Embed Size (px)
Citation preview
/ . Embryol. exp. Morph. Vol. 58, pp. 231-249,1980 2 3 1Printed in Great Britain © Company of Biologists Limited 1980
Changes in cell dimensions andintercellular contacts during cleavage-stage cell
cycles in mouse embryonic cells
By E. LEHTONEN1
From the Zoology Department, University of Oxford
SUMMARYThe 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 wererecorded. The division of an isolated cell results in the formation of a twin-cell pair. Thedivisions of these two cells were always asynchronous. In each division the volume of adaughter cell was approximately half of that of the parental cell but its apparent surface areawas 59-65 % of that of the parental cell. Consequently, the ratio of apparent surface area tovolume increased in each division by 25-30%. The most noticeable changes were observedin the relationship between the two daughter cells of each division. After cytokinesis, theintercellular contact area gradually increased during the following cell cycle in the 2/8- and2/16-cell pairs, whereas it hardly changed in the 2/2- and 2/4-cell pairs. The comparison ofthe behaviour of the daughter cells on different substrates suggested that the zona pellucidaand the mid body might have a role in the contact development at the early stages. Scanningelectron microscopy was used for studying changes in the density of cell surface microvilliin an attempt to explain how the cells regulate their intercellular contacts.
INTRODUCTION
The major factor influencing cell fate during preimplantation mouse develop-ment is the relative position of the cell in the embryonic structure (Tarkowski &Wroblewska, 1967; Hillman, Sherman & Graham, 1972; Kelly, 1977). Themovements of the cells to their different relative positions principally occurduring cell division (Graham & Deussen, 1978; Graham & Lehtonen, 1979) andthe orientation of these divisions apparently depends on continuous cell inter-actions during cleavage-stage development (Graham & Lehtonen, 1979).Furthermore, the cleavage divisions are asynchronous, and cell allocation to theinner cell mass (ICM) and to the trophectoderm of the blastocyst is related tothe division order during the cleavage stages (Kelly, Mulnard & Graham, 1978).
The present study attempts to analyse some of the cell interactions involvedin the formation of the cleavage-stage structure. The movement of the cleavage-stage cells at any moment during development depends on their contacts with
1 Author's present address: Department of Pathology, University of Helsinki, Haart-maninkatu 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 isolatedcleavage-stage cells in vitro. During cell division the spherical cell body narrowsand elongates. This change of form frequently results in the formation of newintercellular contacts as cells extend in division and it probably alters relativecell position because dividing cells tend to maintain their previous intercellularcontacts during this shape change and move in association with the mitotic cell(Graham & Deussen, 1978; Graham & Lehtonen, 1979). After division, thedaughter cells flatten against each other and slight relative cell movementsappear to be the consequence of this process. In order to understand the rangeand efficiency of these possible morphogenetic factors I have quantitated thechanges in cell dimensions and intercellular contacts in cleavage-stage cellsdeveloping in vitro.
MATERIALS AND METHODS
The supply of embryos
The embryos were from natural matings and were (C57BL6 x CBA)F2. Theywere dissected into pre-warmed and pre-equilibrated Whitten's medium (1971).The embryos were removed from the reproductive tract between 2 and 16 hbefore the division stage which was to be studied.
Dissociation of embryos
The zona pellucida was partially dissolved by incubating each embryo inpronase (Calbiochem. Co., U.K., technique of Mintz, 1967) at +37 °C for5-5-6 min and then it was pipetted off 1-cell and 2-cell embryos after 30 minincubation in the medium.
For the dissociation of 4-cell and 8-cell stages, each embryo was culturedinside the pronase-thinned zona for 15 min in Whitten's medium in which thecalcium concentration had been lowered to 0-02 mM. The zona was then pipettedoff the embryo and thereafter the cells were blown apart by pipetting through aflame-polished micro-pipette with a mouthpiece. The cells were replaced innormal Whitten's medium immediately after dissociation. The low calciumtreatment was not needed for 1-cell- and 2-cell-stage embryos.
Culture
The isolated blastomeres were cultured in Whitten's medium in microdropsunder paraffin oil (selected for absence of toxicity to cultured embryos, BootsPure Drug Co., U.K.) in tissue culture flasks (Falcon Plastics, Oxnard, Cali-fornia) pregassed with a gas mixture of 5 % CO2, 5 % O2 and 90 % N2. Themicrodrops 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 con-tinuously in a +37 °C room with a video recording system fitted to a Wildinverted microscope. The recording system consisted of a WV-1350 camera(Matsushita Electric Trading Company, Osaka, Japan), a NV-8030 time-lapsevideo-tape recorder (Matsushita Company) and a VM-172 AV video monitor(Hitachi Denshi Limited, Lodge Road, London, U.K.). The records were madewith 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-thickfilter containing tissue-culture medium; under these conditions twenty 2-cellembryos each formed a morphologically normal blastocyst on the fourth day ofdevelopment, and it is therefore unlikely that the illumination disturbed thecellular events.
