We have compared mouse embryos of three heterozygous, congenic genotypes (with high, medium and low levels of oocyte-coded glucose phosphate isomerase (GPI-1) activity respectively) to test whether 1) the survival time of oocyte-coded GPI-1 activity in the early embryo is affected by its activity level in the oocyte and 2) whether embryo-coded GPI-1 is detected earlier in embryos that inherit low levels of oocyte-coded GPI-1. The oocyte-coded GPI-1 was entirely GPI-1 A allozyme in the high and medium groups but was the less stable GPI-1C allozyme in the low group. We determined total GPI-1 activity and the ratio of different GPI-1 allozymes in early embryos and calculated the activity of oocyte-coded and embryo-coded GPI-1. In all three groups, the oocyte-coded enzyme activity remained at a more or less constant level for the first days. Some oocyte-coded GPI-1 remained in day embryos from the high and medium groups but was gone by days. Very little remained in day embryos that inherited low levels of a less stable form of the enzyme (GPI-1C allozyme). Despite a 4- to 5-fold difference in initial oocyte-coded GPI-1 activity, no differences were seen among the three genotypically distinct groups of embryos in the time of activation of the embryonic Gpi-1s genes. The embryo-coded GPI-1 was first detectable in day compacted morulae in all three groups. The level of oocyte-coded GPI-1, in the high group, when embryo-coded GPI-1 was first detected was higher than the level in the low group at any stage prior to detection of embryo-coded GPI-1. These results imply that embryonic Gpi-1s genes are not activated by a mechanism that depends on oocyte-coded GPI-1 enzyme activity falling below a critical threshold. Once both embryonic Gpi-ls alleles were activated. they were usually equally expressed in Gpi-lsa/Gpi-lsb heterozygous embryos. However, the maternal: paternal expression was closer to 2:1 in 5 % (3/60) of the and day embryos. These could have been digynic triploid embryos.

Embryonic genes are activated during the early stages of mammalian development and the oocyte-coded gene products are replaced by embryo-coded products. The transition from oocyte-coded to embryo-coded glucose phosphate isomerase (GPI-1; E.C. 5.3.1.9) has been extensively studied in mouse embryos (Chapman et al. 1971; Brinster, 1973; West and Green, 1983; Duboule and Burki, 1985; Gilbert and Solter, 1985; West et al. 1986). These studies have shown that embryo-coded GPI-1 is usually first detected at days (or occasionally days) post coitum (p.c.) and that maternally and paternally derived Gpi-ls alleles are probably activated synchronously. (The symbol Gpi-1s; refers to the structural gene encoding GPI-1 enzyme and has replaced the former symbol Gpi-1s; Oocyte-coded GPI-1 activity is relatively stable until days and then declines and is exhausted by to ) days p.c. The total GPI-1 activity falls to a minimum in the to ) day blastocyst even though embryo-coded expression has begun.

These studies have provided a quantitative description of the transition from oocyte-coded to embryo-coded GPI-1 and also imply that oocyte-coded GPI-1 mRNA is not present when the embryonic Gpi-1s genes are expressed (West and Green, 1983; Duboule and Burki, 1985; Gilbert and Solter, 1985). However, the mechanisms involved in activating the embryonic Gpi-1s genes remain unknown. This may involve interaction of genetic elements (e.g. promoter or enhancer sequences) with cellular factors that provide developmental cues (for discussion see Rosenfeld et al. 1987). For example, the activation of the gene encoding the LDH-5 (A4) isozyme of lactate dehydrogenase seems to be triggered by the proliferation of trophoblast after implantation (Auerbach and Brinster, 1967; Monk and Petzoldt, 1977; Speilman et al. 1978).

