Embryos homozygous for the velvet coat mutation, Ve/Ve, were recognized at 6·5 days post coitum by the reduced size of the ectodermal portions of the egg cylinder and the loose, columnar nature of the overlying endoderm. Later in development ectoderm tissues were sometimes entirely absent. Abnormalities appeared in the ectoplacental cone at 8-5 days but trophoblast giant cells and parietal endoderm appeared unaffected. Homozygotes could not be unequivocally identified at 5·5 days nor at the blastocyst stage but were recognized in blastocyst outgrowths by poor development of the inner cell mass derivatives, It has previously been suggested that Ve may exert its action at the blastocyst stage by reducing the size of the inner cell mass, but no evidence for such a reduction was found. Most of the observations on Ve/Ve homozygotes are, however, consistent with the hypothesis that Ve exerts its action primarily on later primitive ectoderm development.

Embryonic lethal mutations in the mouse have been widely studied in an effort to gain insight into the genetic control of early developmental processes (reviewed McLaren, 1976). One such early acting lethal which has not yet been extensively studied is the semidominant mutation, velvet coat (Ve) (Stieler & Hollander, 1972). In the heterozygote, Ve acts on the coat to produce a thick, velvety appearance but Ve/Ve homozygotes die soon after implantation. Initial reports suggested that Ve might act by preventing normal primitive ectoderm formation. Diwan & Stevens reported in Mouse News Letter in 1974 and in McLaren (1976) that Ve/Ve embryos could be recognized by the deficiency or absence of primitive ectoderm cells at the time of implantation, although most other tissues appeared normal. The embryos were resorbed by 9–10 days of pregnancy without forming mesoderm. Diwan and Stevens also suggested that Ve/Ve homozygotes might be recognizable at the blastocyst stage by their reduced inner cell mass (ICM) size, although no data were presented to support this. This possible link between reduced ICM size and reduced or absent primitive ectoderm has made velvet coat an interesting mutation for study. Experimental studies on normal embryos have suggested that primitive ectoderm versus primitive endoderm formation is determined by the position of ICM cells relative to the blastocoel cavity (Gardner & Johnson, 1975; Gardner & Papaioannou, 1975; Rossant, 1975). Cells that are on the outside of the ICM (normally adjacent to the blastocoel) form primitive endoderm while enclosed ICM cells form primitive ectoderm. If this positional interpretation is correct, a reduction in ICM size should lead to a reduction in the proportion of primitive ectoderm to primitive endoderm, since most cells would now be ‘outside’ cells (cf. ICM and trophectoderm formation at the blastocyst stage: Tarkowski & Wroblewska, 1967). It is possible, therefore, that the major defect in Ve/Ve embryos causes small ICM size and that the reduction in primitive ectoderm stems directly from this. If Ve/Ve blastocysts can be recognized, this hypothesis could be readily tested by injection of extra ICM cells, which should allow rescue of the mutation.

Before such manipulative experiments can be undertaken, the reported effects of the Ve mutation must be confirmed. Granholm, Stevens & Theiler (1980) have recently published a description of the post-implantation development of the velvet coat mutation on a C57B1/6J strain background which essentially confirms earlier reported findings. However, they did not look at pre-implantation stages. In the present study, we provide a detailed description of the in vivo and in vitro development of Ve/Ve embryos on a 129J strain background.

Matings

The velvet coat (Ve) mutation (obtained from the Pasteur Institute, Paris) was maintained on a 129J strain background by forced heterozygosity. Mutant embryos were obtained by mating Ve/+ heterozygotes and control litters were produced by reciprocal backcrosses of Ve/+ ♀ × + / + ♂and +/+ ♀ × Ve/ + ♂. No differences were found between embryos from the two backcrosses and so all control results were pooled. Natural mating was used to obtain all postimplantation embryos but pre-implantation blastocysts were recovered after superovulation to increase embryo numbers. Intraperitoneal injection of 5 i.u. of pregnant mare’s serum gonadotrophin (PMS, Organon) was followed 48 h later by 5 i.u. human chorionic gonadotrophin (hCG, Sigma). Ovulation and mating were presumed to occur approximately 12 h after hCG injection.

