In vivo and in vitro studies on the early embryonic lethal oligosyndactylism (Os) in the mouse

The mutation oligosyndactylism (O_S) causes fusion of the second and third digits in all four feet of mice heterozygous for the allele (Grüneberg, 1956, 1961). Anomalies of muscular anatomy in the limbs (Kadam, 1962) and a form of diabetes insipidus (Falconer, Latyszewski & Isaacson, 1964) also usually accompany the mutation, with high renal output and inability to concentrate urine arising from a paucity of nephrons in the kidneys (Stewart & Stewart, 1969).

No satisfactory explanation, however, has been advanced to link these effects with the embryonic lethality which Os causes in the homozygous condition and which is the subject of this study. The time of death of Os/Os embryos in utero has been determined in an histological investigation by Van Valen (1966), who found that abnormalities first occur on day 5 post coitum, leading to death during the following 24 h with subsequent resorption. Affected blastocysts were found to possess increasing numbers of pale-staining cells, lacking a nuclear membrane and nucleolus, and in which the chromatin lay scattered and fragmented in the cytoplasm. The frequent occurrence of these cells in pairs together with the general resemblance of the chromatin to mitotic figures, led the author to postulate that the mutation interferes with cell division. It was suggested that cells undergoing the seventh or eighth cleavage are either unable to complete division or cannot successfully re-enter interphase, with resultant cytostasis and developmental arrest.

The purpose of this study was to confirm Van Valen’s findings and to examine in more detail the nature of the abnormalities described. To this end, the results of an histological study of Os/Os embryos in utero, together with an examination of chromosome morphology in air-dried spreads ‘of the affected cells, are presented here. These are further supported by observations of the behaviour of the mutant embryos when cultured over the lethal phase.

The adult mice used in this study were F₁ hybrids obtained from matings between males heterozygous for oligosyndactylism (Os/ + ) derived from linkage stocks at the Institute of Animal Genetics, Edinburgh, and wild-type females ( + / + ) of the highly fertile CD-I strain (Charles River). This cross was adopted as a rapid method of producing stocks of experimental animals.

Matings between F₁ mice heterozygous for the Os allele (Os/+ × Os/+ ) were employed to produce litters containing homozygous mutant embryos. Control litters were obtained by mating wild-type males with heterozygous females (Os/ + × +/ + ), thereby ensuring that any unforseen effects of the mutation on the maternal reproductive system or on the cytoplasm of the ova would be equalized in both groups.

Copulation was assumed to have occurred at the mid-point of the dark period preceding discovery of a vaginal plug. For convenience of timing, animals were maintained in an artificially reversed diurnal light cycle.

Air-dried spreads of whole embryos were prepared according to an adaptation of the method described by Evans, Burtenshaw & Ford, (1972). Blastocysts were flushed from uteri at approximately 115 h post coitum, this being the latest time at which reasonably intact embryos could be obtained, although yield was often low. These were incubated for 15 min at 37 °C in hypotonic solution (0-8 % sodium citrate in H₂O), followed by a 30 min fixation in three parts ethanol and one part glacial acetic acid. Each embryo was then transferred in a minimum volume of fixative onto a clean microscope slide, and a small drop of 60% acetic acid added immediately. As soon as the embryo started to disaggregate, the drop was drawn across the slide with the tip of a micropipette, leaving a trail of ruptured cells adhering to the surface. Preparations were allowed to dry for 30 min at room temperature before staining with Geimsa or Aceto-orcein.

Blastocysts at approximately 95 h post coitum were flushed from uteri with PBI medium (Whittingham & Wales, 1969). Zonae pellucidae were removed at this stage by gently shearing-off with with a small-bore micropipette. Embryos were washed and transferred into individual droplets of Eagle’s Minimal Essential Medium supplemented with 20% foetal calf serum, 1·6 mM glutamine and 100 units/ml penicillin. Droplets were maintained under paraffin oil in Falcon dishes at 36-5 °C in an atmosphere of 5 % CO₂ in air.

The progress of individual embryos was recorded daily following examination under phase-contrast optics. After 72 h of culture, the number of outgrowing trophoblast cells (all sizes) in each embryo was determined, and the area of outgrowth measured with a calibrated micrometer grid.

A total of 45 embryos in six uteri from Os/ + x Os/ + matings were sectioned at this stage. The majority were present as late blastocysts/early egg-cylinders (Fig. 1C). Twelve embryos (26·7 %), however, were clearly abnormal (Fig. 1a, b). These were presumed to be Os/Os on account of the overwhelming similarity to the mutant homozygotes previously described by Van Valen (1966), and of the excellent statistical agreement with the Mendelian expectation of 25%. Gross morphology was usually somewhat mis-shapen, and frequently the blastocoele had collapsed so that the trophectoderm was folded up around the inner cell mass. Nevertheless, in most cases signs of attachment and initial implantation could clearly be seen. On average, approximately 40 % of the cells in these embryos were in a state resembling metaphase, and they were consequently quite distinct from their litter-mates which typically displayed only one tenth of this proportion of metaphases (Table 1).

