The number of trophoblast giant cells in outgrowths of mouse blastocysts was determined before, during and after egg-cylinder formation in vitro. Giant-cell numbers rose initially but reached a plateau 12 h before the egg cylinder appeared. A secondary increase began 24 h after egg-cylinder formation. Blastocysts whose mural trophectoderm cells were removed before or shortly after attachment in vitro formed egg cylinders at the same time as intact blastocysts but their trophoblast outgrowths contained fewer giant cells at this time. The results support the idea that egg-cylinder formation in vitro is accompanied by a redirection of the polar to mural trophectoderm cell movement which characterizes blastocysts before implantation. The resumption of giant-cell number increase in trophoblast outgrowths after egg-cylinder formation may correspond to secondary giant-cell formation in vivo. It is suggested that a time-dependent change in the strength of trophoblast cell adhesion to the substratum occurs after blastocyst attachment in vitro which restricts the further entry of polar cells into the outgrowth and therefore results in egg-cylinder formation.

The initiation of mouse egg-cylinder formation in vivo occurs soon after implantation due to the development of a multilayered extraembryonic ectodermal tissue from the previously single layer of polar trophectoderm (Snell & Stevens, 1966; Gardner & Johnson, 1975; Gardner & Papaioannou, 1975; Copp, 1979). Consequently, the early egg cylinder consists of two regions: one trophoblastic and the other ectodermal, derived from the inner cell mass (ICM) of the blastocyst. Both regions are surrounded by a layer of primitive endoderm cells. It has been suggested that egg-cylinder morphogenesis depends upon (a) an interaction between ICM (or its derivative the embryonic ectoderm) and overlying polar trophectoderm in the blastocyst, so that a high rate of polar cell division is promoted, (b) resulting movement of cells out of the polar region and (c) mechanical constraints on the direction in which this movement can occur (Gardner & Papaioannou, 1975; Copp, 1978, 1979). Before implantation, polar cells move into the mural region but after the blastocyst has become attached by its mural trophectoderm to the endometrium cell movement appears to be redirected: trophectoderm cells accumulate over the ICM and contribute to the extra-embryonic ectoderm (Copp, 1979). It therefore seems likely that blastocyst attachment during implantation prevents any further influx of polar cells into the mural region and so is responsible for the initiation of egg-cylinder formation.

When mouse blastocysts outgrow in vitro, a trophoblastic giant-cell monolayer is formed on which the ICM can be seen as a compact lump (Gwatkin, 1966). After 3 or 4 days of culture in serum-containing medium the ICM becomes supported above the level of the giant cells by a newly developed ‘proximal embryonic’ region (Hsu, Baskar, Stevens & Rash, 1974; Pienkowski, Solter & Koprowski, 1974; Wiley & Pedersen, 1977). This event may correspond to egg-cylinder formation in vivo, in which case the ‘proximal’ and ‘distal’ regions would represent extraembryonic and embryonic ectodermal components respectively. Continued culture of such outgrowths can lead to the development of apparently normal late egg cylinders in which the neural tube, somites, heart rudiment, amnion, allantois and yolk sac are all formed (Hsu, 1978). Consequently, it seems likely that normal egg-cylinder formation may occur in vitro. It must be concluded, therefore, that either (a) egg-cylinder morphogenesis can occur in the absence of mechanical constraints, (b) the processes of egg-cylinder formation in vivo and in vitro are not related mechanistically or (c) mechanical constraints are present, in the absence of a uterine environment, perhaps due to an interaction between the trophoblastic monolayer and its substratum. If the latter idea is correct, it may be predicted that egg-cylinder formation, following attachment of blastocysts in vitro, will be associated with a restriction in the degree of trophoblast giant-cell outgrowth so that dividing polar cells are forced to accumulate beneath the ICM and form the proximal region of the egg cylinder. This prediction has been tested by studying the kinetics of cell number increase in trophoblast giant-cell outgrowths before, during and after egg-cylinder formation. The results indicate that, as predicted, there is a cessation of cell movement into the trophoblast outgrowth shortly before egg-cylinder formation. The redirection of trophoblast cell movement observed could be a time-related event or alternatively may require the presence of a critical number of cells in the outgrowth. In order to distinguish between these possibilities, the time of egg-cylinder formation has been noted in embryos whose trophoblastic cell number had been altered, experimentally.

