ABSTRACT
Blastomeres isolated from 4- and 8-cell mouse eggs were inserted into empty zonae and transferred to the oviduct. The products of both types of blastomere were capable of inducing decidual formation. One implant produced by an isolated blastomere from a 4-cell egg contained a small, retarded embryo at days but most decidua from blastomeres of either 4- or 8-cell eggs contained only a few trophoblast giant cells. It is suggested that this lack of totipotency is due to insufficient cells being present at cavitation rather than restriction in developmental potential.
INTRODUCTION
The prospective potency of isolated cleavage-stage blastomeres has been studied in several mammalian species in order to discover whether the morphological similarity between early blastomeres is reflected in their developmental potential (see review, Wilson & Stern, 1975). In the mouse, adult animals have been produced from 1/2 blastomeres (Tarkowski, 1959a, b:,Hoppe & Whitten, 1972) but the development of 1 /4 and 1/8 blastomeres has only been studied as far as the blastocyst stage (Tarkowski, 1959a, b;Mulnard, 1965; Tarkowski & Wroblewska, 1967; Sherman, 1975). (The notation used is that of Tarkowski & Wroblewska (1967), where 1/4 blastomere is one blastomere from a 4-cell egg, 1/8 blastomere is one blastomere from an 8-cell egg, and so on.) The present experiments were designed to assess the postimplantation potential of isolated 1/4 and 1/8 mouse blastomeres after development in vivo. If isolated blastomeres are obtained by destroying all other blastomeres within the zona, incompletely destroyed cells and cell debris remaining inside the zone might interfere with development, and so blastomeres were isolated by disaggregating zona-free cleavage stages. The blastomeres were then inserted into empty zonae prior to transfer to the oviduct, since naked blastomeres and cleavage stages do not survive in the oviduct in vivo (Bronson & McLaren, 1970; Mod-linski, 1970).
MATERIALS AND METHODS
Recovery of embryos from donor females
CFLP mice (Anglia Laboratory Animals Ltd) were used throughout this study. Cleaving eggs for preparation of isolated blastomeres were obtained following natural mating. Four-cell eggs were flushed from the oviducts between 8.00 and 9.30 and 8-cell eggs between 12.00 and 15.00 on the 3rd day of pregnancy. Fertilized or unfertilized one-cell eggs for the preparation of empty zonae were obtained from the ampullary region of the oviduct on the morning after mating with fertile or vasectomized males. The cumulus cell mass was removed by incubation in hyaluronidase solution (100 i.u./ml phosphate buffered saline). PB1 +10 % foetal calf serum (Whittingham & Wales, 1969) was used for recovery, storage, microsurgery and transfer of embryos.
Preparation of isolated blastomeres and insertion into zonae
The zonae were removed from 4- and 8-cell eggs by incubation in pronase (Mintz, 1962). Blastomeres were separated by incubation in a 0·25 % solution of trypsin for 20 min at 37 °C. A single 1/4 or 1/8 blastomere was inserted into each empty zona prepared by sucking out the contents of a fertilized or unfertilized egg (Rossant, 1975; Tarkowski & Rossant, 1976). Where possible all blastomeres from each egg were inserted into zonae. A control experiment was also performed to test whether this procedure damaged the blastomeres in any way. Blastomeres from single 4-cell eggs were separated into pairs and one such pair injected into a zona from a fertilized egg and the other into the zona from an unfertilized egg. For brevity, the zonae themselves will be termed fertilized and unfertilized from now on.
Transfer to pseudopregnant recipients
Isolated 1/4 and 1/8 blastomeres in zonae were transferred to the oviducts of recipient females on the first day of pseudopregnancy. Where possible in the control series, 2/4 blastomeres in fertilized zonae were transferred to one oviduct while 2/4 blastomeres in unfertilized zonae were transferred to the opposite oviduct in the same mouse.
Examination of preimplantation development
The oviducts of some recipient females were flushed 2 days after transfer and the flushings examined for zonae containing blastomere derivatives. The oviducts of other recipients were fixed 2 days after transfer, processed and embedded in wax. Serial sections were cut at 5–6 μm and stained with haemalum and eosin. The structures derived from isolated blastomeres were identified in the sections and, where possible, cell counts were made by serial reconstruction from camera lucida drawings or photographs.
Examination of postimplantation development
Recipient females were killed on the 6th day of pregnancy and their uteri fixed and cleared (Orsini, 1962), so that even very small decidual swellings could be detected. All such swellings were embedded in wax and sectioned at 6 μm. The sections were stained with haemalum and eosin and examined for the presence of embryonic derivatives in the decidua.
RESULTS
Development of 2/4 blastomeres in fertilized and unfertilized zonae
Histological analysis at days established that 2/4 blastomeres in both fertilized and unfertilized zonae could induce decidual formation and give rise to normal egg cylinders (Table 1). Some abnormal egg cylinder-like structures were also found. The percentages of normal and abnormal egg cylinders formed are very similar for 2/4 blastomeres in both types of zona (Table 1).
