HEPATOCYTE growth factor/scatter factor (HGF/SF) functions as a mitogen, motogen and morphogen for a variety of cultured cells1–7. The genes for HGF/SF and its receptor (the c-met protooncogene product8) are expressed in many tissues during the embryonic periods and in the adult9–14. HGF/SF is thought to mediate a signal exchange between the mesenchyme and epithelia during mouse development15. To examine the physiological role of HGF/SF, we generated mutant mice with a targeted disruption of the HGF/SF gene. Here we report that homozygous mutant embryos have severely impaired placentas with markedly reduced numbers of labyrinthine trophoblast cells, and die before birth. The growth of trophoblast cells was stimulated by HGF/SF in vitro, and the HGF/SF activity was released by allantois in primary culture of normal but not mutant embryos. These findings suggest that HGF/SF is an essential mediator of allantoic mesenchyme-trophoblastic epithelia interaction required for placental organogenesis.

Placental defect and embryonic lethality in mice lacking hepatocyte growth factor/scatter factor

The development of 181 surplus human embryos, including both normally and abnormally fertilized, was observed from day 2 to day 5, 6 or 7 in vitro. 63/149 (42%) normally fertilized embryos reached the blastocyst stage on day 5 or 6. Total, trophectoderm (TE) and inner cell mass (ICM) cell numbers were analyzed by differential labelling of the nuclei with polynucleotide-specific fluorochromes. The TE nuclei were labelled with one fluorochrome during immunosurgical lysis, before fixing the embryo and labelling both sets of nuclei with a second fluorochrome (Handyside and Hunter, 1984, 1986). Newly expanded normally fertilized blastocysts on day 5 had a total of 58.3 +/- 8.1 cells, which increased to 84.4 +/- 5.7 and 125.5 +/- 19 on days 6 and 7, respectively. The numbers of TE cells were similar on days 5 and 6 (37.9 +/- 6.0 and 40.3 +/- 5.0, respectively) and then doubled on day 7 (80.6 +/- 15.2). In contrast, ICM cell numbers doubled between days 5 and 6 (20.4 +/- 4.0 and 41.9 +/- 5.0, respectively) and remained virtually unchanged on day 7 (45.6 +/- 10.2). There was widespread cell death in both the TE and ICM as evidenced by fragmenting nuclei, which increased substantially by day 7. These results are compared with the numbers of cells in morphologically abnormal blastocysts and blastocysts derived from abnormally fertilized embryos. The nuclei of arrested embryos were also examined. The number of TE and ICM cells allocated in normally fertilized blastocysts appears to be similar to the numbers allocated in the mouse. Unlike the mouse, however, the proportion of ICM cells remains higher, despite cell death in both lineages.

The human blastocyst: cell number, death and allocation during late preimplantation development in vitro

Mouse pre-implantation development gives rise to the blastocyst, which is made up of at least three distinct cell types: the trophectoderm (TE) that surrounds a cavity, and an inner cell mass (ICM) comprising the primitive endoderm (PE) and epiblast (EPI). However, the underlying mechanisms involved in patterning the cleavage-stage embryo are still unresolved. By analyzing the distribution of the transcription factors Oct4 (Pou5f1), Cdx2 and Nanog at precisely defined stages in pre-implantation development, we were able to identify critical events leading to the divergence of TE, EPI and PE lineages. We found that Oct4 is present in all cells until late blastocyst, gradually disappearing from the TE thereafter. The expression patterns of both Cdx2 and Nanog exhibit two specific phases, culminating in their restriction to TE and EPI, respectively. In the first phase, starting after compaction, blastomeres show highly variable Cdx2 and Nanog protein levels. Importantly, the variability in Nanog levels is independent of position within the morula, whereas Cdx2 variability may originate from asymmetric cell divisions at the 8-cell stage in a non-stereotypic way. Furthermore, there is initially no reciprocal relationship between Cdx2 and Oct4 or between Cdx2 and Nanog protein levels. In the second phase, a definite pattern is established, possibly by a sorting process that accommodates intrinsic and extrinsic cues. Based on these results, we propose a model in which early embryonic mouse patterning includes stochastic processes, consistent with the highly regulative capacity of the embryo. This may represent a feature unique to early mammalian development.

Stochastic patterning in the mouse pre-implantation embryo.

Publisher Summary This chapter discusses the functional aspects of satellite DNA and heterochromatin. Despite all attempts to formulate simple rules governing satellite evolution, it is now clear that each case so far analyzed has brought with it its own claims for generalization, none of which have proven sufficiently all-embracing. One initial hypothesis on satellite evolution was that satellites wax and wane with amazing rapidity in evolutionary terms, so that closely related species differ drastically in amount or type of satellite. However, improved methods of DNA sequencing have led to the suggestion that closely related species appear to modulate their satellites from a common library. Because some kinds of heterochromatin are now known to contain satellite DNA and because a large literature exists on the properties of heterochromatin, it is necessary to consider the relationships between heterochromatin and satellite DNA in some detail. The term “heterochromatin” was initially used to define chromosomes or chromosome segments which did not uncoil at mitotic telophase and so maintained a condensed or heteropycnotic state throughout interphase and into the subsequent prophase of the next division cycle.

