development of mouse embryos in hanging drop culture

THE ANATOMICAL RECORD 211:48-56 (1985)

Development of Mouse Embryos in Hanging Drop Culture

SANDRA W. POTTER AND JOHN E. MORRIS Department of Zoology, Oregon State Uniuersity, Coruallis, Oregon 97331

ABSTRACT Mouse blastocysts were cultured in hanging drops for up to 6 days in order to study development under conditions that avoid the distortion of embryos typically seen when they are allowed to attach to a glass or plastic surface. The survival rate of embryos in hanging drops was equal to that of embryos attached to culture dishes and superior to that of embryos suspended in gyrating flasks. Devel- opment of the embryonic portion was similar to that in vivo and on culture dishes but slower than in vivo; the egg cylinder stage was reached after 8-10 equivalent gestation days (4 to 6 days in culture), while that stage is reached at 5.5 to 6 days in vivo. The trophectoderm, however, developed in a unique manner. The cells migrated away from the inner cell mass (ICM), similar to embryos on a culture dish, but without a surface on which to spread they clustered distal to the ICM. In vivo, trophectoderm remained covering the ICM. By 5 days in hanging drop culture the embryos had developed a segmented appearance with trophoblast giant cells a t the abembryonic pole, extraembryonic cells not covered by vacuolated endoderm in the central region, and embryonic endoderm surrounding a developing proamniotic cavity in embryonic ectoderm at the embryonic pole. These observations suggest that the trophectoderm is able to follow a developmental program independent of that in the embryonic portion and that its behavior is dominated by the different adhesive properties of the trophoblastic and embryonic cells.

Implantation is crucial to the continued development of a mammalian embryo in vivo, yet little is known of the molecular events on the cell surface during the adhesion and attachment phases of implantation. In studies of blastocyst adhesion to cell monolayers (Sherman and Salomon, 1975; Salomon and Sherman, 1975; Sherman, 1978; Glass et al., 1979; Chavez and Van Blerkom, 1981), the trophectoderm spreads on the substratum and ex- cludes the monolayered cells, so that it is not clear whether initial adhesion is to plastic or to cells. We have previously described a method for studying blastocyst adhesion (Morris et al., 1982, 1983) in which mouse embryos and vesicles of uterine epithelium are able to adhere to each other when cultured together in hanging drops. This technique permits study of the initial stages of adhesion and attachment in a controlled environment without competition from a plastic or collagen substratum.

There is no evidence, however, that blastocysts cultured without contacting a solid surface will develop normally, because almost all detailed studies of postblas- tocyst development in vitro have been made on embryos attached to a solid substratum over which trophoblast cells can spread (for example, Hsu, 1973, 1979; Wiley and Pedersen, 1977; Shalgi and Sherman, 1979; Solter et al., 1974; Pienkowski et al., 1974; Naeslund et al., 1980). Sherman (1978) has cultured blastocysts in contact with agarose surfaces to which they do not attach but on which they continue to develop, but he did not give details of their development to demonstrate whether it was normal.

0 1985 ALAN R. LISS, INC.

Accordingly, we have investigated the ability of suspension techniques to support development and report here that during 4 to 6 days in culture mouse blastocysts in hanging drops survived as well and developed at the same rate as those attached to dishes. Their development during this time in culture corresponded to development in vivo as reported in the literature (Reinius, 1965; Snell and Stevens, 1966; Solter et al., 1970) but was at a slower rate. The clustering of trophoblast cells at one pole provides a new insight into their adhesive properties.

MATERIALS AND METHODS Blastocyst Isolation and Culture

Blastocysts were obtained a t 3.5 days postcoitum from superovulated 8-10-week-old CF-1 mice using standard techniques (Rafferty, 1970). They were pooled, washed once with Hanks’ balanced salt solution (BSS) that was buffered with 0.01 M morpholinopropane sulfonic acid, pH 7.3, and supplemented with 10% fetal bovine serum (FBS). After removing the zonae pellucidae with 0.5% Pronase (Mintz, 1962) in BSS the embryos were washed through several changes of medium. The zonae were removed so that the conditions would be comparable to those required for studies of adhesion (Morris et al., 1982, 1983), and no systematic study was made of blastocysts allowed to hatch during development in the drops. Blastocysts were cultured under the following

Received September 20, 1983; accepted July 20, 1984

MOUSE EMBRYO DEVELOPMENT IN HANGING DROPS 49

conditions: 1) 10-20 embryos were cultured on a gyra- tory shaker (70 rpm) in 2 ml of medium in a 10 ml Erlenmeyer flask (Morris et al., 1983); 2) embryos were suspended individually in up to 15 hanging drops of 10- 15 pl of medium on the under surface of the cover of a 35-mm bacteriological petri dish (Falcon Plastics) over 1-2 ml of medium to maintain humidity (Morris et al., 1982, 1983); or 3) 10-20 embryos were allowed to settle on the surface of a 35-mm tissue culture petri dish (Fal- con Plastics) (e.g., Cole and Paul, 1965; Hsu, 1979; Pien- kowski et al., 1974; Sherman and Salomon, 1975). The flasks were purged with 5% CO2 and 95% air and tightly stoppered, and the dishes for hanging drop or substrate- attached culture were sealed in humidified chambers (Billups-Rothenberg) and purged with the same gas mix- ture. Incubation was at 37°C.

