Eies
sowie das Verhalten des Trophoblasten der weissen Maus. (Film.) Zool. Anz., Suppl. 9,
187-195.
LEHTONEN, E. & BADLEY, R. A. (1980). Localization of cytoskeletal proteins in preimplantation mouse embryos. /. Embryol. exp. Morph. 55, 211-225.
LEWIS, W. H. & WRIGHT, E. S. (1935). On the early development of the mouse egg. Contr.
Embryol. Carnegie Inst. 148, 115-143.
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.
MINTZ, B. (1967). Mammalian embryo culture. In Methods in Developmental Biology (ed.
F. H. Wilt & N. K. Wessels), pp. 379-400. New York: Crowell.
OGAWA, S., SATOH, K. & HASHIMOTO, H. (1971). In vitro culture of rabbit ova from the single
cell to the blastocyst stage. Nature 233, 422-424.
PORTER, K., PRESCOTT, D. & FRYE, J. (1973). Changes in surface morphology of Chinese
hamster ovary cells during the cell cycle. / . Cell Biol. 57, 815-836.
RECTOR, J. T. & GRANHOLM, N. H. (1978). Differential concanavalin A-induced agglutination of eight-cell preimplantation mouse embryos before and after compaction. /. exp.
Zool. 203, 497-502.
ROWINSKI, J., SOLTER, D. & KOPROWSKI, H. (1976). Change of concanavalin A induced
agglutinability during preimplantation mouse development. Expl Cell Res. 100, 404-408.
TARKOWSKI, A. K. & WROBLEWSKA, J. (1967). Development of blastomeres of mouse eggs
isolated at the 4- and 8-cell stage. / . Embryol. exp. Morph. 18, 155-180.
WHITTEN, W. K. (1971). Nutrient requirements for culture of preimplantation embryos
in vitro (ed. G. Raspe). Advances in the Biosciences 6, 129-140.
WILSON, I. B., BOLTON, E. & CUTTLER, R. H. (1972). Preimplantation differentiation in the
mouse egg as revealed by microinjection of vital markers. J. Embryol. exp. Morph. 27,
467-479.
KNUTTON,
{Received 11 December 1979, revised 10 March 1980)