Embryos and manipulation

Blastocysts were flushed from the uteri of pregnant random-bred CFLP female mice (Anglia Laboratory Animals Ltd) between 3 days 12 h and 3 days 15 h after the estimated time of ovulation (see Copp, 1978). Embryos were recovered and manipulated in PB1 medium (Whittingham & Wales, 1969) plus 10 % heat-inactivated fetal calf serum (FCS). The mural region was separated from the ICM plus its covering polar trophectoderm by cutting blastocysts parallel to the surface of the ICM using a pair of glass needles controlled by a Leitz micromanipulator assembly (Gardner, 1978). Both ICM/polar and trophectoderm vesicle (TV) fragments were cultured separately. The zonae of all embryos were removed, before culture, either by microsurgery or by treatment with acidic Tyrode’s solution, pH 2·5 (Nicolson, Yanagimachi & Yanagimachi, 1975). Trophoblastic giant cells were removed from blastocyst outgrowths (operated outgrowths) under a Wild dissecting microscope by use of glass microneedles. This procedure normally caused the expiants to detach from the plastic substratum, but reattachment usually occurred within about 6 h. In order to confirm the visual classification of outgrowths as containing either ICMs or egg cylinders (see Analysis section), some egg-cylinder-like structures developing in vitro were removed from their giant-cell monolayers usiug glass needles, fixed in formol-acetic-alcohol and were prepared for histological analysis.

Culture

All embryos were cultured in alpha-modification of Eagle’s medium (Flow Laboratories, U.K.) supplemented with 30 μM adenosine, guanosine, cytidine and uridine, 10 μM thymidine and 10 % FCS. Blastocysts were grown in 50 mm plastic tissue-culture dishes (Sterilin), each containing 5 ml culture medium, under an atmosphere of 5 % CO2 in air. Up to ten embryos were cultured in each dish and the culture medium was not changed during the experimental period which ranged from 6 to 8 days. Embryos which attached near to the edge of the dish, or near to each other, were discarded. The day of initiation of culture was designated day 1.

Analysis

Blastocyst outgrowths were examined by inverted phase-contrast microscopy, every 12 h, for the presence or absence of an egg cylinder and, in addition, the area of outgrowth and number of giant-cell nuclei were scored. The positions of expiants were marked so that individual outgrowths could be recognized throughout the period of culture. An egg cylinder was considered to be present when the ICM was clearly supported above the level of the giant cells by a distinct ‘proximal’ region. The presence of a ‘proximal’ region was confirmed histologically for some blastocyst outgrowths. The time of first appearance of an egg cylinder was designated time = 0, although individual outgrowths varied with respect to the day of culture on which an egg cylinder could first be recognized. Subsequently, egg cylinders were classified as ‘organized’ if the ‘distal’, and later the ‘proximal’, regions cavitated and became rounded and smooth (Fig. 1A) while egg cylinders in which the ‘distal’ region lost its rounded regular shape were classified as ‘disorganized’ (Fig. IB). The area of each blastocyst outgrowth was calculated from the formula:
formula
Fig. 1.

Sections of egg cylinders developed in vitro from intact blastocyst outgrowths on day 6 of culture (haemalum and eosin). A. ‘Organized’ egg cylinder consisting of proximal (p) and distal (d) regions surrounding a central cavity which resembles the proamniotic cavity of the day-7 in vivo egg cylinder. Additional resemblances to in vivo egg cylinders include the apparently pseudostratified epithelium of the distal region, showing mitoses at the free surface, and the transition from columnar to squamous cell shape in the visceral endoderm (v). B. ‘Disorganized’ egg cylinder in which a solid lump of cells (l) appears to have overgrown the epithelium of the distal region. The point of attachment of the egg cylinder to the substratum is at the top of the figure in each case.

Fig. 1.