The low percentage for embryo formation and the occasional formation of abnormal, badly organized structures is probably due to use of half-size embryos. Such embryos are known to develop less well than normal embryos, although development to term has been achieved (Tarkowski, 1959a, b; Gardner, 1974). However, production of some morphologically normal egg cylinders showed that the procedure adopted for the study of isolated blastomeres is compatible with continued development. The results also showed that the type of zona used had no effect on blastomere development.
Development of isolated 1/4 blastomeres
The proportion of transferred blastomeres identified after 2 days in the oviduct was rather low. Nevertheless, six morphologically normal ‘1/4’ blastocysts were recovered (Table 2, Fig. 1). Cell counts were made on four of the ‘1/4’ blastocysts recovered. The total number of cells was 13–14, of which 2–3 were inner cell mass (ICM) cells. Three groups of disorganized trophectoderm-like cells were also recovered (Table 2). The number of cells in these was lower than in the blastocysts (ca. 4-10 cells by visual estimation).
Blastocyst developed from 1/4 blastomere in zona after 2 days in the oviduct.
Isolated 1/4 blastomeres in zonae were capable of initiating a decidual reaction in recipient females (Table 3). However, in only 1 out of 13 decidua was a recognizable embryo detected (Fig. 2). This embryo was apparently retarded by approximately 12–15 h since it lacked extra-embryonic ectoderm. It was smaller than normal embryos of similar recipient age or embryonic stage. However, it was not abnormal and contained proliferating polar trophoblast, mural giant cells, proximal endoderm, as well as embyonic ectoderm. Little distal endoderm was present. It is unlikely that it would have survived to term.
Postimplantation development of isolated blastomeres in zonae Recipients killed days p.c., only pregnant females considered

Of the remaining 12 decidua, five had no detectable embryonic cells. In the other seven implants, no ICM derivatives were found, but groups of trophoblast giant cells were detected. The number of giant cells varied widely (range 5-16, average 11) but did not exceed the maximum number of cells that could be derived by four cell divisions from one isolated 1/4 blastomere (24 cells).
Development of isolated 1/8 blastomeres
Developing isolated 1/8 blastomeres were identified 2 days after transfer (Table 2). No true blastocysts containing enclosed cells were found. Six out of eight recovered structures consisted of disorganized trophectoderm-like cells (Fig. 3) while the other two were ‘false blastocysts’ (Tarkowski & Wroblewska, 1967), in which no cells were actually enclosed.
Section of group of cells developed from 1/8 blastomere in zona after 2 days in the oviduct.
Isolated 1 /8 blastomeres were capable of inducing decidua, despite their small size (Table 3). No organized embryo was found in any of the decidua, but four contained trophoblast giant cells. The number of giant cells found was lower than for 1 /4 blastomeres (range 2–9, average 5) and did not exceed (except by one cell in one instance) the maximum number of cells that could be derived by three cell divisions from an isolated 1 /8 blastomere (23 cells). Two further decidua contained small lumps of cells whose nature could not be accurately determined due to poor fixation, but visual observation suggested that they were of embryonic rather than maternal derivation. The remaining three decidua were empty.
Unfortunately, the low implantation rates of both 1 /4 and 1 /8 blastomeres made it impractical to attempt to assess the developmental potential of all blastomeres from one egg.
DISCUSSION
In the present study, isolated 1/4 mouse blastomeres usually formed small blastocysts in vivo as reported by previous workers (Tarkowski, 1959a, b. However, when the blastomeres were allowed to proceed to implantation, only one egg cylinder was found in 13 implants analysed at days p.c. This embryo was small and retarded (Fig. 2). Most other implants contained trophoblast giant cells only. Isolated 1/8 blastomeres never formed true blastocysts in the present experiments (c.f. Tarkowski & Wroblewska, 1967) and, although capable of implanting and forming trophoblast giant cells, never produced egg-cylinder structures. Thus, isolated 1 /4 and 1 /8 blastomeres seem unable to form complete foetuses (c.f. rabbit blastomeres, Moore, Adams & Rowson, 1968). However, Kelly (1975) has shown that 1/4 and 1/8 blastomeres can contribute to both ICM and trophectoderm derivatives in postimplantation conceptuses, when combined with enough ‘carrier’ blastomeres to restore the normal cell number. Thus, the apparent lack of totipotency in the present experiments is probably not due to restriction in developmental potential of the cells. Since blastocyst formation begins at the same time in isolated blastomeres as in intact eggs (Tarkowski & Wroblewska, 1967), insufficient cells may be present at the time of cavitation of 1/4 and 1 /8 blastomeres to produce a functional ICM. Isolated 1 /4 blastomeres normally form blastocysts containing 2–3 ICM cells but only one egg cylinder was produced from such a structure. This suggests that a certain minimum number of ICM cells (probably greater than 3) is required for postimplantation embryonic development (Snow, 1976). On the other hand, the formation of functional trophoblast cells does not seem to be greatly affected by cell number during cleavage since structures containing 1/8 of the normal preimplantation cell number can implant and transform into trophoblast giant cells.
ACKNOWLEDGEMENTS
I should like to thank Dr R. L. Gardner for valuable discussion and Mrs L. Ofer for technical assistance. The author was supported by a Medical Research Council Research Studentship and a Beit Memorial Junior Research Fellowship.