1. Inner cell mass (ICM) and trophoblast were isolated from -day post-coitum mouse blastocysts by microsurgery.

  2. Trophoblastic fragments formed vesicles in culture but did not aggregate with other such fragments. They proved as effective as intact blastocysts in inducing decidua in recipient uteri, but thereafter failed to proliferate.

  3. Isolated ICMs remained as solid balls of cells that readily aggregated in pairs or groups in culture but failed to induce implantation changes in receptive uteri.

  4. Possible explanations for the failure of isolated trophoblast to proliferate after implantation are discussed. It is argued that presence of ICM tissue is necessary for trophoblast proliferation, and suggested that the ICM exerts its effect by controlling development of the ectoplacental cone.

One of the earliest differences to arise between cells of the mammalian embryo is evident at the blastocyst stage with formation of the trophectoderm and inner cell mass (ICM). Attention has been devoted to establishing whether this differentiation is determined in the egg prior to division (Dalcq, 1957; Seidel, 1960; Mulnard, 1965) or by the relations between naïve blastomeres during cleavage (Mintz, 1965; Tarkowski & Wroblewska, 1967; Graham, 1971). The question of the particular properties that serve to distinguish the two tissues in the early blastocyst has so far been largely neglected.

Earlier studies involved disaggregation of morulae and blastocysts of both the rabbit and mouse, and culture of the cells isolated from them (Cole & Paul, 1965; Cole, Edwards & Paul, 1965). Since no attempt was made to separate ICM and trophoblast tissue prior to disaggregation, the cells could not be assigned unequivocally to one or other type.

The present study was thus aimed at separating the blastocyst into its constituent tissues, and comparing their properties both in culture and after transfer to the uterus. This project promised to provide criteria for distinguishing these tissues in investigations into the mechanism of their differentiation, and to clarify their roles in implantation and interrelations in later development. The mouse was chosen because it is the species used in most studies of early mammalian development and because its blastocyst is very resilient to microsurgery (Gardner, 1968, 1971, 1972; Lin, 1969). A preliminary report of part of this work has been published elsewhere (Gardner, 1971).

Previous accounts of the handling and transfer of so-called ‘pure’ trophoblast of murine origin have dealt exclusively with the ectoplacental cone of the implanted egg-cylinder (Grobstein, 1950; Billington, 1965; Simmons & Russell, 1966; Clarke, 1969). Despite the enveloping cell layer of the unimplanted blastocyst being called trophoblast its relationship with the ectoplacental cone is uncertain. This note of caution is justified by the following results.

Preparation of donor females and recovery of blastocysts

Adult or 3-to 4-week-old PDE random-bred or Er inbred albino mice were superovulated by intraperitoneal injection of pregnant mares’ serum gonadotrophin (PMSG -Gestyl, Organon) followed 48 h later by human chorionic gonadotrophin (HCG-Pregnyl, Organon). Five i.u. of each hormone were given to adults and 10 i.u. to the immature female mice. PDE females were paired with males of the same or CBA/H-T6 inbred strain, and Er females exclusively with CBA/H-T6 males after administration of HCG. Females with plugs were killed and their excised uterine horns flushed for blastocysts either between 96–102 or 117–122 h after HCG. Since no consistent differences were noted between material obtained at opposite extremes of each time range, blastocysts recovered during the former interval (and tissue derived from them) will be called -day and those recovered during the latter interval will be called -day blastocysts or fragments.

Conditions for storage, manipulation and culture of blastocysts and fragments

The special conditions imposed by microsurgery required modified methods for keeping blastocysts in vitro. The medium could not be maintained in equilibrium with an atmosphere of 5% CO2 in air. Also, attempts to perform manipulations at 37°C were unsuccessful because of the great increase in adherence of cells and debris to the instruments at higher temperature. Manipulations were thus carried out at room temperature in medium M 199 plus 10% inactivated foetal calf serum (Microbiological Associates Inc., or Flow Laboratories, Scotland), which contained 100 i.u./ml of sodium benzyl penicillin and was buffered at pH 7·0 with 5–8% (v/v) Sorensen’s phosphate buffer (M/ 15). To avoid unnecessary osmotic shocks or changes in pH this medium was used throughout. Under these conditions intact blastocysts remained viable for up to 11 h at room temperature (Table 1).

Table 1

Viability of blastocysts transferred after 812-11 h at room temperature in phosphate-buffered M199 + 10% serum medium

Viability of blastocysts transferred after 812-11 h at room temperature in phosphate-buffered M199 + 10% serum medium
Viability of blastocysts transferred after 812-11 h at room temperature in phosphate-buffered M199 + 10% serum medium

Microsurgery

A Leitz micromanipulator was used. Blastocysts were placed in a drop of medium hanging from the coverslip of a chamber (Puliv, Leitz) filled with heavy liquid paraffin. Microsurgery was carried out at × 125, × 300 or × 500 using bright-field or Heine phase-contrast observation in a Leitz Laborlux microscope.

Trophoblast tissue was obtained free from ICM as follows. Expanded -day blastocysts surrounded by intact zonae were orientated and held by suction pipette against the underside of the coverslip of the hanging drop (Fig. 1). A piece of the cutting edge of a safety razor-blade attached to a mounted needle with

Fig. 1

Schematic side view of hanging drop in manipulation chamber to show arrangement of blastocyst and fragment of razor blade for sectioning blastocyst to obtain pure trophoblast, c.s. = coverslip, i.c.m. = inner cell mass, r.b. = fragment of razor blade, s.p. = suction pipette.

Fig. 1

Schematic side view of hanging drop in manipulation chamber to show arrangement of blastocyst and fragment of razor blade for sectioning blastocyst to obtain pure trophoblast, c.s. = coverslip, i.c.m. = inner cell mass, r.b. = fragment of razor blade, s.p. = suction pipette.

Araldite (CIBA) was arranged vertically on one manipulator unit so that its cutting surface was parallel with the coverslip of the chamber (Fig. 1). By raising the blade slowly the living blastocyst could be severed, parallel to the surface of the ICM, either equatorially (to yield equatorial trophoblastic fragments, Fig. 2 A) or close to the embryonic pole (to yield maximal trophoblastic fragments, Fig. 2B). Sectioning blastocysts through the equator presented no difficulty. The embryo tended to rotate away from the advancing blade when it was cut towards one pole. This could be prevented by increasing suction or by using a needle to help immobilize the embryo. Provided the fragment of razorblade had been cleaned, siliconed and correctly aligned, blastocysts could be severed with only occasional recourse to the needle to clear one or other fragments from the blade. The choice of well-expanded blastocysts enabled trophoblast to be obtained with little risk of inclusion of ICM cells.

