ABSTRACT
Inner cell masses (ICMs) isolated immunosurgically from mouse blastocysts segregating the homozygous lethal mutants t0/t0 and tw5 /tw5 were cultured in vitro. Presumed ICMs fail to grow after three days in culture (equivalent gestational day 7 · 5) when they consist of an outer layer of endoderm cells surrounding about 30 epiblast cells. Presumed homozygous tw5 / tw5 ICMs develop to a more advanced stage in culture and on the seventh day (equivalent gestational day 11 · 5) consist of an inner core of disorganized ectoderm cells with a small proamniotic cavity, surrounded by multiple layers of endoderm cells.
INTRODUCTION
Mouse embryos homozygous for the mutations t0 (or its allele, t6) and Z”‘5 die during the early post-implantation period (Gluecksohn-Schoenheimer, 1940; Bennett & Dunn, 1958) but the mechanism behind the lethality of these and other mutations within the T/t complex is unknown (for reviews see Bennett, 1975; Sherman & Wudl, 1977). The discrete morphological abnormalities seen in the different classes of mutant homozygotes suggested that they have defects in cell-cell interaction, raising the possibility that T/t genes code for cell surface components specifically required during embryogenesis (Bennett, 1975). Indeed, there is some serological evidence that T/t related antigens are present on pre-implantation mouse embryos (Artzt, Bennett & Jacob, 1974; Kemler et al. 1976). Alternatively, it has been argued that the primary effect of t mutations is on the intermediary metabolism of all cells in the embryo, leading to the death of the embryo at certain stages of development (Nadijka & Hillman, 1975; Wudl, Sherman & Hillman, 1977).
One way to distinguish between these hypotheses is to transplant embryos in vivo to ectopic sites. According to the ‘organisational defect’, hypothesis one might expect cells of the mutant embryos to be capable of continued growth under these conditions, and this appears to be the case with homozygous T embryos (Ephrussi, 1935; Bennett, Artzt, Magnuson & Spiegelman, 1977). On the other hand, homozygous t6 and t0 embryos fail to grow in ectopic sites (Wudl et al. 1977; Artzt, 1978 personal communication).
Another approach is to culture embryos segregating homozygous t mutants in vitro. Using this technique two groups (Erikson & Pederson, 1975; Wudl & Sherman, 1976; Wudl et al. 1977) have reported that cells of presumed homozygous t6/t6 and tw5/tw5 embryos fail to grow in culture beyond the 7th or 11th equivalent gestational day respectively. In these experiments intact blastocysts were allowed to attach to the surface of the culture dish and the inner cell mass (ICM) remained in contact with the outgrowing sheet of trophectoderm cells. Under these conditions, interactions between trophectoderm and ICM affecting cell survival are still possible. Recently, Wudl and Sherman (1978; Wudl et al. 1977) have reported abnormalities in the growth of presumed t6/t6 trophectoderm cells when they are cultured in the absence of the ICM. In this paper we have investigated the in vitro development of both t0/t0 and tw5/tw5 ICMs when they are isolated from the trophectoderm by the technique of immunosurgery. Normal ICMs isolated in this way can continue to develop in culture into structures resembling the egg cylinder of normal embryos at around 7 · ·5 days of gestation (Wiley, Spindle & Pederson, 1978; Hogan & Tilly, 1978 a).
MATERIALS AND METHODS
Mice
+ /t0 and + /tw males and females were obtained by crossing T/t0 and T/t5 males with +/+ random bred CF1 females. The F1 hybrid males were progeny tested by mating to T/ + females, and their relative transmission of t to + measured by the ratio of tailless (7/7) to Brachyury (T/ + ) offspring; this ratio was unexpectedly high for the males used in these experiments. For the + /t0 males the ratios were 20:0, 45:2, 46:3, 37:3, 9:2, 22:3, and 29:4, giving an average transmission frequency of 92 · 4%. For the +/tvw5 males the ratios were 12:0 and 11:0 giving a transmission frequency of 100%.