Examination of post-implantation embryos

Female mice from Ve/+ × Ve/+ matings were killed at 5·5, 6·5, 7·5, and 8·5 days post coitum and any implants were fixed in acetic formol alcohol, dehydrated and embedded in wax before sectioning at 7 μm. Sections were stained with haemalum and eosin and examined with the light microscope. Female mice from control matings were killed at the same stages of pregnancy but only 5·5-day implants were processed for histology. Older control embryos were dissected out intact and examined carefully under the dissecting microscope. This procedure was adopted after dissection of 6·5 days and older embryos from Ve/+ × Ve/ + crosses revealed that the major histological abnormalities detected in presumed Ve/Ve embryos could also be recognized in the intact embryo.

Examination, culture and transfer of blastocysts

Blastocysts were obtained from both Ve/ + × Ve/ + matings and control matings 98 h after hCG injection and all were examined under the dissecting or compound microscope for any obvious abnormalities. Some were classified visually into early and late blastocysts and transferred separately to recipient females on the 3rd day of pseudopregnancy. Recipients were killed on the 8th day of pregnancy and examined for normal or abnormal fetuses. Some blastocysts were subjected to immunosurgery (Solter & Knowles, 1975) and the resulting inner cell masses (ICMs) were fixed, air-dried (Tarkowski, 1966) and stained with Giemsa for cell counts. Air-dried spreads of intact blastocysts were also made. Other blastocysts from control and experimental matings were cultured in tissue-culture dishes (Falcon) in a-modified MEM (Gibco) plus 10 % foetal calf serum, gassed with 5 % CO2 in air and maintained at 37 °C for 7 days. A further series of experimental and control blastocysts were cultured on feeder layers of Mitomycin-C-treated STO fibroblast cells (Ware & Axelrad, 1972). After 7 days, most cultures were washed with PBS, fixed in acetic alcohol and air-dried prior to staining with Methylene Blue. This enabled measurements of the diameters of the trophoblast giant cell nuclei to be made. A few cultures were grown in solvent-resistant culture dishes (Permanox, Lux) and then fixed in 2·5 % glutaraldehyde in 0·1 M phosphate buffer, post-fixed in OsO2, dehydrated and embedded in Spurr’s resin (Spurr, 1969). Sections 1 μm thick were cut with a glass knife on a Huxley ultramicrotome and stained with a mixture of toluidine blue and Azur II.

Post-implantation development

The post-implantation development of embryos from experimental and control matings is summarized in Table 1. Most control embryos were entirely normal but a few resorbed or resorbing embryos were detected at each stage of development. These generally consisted of clumps of dead cells with a few trophoblast giant cells. Similar early resorptions were observed in a few cases in the experimental matings and these were not considered further. At 6·5, 7·5 and 8·5 days of development, histological examination of conceptuses from Ve/ + × Ve/ + matings also revealed a class of embryos with distinct sets of abnormalities not observed in control matings. This abnormal class constituted 24·5 % (29/118) of all embryos examined at the three stages which is close to the expected 25% Ve/Ve homozygotes and was therefore presumed to represent these embryos.

Table 1.

Postimplantation development of embryos from Ve/+ × Ve/+ and Ve+ × +/ + matings

Postimplantation development of embryos from Ve/+ × Ve/+ and Ve+ × +/ + matings
Postimplantation development of embryos from Ve/+ × Ve/+ and Ve+ × +/ + matings

At 5·5 days of development, no clear distinction could be drawn between normal and mutant embryos. All embryos recovered were egg cylinders with both extraembryonic and embryonic ectoderm present. The only embryos which showed any sign of abnormality were four embryos which were classified as ‘dumpy’ (Fig. 1). These egg cylinders were smaller than their litter-mates and all showed areas of infolded uterine epithelium beneath the embryo. However, there was no obvious disproportionate growth of any one tissue and no sign of excessive cell death. By 6·5 days of development presumed Ve/Ve embryos were recognized quite readily by their shortened appearance and reduced embryonic and extraembryonic ectoderm. The proximal endoderm overlying the embryonic ectoderm was thickened and columnar instead of squamous. In some cases the ectoderm tissues were so reduced that the proximal endoderm formed a loose sac at the end of the egg cylinder (Fig. 2). In no case were embryonic or extra-embryonic ectoderm completely absent and they always appeared healthy. Mitotic figures were observed and no excessive cell death occurred. No other tissues showed any obvious abnormalities.