The cellular dimensions were measured continuously from the recording, andthe changes in dimensions were traced on transparent film. The cells were rarelycompletely spherical. Therefore the largest and shortest diameters were measuredand the mean diameter was calculated. The development of an isolated cell wasfollowed through at least two consecutive divisions, and data was excluded ifthe 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 anembryo were cultured simultaneously in microdrops on a bacteriological gradedish. 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 cellswere viable only cells with dividing sister cells were included in the data. Thecells were rinsed quickly in 0-1 M cacodylate buffer (pH 7-6) with 2 mM-CaCl2at +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 wereattached 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 substrateand had similar contact to the glass. Thereafter the cells were processed forSEM as described elsewhere (Graham & Lehtonen, 1979).
For quantitative data the surface microvilli were counted over an area ofabout 30 jiim2 on the uppermost part of each cell directly opposite the area incontact with the glass surface. In this area, the cell surface appears to be perpen-dicular to the electron beam and minimal distortion of the surface area of thesecurved objects should occur in this region. The microvilli were counted frommicrographs with final magnifications of about x 15000. Micrographs of agrating replica (2160 lines/mm) were used for calibration.
234 E. LEHTONEN
Control experiments
The following experiments were conducted to discover if the culture conditionsdescribed 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 themedium on an agarose surface (1 % agarose in water) which had previously beenequilibrated with the medium for 30min, or in microdrops on the plasticsurface of bacteriological grade dishes (Sterilin Ltd., Richmond, Surrey, U.K.).In some experiments the plastic surface was scratched with a glass needle beforeculture.
To test the effects of the dissociation methods on cleavage behaviour 1-cell-and 2-cell-stage embryos were subjected to 5-5-7 min pronase treatment andthen cultured undissociated either inside or outside the thinned zona pellucida.
In these preliminary experiments the observations were intermittent and madewith the video recording system or by photography. The data were excludedfrom the quantitative results.
RESULTS
The results are presented in six sections. (1) There is data on cell cycle dura-tions and division asynchrony of isolated cleavage-stage cells in culture. (2) Thephysical dimensions of isolated cleavage-stage cells in culture are given. (3) and(4) The changes in cellular dimensions during and after cell division in cultureare described. (5) There is data on cell cycle differences in the density of surfacemicrovilli of the cleavage-stage cells. (6) There are observations on the effect ofculture substratum on cleavage behaviour.
The quantitative study of cellular dimensions is based on continuous measure-ments made on isolated cleavage-stage cells which divided twice on a sili-conized glass surface producing four descendants. Twenty-five video recordswere made in which the behaviour of cells was clearly recorded. In three of theseexperiments the cell failed to produce four viable descendants and these recordswere discharged.
1. Cell cycle durations and division asynchrony
The cell cycle durations of cleavage-stage cells were calculated by measuringthe time from the division of an isolated cell to the divisions of the cell pair thusformed. 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 considerablywithin each stage studied. The differences between the shortest and the longest2nd, 3rd, 4th and 5th cell cycle were, respectively, 13, 36, 16, and 16% of theshortest cycle (Table 1).
The divisions of the two daughter cells of the isolated cell were alwaysasynchronous. The mean division interval (93 min) between the faster andslower dividing cells of the 2/2-cell pairs represents 6-4 % of the mean cell cycle
Tab
le 1
. C
ell
cycl
e du
rati
ons
and
asyn
chro
ny b
etw
een
divi
sion
s of
tw
in c
ells
Cel
l cy
cle
mea
sure
dSt
ep a
t w
hich
cult
ure
initi
ated
No.
of
obse
rvat
ions
Cyc
le d
urat
ion
of th
efa
ster
cel
l (m
in)
Cyc
le d
urat
ion
of th
esl
ower
cel
l (m
in)
Inte
rval
fro
m f
irst
to
seco
nd d
ivis
ion
(min
)
Mea
nR
ange
Mea
nR
ange
Mea
nR
ange
I a' r 1 x
2/2-
4/4
(2nd
)2/
4-4/
8 (3
rd)
2/8-
4/16
(4t
h)2/
16-4
/32
(5th
)
1/1-
cell
1/2-
cell
1/4-
cell
1/8-
cell
1405 704
786
707
1372-1432
618-801
737-837
673-750
1498 771
810
735
1431-1556
679-843
. 763-853
709-780
93 68 24 29
59-1
4627
-148
16-3
410
-36
The
cel
l cyc
le d
urat
ions
wer
e de
term
ined
by
obse
rvin
g an
isol
ated
zon
a-fr
ee c
ell t
o di
vide
twic
e to
pro
duce
four
des
cend
ants
. The
tim
e w
as c
ount
edfr
om t
he b
egin
ning
of
cyto
kine
sis.