Fig. 2

Scheme of manipulations carried out on the mouse blastocyst to obtain trophoblast and ICM tissue. □, Trophoblast tissue; ◼, inner cell mass tissue. (From Gardner, 1971, with permission of F. Vieweg & Sohn.)

Fig. 2

Scheme of manipulations carried out on the mouse blastocyst to obtain trophoblast and ICM tissue. □, Trophoblast tissue; ◼, inner cell mass tissue. (From Gardner, 1971, with permission of F. Vieweg & Sohn.)

Preliminary attempts to isolate ICM tissue from blastocysts by various chemical and enzymatic treatments were unsuccessful. ICMs were thus obtained by penetrating immobilized expanded blastocysts from opposite sides with a pair of fine, siliconed glass needles, tearing the trophoblast open and pinning it out as a sheet against the coverslip. The exposed ICM was then gently scraped from the trophoblast, but was only used if the sheet of overlying trophoblast left behind was intact. It often proved possible to isolate ICMs relatively free from cells of the enveloping trophoblast. However, the procedure extensively damaged the corresponding trophoblast, so comparison between trophoblast and ICM tissues from the same blastocyst was not possible.

Culture of trophoblastic fragments, isolated ICMs and intact blastocysts

Trophoblastic fragments and intact blastocysts were cultured either in hanging drops in the oil chambers used for microsurgery, or in microdrops on siliconed coverslips placed in culture dishes (Falcon Plastic, 60 × 15 mm) filled with heavy liquid paraffin. Isolated ICMs were generally very small and delicate and were therefore left in the manipulation chambers for culture. Fusion was encouraged by pairing fragments with a blunt glass needle, and with the warm stage of the microscope set at 37°C. This allowed firm contact to be made between fragments before they were moved to the incubator.

Cell counts on blastocysts and isolated ICMs

The material was prepared according to the air-drying method of Tarkowski (1966).

Transfer of tissue and blastocysts to the uteri of recipient mice

The -day blastocysts, trophoblastic fragments or isolated ICMs were transferred to the uteri of mice on the third day of pseudopregnancy, except in one case where a group of trophoblastic fragments were transferred to a recipient on the fourth day. The recipients were albino PDE or Er mice mated with vasectomized males. All ICMs and some trophoblastic fragments were transferred unilaterally, and intact blastocysts introduced into the contralateral horns served as a control for uterine reactivity. Uterine transmigration of embryos is a possible complication in such experiments (Runner, 1951; McLaren & Michie, 1954). However, it was not found among 36 foetuses examined following transfer of blastocysts of pigmented and albino genotypes to opposite horns of 6 recipients, nor among 63 implants developed from blastocysts transferred unilaterally to a further 18 recipients (unpublished observations).

Induction of deciduomata

Intraluminal injection of arachis oil between 15.00 and 16.00 h on the fourth day of pseudopregnancy was employed for this purpose (Finn, 1965).

Examination of the transferred tissue

Recipient females that received trophoblast tissue were examined between 3 and 8 days after transfer. Horns carrying decidual swellings were fixed in Sousa or Bouin’s fixative. The dehydrated tissue was cleared in supercedrol (G. T. Gurr), embedded in paraffin wax and serially sectioned at 6–10 μm. The sections were stained with haemalum and eosin and mounted in DPX (G. T. Gurr).

Females receiving ICM tissue were given i.v. pontamine sky blue 6 BX (G. T.Gurr) approximately 48 h after transfer of the tissue to determine whether implantation changes had been initiated (Psychoyos, 1961; Finn & McLaren, 1967).

Deciduomata and implants derived from intact blastocysts were processed histologically as above.

Assessment of implantation rates by counting decidual swellings -complicating factors

More than one implant may occur within a single decidual swelling (author’s unpublished observations). This phenomenon would not be detected in cases where derivatives of the ICM are poorly represented or totally lacking (as in nearly all implants obtained following transfer of trophoblast). Counts of decidual swellings would thus tend to underestimate the rate of implantation.

Decidualization can also occur from the inevitable trauma accompanying the transfer operation, and inclusion of these swellings would lead to an overestimate of the implantation rate. However, in two cases where such deciduomata were identified with certainty, i.e. in horns where all transferred blastocysts had given rise to decidual swellings containing embryos or embryonic tissues, there was no question of their being confused with normal decidua. They were found at the oviducal end of their respective horns (i.e. at a point corresponding to the transfer site) and were less than half the normal diameter. One was sectioned, and was not only completely without embryonic tissues but very poorly orientated in relation to the mesometrial-antimesometrial axis of the uterus. During this and a succeeding study (in preparation), further examples of probable traumatic deciduomata were encountered that showed all the above features, particularly the small size and poor orientation. Only once did a probable deciduoma approach the size of an adjacent decidual swelling. Yet even here the orientation was atypical, being exactly perpendicular to the mesometrial-antimesometrial axis. So, although these possible complications must be borne in mind their effect on the results will be slight.

1 Observations on trophoblastic fragments in culture

The following remarks apply to both maximal and equatorial fragments (Fig. 2 A, B) unless qualified, and are based on study of 139 maximal and 34 equatorial fragments.

Initially, the fragments were solid balls of cells with no visible cavity (Fig. 3 A). When placed in culture the majority sealed and began to cavitate within 30 min, becoming obviously vesicular within (Fig. 3B). In a typical series 85% (56/66) cavitated. The remaining 10 may have been damaged during cutting or handling. Although the vesiculated fragments could cavitate a second time if torn open, they did not swell to the same size as intact blastocysts.

Fig. 3

A pair of maximal trophoblastic fragments (A) immediately after being placed in contact at 37°C and (B) 4 h later. The area of contact between the two fragments is limited, and there is no tendency for them to fuse together.

Fig. 3

A pair of maximal trophoblastic fragments (A) immediately after being placed in contact at 37°C and (B) 4 h later. The area of contact between the two fragments is limited, and there is no tendency for them to fuse together.

Twenty-four maximal fragments were placed in contact in 12 pairs and the members of each pair held together at 37°C until they adhered. These pairs of fragments remained discrete despite incubation for up to 14 h, though usually cavitating and showing a limited area of mutual adhesion (Fig. 3 A, B). In no case was fusion or aggregation observed, and members of a pair could easily be separated with a glass needle (Table 2, row X).

Table 2

Summary of results of culturing various categories of fragments of mouse blastocysts in contact at 37°C

Summary of results of culturing various categories of fragments of mouse blastocysts in contact at 37°C
Summary of results of culturing various categories of fragments of mouse blastocysts in contact at 37°C

2 Fate of trophoblastic fragments transferred to the uteri of pesudopregnant mice

Sixty-five maximal fragments which had been cultured for were transferred into seven recipients. Thirty-three equatorial fragments, of which only five had been cultured for 2 h, were transferred into a further four recipients. Implantation was assessed by looking for discrete decidual swellings, and only females with one or more such swellings have been included in Table 3. A proportion of both the maximal and equatorial trophoblastic fragments had implanted.

