ABSTRACT
Mouse embryos which are homozygous for the c25H deletion at the albino locus are developmentally arrested at the 2-to 6-cell cleavage stages. This study reveals that the mutant embryos cease development before they can be distinguished ultrastructurally from their normal litter-mates. After prolonged developmental delay (24−48 h), the nuclei of the mutant embryos become extremely aberrant in shape, whereas other subcellular organelles remain normal and there are no signs of pyknosis. Eventually, the mutant embryos do become pyknotic and begin to degenerate. The striking effects of this deletion on nuclear ultrastructure of cleavage-stage embryos are discussed in relation to biochemical and ultrastructural defects caused by other lethal deletions at the albino locus.
INTRODUCTION
Six radiation-induced mutations at the albino locus (chromosome 7) in the mouse have been found to be deletions and homozygous lethals (Erickson, Gluecksohn-Waelsch & Cori, 1968; Erickson, Eicher & Gluecksohn-Waelsch, 1974; Gluecksohn-Waelsch, Schiffman, Thorndike & Cori, 1974; Miller et al. 1974; Jagiello et al. 1976). The longest of the deletions, c25H, is the earliest acting lethal, killing during early embryogenesis (Gluecksohn-Waelsch et al. 1974). Recently, Lewis (1978) found that the homozygous c25H embryos were developmentally arrested at the 3-to 6-cell cleavage stages and that their blastomeres, at the level of light microscopy, appeared to be bi- or multi-nucleated. She suggested, therefore, that the lethal phenotype may include aberrant karyo-kinesis and/or cytokinesis. The results of an ultrastructural study of the morphological effects of the c25H allele which is reported here show that the homozygous mutant embryos have abnormally shaped nuclei rather than bi- or multi-nucleated cells. The ultrastructure of other cellular components is not affected by the mutant genotype.
MATERIALS AND METHODS
Animals heterozygous for albino (c) and the c25H allele (c/c25H) were obtained from the following mating: c/c (AKR albino) × c/c25H (dilute chinchilla). Offspring carrying the mutant allele (c/c25H) are distinguishable as albinos from their litter-mates (c/cch) which are dilute chinchilla. In order to obtain litters containing homozygous lethal embryos, c/c25H animals were mated inter se. Control litters were obtained from the following matings: c/cch × c/cch or c/c25H × c/cch.
Females were time-ovulated by means of intraperitoneal injections of 2·5 i.u. pregnant mare serum (Gestyl, Organon) followed 45 h later by 2·5 i.u. human chorionic gonadotropin (Pregnyl, Organon). After the second injection, females were mated to the genetically appropriate males and copulation plugs were checked the next morning (day 0). All embryonic ages were recorded postovulation. Ovulation was presumed to have occurred 12 h after the injection of human chorionic gonadotropin. The approximate age of the embryos removed from the female or from culture at specific stages are: late 2-cell, 36 h; late 4-cell, 54 h; 8-cell, 66 h; early morula, 76 h; late morula, 88 h; blastocyst, 98 h.
In the first series of studies, 2-cell embryos were removed from experimental and control females, placed into standard egg culture medium (Goldstein, Spindle & Pedersen, 1975) and cultured according to the technique of Brinster (1963). They were allowed to develop in vitro until the embryos from control crosses and normal control littermates in experimental crosses had reached the blastocyst stage. A comparison of the numbers of embryos developmentally arrested and of the stages of developmental arrest in control and experimental litters was used to verify the lethal nature of the mutation as well as the phenocritical period for its expression in homozygotes.
In the second series, experimental and control litters were again obtained at the 2-cell stage. Some litters from both groups were fixed immediately. Others were placed into standard egg culture medium and allowed to develop in vitro. At the late 4-cell stage, all of the developmentally retarded embryos (the putative c25H/c25H embryos) from the experimental litters as well as the developmentally delayed embryos from the control litters were removed from culture and fixed. The remaining viable embryos of the experimental and control litters were allowed to develop to the late 8-cell stage. At this point, all embryos which had failed to advance in development from the 4-cell stage were removed and examined at the ultrastructural level. The embryos which continued development were allowed to remain in culture until they reached either the morula or blastocyst stages when they were fixed and examined for abnormalities. Using this protocol, all of the developmentally arrested embryos were processed for study within a short time period after cessation of development.