Fig. 1.

(a) Section of 5·5-day embryo from Ve/+ × Ve/+ mating, classified as ‘dumpy’. All tissues are present including giant cells and distal endoderm although this is not obvious in this section. Uterine epithelium beneath embryo is highly folded. (b) Section of 5·5-day control embryo from Ve/ + × +/+ mating.

Fig. 1.

(a) Section of 5·5-day embryo from Ve/+ × Ve/+ mating, classified as ‘dumpy’. All tissues are present including giant cells and distal endoderm although this is not obvious in this section. Uterine epithelium beneath embryo is highly folded. (b) Section of 5·5-day control embryo from Ve/ + × +/+ mating.

Fig. 2.

(a) Section of 6·5-day presumed Ve/Ve embryo. Both embryonic and extra-embryonic ectoderm appear reduced, leaving the proximal endoderm as a loose sac at the bottom of the egg cylinder. All other tissues appear normal, (b) Section of control 6·5-day embryo.

Fig. 2.

(a) Section of 6·5-day presumed Ve/Ve embryo. Both embryonic and extra-embryonic ectoderm appear reduced, leaving the proximal endoderm as a loose sac at the bottom of the egg cylinder. All other tissues appear normal, (b) Section of control 6·5-day embryo.

By 7·5 days, the mutant phenotype was more variable. Four of the nine presumed Ve/Ve embryos were still egg cylinders but, as at 6-5 days, the ectoderm tissues were drastically reduced and the proximal endoderm was thickened, loose and columnar. In one of the embryos the embryonic ectoderm was rather misshapen (Fig. 3). The remaining five embryos showed little or no embryonic or extraembryonic ectoderm. Other trophoblast tissues (ectoplacental cone and giant cells) were visible, and so was distal endoderm. However, the embryonic region consisted of a mass of loose, vesiculated cells which were presumed to be collapsed proximal endoderm (Fig. 4). In two cases, there was perhaps a little ectoderm inside the endoderm, but this was not very clear. A large amount of haemorrhage was observed between the parietal endoderm and trophoblast.

Fig. 3.

Fig. 3. Section of 7·5-day presumed Ve/Ve embryo. Ectoderm derivatives (e) are reduced and misshapen and surrounded by loose, columnar proximal endoderm (p). This section is not quite medial so that the ectoplacental cone is not clear but other sections revealed that this tissue appeared healthy. Reichert’s membrane and adhering distal endoderm (d) are visible, surrounded by haemorrhagic tissue. The parietal yolk-sac cavity appears full of debris.

Fig. 4. Section of 7·5-day presumed Ve/Ve embryo. This embryo contains no obvious ectoderm derivatives and consists of a loose lump of proximal endoderm (p) attached to the ectoplacental cone.

Fig. 3.

Fig. 3. Section of 7·5-day presumed Ve/Ve embryo. Ectoderm derivatives (e) are reduced and misshapen and surrounded by loose, columnar proximal endoderm (p). This section is not quite medial so that the ectoplacental cone is not clear but other sections revealed that this tissue appeared healthy. Reichert’s membrane and adhering distal endoderm (d) are visible, surrounded by haemorrhagic tissue. The parietal yolk-sac cavity appears full of debris.

Fig. 4. Section of 7·5-day presumed Ve/Ve embryo. This embryo contains no obvious ectoderm derivatives and consists of a loose lump of proximal endoderm (p) attached to the ectoplacental cone.

At 8·5 days, two presumed Ve/Ve embryos were again represented by ectoplacental cone tissue and giant cells with small vesicles of proximal endoderm. The remaining seven embryos had evidence of some ectoderm derivatives and were presumably formed from Ve/Ve embryos that retained ectoderm at 7·5 days. In two cases the remaining ectoderm was a very small lump surrounded by loose proximal endoderm (Fig. 5). This embryonic remnant was further surrounded by blood, debris and giant cells. Metaphases were observed in the ectoderm lumps. The remaining four embryos contained more ectoderm although this was always very misshapen and surrounded by thick endoderm. These four embryos all showed evidence of some mesoderm formation (Fig. 6). No organized primitive streak or amniotic folds were observed, but cells of mesoderm morphology were found as an intervening layer between endoderm and presumed embryonic ectoderm. Five embryos (four with embryonic ectoderm, one without) also showed abnormal development of the ectoplacental cone, which appeared to be organized into columnar epithelium-like structures, resembling embryonic or extraembryonic ectoderm (Figs. 5, 6). No such structures have ever been reported in normal ectoplacental cone development.