Sections of egg cylinders developed in vitro from intact blastocyst outgrowths on day 6 of culture (haemalum and eosin). A. ‘Organized’ egg cylinder consisting of proximal (p) and distal (d) regions surrounding a central cavity which resembles the proamniotic cavity of the day-7 in vivo egg cylinder. Additional resemblances to in vivo egg cylinders include the apparently pseudostratified epithelium of the distal region, showing mitoses at the free surface, and the transition from columnar to squamous cell shape in the visceral endoderm (v). B. ‘Disorganized’ egg cylinder in which a solid lump of cells (l) appears to have overgrown the epithelium of the distal region. The point of attachment of the egg cylinder to the substratum is at the top of the figure in each case.

where r was half the mean of the diameters, measured by an eyepiece graticule, along the x and y axes, assuming the outgrowth to be circular. Preliminary experiments in which giant-cell numbers obtained from counts of nuclei by phase-contrast microscopy were compared with those obtained from the same outgrowths after fixation and staining indicated that results based on the two methods do not differ significantly (t = 1·56, d.f. = 20, P > 0·10). Consequently, nuclear counts were performed by phase-contrast microscopy at 12 h intervals on each individual explant. Giant-cell nuclei were recognized as being rounded, pale regions of attached cells in which one or more distinct, dense nucleoli were present. The number of nuclei in each outgrowth was counted three times in succession, and the mean of the two nearest scores was used. Both areas of outgrowth and giant-cell numbers varied between individual embryos. Since information concerning time-dependent changes in the relative size of these parameters was required, it was necessary to reduce the inter-embryonic variation by normalizing outgrowth areas and giant-cell numbers. This was done, for any particular outgrowth, by dividing each area or cell number measurement by the value obtained for the same outgrowth at time = 0.

Kinetics of blastocyst outgrowth

Outgrowth areas and trophoblast giant-cell numbers were recorded every 12 h for cultures of intact blastocysts (‘organized’ and ‘disorganized’) and of trophectoderm vesicles (TV). Fig. 2A shows the relationship between mean normalised outgrowth area and the time of egg-cylinder formation. Outgrowth areas increase in an approximately exponential fashion as culture proceeds both in expiants of intact blastocysts and TVs (see McLaren & Hensleigh, 1975).

Fig. 2.

Graphs to show relationship of (A) mean normalized outgrowth area and (B) mean normalized giant-cell number to the time of initiation of egg-cylinder formation in vitro (time = 0). Continuous line represents intact blastocyst outgrowths which gave rise to ‘organized’ egg cylinders (n = 15), dashed line represents ‘disorganized’ egg cylinders (n = 11) and dotted line indicates outgrowths derived from trophectoderm vesicles (n = 7). Standard errors are shown on the graphs.

Fig. 2.

Graphs to show relationship of (A) mean normalized outgrowth area and (B) mean normalized giant-cell number to the time of initiation of egg-cylinder formation in vitro (time = 0). Continuous line represents intact blastocyst outgrowths which gave rise to ‘organized’ egg cylinders (n = 15), dashed line represents ‘disorganized’ egg cylinders (n = 11) and dotted line indicates outgrowths derived from trophectoderm vesicles (n = 7). Standard errors are shown on the graphs.

The relationship between mean normalized outgrowth giant-cell number and the time of egg-cylinder formation is shown in Fig. 2 B. The results demonstrate that, although giant-cell numbers increase during blastocyst attachment, there is a cessation of this increase just before the egg cylinder appears. Giant-cell numbers remain constant for the next 36 h, after which they rise once again. A similar pattern of giant-cell number increase is seen in both ‘organized’ and ‘disorganized’ blastocyst outgrowths. The resumption of giant-cell number increase which occurs 24 h after egg-cylinder formation could be due to cell division within the monolayer, although giant cells are not believed to undergo cytokinesis (see Ansell, 1975), or may reflect the movement of cells from the egg cylinder into the monolayer. Alternatively, the cell number increase might have been an apparent one if binucleation of trophoblast giant cells is a common phenomenon. A test of this latter possibility is provided by the outgrowth of TVs in this experiment (Fig. 2B). There is no secondary increase in mean normalized nuclear number in TV outgrowths, indicating that binucleation alone is not a sufficient explanation for the secondary increase in giant-cell numbers within blastocyst outgrowths. The most likely explanation, therefore, appears to be one based upon cell movement. Support for this idea comes from the observation that, during the period when increase in giant-cell number has resumed (t = 1·5 onwards), numbers of ‘small’ giant cells become visible within the monolayer near the points of insertion of the ‘proximal’ egg-cylinder region. These cells may have recently originated from the trophoblastic part of the egg cylinder.