Int
erva
l fr
om f
irst
to
seco
nd d
ivis
ion
refe
rs t
o th
e as
ynch
rony
bet
wee
n th
e fa
ster
and
slo
wer
div
idin
g ce
ll in
each
ind
ivid
ual
expe
rim
ent.
ZJD
Table
Stage
1-cell (a)
2-cell (b)(a)
4-cell (b)(a)
8-cell (b)(a)
16-cell (b)
2. The
No. ofobser-vations
5
106
125
106
12
E. LEHTONEN
diameters and volumes of cleavage-stage
Cell
Mean
68-8
53054-4
43-243 0
34-534-7
26-9
1 diameterOm)
Range i
67-9-69-5
49-6-54-753-3-55-7
40-8-45-042-2-43-5
33-6-36133-6-35-8
25-8-28-2
Surfacearea
(x 103 /tm2)
14-87
8-829-30
5-865-80
3-753-77
2-28
Volume(xlO3/tm
170-3
77-984-3
All41-6
21-621-8
10-2
cells in culture
Volume difference
A
3) Mean
—
116
51
4-3
6 1
Range
—
50-21-9
1-6-8-2
0-11-8
1-2-10-5
Cell surface areas and volumes are calculated from the mean diameters of the cells. Volumedifference refers to unequal cell division in individual experiments. This difference betweenthe 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-, 2/8-and 2/16-cell pairs represents 9-2, 3-0 and 4-0%, respectively, of the mean cellcycle time (Table 1). This data on asynchrony supplements the less accuratedata previously obtained by intermittent observations (Kelly, Mulnard &Graham, 1978), and confirms the view that asynchronous divisions of daughtercells 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 immedi-ately before an isolated cell divided in culture and immediately after the division.These measurements were used for calculating cell volumes and apparentsurface areas (Table 2). At each cleavage stage, the post-division observations('£'-values in Table 2) are exclusively on the daughter cells formed by cellsisolated at the previous stage ('a'-values in Table 2). Comparison of these pre-division and post-division volumes suggests that a slight volume loss takes placenear the time of the first division of the zona-free zygote: the combined volumeof the two daughter cells is about 91 % of that of the late 1-cell stage. In thefollowing divisions the combined volume of the daughter cells is approximatelyequal to the mother cell volume. Consequently the total cellular volume of the8- 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 differencesof 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
1/1- to 2/2-cell1/2- to 2/4-cell1/4- to 2/8-cell1/8- to 2/16-cell
No. ofobser-
vations
5656
Elongation (/Am)
Mean
37-533-4260181
Range
34-7-38-929-8-36-425-0-27-415-8-19-3
0/
/o
55616052
Width
Mean
15-812-59-381
reduction
Range
150-17111-4-14-080-11-57-0-8-6
Om)
/o
23232223
Elongation is the difference between the mean cell diameter before division and themaximum length of the daughter cell pair at any moment during 150 min after the beginningof division. Width reduction is the difference between the mean cell diameter before divisionand the minimum mean width of the daughter cell pair at any moment during 150 min afterthe beginning of division. The mean changes are expressed also as % of the mean diameter ofthe parental cell. The material is the same as that in Table 2.
During the cleavage-stage development, the ratio of apparent cell surfacearea to volume increases (the apparent surface area is the surface area calculatedfrom these light microscope observations and does not take into account micro-villar projections from the cell surface). The average apparent surface area of adaughter 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 embryois about 2-4 times that of the late 1-cell stage. Similarly, the ratio of apparentsurface area to volume increases in each division by between 25 and 30%; andthis ratio for 16-cell-stage cells (0-224) is about 2-6 times that of the late 1-cellstage (0-087).