Fig. 5.

Fig. 5. Section of 8·5-day presumed Ve/Ve embryo. Ectoderm is reduced, misshapen and surrounded by columnar endoderm. No mesoderm is visible. Ectoderm-like structure (arrow) in ectoplacental cone.

Fig. 6. Section of 8·5-day presumed Ve/Ve embryo. Also misshapen and small but mesoderm (m) apparent between ectoderm and endoderm. No obvious primitive streak. Again ectoderm-like structure (arrow) in ectoplacental cone.

Fig. 5.

Fig. 5. Section of 8·5-day presumed Ve/Ve embryo. Ectoderm is reduced, misshapen and surrounded by columnar endoderm. No mesoderm is visible. Ectoderm-like structure (arrow) in ectoplacental cone.

Fig. 6. Section of 8·5-day presumed Ve/Ve embryo. Also misshapen and small but mesoderm (m) apparent between ectoderm and endoderm. No obvious primitive streak. Again ectoderm-like structure (arrow) in ectoplacental cone.

Blastocyst examination and transfer

Blastocysts from Ve/ + × Ve/ + matings were compared with blastocysts from control matings and no consistent abnormalities were observed in the experimental group. Particular attention was paid to the size of the ICM but no blastocysts with obviously reduced ICMs were observed, nor were there any particularly small ICMs observed after immunosurgery. Cell counts, however, revealed a slight difference between the mean cell number of control and experimental ICMs (Ve/ + × +/+ = 23·2±0·66, n = 79; Ve/+ × Ve /+ = 21·0 ±0·83, n = 86). Histograms of ICM cell number showed that 26·7% of ICMs from Ve/ + × Ve/ + matings had fewer than 15 cells whereas only 11·3 % of control ICMs fell into this class (Fig. 7). These two percentages were significantly different (𝒳2 = 7·4, P < 0·01) but the difference between the two values does not amount to the 25 % expected if all Ve/ Ve embryos fell in this cell number class. If the low ICM cell number apparently shown by some Ve/ Ve ICMs were due to preferential reduction in ICM size, we would have expected to recognize Ve/ Ve ICMs by small size after immunosurgery which we failed to do. We would also have predicted that intact blastocyst cell counts should not reveal much difference between control and experimental embryos, since three quarters of the cells of the blastocyst are trophectoderm which would be unaffected. This prediction was also not fulfilled. There was a large difference in mean cell number between control and experimental groups (Ve/ + × + / + = 66·5±2·6, n = 19; Ve/ + × Ve/+ = 54·0±2·9, n = 22). This difference was due to the presence of seven embryos in the experimental cross with cell numbers in the 30–50 range (Fig. 7) typical of early blastocysts (Rossant & Lis, 1979). Such retarded blastocysts would contain ICMs of normal size but low cell number, explaining the results of ICM cell counts. There was, thus, no conclusive evidence for a specific reduction in ICM size in Ve/Ve homozygotes and even general retardation of growth was not an infallible marker of Ve/Ve embryos. Embryo transfers showed that 3 out of 8 retarded blastocysts from Ve/ + × Ve/ + matings were Ve/Ve homozygotes while 4 out of 18 expanded blastocysts also proved to be Ve/Ve.

Fig. 7.

Histograms of ICM and blastocyst cell numbers from Ve/+ × Ve/+ and Ve/+×+/+ matings. (A) ICM cell numbers: (B) Blastocyst cell numbers. •, Ve/+ × +/ + : ○, Ve/ + × Ve/ + .

Fig. 7.

Histograms of ICM and blastocyst cell numbers from Ve/+ × Ve/+ and Ve/+×+/+ matings. (A) ICM cell numbers: (B) Blastocyst cell numbers. •, Ve/+ × +/ + : ○, Ve/ + × Ve/ + .

