J. Embryol. exp. Morph. 77, 297-308 (1983)
297
Printed in Great Britain © The Company of Biologists Limited 1983
The fifth cell cycle of the mouse embryo is longer for
smaller cells than for larger cells
By HILARY A. M A C Q U E E N 1 AND MARTIN H. JOHNSON 2
From the Department of Anatomy, University of Cambridge, U.K.
SUMMARY
Newly-formed pairs of 16-cell blastomeres were collected by periodic observation of
isolated 8-cell blastomeres. Any pairs formed by division were recovered and classified as
being composed of 1/16 blastomeres that differed in size or were of similar size. All of the
latter and some of the former were then cultured for periods of up to 20 h. The remaining
pairs of different-sized blastomeres were disaggregated to larger or smaller cells. Some of
these were reaggregated as smaller: smaller or larger: larger pairs, and these, together with
the remaining isolated smaller and larger blastomeres were also cultured for up to 20 h. At
hourly intervals, cultured cells were sampled and analysed for incidence of division. It was
found that larger cells divided on average after 12 h whereas smaller cells divided on average
after 14 h.
INTRODUCTION
Blastomeres undergoing the fifth cell cycle (16-cell stage) of mouse development may be divided into two subpopulations occupying different relative
positions within the embryo and having different properties and phenotypes
(Handyside, 1981; Reeve & Ziomek, 1981; Reeve, 1981; Ziomek & Johnson,
1981, 1982; Johnson & Ziomek, 1981, 1983; Randle, 1982; Kimber, Surani &
Barton, 1982). The appearance of phenotypic heterogeneity during development is frequently accompanied by a corresponding heterogeneity in the
lengths of the cell cycles. Indeed it has been demonstrated that inside cells at
the 16-cell stage have a significantly higher thymidine-labelling index than do
outside cells, which has been interpreted as reflecting a faster division rate of
the inside cells (Barlow, Owen & Graham, 1972). We have re-examined this
question more directly using newly formed 16-cell blastomeres of inner and
outer type, and have measured the period over which they divide. We find,
contrary to the earlier interpretations from labelling indices, that as a population larger, outer, polar cells divide two hours earlier than smaller, inner,
apolar cells.
1
Author's address: Biology Department, Open University, Milton Keynes, MK7 6AA,
U.K.
2
Author's address: Department of Anatomy, Downing Street, Cambridge, CB2 3DY,
U.K.
298
H. A. MACQUEEN AND M. H. JOHNSON
MATERIALS AND METHODS
Embryo collection
Embryos were obtained from 3-4 weeks old HC-CFLP female mice (Hacking
& Churchill Ltd., Alconbury) superovulated with 5i.u. pregnant mare's serum
gonadotrophin (PMS, Intervet) followed after 46-50h by 5 i.u. human chorionic
gonadotrophin (hCG, Intervet) and mated with HC-CFLP males. All embryonic
ages are expressed in h post hCG. Early 4-cell embryos were recovered from the
oviducts at 52 h post hCG and cultured at 37 °C for 15 h in vitro in medium 16
(Whittingham, 1971) containing 4mg/ml bovine serum albumin (M16 + BSA)
in 5 % CO2 in air under oil until they were mid- to late-8-cell embryos.
Terminology
A single blastomere from the 8-cell stage is termed 1/8, and the pair of cells
resulting from division of an isolated 1/8 cell is called a natural 2/16 couplet.
Such couplets may be divided into those that contain cells of approximately equal
size and those that contain cells that clearly differ in size. Some couplets in which
there was a size difference were disaggregated to give populations of larger and
smaller 1/16 blastomeres. Some of these dissociated cells were aggregated
together as pairs and these are termed 'reaggregates'. Division of 1/16 or 2/16
cells yielded 2/32 or 4/32 clusters respectively.