Table 3

Implantation of trophoblast and ICM tissue of 312-day mouse blastocysts in the uteri of pseudopregnant mice

Implantation of trophoblast and ICM tissue of 312-day mouse blastocysts in the uteri of pseudopregnant mice
Implantation of trophoblast and ICM tissue of 312-day mouse blastocysts in the uteri of pseudopregnant mice

All but one of the 34 implants examined 4-7 days after transfer of trophoblast appeared similar (Table 4A, B). Embryo, amnion, allantois, yolk sac, Reichert’s membrane and ectoplacental tissue were absent. A chamber, corresponding roughly to the position the embryo would occupy in a normal implant, had almost invariably formed, apparently by pycnosis of the central decidual cells. Occasionally it had a somewhat spongy or reticular appearance due to the persistence of a network of elongated cells (Fig. 4). In other decidua the chamber was virtually acellular, and filled with an homogeneous ground substance broken up by a complex fibrin-like network (Fig. 5), or even with maternal blood cells. A second consistent feature, particularly of the older implants, was a compact mass of eosinophilic material of unknown origin and composition which lay between the mesometrial border of the chamber and the zone rich in maternal sinusoids (Figs. 4, 5). This material sometimes attained large dimensions and small dispersed foci could occasionally be seen in younger implants. Polymorphonuclear leucocytes were present in and around it.

Table 4

Summary of main features of uterine implants developed from trophoblastic fragments

Summary of main features of uterine implants developed from trophoblastic fragments
Summary of main features of uterine implants developed from trophoblastic fragments
Fig. 4

Centre of a decidual swelling 6 days after the transfer of maximal trophoblastic fragments. A large mass of eosinophilic material occupies the top centre of the field. A single pycnotic trophoblast giant cell can be seen in the middle towards the antimesometrial end of the implantation chamber.

Fig. 4

Centre of a decidual swelling 6 days after the transfer of maximal trophoblastic fragments. A large mass of eosinophilic material occupies the top centre of the field. A single pycnotic trophoblast giant cell can be seen in the middle towards the antimesometrial end of the implantation chamber.

Fig. 5

Another decidual swelling from the same recipient that carried the decidual site shown in Fig. 4. A large eosinophilic mass lies mesometrially as before. However, the chamber is larger than the previous one and is virtually free of cells.

Fig. 5

Another decidual swelling from the same recipient that carried the decidual site shown in Fig. 4. A large eosinophilic mass lies mesometrially as before. However, the chamber is larger than the previous one and is virtually free of cells.

The third distinctive feature was the presence in three-quarters of the implants of cells indistinguishable from the trophoblastic giant cells that surround normal conceptuses. Their presence constituted the only difference in cellular composition between decidua induced by trophoblast and deciduomata evoked by oil. The number of giant cells was similar between 4 and 7 days after transfer and was trivial compared with the number found in normal implants of equivalent ages (Table 4A, B; Noyes, 1959; Snell & Stevens, 1966). Some of them were pycnotic, though viable ones persisted in some of the older implants (Fig. 6).

Fig. 6

Centre of a decidual swelling 7 days after transfer of maximal trophoblastic fragments. A rather loosely arranged eosinophilic mass occupies most of the centre. The arrow indicates the position of a viable trophoblastic giant cell.

Fig. 6

Centre of a decidual swelling 7 days after transfer of maximal trophoblastic fragments. A rather loosely arranged eosinophilic mass occupies most of the centre. The arrow indicates the position of a viable trophoblastic giant cell.

The single exceptional implant consisted of a small yolk-sac-like structure surrounded by a thickened Reichert’s membrane, and was framed by a network of healthy giant cells (Fig. 7). This implant was presumably derived from a fragment contaminated with ICM tissue.

Fig. 7

Exceptional trophoblastic implant 7 days after transfer of maximal trophoblastic fragments. Note the abundant giant cells, absence of ectoplacental cone, and the central vesicle consisting of a thick Reichert’s membrane lined presumably with cells of the distal endoderm. (Mesometrium towards the top of the figure.)

Fig. 7

Exceptional trophoblastic implant 7 days after transfer of maximal trophoblastic fragments. Note the abundant giant cells, absence of ectoplacental cone, and the central vesicle consisting of a thick Reichert’s membrane lined presumably with cells of the distal endoderm. (Mesometrium towards the top of the figure.)

The transferred trophoblast thus only gave rise to a very few giant cells which were never observed in division. One explanation for the failure of the trophoblast to proliferate might be that the endocrine status of the recipients differs in these experiments from that of normal pregnancy. Hence the following experiments were undertaken to determine whether trophoblastic fragments developed similarly in recipients made pregnant by simultaneously transferring intact blastocysts.

3 Transfer of trophoblastic fragments and intact blastocysts to opposite horns of pseudopregnant mice

Thirty-six maximal trophoblastic fragments were transferred unilaterally to six females, and 5–6 blastocysts of the same age and genotype were injected into the contralateral horns. Five recipients had implants at autopsy. It was found that trophoblastic fragments devoid of ICM tissue are as effective decidual stimuli as whole blastocysts (Table 5 A).

Table 5

Implantation of maximal trophoblastic fragments in the uteri of pseudopregnant mice that received intact blastocysts in the contralateral horn

Implantation of maximal trophoblastic fragments in the uteri of pseudopregnant mice that received intact blastocysts in the contralateral horn
Implantation of maximal trophoblastic fragments in the uteri of pseudopregnant mice that received intact blastocysts in the contralateral horn

All implants developed from intact blastocysts contained foetuses. With two exceptions, the form of the trophoblastic implants was similar to those described previously as regards central chamber, eosinophilic mass and scarcity of trophoblastic giant cells (Table 4C). The absence of giant cells from most of the oldest implants in this series suggests that death or dispersal of these cells may occur 7–8 days after transfer.

The two implants which were not typical of transferred trophoblastic fragments were adjacent in one horn of a recipient killed 8 days after transfer. One contained a well-developed foetus and the other a mass of giant cells (Fig. 8 A, B). The former might have been due to regulation by a fragment from which all ICM cells had not been excluded or to accidental transfer of a blastocyst. It was probably responsible for the occurrence of the second exception. Thus the giant cells of the first implant were not enclosed by viable decidual tissue on the side bordering the second implant, since a wide track of entirely pycnotic decidual tissue extended to the centre of the latter (Fig. 8B). Though no giant cells were noted actually in the boundary zone they were close enough on both sides to suggest very strongly that they had invaded the second implant rather than having originated there.