Culture Media The media used were Eagle’s minimum essential me-

dium (MEM), Ham’s F-12, medium 199, NCTC 135 (all from M.A. Bioproducts), and Eagle’s basal medium as modified by Spindle and Pedersen (1973; BME/SP). The BME/SP contained amino acids (from Sigma Chemical Co.), Earle’s salts, and Eagle’s vitamins (from M.A. Bio- products). Unless otherwise indicated, each medium was supplemented with 10% heat-inactivated fetal bovine serum (56”C, 30 min; from M.A. Bioproducts or from Sterile Systems, Inc.), 100 IU/ml of penicillin, and 100 pg/ml of streptomycin. The medium was not changed during the course of the experiments, but changing medium daily in one series of experiments did not increase survival rate or appear to influence development.

Microscopy Flask cultures were examined daily by pipetting the

contents of the flasks to a sterile Maximov slide to deter- mine the numbers of embryos developing and their stage of development. Hanging drops were observed from above through the covers of their culture dishes using long focal length objectives on a Zeiss RA microscope. Cultures attached to dishes were examined with an Olympus IMT inverted microscope. For histological study, embryos were fixed in 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.3) for 15 min to 1 hr, rinsed in phosphate buffer, and postfixed in 1% Os04 in 0.1 M phosphate buffer for 15 min to 1 hr. To facilitate han- dling through dehydration and embedding, the fixed embryos were put in a drop of horse serum which was then polymerized by exposure to glutaraldehyde vapors for several hours. The resultant blocks were washed, dehydrated in an alcohol series, and embedded in plastic (Mix E; Spurr, 1969). Thick sections for light microscopy were cut at 0.5 pm and stained with 1% methylene blue in 1% borax. Photographs were taken using a Wratten 22 orange filter. For transmission electron microscopy (TEM), gold to silver sections were stained with uranyl acetate (Watson, 1958) and lead citrate (Reynolds, 1963) and examined with a Philips 200 electron microscope. For scanning electron microscopy (SEMI, fixed embryos were dehydrated in ethanol and trichloro-trifluoro- ethane, subjected to critical point drying (in a Bomar SPC-900 apparatus), coated with 60% gold40% pallad- ium alloy in a modified vacuum evaporator (Elion DV 502), and examined with an AMR 1200 or AMR 1000 scanning electron microscope.

RESULTS Influence of Culture Conditions on Survival Rate

The ability of freshly isolated blastocysts (3.5 days postcoitum) to develop in suspension culture, in either flasks or hanging drops, was determined by comparing their gross morphology with that of blastocysts cultured attached to dishes. Also, comparisons of embryos in suspension culture were made with published descriptions of embryos developing in vivo and in vitro. Embryos that showed blebs or other abnormalities were tallied as not developing. An embryo attached to a dish was not tallied as developing unless both trophectoderm and inner cell mass (ICM) were observed.

In preliminary experiments we examined the development of embryos in a variety of media in gyrating flasks as a method for suspension culture. Rarely did any embryos survive beyond 4 days with 10% fetal bovine serum in NCTC 135, Ham’s F12, or medium 199 while those in BME/SP and MEM showed continued development. Bovine serum albumin (BSA; 4 mg/ml) could not substitute for complete serum in suspension cultures. Embryos in BSA developed very slowly and only 13% survived after 4 days. When both serum and BSA were omitted from the medium no embryos survived beyond the second day. We also found considerable variability between batches of serum in supporting development, and, therefore, we had to test each batch before using it routinely in cultures.

The survival rate of embryos in flasks was lower than that of those attached to a substratum. Blastocysts suspended in hanging drops of BME/SP showed rates of survival superior to those in flasks and equivalent to those attached to a surface (Fig. 11, at least through 5 days, the limit of our comparison. Consequently, all detailed studies made on embryos in suspension culture were in hanging drops and in BME/SP with serum.