To obtain homozygous mutant embryos, + /t F1 hybrid females were super-ovulated and mated with + /t males. For control crosses, either the Brachyury T/ + litter-mates of the + /t females, or +/+ CF1 females, were mated with + /t males.
Embryos
Blastocysts were collected on the fourth day of pregnancy when the day on which the vaginal plug is observed is day 1. The zonae pellucidae were removed with acidic tyrode solution, pH 2 · 5, and the blastocysts incubated overnight in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% foetal bovine serum. Inner cell masses were isolated immunosurgically the following day (equivalent to 4 · 5 days p.c.), and cultured as described by Hogan & Tilly (1978 a, b). The cultures were observed daily and photographed with an inverted phase-contrast microscope.
Inner cell masses were fixed overnight in 2 % glutaraldehyde in 0 · 1M phosphate buffer, post-fixed in osmium tetroxide, dehydrated through acetone, and embedded in Epon. Serial sections from 1 to 2 μM in thickness were stained with 1 % toluidine blue in 1 % borax solution (pH approximately 11), photographed, and the number of interphase nuclei and mitotic cells counted from serial reconstruction of the ICMs.
RESULTS
Inner cell masses from t0/t0 embryos
After one day in culture (equivalent gestation day e.g.d. 5 · 5) there is no significant difference in the size or morphology of the inner cell masses isolated from experimental (+/t0x + /t0) or control (T/ + × +/t0) blastocysts. After two days, all the ICMs have an outer layer of endoderm surrounding an inner core of epiblast cells but some of the ICMs in experimental cultures are significantly smaller than the controls. By day three most of the control ICMs develop a central cavity and begin to expand (Fig. 1,b), but only about half of the ICMs in experimental cultures behave in this way (Fig. 1,a). The remainder fail to increase in size and gradually degenerate over the next few days. As shown in Table 1, of 344 ICMs isolated from experimental blastocysts, 50% died by the seventh day of culture (e.g.d. 11 · 5). This is close to the expected proportion of homozygous mutant embryos in the cross + /t0 × +/t0 (approx. 46%) and suggests that the ICMs which fail to increase in size have the genotype t0/t0.
Phase contrast microscopy of ICMs after three days in culture. ICMs were isolated immunosurgically from 26 blastocysts of either experimental (+ /t0 × + /z0) or control (T/ + × + /t0) crosses and cultured in parallel under the same conditions.
(a) 25/26 ICMs from the experimental cross. The ICMs classified as presumed t0/t0 are labelled with arrows; some of these have remnants of dead trophoblast still attached.
(b) 22/26 ICMs from the control cross; in most of them the inner ectoderm has expanded into a hollow vesicle. Scale bars = 100μm.
Phase contrast microscopy of ICMs after three days in culture. ICMs were isolated immunosurgically from 26 blastocysts of either experimental (+ /t0 × + /z0) or control (T/ + × + /t0) crosses and cultured in parallel under the same conditions.
(a) 25/26 ICMs from the experimental cross. The ICMs classified as presumed t0/t0 are labelled with arrows; some of these have remnants of dead trophoblast still attached.
(b) 22/26 ICMs from the control cross; in most of them the inner ectoderm has expanded into a hollow vesicle. Scale bars = 100μm.
Sectioned ICMs on day 3 of culture. ICMs were prepared from blastocysts of the cross (+/t0 ×+/t0) and cultured in vitro for three days, (a) Presumed t0/r0 ICM showing outer endoderm cells and an inner core of epiblast cells, three of which are in mitosis. (6) Presumed normal (+/t0or + / + ) ICM with the embryonic ectoderm surrounding a small proamniotic cavity. Scale bars =10 μm.