Fig. 8

Two atypical implants obtained following transfer of maximal trophoblastic fragments to a recipient carrying intact blastocysts in the contralateral horn. (A) Low magnification to show relationship between them. (B) Detail of region of contact.

Fig. 8

Two atypical implants obtained following transfer of maximal trophoblastic fragments to a recipient carrying intact blastocysts in the contralateral horn. (A) Low magnification to show relationship between them. (B) Detail of region of contact.

One recipient had implants in the experimental horn only (Table 5) and these conformed to the typical pattern (Table 4D).

In a further experiment a single recipient (R1 (29. 5. 70), Table 5B) was killed only 3 days after transfer of blastocysts to one horn and trophoblastic fragments to the other. Two implants found in the experimental horn already appeared abnormal (Fig. 9A-C). Each consisted of a small localized net or cord of trophoblast cells near the antimesometrial end of a uterine crypt. The uterine epithelium was missing immediately to either side of them. Neither possessed a cavity which, since all the control implants did and the entire tract was fixed as one, is unlikely to have been an artifact of fixation. One implant contained 25 cells and the second roughly 36 (one or two sections being folded in the latter case). The nuclear diameter of these cells was much less than that of fully differentiated giant cells. Their cytoplasmic detail was poor and the viability of some was questionable. No dividing cells were present. Indeed, the number of cells did not exceed that expected in maximal fragments at the time of transfer (Table 6).

Table 6

Number of cells in intact 312-day blastocysts and dissected ICMs of PDE mice

Number of cells in intact 312-day blastocysts and dissected ICMs of PDE mice
Number of cells in intact 312-day blastocysts and dissected ICMs of PDE mice
Fig. 9

Transverse sections of decidual swellings 3 days after transfer of maximal trophoblastic fragments and intact blastocysts to opposite horns of the same recipient. (A) Implanted maximal trophoblastic fragment consisting of only 25 cells, of which four can be identified in the figure. (B) The second implanted maximal trophoblastic fragment consisting of approximately 36 cells of which 4–5 may be seen in the figure. (C) Implanted intact blastocyst in the contralateral horn.

Fig. 9

Transverse sections of decidual swellings 3 days after transfer of maximal trophoblastic fragments and intact blastocysts to opposite horns of the same recipient. (A) Implanted maximal trophoblastic fragment consisting of only 25 cells, of which four can be identified in the figure. (B) The second implanted maximal trophoblastic fragment consisting of approximately 36 cells of which 4–5 may be seen in the figure. (C) Implanted intact blastocyst in the contralateral horn.

4 Observations on ICMs isolated from - and -day blastocysts in culture

ICMs were very variable in shape after they had been scraped from the sheet of overlying trophoblast cells. They rapidly became spherical when cultured, but remained typically as solid balls of cells without developing a central cavity (Fig. 10). When placed in contact in pairs or small groups they almost invariably aggregated to form unitary structures of spherical or near spherical appearance (Fig. 11). Such fusion consistently took place between pairs of -day ICMs of the same or different genotype (Table 2, rows I and II). Fusion was also observed between pairs of -day ICMs despite the endoderm having delaminated in some of these older embryos as discerned by phase-contrast microscopy (Table 2, rows V, VI and VII). In several instances pairs of ICMs differing by nearly 24 h in age behaved similarly (Table 2, row VIII).

Fig. 10

A single isolated ICM from a 312-day blastocyst approximately 212 h after the beginning of culture. It shows neither cavitation nor intracellular accumulation of fluid (cf. Fig. 3B). The smaller fragment is residual trophoblastic debris left behind following dissection of the ICM.

Fig. 10

A single isolated ICM from a 312-day blastocyst approximately 212 h after the beginning of culture. It shows neither cavitation nor intracellular accumulation of fluid (cf. Fig. 3B). The smaller fragment is residual trophoblastic debris left behind following dissection of the ICM.

Fig. 11

Stages in the fusion of ICMs in vitro. Two pairs of 312-day dissected ICMs that had already fused were placed in contact at time zero. (A) +12 h, (B) + 112 h, (C) + 2h, (D) + 434h after contact (darkground illumination). (From Gardner, 1971, with permission of F. Vieweg & Sohn.)

Fig. 11

Stages in the fusion of ICMs in vitro. Two pairs of 312-day dissected ICMs that had already fused were placed in contact at time zero. (A) +12 h, (B) + 112 h, (C) + 2h, (D) + 434h after contact (darkground illumination). (From Gardner, 1971, with permission of F. Vieweg & Sohn.)

Where fusion failed the fragments remained discrete but connected by a narrow waist. Failure was often accompanied by the development of a small eccentric cavity or by accumulation of fluids within peripheral cells (Fig. 12). Both these phenomena suggest that the ICM tissue was contaminated with trophoblast cells. This interpretation is in accordance with the finding that minimal ICM fragments which retain the cells overlying the ICM (see Fig. 2 B) did not aggregate under similar conditions (Table 2, row IX). However, minimal ICM fragments contain more cells than dissected ICMs and hence fusion might simply depend on cell number rather than specifically on cell type. Cell number is probably not the limiting factor because pairs of fused ICMs can fuse with other ICMs, thereby yielding structures containing as many cells as entire blastocysts (Table 2, row III; Table 6; Fig. 11).

Fig. 12

Pair of 312-day dissected ICMs which failed to fuse after 5 h culture. Intra-cellular accumulation of fluid, suggesting contamination by trophoblast cells, can be seen in both fragments.

Fig. 12

Pair of 312-day dissected ICMs which failed to fuse after 5 h culture. Intra-cellular accumulation of fluid, suggesting contamination by trophoblast cells, can be seen in both fragments.

5 Transfer of fused pairs of -day isolated ICMs to the uteri of recipient mice

Aseries of fused pairs of ICMs were transferred unilaterally, intact blastocysts being placed in the opposite horn of each recipient. The results show that of 17 pairs transferred to females with pontamine blue sites in their control horns at autopsy, none induced decidual changes (Table 3C). Also, in no case could the transferred tissue be identified in uterine flushings.

The trophoblast and ICM of the mouse blastocyst are very different. The ability to accumulate or pump fluid (Gamow & Daniel, 1970) and to induce the decidual changes characteristic of implantation are properties peculiar to the trophoblast. Though the trophoblastic fragments can form vesicles, as do cultured fragments of rabbit blastocysts (Daniel, 1961, 1963; Klinger, Kosseff & Plotnick, 1971), they do not aggregate with similar fragments or vesicles. Like-wise, mouse blastocysts cannot be induced to fuse together under conditions that favour fusion of cleaving eggs or morulae in mice (Mintz, 1965; Mulnard, 1971). This may be related to the presence of specialised junctions between trophoblast cells (Enders & Schlafke, 1965).