3 4 5

Days in Culture Fig. 1 . Percentage of embryos surviving in hanging drops (H), at-

tached to petri dishes (A), and in rotating flasks (R) in BME/SP plus 10% serum. Number of embryosinumber of experiments: H, 254/11; A, 8617; R, 23619. Variation lines indicate standard error.

50 S.W. POTTER AND J.E. MORRIS

Morphology After 2 days the blastocoel was no longer fully expanded and 30% of the embryos appeared compact (Ta- ble 1, Figs. 2b, 3b). Although compact and smaller in diameter, the embryos still showed large trophectoderm cells with more obvious microvilli over the entire sw- face (Fig. 3b).

After 1 day (i.e., on day 2) in hanging drop culture most embryos were expanded blastocysts (Table 1, Fig. 2a). With SEM, the cells appeared essentially uniform in size and were covered with short microvilli (Fig. 3a).

TABLE 1. Percentage of mouse blastocysts reaching various stages of development when cultured in hanging drops 4 days

Days in Expanded Collapsed Early egg Abnormal culture blastocyst (compact sphere) Elongated cylinder and dead

1 991 1 0 0 0 2 62 30 4 0 4 3 13 35 36 8 8 4 0 31 23 25 21

'The percentage at each stage of development was calculated from 120 embryos in four experiments of 30 embryos each. Each embryo was cultured singly in a 10-15 p1 drop of BMEi SP plus 10% FBS. Drops were examined daily with covers in place to avoid disturbing them.

Fig. 2. Typical development of embryos in suspension culture (a-e) and attached to petri dish surface (f,g). Scale bars = 100 pm; a-c, d-g to same scale. a) Expanded blastocysts after 1 day in culture. The t,rophect,oderm cells, hlast,woel, and ICM are typical of those in vivo and cultured in dishes. Bright field. b) Collapsed blastocyst stage, 2 days. The blastocoel is no longer fully expanded. Embryos in dishes were attaching a t this stage. Phase contrast. c) Elongating embryo, 2.5-3 days. The mural trophectoderm cells are grouped opposite the embyronic pole. Phase contrast. d) Early egg cylinder, 3.5-4 days. The embryonic region shows endoderm and ectoderm. The extraembryonic region may be seen between ectoderm and large trophectoderm cells.

Phase contrast. e) Advanced egg cylinder, 5 days. This embryo was photographed obliquely from the embryonal end so all trophectoderm cells are not visible. Bright field. f7 Attached, 3 days. Trophectoderm cells have spread on the surface and contain numerous vesicles. Simi- lar vesicles can be seen in trophectoderm cells of embryos in suspension culture. The ICM with distinct endoderm and ectoderm is discernible. Phase contrast. g) Attached, 4.5 days. Note the resem- blance to d above, except that trophectoderm has spread on the surface here. Phase contrast. B, blastocoel; I, inner cell mass; N, endoderm; E, embryonic ectoderm; X, extraembryonic region; T, trophectoderm.


Fig. 3. SEM photographs showing differences in embryo size, cell size, and cell surfaces with development in hanging drop culture. Scale bars = 10 pm. a) Expanded blastocyst, 1 day; trophectoderm cells are large with relatively small microvilli. b) Collapsed blastocyst, 2 days; the surface of the trophectoderm cells protrudes slightly and has more microvilli. Note change in size of blastocyst compared to a. c) Elongat- ing embryo, 2.5-3 days; trophectoderm cells cover entire surface of embryo. d,e,f) Egg cylinder stage embryos, 4-5 days, with distinct

segmentation. In f , trophectoderm cells no longer cover the embryonic region so that distinct areas can be distinguished: trophectoderm cells, a central region, and, at the embryonic end, endodermal cells. Trophec- toderm cells are large and bulging with numerous microvilli while endoderm cells are smaller with many fine microvilli. It is difficult to distinguish individual cells in the central region. T, trophectoderm; Em, embryonic region; EPC, ectoplacental cone region.

After 3 days, larger trophectoderm cells could be distinguished at one end of the embryo, opposite the smaller cells of the ICM (Figs. tLc, 3c). About one third of the embryos in hanging drops were elongated (Table l), and embryos cultured on the surface of petri dishes were at a similar stage of development (Fig. 20, although trophectoderm cells of the latter had spread over the surface of the dish. Embryos of comparable age in vivo

(Snell and Stevens, 1966) are beyond the egg cylinder stage with the ICM beginning to differentiate into ectoderm, endoderm, and extraembryonic ectoderm.