Although embryo-coded GPI-1 activity has not usually been detected before the morula stage, Brinster (1973) detected it in 8-cell-stage embryos. This discrepancy could reflect genetic differences between the strains of mice used in different studies. Genetic differences could affect either the promoter/enhancer sequences or the developmental cues and so alter the time of gene activation. Autoregulation of gene expression occurs in some eukaryotic systems (Serfling, 1989) and it is possible that the falling level of oocyte-coded GPI-1 acts as a developmental cue that triggers expression of the embryonic Gpi-1s genes in the preimplantation embryo. The oocytes of some strains of mice have low levels of GPI-1 (Peterson and Wong, 1978) and it is possible that this could trigger the early expression of embryonic Gpi-1s genes.

In the present study, we have compared the transition from oocyte-coded to embryo-coded GPI-1 in mouse embryos that inherited high, medium or low levels of oocyte-coded GPI-1 activity. These comparisons were made in order to address two questions. First, we asked whether the survival time of oocyte-coded GPI-1 activity, in the early embryo, is affected by the activity level in the oocyte. Second, we asked whether embryo-coded GPI-1 is detected earlier in embryos that inherit low levels of oocyte-coded GPI-1.

Mice

The following strains of mice (with abbreviated strain designations in parentheses) were maintained under conventional conditions in the Centre for Reproductive Biology: inbred strain C57BL/Ola (B); partially congenic strains (8 backcross generations) CS7BL/Ola.129-Gpi-1sa/Ws (B-Gpi-1sa), C57BL/Ola.AKR-c Gpi-1sa/Ws (B-c), C57BL/Ola-Gpi-1sc/ Ws (B-Gpi-lsc) and CS7BL/ Ola-Gpi-1sa-m1h / Ws (B-null). The B-null stock carries the Gpi-1sa-m1h null allele (Peters and Ball, 1986) and was produced by crossing a Gpi-lsa/ Gpi-1sa-m1h (a/null) heterozygote (gift of Dr J.Peters) with C57BL/Ola (B). Heterozygous Ola-Gpi-1sb/ Gpi-1sa-m1h (b/null) offspring were crossed with B-Gpi-lsa and subsequent generations crossed alternately to B and B-Gpi-lsa strains (14 generations). The other stocks are described more fully by West and Fisher (1984).

Egg and embryo collection

Ovulation was induced by intraperitoneal injections of 5i.u. pregnant mares’ serum gonadotrophin (PMS) at approximately 12 noon followed 48 h later by 5i.u. human chorionic gonadotrophin (hCG). Unfertilized eggs were collected from the oviducts of females approximately 21 – 23 h after hCG injection and handled and stored as previously described (West and Fisher, 1984; West et al. 1986). Unfertilized eggs were collected from females of strains B, B-Gpi-1sa, B-c, B-Gpi-1sc and DBA, from various heterozygotes produced by intercrossing these strains and from B-Gpi-1sc/ Gpi-1sa-m1h (a/null) heterozygotes. Samples comprising 4 (B-Gpi-1sc ), 2 (DBA) or individual (all other genotypes) unfertilized eggs were stored for GPI-1 densitometric assays.

Embryos were produced by mating superovulated B-c, B-Gpi-1sa, or B-Gpi-1sc strain females with B males on the day of the hCG injection. Embryos were collected between 9am and 12 noon, so that those collected on the day after the hCG injection (i.e. the day of the vaginal plug) were designated day p.c. Preimplantation embryos ( to days) were flushed from the reproductive tract and postimplantation embryos ( and day egg cylinders) were dissected from the decidual swellings in M2 medium (Quinn et al. 1982) as previously described (West et al. 1986). Samples of eggs or embryos were stored at – 20 ° C as described by West and Green (1983). All samples were frozen and thawed three times in liquid nitrogen vapour before electrophoresis.

Postimplantation embryos for electrophoretic quantification of allozymes were stored individually but samples of to day embryos usually comprised 3 (B-c × B)F1 embryos, 3 (B-Gpi-1sa × B)F1 embryos or 4 to 5 (B-Gpi-1sc × B)F1 embryos. Samples of embryos for GPI-1 activity assays were either individual (B-c × B)F1 or (B-Gpi-1sa × B)F1 embryos ( to daysp.c.) or groups of (B-Gpi-1sc × B)F1 embryos (5 per group at days and 3 per group at to and days p.c.).