Two parallel experiments were done in the third series of studies. In the first, experimental and control litters were removed at the 2-cell stage and separately placed into culture. At each subsequent stage of development, entire experimental and control litters were removed and fixed for microscopic examination. In the second study, entire experimental and control litters were removed from females at the 2-cell, 4-cell, 8-cell, early morula and blastocyst stages and fixed immediately. The ultrastructure of the developmentally arrested mutant embryos from the younger litters was compared with that of developmentally arrested embryos obtained from older litters. This comparison enabled us to assess the ultrastructural effects of the developmental arrest over progressively longer periods of time (ranging from 12 to 60 h). The two sets of studies in this series also allowed us to determine if there were differences between the in vivo and in vitro expression of the mutant genome.
Both experimental and control embryos were fixed in 3 % glutaraldehyde in 0·1 M-P04 buffer (pH 7·4) for 1 h. The embryos were then washed in 0·1 M-PO4 buffer overnight, postfixed in 1 % osmium tetroxide (Millonig’s, pH 7·3), dehydrated through a series of alcohol concentrations and embedded in Epon. Serial sections were stained either with lead citrate for 2 min (Venable & Coggeshall, 1965) or with saturated uranyl acetate (Watson, 1958) for 1 h followed by lead citrate for 2 min. The sections were viewed with a Philips 300 electron microscope. At least one litter in each of the control and experimental series was stained with 0·5% aqueous uranyl acetate before dehydration in order to identify putative viral particles (Chase & Pikó, 1973). These latter embryos were then dehydrated, embedded and sectioned. The sections were stained with both uranyl acetate (1 h) and lead citrate (2 min).
Descriptions of the ultrastructure of normal cleavage-staged mouse embryos are available (Calarco & Brown, 1969; Hillman & Tasca, 1969). Therefore, this report includes descriptions of only those structures of the mutant embryos which differ morphologically from those of control embryos or those which become morphologically altered after the mutant embryos have been in developmental arrest for progressively longer periods of time.
RESULTS
Series 1
The results from the first series of experiments are presented in Table 1. The data show that 96 % of the control embryos developed from the 2-cell to the blastocyst stage in vitro. The finding that approximately one fourth of the experimental embryos die during the early cleavage stages confirms the observation by Lewis (1978) that the c25H deletion acts as an early lethal in homozygotes. Table 1 also shows that the developmental arrest of the putative c25H/c25H embryos occurs from the 2-to 6-cell stages, with the greatest number ceasing development at the 4-to 6-cell stages.
The same frequency of arrested and probably homozygous mutant embryos has been observed in experimental litters developing both in vivo and in vitro. In addition, the phenocritical period for the mutant embryos developing in vivo corresponds to the in vitro range of developmental arrest. No ultrastructural differences have been found between the embryos of the in vitro and in vivo experimental series. For this reason, the two groups of mutant embryos are not distinguished in this report.
Series 2
Comparative ultrastructural observations at the 2-cell cleavage stage of 25 embryos from the control matings and 20 embryos from experimental matings show that the mutant embryos cannot be distinguished from either their correspondingly staged litter-mates or control embryos at this stage of development. All of the embryos contain cellular organelles which are structurally normal for this early cleavage stage.
Additional experimental and control litters were allowed to continue development beyond the 2-cell cleavage stage. Of the 196 experimental embryos which were examined at the late 4-cell cleavage stage, 14 embryos had not advanced beyond the 2-cell stage and 11 remained as 3-cell embryos. Among the corresponding 216 control embryos, only three embryos did not advance to the 4-cell stage. Of the 171, 4-cell experimental embryos, reincubated for a time sufficient for them to attain the 8-cell stage, 18 remained as 4-cell embryos and 17 were developmentally delayed as 5- and 6-cell embryos. Only three of the control embryos failed to advance from the 4-cell stage to the 8-cell stage. All 60 developmentally retarded embryos from experimental litters were examined ultrastructurally. With a few exceptions (two 2-cell, two 4-cell and three 6-cell) the developmentally delayed experimental embryos were structurally normal. The exceptional embryos contained aberrantly shaped nuclei. However, since so few of the control embryos in these early cleavage stage litters show an aberrant structure, the only criterion for a tentative classification of homozygous mutant embryos is retardation of development. An examination of the six developmentally retarded embryos from control crosses revealed that none contained aberrantly shaped nuclei.
All of the remaining experimental and control embryos continued to develop to morula and blastocyst stages and all were structurally normal. As in the studies of Lewis (1978) there are no indications that the deletion affects either the morphology or viability of c/c25H preimplantation embryos.