Isolated ICMs resemble cleaving eggs in the ease with which they fuse together, as also in their inability to accumulate fluid and evoke decidual reactions (Kirby, 1970). The ICMs were fused in pairs before transfer to the uterus to ensure they contained a similar number of cells to trophoblastic fragments. Hence their failure to induce decidual changes could have been due to reduction in viability during the culture period necessary for fusion. Cole & Paul (1965) have argued that presumptive ICM cells show poor survival in culture. However, this explanation is inappropriate here because cultured ICMs transferred into the cavity of blastocysts or trophoblastic vesicles can contribute extensively to the resulting (chimaeric) offspring or give rise to morphologically normal foetuses respectively (Gardner, 1971).

The cells overlying the ICM cannot be studied directly with present techniques because they are either removed together with the ICM (in minimal ICM fragments -see Fig. 2B) or destroyed when the ICM is scraped from the blastocyst (Fig. 2C). Nevertheless, failure of minimal ICM fragments to fuse together while isolated ICMs do, indicates that these cells differ from those of the ICM. Whether they are identical to those of the rest of the trophectoderm with which they are continuous is uncertain at present.

The absence of embryo, amnion, allantois, yolk sac and Reichert’s membrane from implants developed from pure trophoblast accords with the conclusions of histological studies which attribute an ICM origin to these structures (Snell & Stevens, 1966). The similarity of all but 3 of the 61 trophoblastic implants examined 4–8 days after transfer attests to the homogeneity of the tissue transferred. The very few giant cells evident in the majority of implants are almost certainly of embryonic rather than maternal origin (Fawcett, Wislocki & Waldo, 1947; Snell & Stevens, 1966; Gwatkin, 1966). Whether the absence of giant cells from 19 implants was because they did not develop or were too advanced in pycnosis to be identified remains conjectural. No qualitative differences were noticed between implants developed from maximal and equatorial trophoblastic fragments.

There was no evidence that the trophoblast had proliferated normally and subsequently dispersed from the implantation chamber. The low number of trophoblast cells and absence of mitotic figures at all the stages examined leads to the inescapable conclusion that the trophoblastic fragments did not proliferate in utero. This might be because the mass of trophoblast tissue, though able to induce a decidual response, is insufficient to enable subsequent proliferation. This explanation is ruled out by the fact that partial and half blastocysts containing as few trophoblast cells as trophoblastic fragments can give rise to normal conceptuses and young at term (Gardner, 1972).

A further possibility might be that the cells overlying the ICM are required for normal development of the trophoblast. The notion of regional differentiation of trophoblast is implicit in this explanation. Although the ectoplacental cone develops over the ICM in the early egg-cylinder (Snell & Stevens, 1966), ‘reconstituted’ blastocysts specifically lacking embryonic polar trophoblast can develop normally (Gardner, 1971, and unpublished observations).

The remaining and perhaps most obvious difference between trophoblastic fragments that fail to produce significant trophoblastic development and intact, partial or ‘reconstituted’ blastocysts that do, is that the latter all contain some ICM tissue. Hence it is concluded that ICM tissue is required to permit normal proliferation of trophoblast of the day mouse blastocyst. Abundant trophoblast is found in some implants developed from partial blastocysts that do not contain embryos (Gardner, 1972), so proliferation presumably does not depend on development of a definitive embryo. Also, dependence on the ICM is probably limited to an early stage of trophoblast development since ectoplacenta-cones from 6-to -day egg-cylinders can proliferate following transfer ectopically (Grobstein, 1950; Billington, 1965; Simmons & Russell, 1966; Clarke, 1969) or to the uteri of cyclic or pseudopregnant mice (Kirby, 1965, 1970; Kirby & Cowell, 1968).

If the proliferation of trophoblast does indeed depend initially on the ICM one would expect to find derivatives of the latter tissue in implants with abundant trophoblast. This expectation was fulfilled in nearly all implanted partial and ‘reconstituted’ blastocysts that did not develop foetuses (Gardner, 1972). The exceptions may have been more apparent than real since they all showed signs of cellular degeneration. One of the three atypical implants developing from implanted trophoblast in the present experiments carried a foetus, and the second is believed to have acquired its trophoblast from the first. The third showed a continuous Reichert’s membrane enclosing some cells (Fig. 7). This membrane has been found by elegant immunofluorescence studies to be a secretion of the distal endoderm (Pierce, Midgley, Sri Ram & Feldman, 1962; Midgley & Pierce, 1963) and hence presumably an ICM derivative (Snell & Stevens, 1966).

There are few published reports containing detailed descriptions of uterine implants lacking embryos. Tarkowski (1962) found that rat eggs that had undergone blastulation in the oviduct of the mouse showed dispersal of their ICMs and developed poorly after implanting in the rat uterus. In extreme cases no embryo had formed, but profuse trophoblast surrounded a Reichert’s membrane lined with a single layer of cells (Tarkowski, 1962, plate 4, fig. V). The inner cells are presumably those of the distal endoderm. A similar condition was found in another implant following transfer of rat eggs to the oviduct of the rat (Tarkowski, 1962, plate 4, fig. U).

The relevance of descriptions of 20 mouse implantation chambers in which spontaneous embryonic death had occurred is limited both by the uncertainty regarding when development went awry and by the fact that the material was obtained late in pregnancy (Droogleever Fortuyn, 1920). Hence the absence of all ICM derivatives from five chambers that only contained clusters of giant cells does not preclude their being formerly present.

Failure of the trophoblast of mouse blastocysts to grow ectopically in recipients specifically immunized against the donor strain while similar ectoplacental cones succeed has been considered evidence that the former express antigens and the latter do not (Simmons & Russell, 1966). Present considerations raise the possibility that it is the ICM of the blastocyst that is susceptible to damage or immunological attack and that failure of trophoblast development depends secondarily on this factor (Gardner, Johnson & Edwards, 1972).

Critical appraisal of the relationship between the trophoblast and ICM in rodent embryos transferred to ectopic sites is complicated by factors such as abnormal morphogenesis, haemorrhage and degenerative changes. Consequently many authors fail to specify derivatives other than trophoblast or embryo proper (e.g. Jollie, 1961). The general conclusion from numerous studies is that proliferation of the trophoblast is favoured and embryonic development poor and infrequent, especially if oviducal stages rather than blastocysts are transferred (Billington, Graham & McLaren, 1968; Kirby, 1970). Nevertheless, Reichert’s membrane is found frequently in grafts lacking embryos. Even when absent as a discrete membrane it may indeed be present as a homogeneous matrix with cells embedded in it (e.g. Fawcett, 1950, figs. 6 and 7; Whitten, 1958, fig. 2; McLaren & Tarkowski, 1963, fig. 4; and author’s unpublished observations). Such ‘abortive yolk sacs’ closely resemble the undifferentiated murine teratocarcinoma embedded in neoplastic hyalin with which they are almost certainly homologous (Pierce et al. 1962; Midgley & Pierce, 1963).