Glucose phosphate isomerase electrophoresis

Samples of embryos were applied directly from fine Pasteur pipettes to 76 × 60mm Helena, Titan III cellulose acetate electrophoresis plates. Electrophoresis, staining for GPI-1 activity and quantification of the different GPI-1 allozymes was done as previously described (West et al. 1986). When one allozyme is present as a very low proportion of the total GPI-1 the proportion may, for technical reasons, be overestimated (see below and West and Green, 1983).

Glucose phosphate isomerase densitometric assays

The activity of GPI-1 in samples of eggs or preimplantation embryos was estimated relative to that in standard samples that contained a known number of unfertilized C57BL/Ola strain eggs by scanning densitometry as described by West et al. (1989). The activity is expressed relative to the activity in one C57BL/Ola unfertilized egg. (Peterson et al. (1985) have shown that for C57BL/6J eggs this is 1.92 nmol NADPH/egg/ hour at 35 ° C.) Control experiments (West et al. 1989) showed that quantification by this method was excellent when the activity of the test sample was between one and three times that in the standard sample (control results for groups of 1, 2 and 3 eggs were 1.07 ± 0.05, 2.05 ±0.13 and 2.91±0.12, respectively). When the ratio of the two samples was 4:1 or 5:1 the minor component tended to be slightly overestimated (control results were 3.66±0.15 and 4.52±0.24, respectively). In practice, a single B strain egg was used as a standard for all test samples and the number of eggs or embryos in the test sample was adjusted either to approximately match the expected GPI-1 activity of a single B strain egg or fall within the linear range (between one third and three times the standard sample).

GPI-1 activity in unfertilized eggs

Comparison of unfertilized eggs from females of various genotypes (Table 1) revealed a wide range in GPI-1 activity. Three strains, B-c (high activity), B-Gpi-1sa (medium activity) and B-Gpi-lsc (low activity) were selected for the comparative analysis of the transition from oocyte-coded to embryo-coded GPI-1.

Table 1.

Relative glucose phosphate isomerase activity in unfertilized eggs from female mice of different genotypes

Relative glucose phosphate isomerase activity in unfertilized eggs from female mice of different genotypes
Relative glucose phosphate isomerase activity in unfertilized eggs from female mice of different genotypes

Changes in GPI-1 activity in embryos

Table 2 shows the change in total GPI-1 activity during preimplantation development of the three genotypes chosen for study. In both (B-B-Gpi-1sa×B)F1 and (B-Gpi-1scxB)F1 embryos, GPI-1 levels reach a minimum at days (morula and blastocyst stages)and then rise again at days (hatched blastocyst). The GPI-1 activity in (B-c×B)F1 embryos also declines at days. Although there is a 4-fold difference in GPI-1 activity between the three genotypes at iday, by 4idays embryos in all 3 groups have similar total GPI-1 activities.

Table 2.

Change in relative GPI-1 activity during preimplantation development of embryos from three crosses (female×male)

Change in relative GPI-1 activity during preimplantation development of embryos from three crosses (female×male)
Change in relative GPI-1 activity during preimplantation development of embryos from three crosses (female×male)

