Sixteen inner or outer blastomeres from 16-cell embryos and 32 inner or outer blastomeres from 32-cell embryos (nascent blastocysts) were reaggregated and cultured in vitro. In 24 h old blastocysts developed from blastomeres derived from 16-cell embryos the expression of Cdx2 protein was upregulated in outer cells (new trophectoderm) of the inner cells-derived aggregates and downregulated in inner cells (new inner cell mass) of the external cells-derived aggregates. After transfer to pseudopregnant recipients blastocysts originating from both inner and outer blastomeres of 16-cell embryo developed into normal, fertile mice, but the implantation rate of embryos formed from inner cell aggregates was lower. The aggregates of external blastomeres derived from 32 cell embryo usually formed trophoblastic vesicles accompanied by vacuolated cells. In contrast, the aggregates of inner blastomeres quickly compacted but cavitation was delayed. Although in the latter embryos the Cdx2 protein appeared in the new trophectoderm within 24 h of in vitro culture, these embryos formed only very small outgrowths of Troma1-positive giant trophoblastic cells and none of these embryos was able to implant in recipient females. In separate experiment we have produced normal and fertile mice from 16- and 32-cell embryos that were first disaggregated, and then the sister outer and inner blastomeres were reaggregated at random. In blastocysts developed from aggregates, within 24 h of in vitro culture, the majority of inner and outer blastomeres located themselves in their original position (internally and externally), which implies that in these embryos development was regulated mainly by cell sorting.

Cell and developmental studies have clarified how, by the time of implantation, the mouse embryo forms three primary cell lineages: epiblast (EPI), primitive endoderm (PE), and trophectoderm (TE). However, it still remains unknown when cells allocated to these three lineages become determined in their developmental fate. To address this question, we studied the developmental potential of single blastomeres derived from 16- and 32-cell stage embryos and supported by carrier, tetraploid blastomeres. We were able to generate singletons, identical twins, triplets, and quadruplets from individual inner and outer cells of 16-cell embryos and, sporadically, foetuses from single cells of 32-cell embryos. The use of embryos constitutively expressing GFP as the donors of single diploid blastomeres enabled us to identify their cell progeny in the constructed 2n↔4n blastocysts. We showed that the descendants of donor blastomeres were able to locate themselves in all three first cell lineages, i.e., epiblast, primitive endoderm, and trophectoderm. In addition, the application of Cdx2 and Gata4 markers for trophectoderm and primitive endoderm, respectively, showed that the expression of these two genes in the descendants of donor blastomeres was either down- or up-regulated, depending on the cell lineage they happened to occupy. Thus, our results demonstrate that up to the early blastocysts stage, the destiny of at least some blastomeres, although they have begun to express markers of different lineage, is still labile.

We studied the developmental potential of single blastomeres from early cleavage mouse embryos. Eight-and sixteen-cell diploid mouse embryos were disaggregated and single blastomeres from eight-cell embryos or pairs of sister blastomeres from sixteen-cell embryos were aggregated with 4, 5 or 6 tetraploid blastomeres from 4-cell embryos. Each diploid donor embryo gave eight sister aggregates, which later were manipulated together as one group (set). The aggregates were cultured in vitro until the blastocyst stage, when they were transferred (in sets) to the oviducts of pseudopregnant recipients. Eighteen live foetuses or pups were obtained from the transfer (11.0% of transferred blastocysts) and out of those, eleven developed into fertile adults (one triplet, one pair of twins and four singletons). In all surviving adults, pups and living foetuses, only diploid cells were detected in their organs and tissues as shown by analysis of coat pigmentation and distribution of glucose phosphate isomerase isoforms. In order to explain the observed high rate of mortality of transferred blastocysts, in an accompanying experiment, the diploid and tetraploid blastomeres were labelled with different fluorochromes and then aggregated. These experiments showed the diploid cells to be present not only in the inner cell mass (ICM) but also in the trophectoderm. The low number of diploid cells and the predominance of tetraploid cells in the ICM of chimaeric blastocysts might have been responsible for high postimplantation mortality of our experimental embryos.

Spontaneous diploid-triploid chimaeras occur sporadically in various mammalian species including man, but so far have never been produced experimentally. In order to get a deeper insight into the developmental consequences of this anomaly, we have developed two procedures that enabled for the first time to produce routinely diploid-triploid embryos, foetuses, and animals in the mouse. These procedures are: (1) aggregation of cleaving diploid embryos with triploid embryos produced by suppression of the second polar body in zygotes, and (2) fusion of a haploid karyoplast with one blastomere of the two-cell diploid embryos. The first procedure yielded 23 living and 6 dead postimplantation embryos and foetuses (age: 8th-19th day) out of which 22 were chimaeric. In addition, three chimaeric neonates reached adulthood. Two animals were fertile, and one--an overt chimaera--was an infertile male. The rate of postimplantation development of aggregation chimaeras was normal or only slightly retarded, and with one exception the foetuses were morphologically normal. Generally, the highest contribution of the 3n component in extra-embryonic structures was noted in the yolk sac, and usually it was higher than its contribution to the organs of the body. Chimaerism was most often noted in the liver, the heart, the intestine, and the lungs. Participation of triploid cells to all tissues studied, both in the body and in extra-embryonic structures, appeared to decrease slightly as development progressed. The second procedure yielded 10 foetuses and 6 adults. Three foetuses were chimaeric. Six fertile adults were probably non-chimaeras: the triploid component was absent in the coat and in the blood.

Blastocysts obtained from mice differing in pigmentation (albino versus pigmented) and the isoforms of glucose phosphate isomerase (GPI 1A versus 1B) were electrofused and those containing a single chimaeric inner cell mass (ICM) were transferred to the uterus of pseudopregnant recipients. The pups were recovered on the 20 th day by Caesarian section and fostered by females that had littered on the previous night or 24 h earlier. Altogether nine adult animals and two pups, which died soon after delivery, were available for GPI analysis. Between 9 and 13 organs/ tissues were examined and the relative contribution of the GPI 1A and 1B isoforms was estimated using an electrophoretic GPI assay. Eight adult animals were overtly chimaeric and one was chimaeric in some internal tissues only. Eight mice were males: seven were fertile, one was infertile. The ninth adult mouse was a hermaphrodite. The fertile animals produced sperm of one genotype only, i.e. derived either from the albino or from the pigmented component. This is the first report showing that adult chimaeras can be produced from two combined blastocysts, provided that fusion of the adhering trophectoderm cells is first induced and the orientation of blastocysts enables the two ICMs to integrate into a single ICM. Our results suggest that in the preimplantation blastocyst, the organisation of the ICM remains labile thus making it possible for the fused blastocysts to establish new embryonic organisation and to develop into a single organism.