Collection of newly formed 16-cell blastomeres
Late 8-cell embryos were stripped of their zonae pellucidae by brief exposure
to acid Tyrode's solution (Nicholson, Yanagimachi & Yanagimachi, 1975) and
incubated in Ca2+-free medium 16 pre-equilibrated at 37 °C in 5 % CO2 and containing 6 mg/ml BSA (Ca2+-free M16+BSA: Pratt, Ziomek, Reeve & Johnson,
1982) for 10 min prior to disaggregation to single cells by gentle pipetting with a
flame-polished micropipette. Single 1/8 cells were then cultured in M16 + BSA
and examined at hourly intervals for evidence of division to 2/16. Any 2/16
couplets which formed within this hour were pooled together. The couplets were
then sorted into groups of equal-sized or different-sized pairs, and these were
either cultured as pairs in drops of M16+BSA under oil in Sterilin dishes for up to
20 h, or, in the case of some of the different-sized couplets, the pairs were
disaggregated to yield populations of isolated larger cells and smaller cells. These
were further cultured either as single cells or after reaggregation as pairs of
larger + larger or smaller + smaller cells. Reaggregation was facilitated by use of
phytohaemagglutinin (PHA; 1/20 dilution of Gibco stock). In a preliminary experiment we compared single cells cultured with or without brief exposure to
PHA by serial observation, recording any cells which were undergoing division.
No difference was observed between the two groups in this experiment. Since
PHA greatly facilitates adhesion of pairs of cells, it was used throughout.
Cell cycles in mouse embryo
299
Examination of cultured 16-cell blastomeres
At intervals, samples of the cultured cells were placed in 4:6-diamidino-2phenylindole (0-2/ig/ml M16 + BSA; DAPI, Sigma Chemical Co., Johnson &
Ziomek, 1983) and cultured for a further 2h. This labelling procedure i$ not
detrimental to the cells (Reeve & Kelly, 1983). The cells were then mounted in
drops of medium under oil on tissue type slides (Baird & Tatlock) and viewed
on a Zeiss epifluorescence microscope, with incident source HBO 200, filter set
48 77 05 for DAPI and conventional transmission optics for bright field. The
numbers of DAPI-stained nuclei and the cell types containing them were
recorded (see Fig. 1). In this way the number of divisions that had occurred since
the initiation of culture could be calculated (see section 1 of Results).
Photography was on Kodak Tri-X 35 mm film with a 2 min exposure for DAPI
and 1/32-1 /64 s for bright field. Cells that had been cultured for x hours since
the time of harvesting as 2/16 couplets were designated 'x h old' etc, although in
fact they included cells that might be as much as (x + 1) h old, due to the hourly
regime of inspection and sampling.
RESULTS
1. Classification of cell clusters generated by division of 16-cell blastomeres in
vitro
In 96 % of 2/16 natural couplets in which the two cells differ overtly in size,
the larger cell is polar and the smaller is apolar (Johnson & Ziomek, 1981).
Classification of cells by relative size therefore also gives a reliable classification
by phenotype. In order to determine the length of the fifth cell cycle (16-cell
stage), larger polar and smaller apolar cells were cultured as populations of
isolated 1/16 cells, as natural pairs, or as reaggregates of two polar or apolar
cells. Staining with DAPI after varying periods of culture then allowed the
numbers and distribution of nuclei to be determined.
In experiments in which pure populations of polar or apolar cells were cultured, either singly or as homogeneous reaggregates, simply counting the nuclei
after different culture periods allows estimates of division rates of polar or apolar
cells to be made (see Table 1, columns 2, 3, 6 & 7; Fig. la-d). However, for
natural pairs composed of either similar or different sized blastomeres the interpretation is more complex. We have shown previously (Johnson & Ziomek,
1983) that natural 2/16 pairs give rise to 4/32 clusters with four distinct
phenotypes, termed A, B, C & D (Fig. le, f, i, j respectively). Phenotypes A
and B, each consisting of two outer and two inner cells, are derived from 2/16
pairs in which a polar cell has enveloped an apolar cell (as shown in Fig. lg) and
then both have divided. Phenotype C (shown in Fig. 1/) is derived mostly from
a 2/16 pair comprising two polar cells (shown in Fig. Ik) in which both cells have
300
H. A. MACQUEEN AND M. H. JOHNSON
1a
#
C
Fig. 1. Phenotypes of blastomeres isolated from 8-cell embryos and cultured for
various times through two cell cycles. For detailed procedures see Materials and
Methods. DAPI staining reveals nuclear position, (a) Fluorescence and bright-field
photographs of an apolar 1/16 cell (upper left) and a 2/32 pair derived from the
division of an apolar 1/16 cell, (b) Fluorescence and bright-field photographs of a