Sectioned ICMs on day 3 of culture. ICMs were prepared from blastocysts of the cross (+/t0 ×+/t0) and cultured in vitro for three days, (a) Presumed t0/r0 ICM showing outer endoderm cells and an inner core of epiblast cells, three of which are in mitosis. (6) Presumed normal (+/t0or + / + ) ICM with the embryonic ectoderm surrounding a small proamniotic cavity. Scale bars =10 μm.
The morphology of typical presumed t0/t0 and normal ( + /+ or +/t0) ICMs fixed on day three of culture is shown in Fig. 2 a and b. The epiblast cells of the presumed mutants are rounded and disorganized and do not develop the equivalent of a proamniotic cavity. The average number of epiblast cells in eight presumed t0/t0 ICMs fixed on day 3 of culture and serially sectioned was 30 (range 22–45). In some of these ICMs a high percentage of the cells were in mitosis. (This prompted us to determine the percentage of cells in mitosis for eight ICMs and we obtained the values 4, 7, 7, 11, 12, 13, 13 and 36%.) Many of the endoderm cells of the presumed t0/t0 ICMs are highly vacuolated.
However, similar vacuolated cells are seen in the presumed +/+ or +/t0 ICMs (Fig. 2 b), as well as in cultured ICMs from control crosses and normal C3H embryos (unpublished observations; Hogan & Tilly, 1978a).
Inner cell masses from tw5/tw5 embryos
For the first three days in culture there is no significant difference between the development of ICMs from control (T/ + or +/+ × + /tw5) and experimental ( + /tw5× + /tw5) cultures. By four days, however, some of the ICMs in experimental cultures are significantly smaller than in the controls. As shown in Table 1, by seven days (e.g.d. 11· 5) 30–50% of the ICMs from experimental blastocysts can be classified by phase-contrast microscopy as definitely abnormal in their morphology according to the following criteria: small size, either the absence of a central cavity or the presence of a small cavity containing dense material, and multiple outer layers of endoderm cells, which often slough off in clumps into the medium (Fig. 3 a, b). Some of these presumed tw5/tw5 ICMs attached to the culture dish where the endoderm cells extended into flat sheets, but these endoderm cells did not continue to grow beyond about 9 days in culture.
Presumed tw6/tw5 ICMs after 7 days in culture, (a) Phase-contrast microscopy of two presumed homozygous mutant ICMs showing proliferation of rounded endoderm cells on the outside and small, densely filled inner cavities. The presumed mutant ICMs are attached to a third, normal expanded ICM, Scale bar = 50 μm. (b) Section of a presumed tw5/tw8 ICM showing the accumulation of extracellular material, presumably basement membrane material (arrow), between the outer endoderm cells and the inner ectoderm cells. Scale bar = 20 μm. (c) A presumed normal (+/+ or + /tw5) ICM showing the outer epithelial layer of endoderm cells separated by spaces from the inner ectoderm cells which enclose an expanded cavity. Scale bar = 40 μm.
Presumed tw6/tw5 ICMs after 7 days in culture, (a) Phase-contrast microscopy of two presumed homozygous mutant ICMs showing proliferation of rounded endoderm cells on the outside and small, densely filled inner cavities. The presumed mutant ICMs are attached to a third, normal expanded ICM, Scale bar = 50 μm. (b) Section of a presumed tw5/tw8 ICM showing the accumulation of extracellular material, presumably basement membrane material (arrow), between the outer endoderm cells and the inner ectoderm cells. Scale bar = 20 μm. (c) A presumed normal (+/+ or + /tw5) ICM showing the outer epithelial layer of endoderm cells separated by spaces from the inner ectoderm cells which enclose an expanded cavity. Scale bar = 40 μm.