Figure 2 shows that there is no cessation of outgrowth area increase coincident with the cessation in giant-cell number increase at the time of egg-cylinder formation. This suggests that increases in outgrowth area result primarily from growth of the constituent giant cells which are undergoing increase in DNA content (Barlow & Sherman, 1972; Copp, 1980a) and that the addition of new trophoblastic cells to the monolayer is not required for increase in outgrowth area. Support for this conclusion comes from the finding that TV outgrowth areas increase in a similar manner despite their fixed giant-cell numbers.

Relationship between trophoblast cell number and time of egg-cylinder formation

Figure 3 illustrates the experimental design. The time of egg-cylinder formation, and number of cells in the trophoblast monolayer at that time, were determined for intact blastocysts, ICM/polar fragments and operated outgrowths (see Materials and Methods). Of 15 ICM/polar fragments which attached, eight formed egg cylinders plus associated giant-cell outgrowths, while seven gave rise only to giant cells, although in some cases degenerate ICMs were observed on the trophoblast monolayers. Table 1 shows that eggcylinder formation was not markedly delayed in ICM/polar fragment outgrowths relative to intact blastocyst controls, whereas the average number of giant cells in the outgrowth at the time of egg-cylinder formation was significantly lower than for intact blastocysts (t = 4·15, P > 0·001).

Table 1.

The time of initiation and number of trophoblast giant cells present at egg-cylinder formation in outgrowths derived from various blastocyst explant types

The time of initiation and number of trophoblast giant cells present at egg-cylinder formation in outgrowths derived from various blastocyst explant types
The time of initiation and number of trophoblast giant cells present at egg-cylinder formation in outgrowths derived from various blastocyst explant types
Fig. 3.

Diagram illustrating the design of an experiment to investigate the possibility that a critical trophoblast giant-cell number must be present in the blastocyst outgrowth before egg-cylinder formation may occur. (A) Outgrowth of intact blastocyst; (B) Removal of the mural trophectodermal region from blastocysts followed by outgrowth of the resulting ICM/polar fragments; (C) Outgrowth of intact blastocysts followed by removal of trophoblast giant cells prior to egg-cylinder formation (operated outgrowths). Representation of embryonic tissues: trophoblast, white; ICM and embryonic ectoderm, stippled ; endoderm, black.

Fig. 3.

Diagram illustrating the design of an experiment to investigate the possibility that a critical trophoblast giant-cell number must be present in the blastocyst outgrowth before egg-cylinder formation may occur. (A) Outgrowth of intact blastocyst; (B) Removal of the mural trophectodermal region from blastocysts followed by outgrowth of the resulting ICM/polar fragments; (C) Outgrowth of intact blastocysts followed by removal of trophoblast giant cells prior to egg-cylinder formation (operated outgrowths). Representation of embryonic tissues: trophoblast, white; ICM and embryonic ectoderm, stippled ; endoderm, black.

Trophoblast giant cells were removed from nine attached blastocysts on day 3 of culture. All nine operated outgrowths subsequently reattached and formed giant-cell monolayers, while eight additionally gave rise to egg cylinders. As with ICM/polar fragments, the initiation of egg-cylinder formation was not delayed relative to intact blastocyst controls (Table 1). However, average giantcell numbers, at this time, were significantly less than those found in outgrowths of both intact blastocysts (t = 11·38, P < 0·001) and ICM/polar fragments (t = 4·63, P < 0·001).