2/32 pair derived from the division of a polar 1/16 cell, (c) Fluorescence and brightfield photographs of a 2/32 pair, derived from the division of a polar 1/16 cell, in
which one (polar) cell has enveloped the other (apolar). (d) Superimposed
fluorescence and bright-field photographs of a 2/32 pair, derived from the division
of a polar 1/16 cell, in which both cells are polar and envelopment has therefore not
occurred, (e) Fluorescence and bright-field photographs of a 4/32 cluster, phenotype
A, derived from the division-of a 2/16 pair in which a polar cell had enveloped an
apolar cell. Two outer and two inner cells are present, if) Superimposed fluorescence
and bright-field photographs of a 4/32 cluster, phenotype B, derived from the
division of a 2/16 pair in which a polar cell had enveloped an apolar cell. Two outer
and two inner cells are present, (g) Fluorescence and bright-field photographs of a
2/16 pair in which a polar cell has enveloped an apolar cell, thought to give rise to
phenotypes A & B. (h) Superimposed fluorescence and bright-field photographs of
one 1/16 cell surrounded by two 1/32 cells. This may be a transitional stage in the
formation of phenotypes A, B or D (see text for details), (i) Fluorescence and brightfield photographs of a 4/32 cluster containing four outer cells, phenotype C, probably derived from the division of a 2/16 pair of polar cells. (/) Fluorescence and
bright-field photographs of a 4/32 cluster, phenotype D, probably derived from the
division of a 2/16 pair of polar cells. Three outer cells surround one inner cell, (k)
Fluorescence and bright-field photographs of a 2/16 pair of polar cells thought to give
rise to phenotypes C and D. (/) Fluorescence and bright-field photographs of one
outer 1/16 cell and two outer 1/32 cells, thought to be a transitional stage in the
formation of phenotypes C or D.
Cell cycles in mouse embryo
301
EMB77
302
H. A. MACQUEEN AND M. H. JOHNSON
divided such that the cleavage plane bisects the pole (= a conservative division)
thereby generating four polar, outer, 32-cells. Phenotype D (shown in Fig. If) is
also derived mostly from a 2/16 pair of polar cells, but where only one has divided
conservatively through the pole to generate two polar 32-cells, the other dividing
parallel to the pole (= a differentiative division) to generate one polar and one
apolar 32-cell; thus, three polar, outer and one apolar, inner 32-cells result.
The experiments reported here have produced two further phenotypes. That
shown in Fig. 1/z, in which two cells surround a third, can be interpreted in two
ways: it might represent the conservative division of a polar cell in a polar-apolar
couplet (i.e. a transitional stage in the formation of phenotypes A & B), or it
might represent a differentiative division of one polar cell in a polar-polar 2/16
couplet (i.e. a transitional stage in the formation of phenotype D). Since in both
cases the dividing cell would have been polar, the correct assignment of division
will be in each case to the polar cell type (Fig. 2). The other novel phenotype
larger
larger
larger
Fig. 2. Scheme summarizing assumed derivations of the various phenotypes
illustrated in Fig. 1. Note that phenotype (h) can be derived from either type of initial
couplet but that in both cases it results from division of the larger cell, (e/f) represent
the phenotypes A & B and (///) the phenotypes C & D of Johnson & Ziomek, 1983.
In the experiments reported here, of the 4/32 clusters derived from similar-sized
natural pairs 64 % were type A/B, 20 % were type C and 17 % were type D, and of
those derived from different sized natural pairs, 79 % were A/B, 11 % were C and
9% were D. Of the (2/32+1/16) clusters derived from similar-sized natural pairs
80 % were as shown in (h) and 20 % as in (/), and of those from different-sized pairs
88 % were as in (h) and 12 % as in (/)•
0(58)
0(14)
33(6)
19 (16)
54 (13)
77 (13)
86 (14)
84 (88)
<10
10
11
12
13
14
15
>16
In differentsized natural
2/16
0(15)
0(11)
4(27)
79 (14)
91(22)
79 (14)
87 (23)
93 (108)
Aggregated
with a second
larger cell
0 (110)
0(24)
44(16)
67 (12)
72 (14)
72 (14)
70 (10)
93 (116)
In differentsized natural
2/16
0(3)
0(11)
0(21)
0(10)
33 (12)
36 (14)
70 (20)
76 (54)
Aggregated
with a second
smaller cell
0(244)
0(14)
3 (104)
0(10)
44(18)
58 (12)
70 (10)
72 (54)
Alone
0(54)
0(16)
0(5)
0(17)
0(18)
43 (14)
67 (15)
90 (87)
In similarsized natural
2/16
0(196)
0(16)
17 (30)
62 (26)
88 (25)
94 (32)
80 (45)
92 (26)
% of smaller cells that divided
in various cellI combinations
(n)
0(7)
0(14)
0(14)
0(19)
44 (23)
50 (20)
71 (17)
92 (12)
In similarsized natural
2/16
* In fact some blastomeres may be up to one hour older, as cultures were examined at hourly intervals. Each data point represents results from
a minimum of two experiments and in several cases from five or more experiments. Note: values at each time point are not cumulative, since the
analysis terminated the period of culture.