A particularly illuminating study, in which all host kidneys were examined after transfer of -day mouse blastocysts, is reported by Johnson in an appendix to this paper. Only two grafts definitely lacked ICM derivatives and these, like most uterine implants developed from trophoblastic fragments, contained just a few giant cells. Such grafts are likely to have been overlooked in studies where host organs showing macroscopic ‘takes’ were selected for examination.

Any explanation of the relationship deduced between trophoblast and ICM must account for the failure of proliferation of implanted trophoblastic fragments on the one hand, but development of some giant cells in most of them on the other. Available data are embraced by the following hypothesis. The mural trophoblast of the mouse blastocyst which gives rise to the primary giant cells (Dickson, 1963) can do so autonomously, or has already received the appropriate stimulus before it is separated from the ICM. It thus provides the few giant cells found in trophoblastic implants. The development of the ectoplacental cone (and hence the multitude of secondary giant cells derived from it, Snell & Stevens, 1966) is specifically dependent on the presence of ICM tissue in the blastocyst or early egg-cylinder. The role of the ICM might be to promote division or inhibit giant transformation of the overlying trophoblast cells. The notion of an inductive interaction is attractive because the ectoplacental cone invariably develops over the ICM, and because the latter may attain its final mesometrial position by migrating round the inner surface of the trophoblast wall (Kirby, Potts & Wilson, 1967; Jenkinson & Wilson, 1970; Gardner, 1971). However, it is also possible that the ectoplacental cone is a derivative of the ICM (Duval, 1892).

Finally, the morphological similarity between early implants of pure trophoblast and those of homozygous lethal yellow embryos in utero has not escaped notice (fig. 9 A, B; Eaton & Green, 1963). It is of considerable interest to discover whether this resemblance is more than superficial.

I wish to thank Professor C. R. Austin, Dr R. G. Edwards, Dr C. F. Graham, Dr M. H. Johnson and Dr M. I. Sherman for advice and discussion, and Mrs S. C. Barton, Mrs W. J. Gardner and Mrs W. Redmond for help. This work was supported by the Medical Research Council, the World Health Organisation and the Ford Foundation.

APPENDIX

Relationship between inner cell mass derivatives and trophoblast proliferation in ectopic pregnancy

SUMMARY

Forty-two -day mouse embryos were transferred to ectopic sites in the kidneys of male mice. Microscopic investigation of all transfers revealed that trophoblast proliferation had occurred only when inner cell mass derivatives were present. In the absence of inner cell mass derivatives, non-proliferated giant cells were present. This data is taken as evidence compatible with the hypothesis that the inner cell mass is essential for trophoblast proliferation.

The results achieved by the use of elegant microsurgical techniques have led Gardner to formulate the hypothesis that the inner cell mass of the -day mouse embryo is essential for development of normal proliferated trophoblast (see the first part of this paper). In the presence of the inner cell mass, normal ectoplacental cone tissue with its peripheral secondary giant cells develops. In the absence of an inner cell mass, the mural trophoblast of the -day blastocyst merely differentiates into primary giant cells without prior division.

This hypothesis should also apply to ectopically transferred -day embryos, but appears at first sight to be refuted by the reports emphasizing substantial trophoblast proliferation with little or no embryonic development in such conditions (Fawcett et al. 1947; Runner, 1947; Fawcett, 1950; Kirby, 1960, 1962). In fact, careful examination of these reports in most cases reveals a description of some derivatives of inner cell mass such as an abortive yolk sac (Fawcett, 1950) or some more complex, but imperfect structure of proven inner cell mass origin (Snell & Stevens, 1966). With the hypothesis of Gardner in mind, careful histological analysis was made of 42 mouse embryos (-day, C3H or C57BL) transplanted ectopically to the kidneys of male mice of homologous strains. The transfer technique has been described elsewhere (Johnson & Dharmawardena, 1972). Seven days after transfer, all kidneys were removed, fixed and serial sections were cut and examined regardless of whether the embryo appeared to have ‘taken’ macroscopically. Each embryo found was classified into one of the five categories described in Table 7. The classification of group 4 transplants (Fig. 13) as Reichert’s membrane plus endoderm-like nuclei was made by others (Fawcett, 1950; Stevens, 1968) and has been discussed in detail by Gardner in the first part of this paper.

Table 7
graphic
graphic
Fig. 13

Nodule of Reichert’s membrane and distal endoderm nuclei embedded in proliferating trophoblast, × 720.

Fig. 13

Nodule of Reichert’s membrane and distal endoderm nuclei embedded in proliferating trophoblast, × 720.

The data show that, with one exception, trophoblastic proliferation and the presence of secondary type giant cells were detected only where some evidence of inner cell mass derivatives existed. For the exceptional transplant, an area of necrosis was present in the core of the ectoplacental cone tissue normally occupied by the embryo. For two transplants, in which inner cell mass derivatives were not present, small nodules of primary giant cells were detected (Fig. 14). No proliferating trophoblast of the type seen in the ectoplacental cone of uterine or other ectopic transplants was present, suggesting that the giant cells were primary rather than secondary. The number of cells in each nodule, 9 and 21, were of the order expected for primary trophoblast after transfer of the 30–40 mural trophoblast cells in the -day embryo (Gardner, above). The giant cells were not surrounded by any extensive necrotic or scar tissue, thereby negating the possibility that they were the residuum of a proliferated transplant that subsequently died. In fact, the two clusters of giant cells bore a remarkable resemblance, both histological and in terms of cell numbers, to those occurring in utero after transfer of pure trophoblastic vesicles (Gardner, above). Presumably the inner cell masses of these two embryos had been damaged on transfer, resulting in the absence both of inner cell mass derivatives, and, according to the hypothesis of Gardner, of a stimulus to the proliferation of ectoplacental trophoblast. The trophectoderm of the transplanted blastocyst therefore merely differentiated into primary giant cells without prior division.

Fig. 14

Non-proliferated trophoblastic giant cells embedded in kidney tissue, × 180.

Fig. 14

Non-proliferated trophoblastic giant cells embedded in kidney tissue, × 180.