Changes in GPI-1 allozymes in embryos

Representative densitometric tracings of electrophoresis plates are shown in Fig. 1. Table 3 shows the changes in the proportions of GPI-1 allozymes during the first week of development of (B-c×B)F1 embryos (high oocyte-coded GPI-1 activity). Only the oocyte-coded A allozyme (AA dimer) was detected at to days and in about half of the samples at days p.c. One sample of two day blastocysts produced an additional weak AB band but no detectable B allozyme. The remaining to day samples all produced all three allozymes. All but one of the postimplantation samples ( and days) produced allozymes close to the 1:2:1 ratio expected if no oocyte-coded GPI-1 remained and GPI-1 was entirely embryo-coded. The exceptional day embryo produced allozymes closer to the ratio 4A:4AB:1B. The proportion of total GPI-1 activity that is oocyte-coded was calculated in two ways (Table 3). When no oocyte-coded GPI-1 remains the expected allozyme ratio is IA: 2AB: IB (as seen in most and day embryos), so the proportion of embryo-coded GPI-1 is approximately twice the proportion of the AB allozyme; the proportion of oocyte-coded GPI-1 is then 1 minus the embryo-coded proportion. The second calculation makes use of the proportions of all three allozymes to derive the proportion of oocyte-coded GPI-1 from the formula m/(m+p+q) as noted in Table 3 and explained by West et al. (1986). The two calculations produced similar estimates of the proportion of residual oocyte-coded GPI-1. The final column of Table 3 shows the estimated contributions of the maternally derived Gpi-lsa allele and the paternally derived Gpi-lsb allele to the embryo-coded GPI-1 activity (the calculation is explained by West et al. 1986). Apart from the one anomolous day embryo, mentioned above, the two alleles are expressed to a similar extent. However, in these experiments, it was not possible to test whether the maternally derived, embryo-coded Gpi-1s allele was activated before the paternally derived allele.

Table 3.

Change in GPI-1 allozyme ratios during preimplantation development of (B-c×B)F1 embryos (a/a females×b/b males)

Change in GPI-1 allozyme ratios during preimplantation development of (B-c×B)F1 embryos (a/a females×b/b males)
Change in GPI-1 allozyme ratios during preimplantation development of (B-c×B)F1 embryos (a/a females×b/b males)
Fig. 1.

Representative densitometric tracings of GPI-1 allozymes in (B-Gpi-1sa×B)F1 and (B-Gpi-1sc× B)F1 embryos showing the transition from oocyte-coded to embryo-coded enzyme. The peaks represent the proportions of the different GPI-1 allozymes (A, AB and B or B, BC, and C) in embryos at (a) 212 days, (b) 312 days, (c) 412 days and (d) 512 days post coitum.

Fig. 1.

Representative densitometric tracings of GPI-1 allozymes in (B-Gpi-1sa×B)F1 and (B-Gpi-1sc× B)F1 embryos showing the transition from oocyte-coded to embryo-coded enzyme. The peaks represent the proportions of the different GPI-1 allozymes (A, AB and B or B, BC, and C) in embryos at (a) 212 days, (b) 312 days, (c) 412 days and (d) 512 days post coitum.

Table 4 shows the equivalent analysis for the (B-Gpi-1sa×B)F1 embryos (medium oocyte-coded GPI-1 activity). The overall pattern is similar to the (B-cxB)Fj embryos (Table 3) but embryo-coded enzyme was detected in a higher proportion of day samples (8/10 compared to 11/21). Again, one day sample produced a strong A band, a weak AB band but no B allozyme, but all three allozyme bands were visible in the majority of samples. Most and day embryos produced allozymes close to the expected 1:2:1 ratio but, as for the (B-c×B)F1 embryos, there were exceptions. Two day (B-Gpi-1sa×B)F1 embryos had allozyme ratios close to 4A:4AB:1B. Apart from the exceptions noted, the expression of the maternally derived, embryo-coded Gpi-1s allele was approximately equal to that for the paternally derived, embryo-coded allele (p:q ratios in Tables 3 and 4).

Table 4.

Change in GPI-1 allozyme ratios during preimplantation development of (B-Gpi-lsa×B)F1 embryos (a/a females×b/b males)

Change in GPI-1 allozyme ratios during preimplantation development of (B-Gpi-lsa×B)F1 embryos (a/a females×b/b males)
Change in GPI-1 allozyme ratios during preimplantation development of (B-Gpi-lsa×B)F1 embryos (a/a females×b/b males)

Table 5 shows the results for the (B-Gpz-1sc×B)F1 embryos (low oocyte-coded GPI-1 activity). Some day samples and two day samples produced a strong C allozyme, a weak B allozyme but no CB allozyme. (It was not certain whether the weak band was a genuine B allozyme because it often seemed to be a little further from the C allozyme than is usual for a B allozyme.) This could be either a technical artefact or, for example, a result of transient expression of sperm GPI-1 mRNA in day embryos or early expression of the paternally derived Gpi-1sb allele in the day embryos. To test-this, samples containing five unfertilized B-B-Gpi-1sc eggs or five day (B-Gpt-1sc×B)F1 embryos were run in alternate positions on the same electrophoresis plates. Some of both the unfertilized eggs (3/11 samples) and the embryos (4/11 samples) produced a weak extra band. It therefore seems probable that the extra band in seven of the eighteen day embryos and in two of the fourteen day embryos, shown in Table 5, was a trailing sub-band of GPI-1C or another technical artefact rather than a genuine GPI-1B allozyme.