After 4 days in culture 25% of the embryos in hanging drops had developed a segmented appearance (Table 1, Figs. 2d, 3d). It was possible to distinguish two cell layers, which probably were endoderm and ectoderm on the basis of their position (Fig. 2d) and their histology


(see below). About one third of the embryos were still at the compact sphere stage (Table l), despite the fact that they otherwise appeared to be healthy. When trans- ferred to dishes, such embryos had attached and continued development when examined the next day. These secondarily attached embryos were not followed further. Embryos surviving 5 days in hanging drop culture typically developed a more distinct segmentation (Figs. 2e, 3d-0. After 4 to 5 days trophectoderm cells no longer covered the ICM, leaving the underlying smaller endodermal cells exposed (Figs. 2d,e, 30. Trophectoderm cells were three to four times larger in diameter than the endodermal cells. Despite the flattening of the trophectoderm cells when cultured on the surface of the dish, the embryonic regions were similar to those in suspension (compare Fig. 2d and g).

Histology After 1 day in hanging drop culture the cells of the

trophectoderm were uniformly thin around the blastocoel (Fig. 2a). As development continued the trophectoderm cells thickened and vacuoles were visible in their cytoplasm (Fig. 4a,b), as also occurs on culture dishes (Wiley and Pedersen, 1977; Chavez and Van Blerkom, 1981); this was first noticeable in cells a t the abembryonic pole. Differentiation of cells of the ICM into an inner layer and a thicker and vacuolated outer layer of cells was discernible (Fig. 4b). The relative position, size, and appearance of these cell layers indicate that they are, respectively, presumptive ectoderm and endoderm (Snell and Stevens, 1966; Wiley and Pedersen, 1977). Parietal (or distal) endoderm cells were identified by their position on the inner surface of the mural trophectoderm (Fig. 4b).

The blastocoel was reduced in size in compact embryos that had been in culture for 2 days (Fig. 4c). Larger cells of the trophedoderm could be distinguished from slightly smaller cells of the ICM in which vacuolations were seen in some outer endodermal cells (Fig. 4c).

In hanging drops, after 3 days in culture only a small remnant of the blastocoel was detectable in the elongated embryo (Fig. 4d). Cells of the mural trophectoderm were clustered at the abembryonic end while the polar trophectoderm cells formed a single flattened layer over the ICM (Fig. 4d). More densely staining, vacuolated endoderm cells were recognizable on either side of the blastocoel.

Embryos in culture 4 or 5 days developed an obvious segmented appearance (Fig. 4e,0. When sectioned, the bulging cells of the abembryonic portion were seen to have large nuclei, with nuclear diameters averaging two to three times those in the embryonic portion. This was taken as evidence of their transformation to trophoblastic (primary) giant cells. The cells in the central portion probably represented polar trophectoderm that had migrated to this position and thus can be considered extraembryonic cells. Others have deduced a similar migration of polar trophectoderm from observations of embryos attached to dishes in vitro (Wiley and Pedersen, 1977; Hsu, 1979) and marking experimcnts of cmbryos in vivo (Copp, 1979). In some instances endoderm was exposed (Fig. 40 and in others there appeared to be some thinly stretched cells, probably polar trophectoderm, still covering endoderm (Fig. 4e). In most embryos the con- nection between embryonic and extraembryonic regions

became a relatively thin neck of cells (Figs. 2e, 40, but in other instances the two regions were completely sev- ered at this point. A proamniotic cavity (Figs. 4f, 5c) indicated that they had reached the equivalent of at least 5 days development in vivo (Snell and Stevens, 1966).

Although not done routinely, several embryos were followed in culture for 6 days. These embryos showed additional growth and development (Fig. 5a-c). The seg- ment closest to trophoblast cells was composed of elongated central cells arranged concentrically and a single layer of flattened cells at the outer surface (Fig. 5a,b). The central cells had large nuclei, with relatively little cytoplasm, and were loosely arrayed. These characteris- tics were consistent with descriptions of ectoplacental cone cells (Reinius, 1965; Snell and Stevens, 1966; Wiley and Pedersen, 1977). We wish to emphasize, however, that since identification of the central cells was based on histological criteria it should be considered tentative until verified by cell lineage studies in which polar trophectoderm cells are marked and traced during development. The central cells were not covered with endoderm in embryos in hanging drop culture (Figs. 4f, 5a-c). In vivo, ectoplacental cone cells were not covered by endoderm (Snell and Stevens, 1966), in contrast to embryos cultured attached to dishes in which Wiley and Pedersen (1977) noted that ectoplacental cone cells were covered with endoderm. Attachment of ectoplacental cone cells to trophoblast cells was made by a relatively small number of cells (Fig. 5a). The portion distal to the ectoplacental cone was elongated (Fig. 5b,c) and showed interior embryonic ectoderm cells with a central proamniotic cavity (Fig. 5c). These cells were covered by vacuolated endoderm cells, just as in embryos cultured for 5 days in hanging drops. Elongation and slight segmentation of the endoderm-covered region suggested the differentiation of extraembryonic ectoderm (Figs. 3f, 5b,c), but definite identification of these cells cannot be made without further study.