In vitro culture

Although Ve/Ve homozygotes could not be unequivocally identified at the blastocyst stage, they were recognizable after a few days in blastocyst outgrowths (Table 2). In a medium alone, 30·8% of Fe/ + × Ve/ + blastocysts failed to show normal egg-cylinder development (ectoderm and endoderm formation) and produced spreads of trophoblast giant cells with occasionally a few disperse cells which were presumed to be ICM derivatives. It was thought likely that the Ve/Ve embryos were included in this class but the rate of abnormal development in the control cultures was too high (17·7 %) to allow any firm conclusions to be drawn. Improved development of control cultures was achieved by using a feeder layer of fibroblasts; the percentage of control embryos showing poor ICM development was reduced to 6-4%. However, under the same culture conditions, 27·4% of Ve/+ × Ve/ + blastocysts failed to develop normal ICM structures. If the control rate of abnormal development is subtracted from this figure, 21% of the blastocysts from Ve/ + × Ve/ + matings show unexplained abnormal development. This is close to the expected figure of 25 % Ve/Ve homozygotes. Trophoblast giant cell formation appeared normal in all outgrowths from the experimental embryos. Comparison of the nuclear diameters of all the peripheral trophoblast cells of each outgrowth (Wudl & Sherman, 1978) revealed that the extent of endoreduplication was similar in both normal and presumed mutant embryos (Table 3).

Table 2.

Development of blastocysts from Ve/ + × Ve/+ and Ve/ + × + / + matings after 7 days in vitro

Development of blastocysts from Ve/ + × Ve/+ and Ve/ + × + / + matings after 7 days in vitro
Development of blastocysts from Ve/ + × Ve/+ and Ve/ + × + / + matings after 7 days in vitro
Table 3.

Nuclear diameters of trophoblast cells from presumed mutant and wildtype blastocyst outgrowths

Nuclear diameters of trophoblast cells from presumed mutant and wildtype blastocyst outgrowths
Nuclear diameters of trophoblast cells from presumed mutant and wildtype blastocyst outgrowths

The earliest stage at which embryos homozygous for the velvet coat mutation could be unequivocally recognized in vivo in this study was 6-5 days post coitum, when presumed Ve/Ve embryos showed reduced egg-cylinder growth. Both embryonic and extraembryonic ectoderm were apparently reduced in size and proximal endoderm overlying the embryonic region was loose and columnar rather than squamous. We were unable to identify unequivocally Ve/Ve embryos at 5·5 days of development as reported by other workers using mice of a different genetic background (Granholm et al., 1980), although some embryos appeared normal but retarded. We were also unable to identify Ve/Ve homozygotes prior to implantation. No class of blastocyst showing consistent abnormalities, including reduction in ICM size, was detected in Ve/+ × Ve/ + matings. There were more normal but retarded blastocysts in experimental matings than in control crosses, but embryo transfer studies showed that this was not an absolute criterion for identifying Ve/Ve homozygotes. Thus, on our strain background, the velvet coat mutation does not seem to exert any marked effect on morphogenesis until well after implantation, although it may contribute to general retardation of growth at earlier stages. Our initial hypothesis that the velvet coat mutation caused a reduction in ICM size and thence a reduction in the ratio of primitive ectoderm to primitive endoderm does not seem to be correct.

However, velvet coat might still provide some useful insights into normal developmental processes if it could be shown to be tissue-specific, affecting only the primitive ectoderm and its derivatives. Most early developmental lethals do not appear to be restricted to one cell type or another (McLaren, 1976). All the abnormalities observed in Ve/Ve homozygotes at 6·5 days could be interpreted as secondary effects of a primary failure of primitive ectoderm growth, although they are not confined to this tissue. Reduction in the size of the extraembryonic ectoderm could be brought about by failure of the primitive ectoderm to provide the normal stimulus for proliferation (Gardner & Papaioannou, 1975; Rossant & Lis, 1981) and the loose, columnar nature of the proximal endoderm might be a response to reduced pressure from the underlying ectoderm. Later than 6·5 days, other abnormalities occurred that are harder to reconcile with tissue specificity of Ve action, but their late appearance suggests that they too could be due to secondary effects. In vitro studies also showed poor development of the primitive ectoderm from Ve/Ve blastocysts, although trophoblast giant cell formation was not affected and outgrowths survived beyond the normal time of death in vivo.