In the electron microscope presumed + / + and + /tw5 ICMs fixed on day 7 of culture (e.g.d. 1T5) have a single epithelial layer of endoderm cells separated by large spaces containing mesodermal cells from the embryonic ectoderm, which usually consists of a single layer of cuboidal or columnar cells surrounding a large cavity (Fig. 3 c). In the case of presumed tw5/tw5 ICMs fixed on day 7 of culture, both light and electron microscope sections reveal that the outer endoderm cells are producing large amounts of extracellular material which accumulates between the endoderm and the epiblast (Fig. 3 b). The endoplasmic reticulum of the endoderm cells is engorged with this secretory material. The inner cells appear quite healthy and have many desmosomes and other types of junctions, even when the cells show no obvious organization into an epithelial layer.
DISCUSSION
In these experiments we have shown abnormal development of presumed t0/t0 inner cell masses cultured in vitro in the absence of trophectoderm (which was removed during the immunosurgery procedure). By three days in culture (e.g.d. 7· 5) the presumed homozygous t0 ICMs resemble the egg cylinders of presumed t0/t0 embryos of approximately six days of development in utero (Gluecksohn-Schoenheimer, 1940). In both cases the presumed homozygous mutants are much smaller than normal and have an irregular outer layer of endoderm cells surrounding a disorganized mass of epiblast cells which do not form a proamniotic cavity. In other experiments, Wudl & Sherman (Wudl et al. 1977; Wudl & Sherman, 1978) have reported that trophectoderm cells of presumed t6/t6 blastocysts fail to endoreduplicate their DNA in culture to the same extent as controls, although, other properties of the trophectoderm cells (such as the synthesis of specific enzymes) are apparently unaffected. Reduced endoreduplication of DNA was also found in cultures of presumed homozygous mutant trophectoderm vesicles, which had developed from dissociated blastomeres in the absence of an inner cell mass. Whatever the common factor responsible for the defect in trophectoderm and ICM cells of cultured t0/t0 embryos, these results suggest that the t0 mutation can be expressed independently in more than one cell type.
Inner cell masses from the cross + /tw5 × + /tw5 develop in culture to a more advanced stage than t0 homozygotes. Subsequently, however, some of them apparently fail to form either a fully expanded proamniotic cavity or mesoderm cells, and show relatively luxuriant growth of endoderm cells and the accumulation of extracellular material between the endoderm and ectoderm (Fig. 3 b), all features consistent with the morphology of presumed homozygous mutants in utero (Bennett & Dunn, 1958; Babiarz, personal communication). ICMs with the abnormal morphology we have associated with the tw5/tw5 genotype are not seen in cultures of ICMs from other late-acting t mutants (twl8 and tw73, Hogan, unpublished observations). We therefore feel confident that the abnormalities we have imputed to presumed tw5/t5 ICMs are due to their genotype and not to the culture conditions per se. However, the proportion of ICMs scored as presumed homozygotes after 7 days in culture (average 37%) is somewhat lower than the proportion of homozygotes expected in the experimental cross ( + /Zw5x + /zw5) (approximately 50%). This suggests that under our conditions some homozygotes are developing at least to the same extent as some of the normal ( + /+ or + /t5) ICMs. In this context it is interesting to note that Bennett and Dunn (1958) described a class of homozygotes that maintained sufficient viable embryonic ectoderm cells to permit some mesoderm formation. At present we know nothing about the failure of presumed tw5/tw5 ICMs to develop normally in culture. It has been suggested that the tw5 mutation is expressed initially in embryonic ectoderm cells (Bennett & Dunn, 1958). Our observations, as well as the finding that trophectoderm cells of tw5/tw5 blastocysts behave the same in culture as control trophectoderm cells (Wudl & Sherman, 1978), are not inconsistent with this hypothesis.
Acknowledgements
This work was carried out in the Sloan-Kettering Cancer Center, and B.H. would like to thank Dorothea Bennett for her generous support and encouragement, and for making the facilities of her laboratory so freely available. All the members of the group, in particular Karen Artzt, contributed to the work through many informative and enthusiastic discussions. We would like to thank Cathy Calo and Steven Dunn for technical assistance.
Supported by NSF grant PCM 77-17835 and NIH CA-08748-13.