Table 5.

Change in GPI-1 allozyme ratios during preimplantation development of (B-Gpi-1sc×B)F1 embryos (c/c females ×b/b males)

Change in GPI-1 allozyme ratios during preimplantation development of (B-Gpi-1sc×B)F1 embryos (c/c females ×b/b males)
Change in GPI-1 allozyme ratios during preimplantation development of (B-Gpi-1sc×B)F1 embryos (c/c females ×b/b males)

The other to day (B-Gpz-1sc×B)F1 samples only produced the oocyte-coded C allozyme and the other day samples (12/14) produced all three allozyme bands. Similarly all and and all but one of the day embryos produced all 3 allozymes. (The anomalous day embryo had no detectable C allozyme and this probably reflects the greater lability of this allozyme.) Most of the and day samples produced allozymes close to the ratio 1C:4BC: IB but two day embryos had allozymes that were closer to a ratio of 4C:4BC:1B.

The (B-Gpi-1sC×B)F1 embryos produced C, BC and B allozymes rather than A, AB and B. The C allozyme is known to be heat labile in vitro (Padua el al. 1978) and the stability of the BC allozyme is intermediate between B and C (West and Flockhart, 1989). Although unfertilized eggs from the B-Gpi-1sC strain have low GPI-1 activity (Table 1) it has been suggested (West et al. 1987) that the activity may be initially higher than normal but declines rapidly in vivo to a low level in some tissues. Furthermore, the ratio of C:BC:B allozymes in (B-Gpz-1sc×B)F1 hybrids differs from the more usual 1:2:1. We have suggested (West et al. 1987) that the initial ratio may be close to 4C:4BC:1B (perhaps as a result of increased synthesis of GPI-1C monomers) and that this 4:4:1 ratio is subsequently modified, in a tissue-specific way, by differential stabilities of the three allozymes. This may account for the 1C:4BC:1B ratios seen in most of the and day embryos.

The unusual allozyme ratios, coupled with the differences in stability of the allozymes, introduces some complications to the analysis of the proportion of GPI-1 that is oocyte-coded. Three calculations, based on different assumptions, are shown in Table 5. The most accurate is probably that shown in the last column. This assumes that when no oocyte-coded GPI-1 remains the allozyme ratio will be close to 1C:4BC: IB, as seen in most and day samples. The proportion of embryo-coded GPI-1 would be the proportion of the BC allozyme multiplied by 6/4; the proportion of oocyte-coded GPI-1 would then be 1 minus the embryo-coded proportion.

Although, in Table 2, the proportion of day (B-c×B)F1 embryos that had reached the blastocyst stage was lower than those in the other two groups, this was not true for embryos collected for allozyme analysis (Tables 3 to 5). It seems unlikely that there are significant differences in rate of development among the three groups of heterozygous, congenic embryos.

Changes in oocyte-coded and embryo-coded GPI-1 activities in embryos

The calculated proportions of oocyte-coded GPI-1 (l-2×2 in Tables 3 and 4, and l-(6×2/4) in Table 5) were used to calculate the activities of oocyte-coded and embryo-coded GPI-1 from the total GPI-1 activities shown in Table 2. (Weighted means were used where there were different sub-groups of samples at a particular age. For example, the mean proportion of oocyte-coded GPI-1 in day (B-c×B)F1 embryos was calculated as 0.90 from the five values (21 samples) given in Table 3.)