Examination of the literature on ectopic transfers has revealed only one study in which ectopic transfer sites have been examined histologically regardless of whether a macroscopic haemorrhage was present (Stevens, 1968). In that study also, non-proliferated primary trophoblastic giant cells only evidently occurred in the absence of any inner cell mass derivatives. Careful analysis of ectopic transfers in other studies would be worth while. The absence of such analyses in ectopic transfers of blastocysts to recipients preimmunized against donor histo-compatibility antigens may have led to misleading conclusions about the distribution of these antigens on the early embryo. The failure to detect macroscopic haemorrhage in these studies could have been due to failure of trophoblast to proliferate secondary to immunological destruction of the inner cell mass rather than to a primary immunological destruction of the trophoblast (Gardner et al. 1972).

In conclusion, the findings that trophoblastic proliferation is invariably associated with the presence of inner cell mass derivatives and that a failure of trophoblastic proliferation is associated with the absence of inner cell mass derivatives are not in conflict with the hypothesis of Gardner (above) and indeed offer positive support for it.

I wish to acknowledge the stimulation provided by discussion with Richard Gardner and the technical assistance of Vinitha Dharmawardena. The work was done whilst the author was in receipt of an M.R.C. Junior Fellowship, and was supported by a grant to Professor C. R. Austin from the Ford Foundation.