DISCUSSION We have shown that when the mouse blastocysts are

cultured in hanging drops to the early egg cylinder stage the embryonated portion (ICM) can grow and develop with a morphology that corresponds to normal development for at least 6 days of culture. The trophectoderm, by contrast, displays a developmental path that is different from that in vivo and in cultures on dishes.

It is clear that during this period, attachment to a substratum was not required for the normal development of the ICM in those embryos that survived. The percentage of embryos cultured in flasks surviving beyond 2 days was much lower than that of embryos in hanging drops or attached to a substratum (present study and Buckley et al., 1978; Hsu, 1973; Hsu et al., 1974; McLaren and Hensleigh, 1975). It is possible that the rotating motion of the flasks coupled with the fragile structure of 4 and 5-day embryos (Figs. 2e, 40 resulted in their destruction.

Fetal bovine serum was necessary for the development of embryos in hanging drop culture. Bovine serum albumin was not able to substitute fully for fetal bovine serum, as had been also found by McLaren and Hen- sleigh (1975) for blastocysts attached to dishes. They found that with BSA blastocysts hatched but remained


Fig. 4. Histological sections of embryos cultured in hanging drops; a<, thick sections. Bars = 10 pm. a) Expanded blastocyst, 1 day. Both mural and polar trophectoderm cells are stretched and flattened around the blastocoel; early differentiation of endoderm and ectoderm is ap- parent. b) Expanded blastocyst, 1-2 days. Visceral endoderm, embryonic ectoderm, parietal endoderm, and mural and polar trophectoderm are visible. c) Collapsed blastocyst, 2 days. Both mural and polar trophectoderm cells are rounded and only a small blastocoel remains. Part of a cocultured epithelial vesicle (ep) lies next to the embryo. d) Elongated embryo, 2.5-3 days. Polar trophectoderm cells still cover ICM; endoderm and ectoderm can be distinguished but blastocoel is nearly obliterated. e) Early egg cylinder embryo, 4 days. A thin layer

of polar trophectoderm cells covers ICM. Because of the embryo’s curvature this section is not completely midsagittal. D Egg-cylinder embryo, 4.5-5 days, TEM. Endoderm, ectoderm, proamniotic cavity, and the extraembryonic region (not covered with vacuolated endoderm) are shown. Trophectoderm portion of the embryo is not shown. Endo- derm cells have many microvilli and vacuoles and in places (arrow) are loosely attached to embryonic ectoderm. B, blastocoel; Pr, polar trophectoderm; MT, mural trophectoderm; N, endoderm; E, embryonic ectoderm; VN, visceral endoderm; PN, parietal endoderm; T, trophectoderm; X, extraembryonic region; P, proamniotic cavity; I, inner cell mass.


Fig. 5. Histological sections of embryos cultured 6 days in hanging drops. Bars = 10 jm. a) Only part of the embryo appears in this section because of its curvature. Concentrically arrayed cells of ectoplacental cone region and large trophectoblast cells are shown. Note limited amount of contact between the two regions. b,c) Semiserial sections lateral (b) and medial (c). Note proamniotic cavity (PI, ectoplacental

cone region, and endodermal cells covering elongated embryonic ectoderm region (b,c); slight segmentation visible (*I may indicate early differentiation of extraembryonic ectoderm (see also Fig. 30. T, trophectoderm; EPC, ectoplacental cone region; E, embryonic ectoderm; N, endoderm.

as expanded blastocysts with no further development. In our hanging drop cultures with BSA, the blastocysts showed some development but at a much slower rate and with lower survival than in the presence of serum. Longer development might be possible using human cord serum (Hsu, 1973, 1979; Hsu et al., 1974; Juurlink and Federoff, 1977) or rat serum (Wu et al., 1981) with or without fetal bovine serum, but these were not tested. Juurlink and Federoff (1977) found that no embryos developed beyond the egg cylinder stage when fetal bovine serum was used for 4 days before adding human cord serum.

The morphology of embryos that had developed in hanging drop culture resembled that of embryos developed attached to dishes (Hsu, 1973; Pienkowski et al., 1974; Wiley and Pedersen, 1977) and that of embryos in vivo (Snell and Stevens, 1966; Poelmann, 1975; Shalgi and Sherman, 1979) except in one detail. In suspension culture the trophectoderm cells remained clustered distal to the embryo proper, while in dishes the trophectoderm cells spread out on the surface of the culture dish (Hsu, 1973; Wiley and Pedersen, 1977; see also Fig. lf,g). Despite this difference the trophectoderm cells retained their full potential for attachment; embryos that had

been cultured in suspension for 3 and 4 days were able to attach to dishes within an hour.