The results (Fig. 2) show that in all three groups the embryo-coded GPI-1 is first detected at days. At this time the oocyte-coded GPI-1 activity differs between the groups but the embryo-coded activity is the same in all groups. By days, similar levels of oocyte-coded GPI-1 remain in (B-c×B)F1 embryos and (B-Gpi- 1sa×B)F1 embryos but this has almost all disappeared from the (B-Gpi-1sc×B)F1 embryos. The total GPI-1 activity is similar in each group because the embryo-coded GPI-1 activity is higher in the (B-Gpt-1sc×B)F1 embryos.

Fig. 2.

Loss of oocyte-coded GPI-1 activity and onset of embryo-coded activity in mouse embryos of three different genotypes.

Fig. 2.

Loss of oocyte-coded GPI-1 activity and onset of embryo-coded activity in mouse embryos of three different genotypes.

The oocyte-coded GPI-1 activity remains at a more or less constant level during the first days of development in all three genotypes, although this level is different for each genotype. Maintenance of oocyte-coded enzyme levels during the first days of development suggests either that oocyte-coded GPI-1 enzyme is stable during this phase of development or that enzyme levels are maintained by translation of oocyte-coded GPI-1 mRNA present in the early embryo, as previously discussed for GPI-1 (West et al. 1986) and other enzymes (Harper and Monk, 1983). Biochemical and molecular studies suggest that most oocyte-coded mRNA is inactivated in the late 2-cell mouse embryo (Flach et al. 1982; Bolton et al. 1984) but some mRNA may persist until the early blastocyst stage (Bachvorova and De Leon, 1980).

The oocyte-coded GPI-1 enzyme level begins to fall after days and is virtually absent by days in all three genotypes. Some oocyte-coded activity remains in day (B-Gpi-1sa×B)F1 and (B-c×B)F1 embryos, which both inherit the A allozyme as their oocyte-coded GPI-1, but there is virtually none left in (B-Gpi-1sa: ×B)F1 embryos, which inherit the more labile C allozyme.

Despite a 4- to 5-fold difference in initial oocyte-coded GPI-1 activity, no differences were seen among the three genotypically distinct groups of embryos in the activation of the embryonic Gpi-1s genes. Although embryo-coded GPI-1 was detected in a lower proportion of day samples of (B-c×B)F] embryos (11/ 21) than in the two groups with lower oocyte-coded activity (8/10 and 12/14), both the time when embryo-coded GPI-1 was first detected, and its mean activity at this time, were remarkably consistent in all three groups. Moreover the embryos with the lowest oocyte-coded GPI-1 activity had a form of oocyte-coded GPI-1 that is known to be heat labile in vitro and is probably also less stable in vivo. The oocyte-coded GPI-1 activity was lower in these embryos, before gene activation, than in the high group once embryo-coded GPI-1 was detected. The similarity, in the timing of embryonic gene expression, between (B-Gpi-1sc×B)F1 embryos and the other two groups implies that the embryonic Gpi-1s genes are not activated by a mechanism that depends on a critical threshold of oocyte-coded GPI-1 enzyme activity, but a threshold based on oocyte-coded GPI-1 mRNA remains a possibility.

The occurrence of atypical allozyme ratios suggests that the two embryonic Gpi-1sa alleles are not always expressed identically. One day (B-c×B)F1 embryo and two day (B-Gpi-1sa×B)Fi embryos each produced allozymes closer to the ratio 4A:4AB:1B than the typical 1:2:1 ratio (Tables 3 and 4). Residual oocyte-coded GPI-1 or maternal contamination (A allozyme) could explain the high proportion of A allozyme but would not account for the elevated AB/B ratio. It therefore seems more likely that the maternally derived Gpi-1sa allele is over-expressed, relative to the paternally derived Gpi-1sb allele, in these three atypical embryos. (The expression ratio is approximately 2:1). The overall frequency of this unusual ratio among and day embryos of these two genotypes was 3/60 (5%). In all other samples of Gpi-1sa/Gpi-1sb heterozygous embryos, the expression of the a and b alleles appeared to be equal. However, as pointed out earlier, the analysis would not detect the expression of the maternally derived, embryo-coded Gpi-1s allele until the paternally derived allele is expressed. It is possible that the atypical embryos were either trisomic for chromosome 7 or, more likely, digynic triploids. Takagi and Sasaki (1976) reported a 5-fold increase in the frequency of digynic triploid embryos after A/He strain females were superovulated. At -days, the controls produced 4.5 % and the superovulated A/He females produced 19.7 % triploid embryos.