Functional aspects of satellite DNA and heterochromatin.

Appearance and distribution of collagens and laminin in the early mouse embryo.

1. Trophoblast development in utero in the first half of gestation is accompanied by a rise in the ploidy levels in giant cell nuclei. Giant cells constitute up to one-third of the trophoblast population over this period. 2. The expression of polyploidy is independent of the maternal environment; the extent of polyploidy and the degree to which trophoblast develops morphologically under various conditions appear to be related. 3. The following events are not necessary for the initiation of polyploidization in trophoblast: (a) hatching from the zona pellucida, (b) oestrogen sensitization, (c) implantation, (d) postimplantation changes in topology. 4. That earlier developmental events such as the formation of the blastocoel cavity may in some way trigger polyploidization of trophoblast is not inconsistent with the studies described.

The biochemistry of differentiation of mouse trophoblast: studies on polyploidy

Strain PO preimplantation embryos were labelled with [ 3 H]thymidine. The incorporation of the label was studied by autoradiography of air-dried and serially sectioned embryos. DNA amounts were measured with a microdensitometer. The following observations were made at the 5- to 16-cell stages. All nuclei contained 2 C –4 C amounts of DNA and all could eventually synthesize DNA. However, after short labelling intervals, unlabelled nuclei were found with 2C and AC amounts of DNA. We concluded that both the G 1 and the G 2 phases of the cell cycle were present at this time. Embryos were found in which the S phase of the 4th and the 5th cell cycles post fertilization overlapped. In 12- to 15-cell embryos which contained inside cells it was found that the inside cells were produced by one of the first four 8-cell-stage blastomeres to divide. The inside cells of 9- to 256-cell embryos had a significantly higher labelling index than the outside cells and the number of inside cells increased faster than the number of outside cells during development. We concluded that the inside cells were dividing faster than the outside cells. Blastocysts which had developed in vivo or in vitro contained nuclei with greater than 4 C amounts of DNA. We concluded that the development of excess DNA amounts does not depend on the maternal environment. These nuclei which contained greater than 4 C amounts of DNA were labelled after short exposures to radioactivity. We concluded that it was likely that they were becoming polyploid.

DNA synthesis in the preimplantation mouse embryo.

SUMMARYIn Listera ovata the chiasma frequency is higher in embryo sac mother cells than in pollen mother cells. In both types of cell the chiasma frequency is raised by the presence of euchromatic B-chromosomes. The rise is not linear but zig-zag with odd numbers of the B-chromosome having a greater effect than even numbers. It is suggested that B-chromosomes prolong the time available for chiasma formation allowing the utilization of more sites for chiasma formation so accounting for the rise in chiasma frequency.

Meiosis and B-Chromosomes in Listera Ovata (Orchidaceae)

The presence of the partly heterochromatic supernumerary chromosomes in pollen mother cells of Puschkinia libanotica raises the chiasma frequency in the A chromosome bivalents. The pattern of chiasma distribution along each of the five A bivalents was related to the DNA labelling pattern of mitotic chromosomes. Regions that showed heavy labelling at the end of the DNA synthetic phase had fewer chiasmata than lightly labelled regions. As this relation is the opposite to that found by Rees and Evans in another species we regard any correlation between labelling pattern and chiasma distribution as fortuitous.

The effect of supernumerary chromosomes on meiosis in Puschkinia libanotica (Liliaceae)

The chromosome complement of Puschkinia libanotica is described. In addition to five pairs of A chromosomes plants may possess up to 7 B chromosomes. Part of the long arm of the B chromosome gives rise to a heterochromatic mass in interphase nuclei and this can be seen to be a double structure in G1 nuclei and a quadruple structure in G2 nuclei. It is believed that these configurations represent the pre- and post-replication forms of subchromatids in the heterochromatic segment of the B chromosome. Microdensitometry of metaphase chromosomes shows that the segment of the B chromosome that is heterochromatic during interphase has no more DNA per unit volume than any of the euchromatic A chromosomes.

Peter W. Barlow

Papers

The biochemistry of differentiation of mouse trophoblast: studies on polyploidy

DNA synthesis in the preimplantation mouse embryo.

Meiosis and B-Chromosomes in Listera Ovata (Orchidaceae)

The effect of supernumerary chromosomes on meiosis in Puschkinia libanotica (Liliaceae)

The chromosomes of Puschkinia libanotica during mitosis