It has been suggested that egg-cylinder morphogenesis in vivo depends on a redirection of polar trophectodermal cell movement due to a restriction on the ability of cells to enter the mural trophectodermal region after blastocyst attachment (Copp, 1979). If a similar mechanism underlies egg-cylinder morphogenesis in vitro, it might be expected that the number of giant cells within each outgrowth would become fixed at the time of egg-cylinder formation. This has been confirmed by the present experiments (see Fig. 2). When blastocysts attach in vitro, giant-cell numbers initially rise, presumably as a result of spreading of mural trophectoderm cells, mural cell division and entry into the outgrowth of cells which originate in the polar region. However, 12 h before egg-cylinder formation occurs, there is a cessation of the increase in giant-cell numbers. This phase lasts for about 36 h, after which time giant-cell numbers rise again, probably due to the addition of cells originating in the proximal region of the egg cylinder. It is possible that cells entering the giant-cell monolayer at this time are equivalent to secondary giant cells which arise in vivo from the ectoplacental cone (Duval, 1892). The possibility that the number of cells entering the giantcell monolayer at the time of egg-cylinder formation is balanced by cell death cannot be excluded, but cellular debris was only rarely observed within giantcell monolayers, and outgrowths derived from TVs showed no marked fall in giant-cell numbers (Fig. 2) contrary to expectation if cell death was occurring. It is clear from Fig. 2 that increase in area of giant-cell outgrowth in vitro does not reflect the entry of new trophoblast cells into the giant-cell monolayer. It seems likely that growth of individual giant cells within the monolayer causes the continued increase in area of trophoblastic outgrowths.

Whether giant cells are removed at the blastocyst stage (i.e. the mural trophectoderm) or after blastocyst attachment, the time of initiation of eggcylinder formation is not delayed in the resulting outgrowths. However, there are fewer giant cells in outgrowths derived from ICM/polar fragments than from intact blastocysts, and still fewer in operated outgrowths whose giant cells are removed after attachment. It seems likely, therefore, that the time of eggcylinder formation depends not on the presence of a critical number of giant cells in the monolayer, but rather on some other factor which acts at a particular embryonic age and requires only that trophoblastic giant cells are in contact with a substratum at that time.

It is possible that the in vitro morphogenetic events described in this paper result from progressive changes in the strength of adhesion of trophoblast cells to each other, and to their substratum, during blastocyst outgrowth. These changes presumably reflect alterations in the nature of the trophoblast cell surface which have been described in blastocysts undergoing implantation (see Schlafke & Enders, 1975; Sherman & Wudl, 1976; Jenkinson, 1977; Copp, 1980Z>). The initiation of egg-cylinder formation in vivo, therefore, may depend on changes in adhesion between trophectoderm and uterine epithelial cells as well as clamping of the blastocyst within the uterine luminal crypt (Copp, 1978).

I thank Ginny Papaioannou and Richard Gardner for reading the manuscript. Much of this work was carried out in the Department of Zoology, University of Oxford, and was supported by a Christopher Welch Scholarship and by the Medical Research Council.