Cell types were identified by comparison of the structure and position of cells in embryos cultured in hanging drops in vivo and in vitro. Embryos cultured in hanging drops progressed from expanded blastocysts to compact embryos to those that grew and developed a form resem- bling that described for egg-cylinder stage embryos (Snell and Stevens, 1966). Large cells clustered at one end of the embryo were identified as primary trophoblastic giant cells based on their large nuclear and cell size (Snell and Stevens, 1966). Cells in the central region were tentatively identified as ectoplacental cone cells that had developed from polar trophectoderm that had migrated to this central position (Snell and Stevens, 1966; Wiley and Pederson, 1977; Copp, 1979; Hsu, 1979). This identification must be considered tentative until cell lineage studies of the trophectoderm are done. Cells of the adjacent embryonic portion were identified as vacuolated visceral endoderm surrounding interior ectoderm, containing a proamniotic cavity (Reinius, 1965; Snell and Stevens, 1966; Solter et al., 1970; Juurlink and Federoff, 1977; Wiley and Pedersen, 1977). Exami- nation of embryos cultured in hanging drops for 6 days


suggested the possible development of extraembryonic edoderm but further studies are necessary to confrm this.

Several aspects of behavior of the differentiating trophoblast giant cells provide an insight into their adhesive specificity. 1) In vivo these cells phagocytose sloughed uterine epithelium (Finn and Hinchliffe, 1968; Smith and Wilson, 1971). 2) In hanging drops they phagocytose degenerating embryos and other moribund cells that may be present (Morris et al., 1982, 1983). 3) On dishes they spread, which may be behaviorally identical to cells engaged in phagocytosis (Grinnell, 1984). 4) In hanging drops the differentiating trophoblast cells group together in a loose, grape-like cluster at the abembryonic pole (present work). 5) This cluster becomes in- creasingly separate from the embryonated region, joined only by a narrow isthmus of extraembryonic or ectoplacental cone cells (present work). 6) In both hanging drops and on dishes the polar trophectoderm tends to migrate away from the embryonic surface to expose the underlying endoderm (present work; Wiley and Pedersen, 1977). Embryos in utero are continually covered with trophectoderm and a thin layer of parietal endoderm cells (Snell and Stevens, 1966). Because the relative position of cells in the early cleavage-stage mouse embryo determines whether the cells become trophectoderm or ICM, it is clear that interaction between these cells is essential in early development (Gardner and Johnson, 1972; Adamson and Gardner, 1979). However, a t the time that trophoblast giant cell differentiation into phagocytic cells begins, these interactions appar- ently cease. The trophectoderm shows preferred adhesion to a nonliving substratum, whether it is an inert surface like a culture dish or dead tissue (Morris et al., 19831, and the phagocytic trophoblast cells show this property even more strongly. The tendency for these cells to cluster at one pole and pull away from the embryo in hanging drops and for the polar trophectoderm to migrate in the direction of the cluster suggests that in the absence of a preferred substratum, homo- typic adhesion between the trophoblastic cells is allowed to dominate. Following this line of reasoning, the fact that the trophectoderm does not migrate away from the embryonic surface in vivo may be explained by the fact that the embryo is exposed equally on all surfaces to sloughing uterine epithelium and to underlying acellu- lar stromal matrix. Thus, it is likely that the distinctly different behavior of the trophectoderm and its inde- pendence from the behavior of the embryonic portion in vivo, in hanging drops, and on a culture dish may be attributable simply to the differences in availability of preferred substratum for the trophoblast giant cells. This hypothesis leads to the testable prediction that surrounding the embryo with an adhesive substratum should prevent the migration of trophectoderm in vitro.

Other groups have used culture systems in which embryos do not attach to a substratum, but in most instances such culture was incidental to other research and not investigated systematically. Postimplantation rat embryos dissected free of membranes (7.5 to 11.5 days of gestation) have been cultured in suspension in roller bottles with 100% rat serum (New et al., 1973; Buckley et al., 1978). The authors report better development in roller bottles than in watch glass culture. Sher- man (1978) has cultured mouse blastocysts for about 6

days (11 equivalent gestation days) over agarose pads to delay the attachment to a substratum and did not report abnormalities of development during this period. Fi- nally, Monk and Petzoldt (1977) used hanging drop cultures for mouse blastocysts as a means of coordinately inhibiting attachment and the appearance of develop- mentally linked lactic dehydrogenase (LDH-5) in the ICM. Whether or not their embryos were healthy is not clear, since Spielmann et al. (19781, using cultures in dishes, showed that developmental changes in LDH oc- curred even in blastocysts that failed to develop morpho- logically or hatch from their zonae pellucidae.