Series 3
Consistent ultrastructural abnormalities are found only after the abnormal embryos from experimental litters have been in developmental arrest for an extended period of time. For example, an 8-cell experimental litter contains developmentally arrested embryos ranging from the 2-to 6-cell stages. While most of the 4-, 5- and 6-cell embryos appear structurally normal, the 2- and 3-cell embryos which have been in developmental arrest for a longer period of time contain irregularly shaped nuclei. Three of the four 2-cell embryos and all four of the 3-cell embryos were examined ultrastructurally by serial sections and found to contain nuclei with aberrant shapes. The degree of the aberrancy varies from cell to cell both within and between embryos. While some nuclei are only moderately misshapen, e.g. crescent-shaped or elongated, others are more abnormal in configuration (Figs. 1, 2). Although the examination of randomly selected sections suggests that cells may be bi- or multi-nucleated, a study of serial sections reveals that the apparently separate nuclear lobes are either attached to each other or to larger nuclear areas. None of the developmentally arrested embryos contain binucleated or multinucleated cells. All 13 of the developmentally arrested embryos from control animals have been studied. None of them contain aberrantly shaped nuclei.
Section of a developmentally arrested 2-cell embryo from an experimental litter at the 8-cell stage. Note the horseshoe-shaped nucleus. Primary nucleoli (P) and cytoplasmic organelles appear normal, × 6800.
Section of part of an arrested embryo in the 3-cell stage from an experimental litter at the 8-cell stage. Note the aberrantly shaped nucleus, × 6500.
Twenty-six of the arrested 2-to 6-cell embryos obtained from experimental litters at the early morula stage have the same abnormally shaped nuclei as those noted in the youngest, developmentally arrested mutant embryos from litters in the 8-cell stage. The morphology of other organelles is not adversely affected and none of the putative mutant embryos exhibit degenerative effects from the extended time in arrest. Conversely, developmentally arrested embryos from control litters are in varying stages of degeneration.
Developmentally retarded embryos in experimental litters examined at the late blastocyst stage present a range of degenerative defects. Some show no signs of degeneration while others are totally pyknotic. Those with the lowest grade of degeneration (a total of 28 embryos) have nuclei which are multilobed and have narrow fingerlike projections (Fig. 3). When these cells, which appear to be multinucleated in single sections, are traced serially, small bridges of nuclear material are found connecting the almost severed nuclear lobes to the main body of the nucleus. No binucleated or multinucleated cells have been observed. Other cellular organelles are not affected by the prolonged period of developmental arrest.
Section of a developmentally arrested embryo in the 2-cell stage from an experimental litter in the blastocyst stage. Note the nuclear projections and lobations. × 6600.
The remainder of the arrested embryos contain both pyknotic nuclei and pyknotic cells in varying numbers. In the most degenerated state, the plasma and nuclear membranes are discontinuous and organelles are indistinct. The small number of control embryos which are developmentally arrested at the 2-to 6-cell stages and which have been observed when their litter-mates are in the early morula or blastocyst stages, are either completely pyknotic or totally decomposed.
Virus observations
Viral aggregates, judged to be A particles by both their size and perinuclear location (Biczysko, Solter, Graham & Koprowski, 1974; Solter, personal communication) have been found in both experimental and control litters at the 2-cell stage (Fig. 4A, B). In control embryos and in the normal litter-mates of the mutant embryos, these viral aggregates disperse after the 2-cell stage. In later stages, single viruses are randomly distributed in the cytoplasm. Among the arrested embryos, removed from litters in 4-cell, 8-cell, early morula and blastocyst stages, all of the 2-cell and a few of the 3- and 4-cell embryos retain their perinuclear viral clusters. In the remainder of the 3- and 4-cell embryos and in all mutant embryos in later stages the viruses are dispersed in the cytoplasm.
(A) Portion of developmentally arrested 2-cell embryo from an experimental, litter in morula stage. Note perinuclear clusters of viral particles, x×8400. (B) A higher magnification of a type A virus particle found in both control and experimental embryos, x×130000.
DISCUSSION
Of the six radiation-induced deletions at the albino locus in the mouse, two are lethal prenatally (c6H, c25H) and four perinatally (c14Cos, c3H, c65K, c112K). The perinatal syndrome covers a multiplicity of biochemical and morphological effects. The biochemical abnormalities include deficiencies in the activities of glucose-6-phosphatase, tyrosine aminotransferase, serine dehydratase (Gluecksohn-Waelsch et al. 1974), glutamine synthetase (Gluecksohn-Waelsch, Schiffman & Moscona, 1975) and UDP-glucuronyltransferase (Thaler, Erickson & Pelger, 1976). The mutant neonates also have decreased concentrations of albumin, a fetoprotein, and transferrin, the principal plasma proteins (Garland, Satrustegui, Gluecksohn-Waelsch & Cori, 1976).