3. Changes in cellular dimensions during cell divisions in culture
The cellular dimensions of isolated 1- to 8-cell-stage cells were measuredcontinuously before and during cell division, and during subsequent develop-ment 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 notdevelop an intercellular membrane contact; in each experiment the mid bodywas seen separating the cells during the 2-cell stage. One cell pair out of sevenproduced by the division of an isolated 2-cell-stage cell did not develop anintercellular membrane contact. In this case, the mid body was seen separatingthe daughter cells during the following cell cycle and this record was not includedin the cell dimension data. In all other cases, the daughter cells formed a con-tiguous pair with contact between the membranes of the daughter cells.
In cell division the spherical cell body elongates and narrows to produce adaughter cell pair. The relative magnitude of these dimension changes wassimilar at all stages studied. On average, the maximum division associatedelongation of the cell body was 52-61 % of the mean diameter of the parental
16-2
238 E. LEHTONEN
90
80
70
60
50
40
30
20
10
703<D
"5e
•3
h or
60
50
40
30
20
10
50
« 40
•2 30
I 2 0
"So
J 10
-1 • • • • • • a .
• * • - • • • • . • » % .
Q i i i i i
o I 3 4 7 8 9 10 11 12 13 14
- 2
. . I
n, 8 J I J I I I I 1 I 1
0 1 8 9 10 11 12 13 14
8 Ri i i i n i i i i i i i i i
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14Time after cytokinesis (h)
FIGURES 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 diagramof each figure represents the maximum length of the undivided parental cell or thatof the daughter cell pair after division. Each of the lower diagrams shows thediameter of the cleavage furrow or that of the contact area between the daughtercells in the same experiments. The beginning of cleavage furrow formation waschosen for time point 0. The diagram pairs are from individual experiments; thedata from these diagrams are included in the Tables.
Dimensional changes in cleavage-stage cells 239
Table 4. The plateau diameters of contact areas betweentwo sister cells
Stage
2/4-cell2/8-cell2/16-cell
No. ofobser-
vations
656
Contact
Mean
14-535-627-2
diameter (jim)
Range
12-9-16132-5-38-926-8-27-6
0/
/o
Mean
359598
of width
Range
31-3890-9895-100
% of cellsurface
2-826-525-5
For the plateau diameter of the contact area, the peak value of the apparent contactplateau was averaged with values 60 min on either side of that point. In two experiments the2/16-cell structure turned so that the cells were lying on top of each other. This appeared tohappen soon after the plateau contact was reached. In these occasions the plateau contactdiameter was calculated by averaging the last measurable contact value with the value 30 minbefore that. The plateau contact diameters are expressed also as % of the simultaneousmean widths of the cell pairs. The percentage of the contact area compared to the total cellsurface area ('b '-values from Table 2) is calculated assuming that the contact area is sphericaland that the total cell surface area does not change during the cell cycle. The material is thesame as that in Table 2.
cell (Figs 1-3, 9; Table 3). Correspondingly, the division-associated widthreduction of the cell body was 22-23 % of the mean diameter of the parentalcell (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 thefollowing cell cycle. Immediately after cytokinesis the contact area is relativelysmall. Subsequently it increases, reaches an apparent plateau and then decreasesto some extent just before the next cell divisions (Figs 1-3). The maximumcontact 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 diametersrepresent 2-8, 26-5 and 25-5 %, respectively, of the corresponding total cellsurface areas (lb '-values in Table 2).
The changes in the intercellular contact area are closely connected withchanges in the total length of the cell pair: increase in the contact diameter isaccompanied 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-cellpairs were, respectively, 5-4, 30-5 and 24-0% of the mean peak lengths of thecell pairs (Table 5).
The rate of changes in the above dimensions is not constant, and the kineticsof the changes appears to differ from stage to stage (Figs 1-3). In order toquantify this impression, the time which elapsed from the beginning of cyto-kinesis 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 pairsin culture
No. of Length reduction (/tm) % of peak lengthobser- , A N , *
Stage vations Mean Range Mean Range
2/4-cell 6 4-8 1-6- 6-8 5-4 1-9- 7-92/8-cell 5 21-1 18-9-22-5 30-5 27-1-33-62/16-cell 6 12-5 11-4-14-3 240 21-8-26-9
The peak length of the newly divided sister cell pair was compared with the bottom lengthof the pair. The length reduction is expressed also as % of the peak length. The material isthe same as that in Table 2.
determined. To calculate the onset of the latter period, the rate of increase of thecontact diameter was studied: the midpoint of the last 30 min period duringwhich the intercellular contact diameter still increased by at least 5 % of thetotal increase was used. In the case of 2/4-cell pairs the dimension changes wereso small that reliable calculations were not possible although it is obvious thatcell contact diameter does not increase markedly at times later than 60 min afterdivision (Fig. 1). In the case of 2/8-cell pairs this time period ranged from 453to 525 min, and the mean was 482 min. In the case of 2/16-cell pairs this timeperiod ranged from 129 to 177 min, the mean being 155 min.