Alone
GO*
Time since
formation
% of larger cells that divided
in various cellI combinations
(n)
Table 1. Division frequency of larger and smaller 16-cell blastomeres at various times after their formation and in various cell
combinations
304
H. A. MACQUEEN AND M. H. JOHNSON
observed, shown in Fig. 11, in which there are three outer cells, also appears to
be a transitional stage in the formation of phenotypes C & D, i.e. a conservative
division of one polar cell in a polar-polar couplet (Fig. 2).
However, we showed previously (Johnson & Ziomek, 1983) that a few C & D
phenotypes arise not from polar-polar couplets, but from 2/16 pairs in which the
polar cell fails to envelop the apolar cell. In these pairs the apolar cell is induced
to polarize by the asymmetry of cell contacts (Ziomek & Johnson, 1981). The
proportion of polar-apolar couplets showing this pattern is about 10 %. In this
No of Cells
379
65
79
68
74
73
92
338
O Larger
308
55
144
56
71
60
62
207
•
Smaller
O
90-
°
80-
o •
w
O
70-
60-
#
o
504030-
20-
•
o
10-
0
—*—*—?—f—.—,—,—.
<10
10
11
12
13 14
15
Time since formation (h)
Fig. 3. Graph showing division frequency (expressed as a percentage of the total
number of cells analysed) amongst smaller and larger 1/16 cells with time after their
formation. The data are drawn from all smaller and larger cells examined (as recorded in Table 1) pooled regardless of their cell association. The number of larger and
smaller cells examined at each time point is recorded above the graph. Note: each
data point represents the sampling at the termination of the culture period of that
group of cells (or their progeny), and is not part of a sequence of sampling of
continuously cultured cells.
Cell cycles in mouse embryo
305
study we have ignored this 10 % and have assumed that all phenotypes shown in
Fig. li, j & / have resulted from polar cell divisions. This will not lead to any
distortion of the numbers if the cell cycles of polar and apolar cells are of similar
length. However, if there is a difference between them it will be underestimated
by this classification.
2. Division frequency of larger and smaller 1/16 cells at different times after their
formation
A total of 1168 larger and 963 smaller 1/16 cells were scored for division over
the period 4 to 20 h after their formation from 8-cell blastomeres. The results are
summarized in Table 1 for the different cell combinations and for the total
populations of each cell type in Fig. 3. It is clear that division of smaller cells does
not occur to a significant extent until cells are 13-14 h old, whereas significant
division of larger cells is seen two hours earlier, when they are 11-12 h old.
DISCUSSION
We have used the relative size difference of polar and apolar 16-cell blastomeres (Johnson & Ziomek, 1981), in conjunction with the results of our previous
analyses of the behaviour of these cells when aggregated in pairs (Ziomek &
Johnson, 1981; Johnson & Ziomek, 1983), to identify and follow cells through
the fifth cell cycle and into the sixth cell cycle of mouse development. We
estimate that approximately 5 % of newly formed 16-cell blastomeres are likely
to have been misidentified (Johnson & Ziomek, 1981), and the progeny of a
further 10 % of smaller, apolar cells may have been misclassffied as being derived
from larger, polar cells (see Results). These errors would be likely to reduce only
marginally any natural differences between cell populations, and would be
manifested by two subgroups of cells. A subgroup of 'smaller' cells should show
early division, and may well be represented in the 13 h sample of smaller cells
(Fig. 3). A subgroup of 'larger' cells should show delayed division, and may well
be represented in the 14-15 h samples of larger cells. We conclude therefore that
our estimates of a median cell cycle time of 12 h for isolated polar cells and of 14 h
for isolated apolar cells are unlikely to be distorted significantly. Lehtonen
(1980) examined by time-lapse videomicrography four 1/8 cell blastomeres cultured in vitro to the 32-cell stage. He recorded also values for the 5th cell cycle
ranging from 11-2 h to 13 h which is consistent with the data reported here.