Billington
,
W. D.
(
1965
).
The invasiveness of transplanted mouse trophoblast and the influence of immunological factors
.
J. Reprod. Fert.
10
,
343
352
.
Billington
,
W. D.
,
Graham
,
C. F.
&
McLaren
,
A.
(
1968
).
Extra-uterine development of mouse blastocysts cultured in vitro from early cleavage stages
.
J. Embryol, exp. Morph.
20
,
391
400
.
Clarke
,
A. G.
(
1969
).
Factors affecting the growth of trophoblast transplanted to the testis
.
J. Reprod. Fert.
18
,
539
541
.
Cole
,
R. J.
,
Edwards
,
R. G.
&
Paul
,
J.
(
1965
).
Cytodifferentiation in cell colonies and cell strains derived from cleaving ova and blastocysts of the rabbit
.
Expl Cell Res
.
37
,
501
504
.
Cole
,
R. J.
&
Paul
,
J.
(
1965
).
Properties of cultured preimplantation mouse and rabbit embryos, and cell strains derived from them
. In
Preimplantation Stages of Pregnancy
(ed. G. E. W. Wolstenholme & M.O’Connor), pp.
82
112
.
London
:
Churchill
.
Dalcq
,
A. M.
(
1957
).
Introduction to General Embryology.
Oxford University Press
.
Daniel
,
J. C.
Jr
. (
1961
).
Reconstitution of rabbit blastocysts from fragments in vitro
.
Am. Zool.
1
,
444
.
Daniel
,
J. C.
Jr
. (
1963
).
Some kinetics of blastocyst formation as studied by the process of reconstitution
.
J. exp. Zool.
154
,
231
237
.
Dickson
,
A. D.
(
1963
).
Trophoblastic giant cell transformation of mouse blastocysts
.
J. Reprod. Fert.
6
,
465
466
.
Droogleever Fortuyn
,
A. B.
(
1920
).
The involution of the placenta in the mouse after the death of the embryo
.
Archs Biol., Paris
30
,
323
355
.
Duval
,
M.
(
1892
).
Le Placenta des Rongeurs. (Extrait du Journal de F Anatomie et de la Physiologie Années 1889-1892)
(ed.
F.
Alcan
). Paris: Acienne Librarie Germer Bailliere et Cie.
Eaton
,
G. J.
&
Green
,
M. M.
(
1963
).
Giant cell differentiation and lethality of homozygous yellow mouse embryos
.
Genética
34
,
155
161
.
Enders
,
G. J.
&
Schlafke
,
S. J.
(
1965
).
The fine structure of the blastocyst: some comparative studies
. In
Preimplantation Stages of Pregnancy
(ed. G. E. W. Wolstenholme & M. O’Connor), pp.
29
-
54
. London: Churchill.
Fawcett
,
D. W.
(
1950
).
The development of mouse ova under the capsule of the kidney
.
Anat. Rec.
108
,
71
91
.
Fawcett
,
D. W.
,
Wislocki
,
G. B.
&
Waldo
,
C. M.
(
1947
).
The development of mouse ova in the anterior chamber of the eye and in the abdominal cavity
.
Am. J. Anat.
81
,
413
443
.
Finn
,
C. A.
(
1965
).
Oestrogen and the decidual cell reaction of implantation in mice
.
J. Endocr.
32
,
223
229
.
Finn
,
C. A.
&
McLaren
,
A.
(
1967
).
A study of the early stages of implantation in mice
.
J. Reprod. Fert.
13
,
259
267
.
Gamow
,
E.
&
Daniel
,
J. C.
Jr.
(
1970
).
Fluid transport in the rabbit blastocyst
.
Wilhelm Roux Arch. EntwMech. Org.
164
,
261
278
.
Gardner
,
R. L.
(
1968
).
Mouse chimaeras obtained by the injection of cells into the blastocyst
.
Nature, Lond.
220
,
596
597
.
Gardner
,
R. L.
(
1971
).
Manipulations on the blastocyst
. In
Advances in the Biosciences, no. 6: Schering Symposium on Intrinsic and Extrinsic Factors in Early Mammalian Development
(ed. G. Raspé), pp.
279
296
.
Oxford
:
Pergamon
.
Gardner
,
R. L.
(
1972
).
Microsurgical approaches to the study of early mammalian development. In the Seventh Harold C. Mack Symposium: Birth Defects and Fetal Development, Endocrine and Metabolic Factors. (In the Press.)
Gardner
,
R. L.
,
Johnson
,
M. H.
&
Edwards
,
R. G.
(
1972
).
Are H-2 antigens expressed in the preimplantation blastocyst?
In
Proceedings of the Second International Symposium on the Immunology of Reproduction, Varna, Bulgaria. (In the Press.)
Graham
,
C. F.
(
1971
).
The design of the mouse blastocyst
. In
Control Mechanisms of Growth and Differentiation. Symposium Society for Experimental Biology, no. 25
(ed.
D. D.
Davis
&
M.
Balls
), pp.
371
378
.
Cambridge University Press
.
Grobstein
,
C.
(
1950
).
Production of intra-ocular haemorrhage by mouse trophoblast
.
J. exp. Zool.
114
,
359
373
.
Gwatkin
,
R. B. L.
(
1966
).
Amino acid requirements for attachment and outgrowth of mouse blastocysts in vitro
.
J. Cell Physiol.
68
,
335
345
.
Jenkinson
,
E. J.
&
Wilson
,
I. B.
(
1970
).
In vitro support system for the study of blastocyst differentation in the mouse
.
Nature, Lond.
228
,
776
778
.
Johnson
,
M. H.
&
Dharmawardena
,
V.
(
1972
).
Effect of preimmunisation with spermatozoa on survival of ectopically transferred mouse blastocysts
.
Fert. Steril.
23
,
353
356
.
Jollie
,
W. P.
(
1961
).
The incidence of experimentally produced abdominal implantations in the rat
.
Anat. Rec.
141
,
159
167
.
Kirby
,
D. R. S.
(
1960
).
Development of mouse eggs beneath the kidney capsule
.
Nature, Lond.
187
,
707
708
.
Kirby
,
D. R. S.
(
1962
).
The influence of the uterine environment on the development of mouse eggs
.
J. Embryol. exp. Morph.
10
,
496
506
.
Kirby
,
D. R. S.
(
1965
).
The ‘invasiveness’ of the trophoblast
. In
The Early Conceptus, Normal and Abnormal
(ed.
W. W.
Park
), pp.
68
73
.
Dundee
:
University of St Andrews
.
Kirby
,
D. R. S.
(
1970
).
Immunological aspects of implantation
. In
Ovo-Implantation. Human Gonadotrophins and Prolactin
(ed.
P. O.
Hubinont
,
F.
Leroy
,
C.
Robyn
&
P.
Leleux
), pp.
86
97
.
Basel
:
S. Karger
.
Kirby
,
D. R. S.
&
Cowell
,
T. P.
(
1968
).
Trophoblast/host interactions
. In
Epithelial-Mesenchyme Interactions
(ed. R. Fleischmajer &
R. E.
Billingham
), pp.
64
-
77
. Baltimore: Williams & Wilkins.
Kirby
,
D. R. S.
,
Potts
,
D. M.
&
Wilson
,
I. B.
(
1967
).
On the orientation of the implanting blastocyst
,
J. Embryol. exp. Morph.
17
,
527
532
.
Klinger
,
H. P.
,
Kosseff
,
A. L.
&
Plotnick
,
F.
(
1971
).
Sex chromatin formation and RNA and protein synthesis during preimplantation development of the rabbit
. In
Advances in the Biosciences, no.6: Schering Symposium on Intrinsic and Extrinsic Factors in Early Mammalian Development
(ed. G. Raspé), pp.
207
-
224
. Oxford: Pergamon Press.
Lin
,
T. P.
(
1969
).
Microsurgery of the inner cell mass of mouse blastocysts
.
Nature, Lond.
222
,
480
481
.
McLaren
,
A.
&
Michie
,
D.
(
1954
).
Transmigration of unborn mice
.
Nature, Lond.
173
,
630
.
McLaren
,
A.
&
Tarkowski
,
A. K.
(
1963
).
Implantation of mouse eggs in the peritoneal cavity
.
J. Reprod. Fert.
6
,
384
392
.
Midgley
,
A. R.
&
Pierce
,
G. B.
(
1963
).
Immunohistochemical analysis of basement membranes of the mouse
.
Am. J. Path.
43
,
929
943
.
Mintz
,
B.
(
1965
).
Experimental genetic mosaicism in the mouse
. In
Preimplantation Stages of Pregnancy
(ed.
G. E. W.
Wolstenholme
&
M.
O’connor
), pp.
194
-
207
.
London
:
Churchill
.
Mulnard
,
J. G.
(
1965
).
Studies of regulation of mouse ova in vitro. In Preimplantation Stages of Pregnancy
(ed.
G. E. W.
Wolstenholme
&
M.
O’connor
), pp.
123
-
138
. London: Churchill.
Mulnard
,
J. G.
(
1971
).
Manipulation of cleaving mammalian embryos with special reference to a time-lapse cinematographic analysis of centrifuged and fused mouse eggs
. In
Advances in the Biosciences, no. 6: Schering Symposium on Intrinsic and Extrinsic Factors in Early Mammalian Development
(ed. G. Raspé), pp.
255
-
277
. Braunschweig: Pergamon Press.
Noyes
,
R. W.
(
1959
).
Trophoblast: problems of invasion and transport
.
Ann. N.Y. Acad. Sci.80, 54-61
.
Pierce
,
G. B.
,
Midgley
,
A. R.
,
Sri Ram
,
J.
&
Feldman
,
J. D.
(
1962
).
Parietal yolk sac carcinoma: clue to the histogenesis of Reichert’s membrane of the mouse embryo
.
Am. J. Path.
41
,
549
566
.
Psychoyos
,
A.
(
1961
).
Perméabilitié capillaire et décidualisation utérine
.
C. r. hebd. Séanc. Acad. Sci. Paris
252
,
1515
.
Runner
,
M. N.
(
1947
).
Development of mouse eggs in the anterior chamber of the eye
.
Anat. Rec.
98
,
1
17
.
Runner
,
M. N.
(
1951
).
Differentiation of intrinsic and maternal factors governing intrauterine survival of mammalian young
.
J. exp. Zool.
116
,
1
20
.
Seidel
,
F.
(
1960
).
Die Entwicklungsfahigkeiten isolierter Furchungszellen aus dem Ei des Kaninchens, Oryctolagus cuniculus
.
Wilhelm Roux Arch. EntwMech. Org.
152
,
43
130
.
Simmons
,
R. L.
&
Russell
,
P. S.
(
1966
).
The histocompatibility antigens of fertilized mouse eggs and trophoblast
.
Ann. N.Y. Acad. Sci.
129
,
35
45
.
Snell
,
G. D.
&
Stevens
,
L. C.
(
1966
).
Early embryology
. In
Biology of the Laboratory Mouse,2nd
ed. (ed.
E. L.
Green
), pp.
205
245
.
New York
:
McGraw-Hill
.
Stevens
,
L. C.
(
1968
).
The development of teratomas from intratesticular grafts of tubal mouse eggs
.
J. Embryol. exp. Morph.
20
,
329
341
.
Tarkowski
,
A. K.
(
1962
).
Inter-specific transfers of eggs between rat and mouse
.
J. Embryol. exp. Morph.
10
,
476
495
.
Tarkowski
,
A. K.
(
1966
).
An air-drying method for chromosome preparations from mouse eggs
.
Cytogenetics
5
,
394
400
.
Tarkowski
,
A. K.
&
Wroblewska
,
J.
(
1967
).
Development of blastomeres of mouse eggs isolated at the 4- and 8-cell stage
.
J. Embryol. exp. Morph.
18
,
155
180
.
Whitten
,
W. K.
(
1958
).
Endocrine studies on delayed implantation in lactating mice. Role of the pituitary in implantation
.
J. Endocr.
16
,
435
440
.