In addition to the cell cycle changes in dimensions, the cells regularly showedother motile features. During the first 60-90 min after division cell membraneblebbing 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 thecells on the glass surface. Moreover, the cells showed contraction and expan-sion movements of varying amplitudes and frequencies. These rhythmicalpumping movements were most pronounced in the 8- and 16-cell-stage cellswhere their amplitude was often about 10% of one dimension. In the 8-cell-stage cells the pumping movements usually started about 2 h after division andcontinued at a frequency of about 1 cycle of movements in 1 min through mostof the interphase. The result is based on two experiments, in each of which about10 cycles were followed. About 30-60 min before next cell division the move-ments slowed down and stopped about 10 min before division. The abovemovements were less striking in the 4-cell-stage cells and usually very mild ornegligible in 1- and 2-cell-stage cells. Comparable rhythmical contraction andexpansion have been reported in rabbit ova (Ogawa, Satoh & Hashimoto, 1971)and in mouse blastocyst (Kuhl & Friedrich-Freksa, 1936). After the second roundof cell divisions the above described cell movements resulted in the establishmentof 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) andabout 5 h (Figs 6-7) after the division of the isolated 4-cell-stage cell. The distributionof the microvilli is relatively even, although there seems to be some concentrationin the intercellular contact area and in the polar regions. The microvillar densityappears lower 20 min after cytokinesis than it does 5 h after cytokinesis. MB, midbody.
242 E. LEHTONEN
12
10
QO
6
4
2
4-cellstage
-
-
-
8-cellstage
•
' :
•
i i
* •
• {
i i i i
0 1 2 3 4 5 6 7
Time after cytokinesis (h)
Fig. 8. Microvillar densities of isolated late 4-cell-stage cells and of 8-cell-stagecells derived from the division of isolated 4-cell-stage cells in culture. The diagramis based on scanning electron micrographs.
5. Surface microvilli during cell cycle
In order to explain the cell behaviour described above, the surface charac-teristics of 4- to 8-cell-stage cells were studied with SEM. It was supposed thatthe differences in contact area between cells at different stages of developmentmight correlate with the number of microvilli on the blastomeres; these surfaceorganelles are known to make contact between adjoining cells.
In isolated late 4-cell-stage cells the distribution of microvilli was relativelyeven. During cytokinesis concentration of microvilli to the cleavage furrow areawas regularly observed, but already about 20 min after cytokinesis the distri-bution of the microvilli appeared quite even (Figs 4 and 5). Later during the8-cell-stage cell cycle some polar microvillus concentration was seen occasion-ally (Figs 6 and 7), but usually the distribution appeared relatively even. Themicrovillar densities were determined from an area about 1/4 of a cell circum-ference away from the intercellular contact point both because this is the areawhich would probably touch other cells in the intact embryo and because it wasimpossible to obtain accurate counts at the junction between the two daughtercells. At the late 4-cell stage and the late 8-cell stage the density of surface micro-villi is similar (Fig. 8). The microvillar density drops for 60 to 90 min after 4-to 8-cell division, but it then again reaches the level scored for late 4-cell-stagecells. The microvillar densities which were measured within 1 h after cytokinesisare 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 thoseof the late 2/8-cell pairs.
Dimensional changes in cleavage-stage cells 243
Table 6. The effect of culture substratum-on cleavage behaviour
Contiguous cell pairs/number ofexperiments
Division observed
1/1-to 1/8-to2/2-cell 2/16-cell
Siliconized glass 0/8 11/11Bacteriological dish 5/11 8/8Scratched bacteriological dish 8/8 8/8Agarose 10/11 12/12
Isolated 1-cell-stage and 8-cell-stage cells were cultured on different substrata. The possiblecontiguity 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 behaviourfour different culture substrata were used, viz. siliconized glass, bacteriologicalgrade 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 produceda contiguous cell pair with an intercellular membrane contact on siliconizedglass surface (none out of eight experiments); in each experiment the mid bodywas seen separating the cells during the 2-cell stage. On bacteriological gradeplastic surface five experiments out of eleven resulted in a contiguous cell pairwith an intercellular membrane contact; on scratched plastic surface or onagarose 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. Theresulting four 4-cell-stage cells always established at least three intercellularmembrane contacts even if the previous 2/2-cell pair appeared to be separatedby the mid body. During further cultivation most of these embryos formedblastocysts. The ability to develop to this stage was not dependent on culturesubstratum or membrane contact at the 2/2-cell stage. The division of an isolated8-cell-stage cell always produced a contiguous cell pair regardless of the culturesubstratum (Table 6). During further culture these experiments regularlyresulted in the formation of blastocyst vesicles. Clearly, the substratum caninfluence the division behaviour of blastomeres from early stages of development(see Discussion).