We have shown that polar cells occupy an outside position and apolar cells an
inside position within the 16-cell embryo (Handyside, 1981; Reeve & Ziomek,
1981). Our conclusion that the cell cycles of inside apolar cells are longer than
those of outside polar cells at the 16-cell stage therefore conflicts with inferences
drawn previously from studies using thymidine-labelling indices (Barlow et al.
1972). There are at least two reasons why this might be so. (i) Our experiments
have been conducted on isolated cells or pairs of cells, whereas the labelling
306
H. A. MACQUEEN AND M. H. JOHNSON
indices were measured on cells within intact embryos. Isolation and culture
procedures might be deleterious and might affect one cell subpopulation differentially, eg. apolar cells being more affected than polar cells. Of cells aged
16 h or more, 90-5 % of larger, polar and 82-5 % of smaller, apolar cells had
divided, suggesting little difference in the final division frequency. Moreover, we
have shown that smaller, apolar cells also divide more slowly when reaggregated
into clusters of 16 cells in various combinations (Ziomek, Johnson & Handyside,
1982). Finally, examination of inside: outside cell ratios in newly expanding
blastocysts reveals evidence of earlier division of outside cells in situ
(unpublished results with T. Fleming and P. Warren). None of these pieces of
evidence is in itself conclusive, but taken together they suggest that the use of
isolated cells has not led to distortion.
(ii) The conclusion that the higher labelling index for inside cells reflects a
faster division rate than that of outer cells requires a number of assumptions.
First, the cells in section must be identified correctly as being inside or outside,
and fixation and embedding could lead to artefactually low estimates of inside
cell numbers. Second, it is necessary to assume that the proportion of the cell
cycle occupied by the S phase is the same in both cell subpopulations, and there
is no evidence to suggest that this is necessarily the case. Third, it is necessary to
assume that the S phase occurs in the same relative position in the cell cycle;
again there is no evidence on this question. Finally it is necessary to assume that
the level of incorporation of radioactive thymidine as a result of DNA repair
(perhaps of damage induced by the radiolabel itself; MacQueen, 1979) is the
same in both cell subpopulations. There is some evidence that this is not the case
(Snow, 1973; Goldstein, Spindle & Pedersen, 1975).
Since each of the methods used thus far to measure the length of the fifth cell
cycle is open to criticism, a final decision on differences between the two cell
subpopulations must await a more definitive experimental approach.
It is of interest that in a number of cell types an inverse relationship between
the size of a new-born cell and the length of the ensuing cycle has been observed
(reviewed Fantes & Nurse, 1981). These studies have led to the proposal of sizecontrol theories for the cell cycle in which the size of a cell is a principal determinant of its division time. Thus, in the steady state large cells divide earlier than
small cells since they reach the critical size sooner (assuming growth rates to be
independent of initial size). In this way mean cell size in a population would be
stabilized over many generations. However, our observations on the cells of the
mouse early embryo cannot be explained simply in this way, since the interphase
of the mouse blastomere cycle is characterized by zero growth in mass; indeed
a slight reduction in cell size may occur. Thus size of blastomeres must be
assessed, and related to division time, by some feature that is independent of
growth. Each daughter 16-cell blastomere receives an equal endowment of
genetic material, but an endowment of non-genetic material that differs both
qualitatively and quantitatively. There is considerable evidence from other cell
Cell cycles in mouse embryo
307
types that cytoplasmic components can indeed influence the relationship between size and division time (Prescott, 1956; Galavazi & Bootsma, 1966; Sachsenmaier, 1981). These components have been proposed to be either repressors,
that become diluted with growth, or activators that are inherited in different
amounts or synthesized at a rate dependent upon size. Our results would be more
consistent with activator models and we are now investigating this possibility.
We wish to thank Gin Tampkins for her technical help, Ian Edgar and Raith Overhill for
help in the preparation of figures and Drs Hester Pratt, Harry Goodall and Tom Fleming for
constructive criticism and advice. The work was supported by grants from the Cancer
Research Campaign and the Medical Research Council to M. H. J.