Cell number was estimated according to the formula of Abercrombie (1946) for nucleated cells, and by direct counting for mitotic cells. The mean number of mitotic-type cells per embryo was thus calculated as 38·0 (±2·8) for Os homozygotes and as 5·6 (±0·7) for litter-mates, while mean total cell number per embryo was estimated as 96·0 (±9·5) and 130·0 (±6·7) for each group respectively.

Student’s t-test confirms that cell number in the Os homozygotes is significantly lower than in litter-mates (P < 0-01).

Normal embryos at this time possessed well-formed egg cylinders. Attachment was now complete, and each implantation site was surrounded by a marked decidual swelling. A notable feature was the presence of large invasive trophoblastic giant cells.

Ten (28·6 %) out of 35 embryos sectioned, however, were in a state of advanced degeneration. This proportion is not significantly different from 25 % (χ² = 0-24). In all cases apparently normal decidual responses had been induced, although the embryos themselves had become little more than disorganized masses of necrotic cells and cellular debris.

Of particular interest, however, was the presence of trophoblastic giant cells embedded in the decidual tissue immediately around the embryonic remains (Fig. 1d). These giant cells were a feature of all the degenerate implantation sites, and the majority appeared normal in respect of both size and morphology when compared with those of litter mates.

The highly disorganized state of the great majority of cells in these embryos precluded any meaningful analysis of cell number or mitotic index; however, their comparatively advanced state of degeneration makes it seem unlikely that any significant amount of growth or successful cell division had taken place during the preceding 24 h.

Forty three air-dried preparations of -day blastocysts from Os/ + x Os/ + matings were examined. In general, these consisted of large numbers of scattered nuclei, interspersed by a few metaphase spreads.

Nine (21 %) of the preparations were very different in respect of two criteria. Firstly, they displayed an extremely high proportion of metaphase-type spreads sometimes in such abundance that it was almost impossible to distinguish one set of chromosomes from another (Fig. 2a). Of those preparations sufficiently well spread to be countable, a mean mitotic index of 34% was calculated, in contrast to the figure of 2 % obtained for their normal litter mates. (It should be noted that mitotic indices may be prone to underestimation in these preparations, since vigorous cell spreading is apt to lead to the loss of integrity of chromosome groups.) Secondly, within each of these embryos, the morphology of the chromosomes differed widely, ranging through varying degrees of condensation from normal-looking metaphase spreads to tiny blobs of chromatin scarcely larger than a centromere. Chromosomes were, however, of uniform appearance within any single metaphase. The more normal types of spread contained approximately 40 telocentric pairs of sister chromatids (mean = 38-6), whereas the chromatin blobs of the very highly condensed types were nearly always present in groups of about 80 (mean = 76-4). Close scrutiny of this latter kind suggested strongly that they had arisen from the separation and extreme spiralization of previously paired sister chromatids, since in some cases the chromatin blobs were still loosely associated in pairs (Fig. 2b). This interpretation is also supported by the fact that chromosome number never exceeded 40 in the less condensed spreads containing clearly intact chromatid pairs.

As with the sectioned material, morphology of interphase nuclei in these presumed mutant embryos appeared normal.

The developmental behaviour of 149 zona-free blastocysts from Os/ + × Os/ + matings, and of 96 from Os/ + × + / + matings was observed during culture over the lethal phase.

No normal class was visible in either experimental or control groups at the commencement of culture (4 days p.c). Typically, during the next 24 h, embryos had become attached to the surface of the culture dish, and by 48 h (6 days p.c) were surrounded by a monolayer of outgrowing trophoblast. The inner cell masses (ICMs) also became enlarged, and after 72 h of culture (7 days p.c.) could be seen protruding upwards from the centre of the trophoblast layer.

In 36 embryos (24-2%) from inter-heterozygote matings, however, this enlargement of the ICM failed to occur (Table 2). Indeed, after 48 h of culture the ICM appeared markedly degenerate, and by 72 h had disintegrated almost completely. In some cases clumps of necrotic cells remained clinging to the sur-face of the trophoblast, but in others only a naked layer of trophoblast giant cells remained. The morphology of these trophoblast cells seemed normal, although there was a greater uniformity of size, arising from the absence of smaller diploid or less giant, trophoblast cells which were present in the central area of litter mates and controls. Viability too seemed unaffected since the cells remained healthy and were of normal appearance after a further 24 h of culture.