The following experiments were performed in order to test the effects of thedissociation methods on cleavage behaviour. (1) 2-cell-stage embryos werepronased for 5-5-7 min, the zona pellucida were removed, and the embryos werecultured on bacteriological dish undissociated. In this experiment fourteen out offourteen 2/2-cell pairs remained contiguous; and in a typical experiment five out
244 E. LEHTONEN
Length reduction
1 Width reduction ' —
IU— 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 ofpronased for 5-5-7 min and cultured inside the thinned zona pellucida. About10-12 h after division the zona pellucida were mechanically removed. In sub-culture on bacteriological dishes eight out of eight 2/2-cell pairs remainedcontiguous; and in a typical experiment three out of four embryos developed tothe blastocyst stage. In neither of these experiments was the capacity to developto the blastocyst stage dependent on the length of the pronase treatment.
Throughout the present study no difference could be observed in the behaviourof cells which were either a long (8-9 h) or short (30 min) time in culture beforedivision. Therefore the recovery period after dissociation is not critical for thedevelopment of these dimensions and any possible effects of the dissociationprocedures 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 apparentlydepend on continuous interaction between the cells of the embryo (Graham &Lehtonen, 1979). The mechanism of these interactions is unknown, but it seemsto be connected with basic cellular relationships such as cell contacts and timingof cell divisions (Graham & Lehtonen, 1979; Kelly et al. 1978). The presentstudy 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 summarizedin Fig. 9 and Table 7. The main observations of this paper are from isolatedzona-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 duringcleavage-stage cell cycles (from Tables 2-5)
Dimension (jim)
DiameterElongationWidth reductionPlateau contactdiameter
Length reduction
1/1-2/2(1st)
68-837-5(55%)15-8(23%)
—
—
Cell
1/2-2/4(2nd)
53033-4(61%)12-5(23%)14-5(35%)
4-8(5%)
cycle measuredA.
1/4-2/8(3rd)
43-226-0(60%)9-3(22%)
35-6(95%)
21-1 (30%)
1/8-2/16(4th)
34-518-1 (52%)8-1(23%)
27-2(98%)
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 theextent of active intercellular contact is difficult to distinguish from passivemembrane apposition possibly resulting from constraints imposed by the zonapellucida. Second, there are variations in the cell patterns of the cleavage-stageembryos (Lewis & Wright, 1935; Graham & Deussen, 1978). Therefore thepossible effects of other cells on a particular cell-to-cell contact must beexcluded. Third, the measurement of cellular relationships from undissociatedembryos is practically impossible after compaction at the 8-cell stage.
Cell size. The cleavage-stage cell dimensions measured here for isolatedblastomeres developing in culture are similar to the values previously measuredfor blastomeres in undissociated embryos which were in culture medium forshort periods (Lewis & Wright, 1935; Abramczuk & Sawicki, 1974). It istherefore improbable that these dimensions were greatly influenced by theculture conditions used in this study. It is well known that the cell size of daughtercells may vary considerably in cleavage-stage embryos. In the present materialthe volume differences between newly divided twin cells ranged up to around10 % 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 anycleavage division. This contrasts with the situation in many vertebrate andinvertebrate embryos where unequal cell divisions characteristically occur atparticular cleavage stages (e.g. Biggelaar & Guerrier, 1979; Hara, 1977).
Cell contact. There are no previous data on cell cycle changes in cellulardimensions or contacts during cleavage stages. In zona, the cells of the cleavage-stage embryo always appear to be in contact with each other. In vitro, however,the contiguity of the cleavage-stage cells depends on culture environment. Forthe main observations in this paper, culture conditions were used in which itappeared that there was little interaction between the cells and the substratum(see Results, section 6). The comparison of the behaviour of daughter cells ondifferent substrates suggests that interaction with substrate may promote celladhesion, 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-cell-stage cell falls apart or remains contiguous depending on the culture sub-stratum. On the other hand, division inside an intact or a pronase-thinned zonaresults in 2/2-cell pairs that always remain contiguous even after the removal ofthe zona. These observations suggest that the initiation of contact formation isan active cellular function for which the zona pellucida provides a necessarysubstratum. After the 2-cell stage, it appeared that cell behaviour was quali-tatively similar on a variety of substrata, and the zona pellucida may not play arole in blastomere adhesion after this stage.