We conclude that under several conditions known to support the development of mouse blastocysts attached to substrata, development will also occur in suspension culture. While the survival rate of blastocysts in hanging drops was comparable to that of those attached to a substratum, that of blastocysts in gyrating flasks was much lower. We are unable to say whether the poorer development in flasks was due to the constant motion and fragility of the embryos, the continued exchange of medium, or the lack of a substratum. Blastocysts in hanging drops may have been supplied with an effective “substratum” by serum proteins denatured at the medium-gas interface on which they rested (Giaever and Kesse, 1983), although we did not observe any spreading of cells on this interface. We showed previously that coculture of blastocysts and uterine epithelium in hanging drops appears to offer a promising technique for studying cellular interactions at implantation without the difficulties of interpretation inherent in culture systems that employ competing artificial substrata (Morris et al., 1982, 1983). It is now clear that culture of blastocysts in hanging drops, at least for 6 days, is an impor- tant alternative to culture in dishes when it is desirable to prevent trophoblast spreading. Because single embryos in individual drops can be observed with standard optics the technique also should offer advantages for the repeated monitoring of specific embryos in studies of the effects of drugs, hormones, or other agents on development.

ACKNOWLEDGMENTS This work was supported by grants to S.W.P. from NIH

Biomedical Research Support Grant RR07079 and to J.E.M. from the Population Council (B77.70X) and the National Institutes of Health (HD 14071). We wish to thank Suzi Sargent and Julie Valenter for typing the manuscript, A1 Soeldner for his help with scanning electron microscopy, and USDA Forest Service Pacific Northwest Forest and Range Experiment Station for use of the Philips 200 and AMR 1200 microscopes.

LITERATURE CITED

Adamson, E.D., and R.L. Gardner (1979) Control of early development.

Buckley, S.K.L., C.E. Steele, and D.A.T. New (1978) In vitro develop- Br. Med. Bull., 35113-119.

ment of early postimplantation rat embryos. Dev. Biol., 65:396- 403.

Chavez, D.J., and J. Van Blerkom (1981) In vitro attachment and outgrowth of mouse trophectoderm. In: Cellular and Molecular Aspects of Implantation. S.R. Glasser and S.W. Bullock, eds. Plenum Publ. Corp., New York, pp. 457-460.

Cole, R.J., and J. Paul (1965) Properties of cultured preimplantation mouse and rabbit embryos, and cell strains derived from them. In: Preimplantation Stages of Pregnancy. G.E.W. Wolstenholme and M. OConnor, eds. Churchill, London, pp. 82-112.


Copp, A.J. (1979) Interaction between inner cell mass and trophectoderm of the mouse blastocyst. 11. The fate of the polar trophectoderm. J. Embryol. Exp. Morphol., 51:109-120.

Finn, C.A., and J.R. Hinchliffe (1968) Histological and histochemical analysis of the formation of implantation chambers in the mouse uterus. J. Reprod. Fertil. 9r301-309.

Gprdner, R.L., and M.H. Johnson (1972) An investigation of innercell mass and trophoblast tissues following their isolation from the mouse blastocyst. J. Embryol. Exp. Morphol., 28:279-312.

Giaever, I., and C.R. Keese (1983) Behavior of cells at fluid interfaces. Proc. Natl. Acad. Sci. U.S.A., 80:219-222.

Glass, R.H., S.I. Spindle, and R.A. Pedersen (1979) Mouse embryo attachment to substratum and interaction of trophoblast with cultured cells. J . Exp. Zool., 208r327-335.

Grinnell, F. (1984) Fibroblast spreading and phagocytosis: Similar cell responses to different-sized substrata. J. Cell. Physiol., 119:58-64.

Hsu, Y:C. (1973) Differentiation in vitro of mouse embryos to the stage of early somite. Dev. Biol., 33:403-411.

Hsu, Y X . (1979) In vitro development of individually cultured whole mouse embryos from blastocyst to early somite stage. Dev. Biol., 68t453-461.

Hsu, YX., J. Baskar, L.C. Stevens, and J.E. Rash (1974) Development in vitro of mouse embryos from the two-cell stage to the early somite stage. J. Embryol. Exp. Morphol., 31:235-245.

Juurlink, B.H., and S. Federoff (1977) Effects of culture milieus on the development of mouse blastocysts in vitro. In Vitro 13390-802.

McLaren, A., and H.C. Hensleigh (1975) Culture of mammalian embryos over the implantation period. In: Early Development of Mam- mals. M. Balls and A.E. Wild, eds. Cambridge University Press, Cambridge, pp. 45-60.