Ansell
,
J. D.
(
1975
).
The differentiation and development of mouse trophoblast
.
In Early Development of Mammals
(ed.
M.
Balls
&
A. E.
Wild
), pp.
133
-
144
. Cambridge:
Cambridge University Press
.
Barlow
,
P. W.
&
Sherman
,
M. I.
(
1972
).
The biochemistry of differentiation of mouse trophoblast: studies on polyploidy
.
J. Embryol. exp. Morph
.
27
,
447
465
.
Copp
,
A. J.
(
1978
).
Interaction between inner cell mass and trophectoderm of the mouse blastocyst. T. A study of cellular proliferation
.
J. Embryol. exp. Morph
.
48
,
109
125
.
Copp
,
A. J.
(
1979
).
Interaction between inner cell mass and trophectoderm of the mouse blastocyst. II. The fate of the polar trophectoderm
.
J. Embryol. exp. Morph
.
51
,
109
120
.
Copp
,
A. J.
(
1980a
).
The development of field vole (Microtus agrestis) and mouse blastocysts in vitro: a study of trophoblast cell migration
.
Placenta
,
1
,
47
60
.
Copp
,
A. J.
(
1980b
).
Problems of early embryogenesis
.
In Biochemical Development of the Fetus and Neonate
(ed.
C. T.
Jones
).
Elsevier/North Holland
(in press).
Duval
,
M.
(
1892
).
Le placenta des rongeurs
.
J. Anat. Physiol., Paris
27
,
279
476
.
Gardner
,
R. L.
(
1978
).
Production of chimaeras by injection of cells or tissues into the blastocyst
.
In Methods in Mammalian Reproduction
(ed,
J. C.
Daniel
), pp.
137
165
.
New York
:
Academic Press
.
Gardner
,
R. L.
&
Johnson
,
M. H.
(
1975
).
Investigation of cellular interaction and deployment in the early mammalian embryo using interspecific chimaeras between the rat and mouse
.
In Cell Patterning, Ciba Foundation Symposium
, pp.
183
200
.
Amsterdam
:
Associated Scientific Publishers
.
Gardner
,
R. L.
&
Papaioannou
,
V. E.
(
1975
).
Differentiation in the trophectoderm and inner cell mass
.
In The Early Development of Mammals
, (eds.
M.
Balls
&
A. E.
Wild
), pp.
107
132
Cambridge
:
Cambridge University Press
.
Gwatkin
,
R. B. L.
(
1966
).
Amino acid requirements for attachment and outgrowth of the mouse blastocyst in vitro
.
J. cell Physiol
.
68
,
335
344
.
Hsu
,
Y.-C.
(
1978
).
In vitro development of whole mouse embryos beyond the implantation stage
.
In: Methods in Mammalian Reproduction
, (ed.
J. C.
Daniel
), pp.
229
245
.
New York
:
Academic Press
.
Hsu
,
Y.-C.
,
Baskar
,
J.
,
Stevens
,
L. C.
&
Rash
,
J. E.
(
1974
).
Development in vitro of mouse embryos from the two cell stage to the early somite stage
.
J. Embryol. exp. Morph
.
31
,
235
245
.
Jenkinson
,
E. J.
(
1977
).
The in vitro blastocyst outgrowth system as a model for the analysis of peri-implantation development
.
In: Development in Mammals
, vol.
2
. (ed.
M. H.
Johnson
), pp.
151
172
.
Amsterdam
:
North Holland
.
McLaren
,
A.
&
Hensleigh
,
H. C.
(
1975
).
Culture of mammalian embryos over the implantation period
.
In: The Early Development of Mammals
. (
M.
Balls
&
A. E.
Wild
), pp.
45
60
.
Cambridge
:
Cambridge University Press
.
Nicolson
,
G. L.
,
Yanagimachi
,
R.
&
Yanagimachi
,
H.
(
1975
).
Ultrastructural localisation of lectin-binding sites on the zonae pellucidae and plasma membranes of mammalian eggs
.
J. Cell Biol
.
66
,
263
274
.
Pienkowski
,
M.
,
Solter
,
D.
&
Koprowski
,
H.
(
1974
).
Early mouse embryos: growth and differentiation in vitro
.
Expl. Cell Res
.
85
,
424
428
.
Schlafke
,
S.
&
Enders
,
A. C.
(
1975
).
Cellular basis of interaction between trophoblast and uterús at implantation
.
Biol. Reprod
.
12
,
41
65
.
Sherman
,
M. I.
&
Wudl
,
L. R.
(
1976
).
The implanting mouse blastocyst
.
In The Cell Surface in Animal Embryogenesis and Development
(ed.
G.
Poste
&
G. L.
Nicolson
), pp.
81
125
.
Amsterdam
:
North Holland
.
Snell
,
G. D.
&
Stevens
,
L.C.
(
1966
).
Early embryology
.
In Biology of the Laboratory Mouse
(ed.
F. L.
Green
).
New York
:
McGraw-Hill
.
Whittingham
,
D. G.
&
Wales
,
R. G.
(
1969
).
Storage of two-cell mouse embryos in vitro
.
Aust. J. biol. Sci
.
22
,
1065
1068
.
Wiley
,
L. M.
&
Pederson
,
R. A.
(
1977
).
Morphology of mouse egg cylinder development in vitro: a light and electron microscopic study
.
J. exp. Zool
.
200
,
389
402
.