The mid body may also have a role in the development of cell contacts as itwas always seen between the noncontiguous twin cells. Furthermore, thedevelopment of contiguous cell pairs was frequently seen to involve gradualtilting of the two cells round the mid body to reach a membrane contact. Thissuggests that the mid body may have some influence on the orientation andpositioning of the cell contacts. Presumably, after the 2-cell stage also the othercells of the embryo may influence the development of a particular cell contact.
The data we have obtained describes the properties of individual cells and cellpairs which, when acting in concert with the other cells, may account formorphogenesis of the blastocyst. Although this analysis is not yet complete, it isalready possible to suggest that one of the most obvious features of mousedevelopment, 'compaction', is a progressive event which is cell cycle dependent.Towards the end of the 8-cell stage the blastomeres are known to undergo theprocess of compaction whereby their intercellular contacts are maximized. Atearlier stages, however, the cells appear to lie in a looser arrangement withoutclear cell cycle changes in the contact areas. In the present study the 2/8- and2/16-cell pairs progressively increased their intercellular contacts up to around8 and 2-5 h after division, respectively, whereas there was little change in the2/2- and 2/4-cell pairs during the cell cycle. Thus the ability to compact charac-terizes the developmental stage of the cell and is not dependent on zona pellucidaor on the cell number.
2. Surface microvilli during cell cycle
The mechanism by which the cleavage-stage cells regulate their intercellularcontacts is unknown. The cells have microfilament containing surface microvilli,and it has been suggested that the microvilli may approximate the cell membranesin compaction (Ducibella, Ukena, Karnovsky & Anderson, 1977). The presentstudy confirms the impression (Graham & Lehtonen, 1979) that the density ofthe microvilli in 8-cell-stage cells is reduced for a short period of time aftercytokinesis, but it then again reaches a steady level in about 60-90 min. In arather similar way, somatic cells early in the cell cycle have less microvilli thancells late in the cycle (Porter, Prescott & Frye, 1973; Knutton, Sumner &Pasternak, 1975; Knutton, 1976). This time period of increasing microvillardensity 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 subsequentdevelopment of cell contacts as the microvillar densities of the late 4-cell- andlate 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 95% of the cell diameter.The failure to discover a correlation between microvillar density and extent ofcell contact may be due to the fact that the density of these organelles wasmeasured 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-cell-stage 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 embryoand eventually into the ICM (Kelly et ah 1978; Graham & Deussen, 1978;Graham & Lehtonen, 1979). The present quantitative data show that during thecell cycle a twin-cell pair increases its inherited intercellular contact. Additionalcell contacts result from cytokinesis associated and cell cycle changes in cellulardimensions (see for instance fig. 2 in Graham & Lehtonen, 1979); and ourpreliminary observations suggest that these contacts behave in a similar way asthe inherited ones. This apparently implies that during the early phases of eachcleavage stage the descendants of the first dividers tug the other cells aroundthem and therefore tend to segregate into an internal position. However, if thedata on division asynchrony (this study; Kelly et al. 1978) are compared to thetiming of changes in cell contacts, it appears that even the last dividers have thepossibility to maximize their contacts and to reverse the morphogenetic effectsof the earlier contacts before next cell divisions. The implication is that changesin stage-dependent and cell-cycle-dependent cell contacts observed in isolatedcell pairs cannot by themselves account for the internalization of fast dividingcells.
There are at least two possible mechanisms for preventing this deshapingeffect of the last dividing cells and each of them would depend on interactionsbetween all the cells of the embryo. First, the number of intercellular contactstends to increase during and soon after cytokinesis. As a consequence, cellcontacts occupy a rapidly growing proportion of the surface of the fast dividers'descendants. The slow dividers therefore tend to segregate into an outsideposition, and as they divide their descendants have a reduced probability toestablish new contacts, especially as the earlier cell contacts tend to directcytokinesis towards the exterior of the embryonic structure (Graham & Leh-tonen, 1979). The relatively low deformability of the spherical cells results insimple physical limits for the area of a cell contact and therefore the slowdividers' descendants with fewer contacts are likely to stay in an outside positionand to form the outer surface of the embryo. Second, after reaching the maximumcontact area, the cell contacts may change so that they resist the reversing forcesthat 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 agglutin-ability (Rector & Granholm, 1978; Rowinski, Solter & Koprowski, 1976) in8-cell-stage embryos.