REFERENCES
D. A. J. & GRAHAM, C. F. (1972). DNA synthesis in the pre-implantation
mouse embryo. /. Embryol. exp. Morph. 27, 431-445.
FANTES, P. A. & NURSE, P. (1981). Division timing: controls, models and mechanisms. In The
Cell Cycle (ed. P. C. L. John), Cambridge University Press, pp. 11-34.
GALAVAZI, G. & BOOTSMA, D. (1966). Synchronisation of mammalian cells in vitro by inhibition of the DNA synthesis. II. Population dynamics. Expl Cell Res. 41, 438-451.
GOLDSTEIN, L. S., SPINDLE, A. I. & PEDERSEN, R. A. (1975). X-ray sensitivity of the preimplantation mouse embryo in vitro. Radiat. Res. 62, 276-287.
HANDYSIDE, A. H. (1981). An immunofluorescent technique for detection of the numbers of
inside and outside cells from mouse morulae. /. Reprod. Immunol. 2, 339-350.
JOHNSON, M. H. & ZIOMEK, C. A. (1981). The foundation of two distinct cell lineages within
the mouse morula. Cell 24, 71-80.
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.
KIMBER, S. J., SURANI, M. A. H. & BARTON, S. C. (1982). Interactions of blastomeres suggest
changes in cell surface adhesiveness during the formation of inner cell mass and trophectoderm in the preimplantation mouse embryo. /. Embryol. exp. Morph. 70, 133-152.
LEHTONEN, E. (1980). Changes in cell dimensions and intercellular contacts during cleavagestage cell cycles in the mouse embryonic cells. /. Embryol. exp. Morph. 58, 231-249.
MACQUEEN, H. A. (1979). Lethality of radioisotopes in early mouse embryos. /. Embryol.
exp. Morph. 52, 203-208.
NICHOLSON, G. A., YANAGIMACHI, R. & YANAGIMACHI, H. (1975). Ultra-structural localisation of lectin binding sites on the zonae pellucidae and plasma membranes of mammalian
eggs. /. Cell Biol. 66, 263-274.
PRATT, H. P. M., ZIOMEK, C. A., REEVE, W. J. D. & JOHNSON, M. H. (1982). Factors affecting
compaction of the 8-cell mouse embryo. /. Embryol. exp. Morph. 70, 113-132.
PRESCOTT, D. M. (1956). Relation between cell growth and cell division. Expl Cell Res. 11,
86-98.
RANDLE, B. (1982). Cosegregation of monoclonal reactivity and cell behaviour in the mouse
preimplantation embryo. J. Embryol. exp. Morph. 70, 261-278.
REEVE, W. J. D. (1981). Distribution of ingested horseradish peroxidase by cells of the 16-cell
mouse morula. /. Embryol. exp. Morph. 66, 191-207.
REEVE, W. J. D. & KELLY, F. P. (1983). Nuclear position in cells of the mouse early embryo.
/. Embryol. exp. Morph. 75, 117-139.
REEVE, W. J. D. & ZIOMEK, C. A. (1981). Distribution of microvilli on dissociated blastomeres
from mouse embryos: evidence for surface polarization at compaction. J. Embryol. exp.
Morph. 62, 339-350.
BARLOW, P., OWEN,
308
H. A. MACQUEEN AND M. H. JOHNSON
W. (1981). The mitotic cycle in Physarwn. In The Cell Cycle, (ed. P. C. L.
John), pp.139-160.
SNOW, M. H. L. (1973). Abnormal development of preimplantation mouse embryos grown
in vitro with [3H]thymidine. J. Embryol. exp. Morph. 29, 601-615.
WHITTINGHAM, D. G. (1971). Culture of mouse ova. /. Reprod. Fert. (Suppl.) 14, 7-21.
ZIOMEK, C. A. & JOHNSON, M. H. (1981). Properties of inside and outside cells from the 16cell mouse morula. Wilhelm Roux' Arch, devl Biol. 190, 287-296.
ZIOMEK, C. A. & JOHNSON, M. H. (1982). The roles of phenotype and position in guiding the
fate of 16-cell mouse blastomeres. Devi Biol. 91, 440-447.
ZIOMEK, C. A., JOHNSON, M. H. & HANDYSIDE, A. H. (1982). The developmental potential
of mouse 16-cell blastomeres. /. exp. Zool. 221, 345-355.
SACHSENMAIER,
{Accepted 16 May 1983)