Morphologically, the perinatal mutants can be distinguished from their litter-mates by the abnormal ultrastructure of specific subcellular membrane systems in kidney and liver parenchymal cells. The affected cells have dilated and vesiculated rough endoplasmic reticulum, dilated Golgi apparatus and abnormal nuclear membranes (Trigg & Gluecksohn-Waelsch, 1973). Neonatal litter-mates heterozygous for the mutant lethal alleles are both biochemically and morphologically normal (Gluecksohn-Waelsch & Cori, 1970; Trigg & Gluecksohn-Waelsch, 1973).
There have been no previous comparative biochemical or ultrastructural studies of embryos homozygous for either of the prenatal mutations, c6H and c25H. However, enzyme and serum protein studies have been reported of complementing genotypes heterozygous for c6H or c25H and one of the perinatal lethal mutations (Gluecksohn-Waelsch et al. 1974). Morphological studies of c6H/c6H embryos have been limited to a light microscopic examination of the embryos during the phenocritical and lethal periods. They can be distinguished from their phenotypically normal litter-mates at 6-5-7 days of gestation by their severely reduced size and by abnormalities of the ectoplacental cone and parietal endoderm (Lewis, Turchin & Gluecksohn-Waelsch, 1976).
The albino deletion described in this report (c25H) causes developmental arrest of homozygotes prior to the second, or during the third, cleavage division of the fertilized ovum. It is, therefore, the earliest acting lethal mutation reported in mammalian embryos. c25H is the largest of the albino deletions and in contrast to c6H fails to complement fully with any of the other deletions at the albino locus (Gluecksohn-Waelsch et al. 1974). However, c25H, like the other five lethal deletions, acts as a true recessive.
The ultrastructure of the homozygous c25H embryos resembles that of their litter-mates and of other control embryos when these mutant embryos are examined soon after developmental arrest. Such embryos appear to be viable and non-necrotic both in vitro and in vivo until their litter-mates reach the blastocyst stage. Bizarrely shaped nuclei are not found until the embryos have been in developmental arrest for some time. It is of interest to note that the c25H deletion, unlike the perinatal lethal deletions, does not appear to affect the subcellular membrane system of cleavage stage embryos. Membranes retain their ultrastructural integrity until the embryos are in advanced stages of degeneration. Nevertheless, the effect of this deletion on membranes cannot be entirely ruled out since complementing genotypes of c25H and one of the perinatal lethals show ultrastructural defects of rough endoplasmic reticulum, Golgi apparatus and nuclear membrane (Gluecksohn-Waelsch et al. 1974), and the aberrant nuclear configuration of c25H homozygotes could result from a defect of the nuclear envelope. Also, because of the paucity of both rough endoplasmic reticulum and Golgi apparatus in the early embryonic stages, subtle effects would not be necessarily observable. Our studies, however, have failed to show bi- or multi-nucleated cells in the arrested, and presumably homozygous mutant embryos.
In our studies the phenocritical period of c25H homozygotes occurs at the 2-to 6-cell stage. Lewis (1978) reported it to be between the 3- and 6-cell stages. This minor discrepancy may be due to differences of genetic backgrounds used in the two studies. Whereas Lewis used heterozygous parents from a brothersister strain inbred for eight to nine generations, we obtained heterozygous c/c25H animals from outcrosses of cch/c25H to albino AKR.
The AKR genetic background may also account for the viral aggregates which we observed in both control and experimental litters and which bear no relation to the presence or absence of c25H. Biczysko, Pienkowski, Solter & Koprowski (1973) compared the concentration of A particles in blastocysts obtained from eight different strains of mice. They found AKR embryos to harbour the highest concentration of A-type particles, and suggested that A particle replication and maturation may be expressed in only specific strains of mice, the AKR background genotype being one which supports this expression.
ACKNOWLEDGEMENTS
This research was supported by United States Public Health Service Grants HD 00827, GM 19100 and HD 00193, and by a grant from the American Cancer Society VC-64. The authors would like to thank Marie Morris and Geraldine Wileman for their technical assistance.