Although most of the description of Ve/Ve homozygotes is consistent with the hypothesis that Ve acts specifically on the primitive ectoderm, morphology alone cannot be used to establish the primary site of action of the gene. The hypothesis must be tested experimentally by assessing the ability of various tissues from Ve/Ve embryos to form chimeras with normal embryos. The hypothesis predicts that both trophectoderm and primitive endoderm from Ve/Ve embryos will be able to survive when combined with normal embryonic cells but primitive ectoderm will not be able to contribute to a chimera. Our inability to recognize unequivocally Ve/Ve homozygotes at the blastocyst stage and the absence of any suitable genetic marker closely linked to Ve will make such experimental analysis difficult (Papaioannou & Gardner, 1979).

We should like to thank Drs V. E. Papaioannou, L. C. Stevens, N. H. Granholm and D. A. Hickey for useful comments on the manuscript. This work was supported by the Canadian Natural Sciences and Engineering Research Council.

Gardner
,
R. L.
&
Johnson
,
M. H.
(
1975
).
Investigation of cellular interaction and deployment in the early mammalian embryo using interspecific chimeras between the rat and mouse
.
In Cell Patterning (Ciba Found. Symp
.), pp.
183
200
.
Amsterdam
:
A.S.P
.
Gardner
,
R. L.
&
Papaioannou
,
V. E.
(
1975
).
Differentiation in the trophectoderm and inner cell mass
.
In The Early Development of Mammals: 2nd Symposium of the British Society for Developmental Biology
(ed.
M.
Balls
and
A. E.
Wild
), pp.
107
132
.
London
:
Cambridge University Press
.
Granholm
,
N. H.
,
Stevens
,
L. C.
&
Theiler
,
K.
(
1980
).
Development of velvet coat (Pe/ Ve), another early lethal mutation in the house mouse
.
Anat. Embryol
,
157
,
237
242
.
McLaren
,
A.
(
1976
).
Genetics of the early mouse embryo
.
Ann. Rev. Genet
.
10
,
361
388
.
Papaioannou
,
V. E.
&
Gardner
,
R. L.
(
1979
).
Investigation of the lethal yellow A3/A3 embryo using mouse chimaeras
.
J. Embryol. exp. Morph
.
52
,
153
163
.
Rossant
,
J.
(
1975
).
Investigation of the determinative state of the mouse inner cell mass II. The fate of isolated inner cell masses transferred to the oviduct
.
J. Embryol. exp. Morph
.
33
,
991
1001
.
Rossant
,
J.
&
Lis
,
W. T.
(
1979
).
Potential of isolated mouse inner cell masses to form trophectoderm dérivâtes in vivo
.
Devl Biol
.
70
,
255
261
.
Rossant
,
J.
&
Lis
,
W. T.
(
1981
).
Effect of culture conditions on diploid to giant cell transformation in postimplantation mouse trophoblast
.
J. Embryol. exp. Morph
.
62
,
217
227
.
Solter
,
D.
&
Knowles
,
B. B.
(
1975
).
Immunosurgery of mouse blastocyst
.
Proc. natn. Acad. Sci. U.S.A
.
72
,
5099
5102
.
Spurr
,
A. R.
(
1969
).
A low-viscosity epoxy resin embedding medium for electron microscopy
.
J. Ultrastruct. Res
.
26
,
31
43
.
Stieler
,
C.
&
Hollander
,
W. F.
(
1972
).
Location of the velvet coat mutant in linkage group VI of the mouse
.
J. Hered
.
63
,
212
214
.
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 four- and eight-cell stage
.
J. Embryol. exp. Morph
.
18
,
155
180
.
Ware
,
L. M.
&
Axelrad
,
A. A.
(
1972
).
Inherited resistance to N- and B-tropic murine leukemia viruses in vitro: evidence that congenie mouse strains SIM and SIM.R. differ at the Fv-1 locus
.
Virology
50
,
339
348
.
Wudl
,
L.
&
Sherman
,
M. I.
(
1978
).
In vitro studies of mouse embryos bearing mutations in the T complex: t6
.
J. Embryol. exp. Morph
.
48
,
127
151
.