Mintz, B. (1962) Experimental study of the developing mammalian egg: Removal of the zona pellucida. Science, 138:594-595.

Monk, M., and U. Petzoldt (1977) Control of inner cell mass development in cultured mouse blastocysts. Nature, 265338-339.

Morris, J.E., S.W. Potter, and P.M. Buckley (1982) Mouse blastocysts and uterine epithelia show implantation interactions in culture. J. Exp. Zool., 222:195-198.

Morris, J.E., S.W. Potter, L.S. Rynd, and P.M. Buckley (1983) Adhesion of mouse blastocysts to uterine epithelium in culture: A require- ment for mutual surface interactions. J. Exp. Zool., 225467-479.

Naeslund, G., 0. Lundkvist, and B.O. Nilsson (1980) Transmission electron microscopy of mouse blastocysts activated and growth- arrested in vivo and in vitro. Anat. Embryol., 159r33-48.

New, D.A.T., P.T. Coppola, and S. Terry (1973) Culture of explanted rat embryos in rotating tubes. J. Reprod. Fertil., 351355138.

Pienkowski, M., D. Solter, and H. Koprowski (1974) Early mouse embryos: Growth and differentiation in vitro. Exp. Cell Res., 85:424- 428.

Poelmann, R.E. (1975) An ultrastructural study of implanting mouse

blastocysts: Coated vesicles and epithelium formation. J. Anat., 119:421-434.

Rafferty, K.A. Jr. (1970) Methods in Experimental Embryology of the Mouse, Johns Hopkins Press, Baltimore, pp. 58-59.

Reinius, S. (1965) Morphology of the mouse embryo, from the time of implantation to mesoderm formation. Z. Zellforschung, 68:711-723.

Reynolds, E.S. (1963) The use of lead citrate at high pH as an electron- opaque stain in electron microscopy. J. Cell Biol., 17:208-212.

Salomon, D.S., and M.I. Sherman (1975) Implantation and invasive- ness of mouse blastocysts on uterine monolayers. Exp. Cell Res., 90:261-268.

Shalgi, R., and M.I. Sherman (1979) Scanning electron microscopy of the surface of normal and implantation-delayed mouse blastocysts during development in vitro. J. Exp. Zool., 2lOt69-80.

Sherman, M.I. (1978) Implantation of mouse blastocysts in vitro. In: Methods in Mammalian Reproduction. J.C. Daniel Jr., ed. Aca- demic Press, New York, pp. 247-257.

Sherman, M.I., and D.S. Salomon (1975) The relationships between the early mouse embryo and its environment. In: The Developmental Biology of Reproduction. C.L. Markert and J. Papaconstantinou, eds. Academic Press, New York, pp. 277-309.

Smith, M.S.R., and I.B. Wilson (1971) Histochemical observations on early implantation in the mouse. J. Embryol. Exp. Morphol., 26165-174.

Snell, G.D., and L.C. Stevens (1966) Early embryology. In: Biology of the Laboratory Mouse. E.L. Green, ed. McGraw-Hill Book Co., New York, pp. 205-246.

Solter, D., W. Biczysko, M. Pienkowski, and H. Koprowski (1974) Ultra- structure of mouse egg cylinders developed in vitro. Anat. Rec., 180:263-280.

Solter, D., I. Damjanov, and N. Skreb (1970) Ultrastructure of the mouse egg-cylinder. Z. Anat. Entwicklungsgesch., 132t291-298.

Spielman, H., H.-G. Eibs, U. Jacob-Muller, and R. Bischoff (1978) Expression of lactate dehydrogenase isozyme 5 (LDH-5) in cultured mouse blastocysts. Biochem. Gen., 16:191-202.

Spindle, A.I., and R.A. Pedersen (1973) Hatching, attachment, and outgrowth of mouse blastocysts in vitro: Fixed nitrogen require- ments. 6. Exp. Zool., 186305-318.

Spurr, A.R. (1969) A low-viscosity epoxy resin embedding medium for electron microscopy. J. Ultrastruct. Res., 26t31-43.

Watson, M.L. (1958) Staining of tissue sections for electron microscopy with heavy metals. J. Biophys. Biochem. Cytol., 4:475-478.

Wiley, L.M., and R.A. Pedersen (1977) Morphology of mouse egg cylinder development in vitro: A light and electron microscopic study. J . Exp. Zool., 200:389-402.

Wu, T.-C., Y.-J. Wan, and I. Damjanov (1981) Rat serum promotes the in vitro development of mouse blastocysts during early somitic stages of embryogenesis. J. Exp. Zool., 217t451-453.

development of mouse embryos in hanging drop culture

Documents