By contrast, similar degeneration of the ICM was observed in only two of the control embryos (2·1 %).

After proportionate correction for non-specific abnormality from the control data, the χ² test indicates that the number of outgrowths with degenerate ICMs in the experimental group (22·1%) is not significantly different from 25% (χ² = 0·66). It is thus concluded that these were Os homozygotes.

Since initial observations suggested that the trophoblastic outgrowth of these mutant embryos was smaller than normal, direct measurements were made both of the number of outgrowing trophoblast cells and of the area of outgrowth after 72 h culture. This procedure also enabled a more accurate assessment to be made of the average size of individual trophoblast cells in each group. A total of 91 outgrowths from the experimental group (of which 21 were presumptive Os/Os), together with the 96 outgrowths from the control group were thus analysed. The results show (Table 3) that the Os homozygotes are signi ficantly smaller than litter mates and controls, both in terms of total size and in number of trophoblast cells. There is, however, no evidence for a reduction in mean trophoblast cell size in the mutant outgrowths.

The results of the histology in this study substantially corroborate those of the original investigation performed by Van Valen (1966). In addition, the air-dried preparations of Os/Os embryos confirm her suspicion that the abnormal cells containing free chromatin bodies are blocked in mitosis; more precisely, they appear unable to progress beyond metaphase.

Cell counts of day-5 embryos rule out any possibility that the high proportion of metaphase cells seen in the mutants represents an increased rate of cell division, since cell number is found to be substantially reduced. The morphology of the chromosome spreads also opposes this; extreme chromosomal condensation as observed in Os/Os embryos is not seen in normally dividing cells but may be produced by prolonged treatment with spindle poisons such as colcemid. In particular it is significant that the discrepancy in mean cell number between mutant embryos and litter mates (96 as opposed to 130) can almost exactly be accounted for if the 40 % mitotic cells observed in the former group have been blocked during division. These figures support Van Valen’s conclusion that the abnormality sets in during the seventh or eighth cleavage; for, if the mutation affects all cells which have undergone a specific number of cycles, then allowing for the accumulation of asynchrony during cleavage, a mean cell number of 96 indicates a blockage occurring at the eighth mitotic division.

A further significant feature of the histology was seen in uteri sectioned at 6 days. Whilst it is not surprising that the mutant embryos have induced a decidual response, since some initial stages of implantation have taken place on the previous day, the presence of apparently healthy trophoblastic giant cells surrounding the implantation site of an otherwise necrotic embryo is of considerable interest. The behaviour of Os/Os embryos in vitro mirrors the events which take place in utero with respect to the persistence of trophoblastic giant cells after the degeneration of the ICM. This refractoriness of giant cells to the action of the lethal gene is most readily explained in terms of their cell cycle, for these cells almost certainly attain their highly polyploid state by endoreduplication (Chapman, Ansell & McLaren, 1972; Gearhart & Mintz, 1972). Trophoblast cells which have initiated endoreduplication prior to the onset of Os gene action will consequently remain unaffected by a lesion which exerts its principal influence on dividing cells.

The mitotic apparatus is a probable site of disruption in Os/Os cells for two reasons. First, as seen in sectioned embryos, the chromosomes of affected cells often exhibit a much more disorganized appearance than is observed in normal metaphases, indicating that the spindle (if present) is incapable of organizing the chromosomes into an orderly metaphase plate. Secondly, cultured wild-type embryos, in which spindle formation is inhibited with colcemid, develop in a very similar fashion to Os/Os embryos, exhibiting selective degeneration of the ICM (Sherman & Atienza, 1975).

A singular aspect of the Os allele is that it interferes with a fundamental function which has proceeded normally for several days prior to the lethal phase. One explanation of such an effect occurring in early development is that the mutant embryo’s own genome may be unable to manufacture a substance required for cell division, which has hitherto been present in sufficient reserves within the cytoplasm. Such a substance would be subject, during successive cleavages, to rapid and predictable dimunition in quantity per cell, thus accounting satisfactorily for the pronounced phase specificity exhibited by the Os allele.

EM studies have demonstrated microtubular abnormalities associated both with the somewhat analogous lethal-mitotique (lm) mutation in the newt, which blocks cells in metaphase shortly after hatching (Gounon & Collenot, 1974), and with lethal-polyploid (l(3)pl) in Drosophila hydei, which causes mitotic disturbances in certain groups of cells during metamorphosis (Rungger-Brandle, 1977 a, b). A similar approach may prove able to reveal any structural peculiarities of the mitotic apparatus in Os homozygotes during the lethal-phase.

The author wishes to express his appreciation of the advice and encouragement given by Dr Michael Snow and Dr Anne McLaren.

The work was supported by a Medical Research Council studentship.