The present and our previous results suggest that the mouse cell lineage is aresult of continuous cell interactions starting from the 2-cell stage. This epi-genetic control of the lineage can be partly understood in terms of simple spatialand temporal relationships of the cells and especially their intercellular contactsduring the cleavage stages. The molecular mechanism by which the developmentof cell contacts is controlled is unknown, but changes in cell surface microvilli(this study; Ducibella et al. 1977; Graham & Lehtonen, 1979) and the organiza-tion of cytoskeletal structures (Lehtonen & Badley, 1980) may be involved inthis process. Furthermore, as suggested by Lewis & Wright (1935), the internalpressure of the cells and the surface tension or deformability of the cell cortexmay 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 amgrateful to Ms Margaret K. Arnold and Dr K. A. Harrap of the Natural Environment Unitof Invertebrate Virology, Oxford, for the use of the Jeol 100 CX Temscan electron microscope.
REFERENCES
ABRAMCZUK, J. & SAWICKI, W. (1974). Variation in dry mass and volume of nonfertilizedoocytes 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 mesento-blast determination as concomitant results of cellular interactions in the mollusc Patellavulgata. Devi Biol. 68, 462-471.
DUCIBELLA, T. & ANDERSON, E. (1975). Cell shape and membrane changes in the eight-cellmouse embryo: prerequisites for morphogenesis of the blastocyst. Devi Biol. 47, 45-58.
DUCIBELLA, T., ALBERTINI, D. F., ANDERSON, E. & BIGGERS, J. D. (1975). The preimplanta-tion mammalian embryo: characterization of intercellular junctions and their appearanceduring development. Devi Biol. 45, 231-250.
DUCIBELLA, T., UKENA, T., KARNOVSKY, M. & ANDERSON, E. (1977). Changes in cell surfaceand cortical cytoplasmic organization during early embryogenesis in the preimplantationmouse embryo. / . Cell Biol. 74, 153-167.
GRAHAM, C. F. & DEUSSEN, Z. A. (1978). Features of cell lineage in preimplantation mousedevelopment. / . Embryol. exp. Morph. 48, 53-72.
GRAHAM, C. F. & LEHTONEN, E. (1979). Formation and consequences of cell patterns inpreimplantation mouse development. / . Embryol. exp. Morph. 49, 277-294.
HARA, K. (1977). The cleavage pattern of the axolotl egg studied by cinematography andcell counting. Wilhelm Roux's Arch. Devi Biol. 181, 73-87.
HILLMAN, N., SHERMAN, M. I. & GRAHAM, C. (1972). The effect of spatial arrangement on celldetermination during mouse development. / . Embryol. exp. Morph. 28, 263-278.
KELLY, S. J. (1977). Studies on the developmental potential of 4- and 8-cell stage mouseblastomeres. / . exp. Zool. 200, 365-376.
KELLY, S. J., MULNARD, J. G. & GRAHAM, C. F. (1978). Cell division and cell allocation inearly mouse development. / . Embryol. exp. Morph. 48, 37-51.
KNUTTON, S. (1976). Structural changes in the plasma membrane of synchronized P815Ymastocytoma cells. Expl Cell Res. 102, 109-116.
Dimensional changes in cleavage-stage cells 249KNUTTON, 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 preimplanta-tion 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 intercellularjunctions in the preimplantation mouse embryo by freeze-fracture and thin-sectionelectron 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 singlecell to the blastocyst stage. Nature 233, 422-424.
PORTER, K., PRESCOTT, D. & FRYE, J. (1973). Changes in surface morphology of Chinesehamster ovary cells during the cell cycle. / . Cell Biol. 57, 815-836.
RECTOR, J. T. & GRANHOLM, N. H. (1978). Differential concanavalin A-induced aggluti-nation 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 inducedagglutinability during preimplantation mouse development. Expl Cell Res. 100, 404-408.
TARKOWSKI, A. K. & WROBLEWSKA, J. (1967). Development of blastomeres of mouse eggsisolated at the 4- and 8-cell stage. / . Embryol. exp. Morph. 18, 155-180.
WHITTEN, W. K. (1971). Nutrient requirements for culture of preimplantation embryosin vitro (ed. G. Raspe). Advances in the Biosciences 6, 129-140.
WILSON, I. B., BOLTON, E. & CUTTLER, R. H. (1972). Preimplantation differentiation in themouse egg as revealed by microinjection of vital markers. J. Embryol. exp. Morph. 27,467-479.
{Received 11 December 1979, revised 10 March 1980)