Although most studies, like the present one, have first detected embryo-coded GPI-1 in day morulae or blastocysts, Brinster (1973) detected embryo-coded GPI-1 activity in a group of 500 (SJL/J×C57BL/6)F1 embryos at the 8-cell stage of development (days). This discrepancy could reflect technical differences, differences in the number of embryos per group, genetic differences between strains of mice that alter the timing of gene expression or, perhaps most likely, the inclusion of precocious embryos (e.g. morulae) among the day embryos (this could involve genetic differences between strains of mice that affect the rate of development). West et al. (1986) also found traces of embryo-coded enzyme in some samples of (129×C57BL)F1 embryos at days but these samples included embryos that had advanced beyond the 8-cell stage.

The results of this study answers the two questions originally posed. First, a relatively labile form of oocyte-coded GPI-1 was exhausted before a more stable form of the enzyme. In contrast, the comparison of the high and medium genotypes suggests that the initial GPI-1 activity in the oocyte had relatively little effect on the duration of oocyte-coded activity in these embryos. Second, embryo-coded GPI-1 was first detectable at the same developmental stage ( day compacted morulae) in all three groups. The level of oocyte-coded GPI-1 in the high group, when embryo-coded GPI-1 was first detected, was higher than the level in the low group at any stage prior to detection of embryo-coded GPI-1. Thus, the two embryonic Gpi-1s alleles cannot be activated by a mechanism that depends on a critical threshold of oocyte-coded GPI-1 enzyme activity. Once both embryonic Gpi-1s alleles were activated, they were usually equally expressed in B-Gpi-1sa/Gpi-1sb heterozygotes but in a few (possibly digynic triploid) embryos the maternally derived Gpi-1s allele was expressed to a greater extent. In conclusion we have revealed a genetic effect on the duration of oocyte-coded enzyme that is probably related to enzyme lability but we have not found a genetic effect on the timing of embryonic gene expression.

We thank Jo Peters for supplying mice with the null allele of Gpi-ls, Denis Doogan and Jan Speed for expert mouse husbandry, Tom McFetters and Ted Pinner for preparing the figures and the Wellcome Trust for financial support (grant to JDW). We are also grateful to both referees for constructive criticism of the manuscript.

Auerbach
,
S.
and
Brinster
,
R. L.
(
1967
).
Lactate dehydrogenase isozymes in the early mouse embryo
.
Expl Cell Res.
46
,
89
92
.
Bachvorova
,
R.
and
De Leon
,
V.
(
1980
).
Polyadenylated RNA of mouse ova and loss of maternal RNA in early development
.
Devi Biol
.
74
,
1
8
.
Bolton
,
V. N.
,
Oades
,
P. J.
and
Johnson
,
M. H.
(
1984
).
The relationship between cleavage, DNA replication and gene expression in the mouse 2-cell embryo
.
J. Embryol. exp. Morph
.
79
,
139
162
.
Brinster
,
R. L.
(
1973
).
Parental glucose phosphate isomerase activity in three-day mouse embryos
.
Biochem. Genetics
9
,
187
191
.
Chapman
,
V. M.
,
Whitten
,
W. K.
and
Ruddle
,
F. H.
(
1971
).
Expression of paternal glucose phosphate isomerase (Gpi-1) in preimplantation stages of mouse embryos
.
Devi. Biol
.
26
,
153
158
.
Duboule
,
D.
and
Burki
,
K.
(
1985
).
A fine analysis of glucose-phosphate-isomerase patterns in single preimplantation mouse embryos
.
Differentiation
29
,
25
28
.
Flach
,
G.
,
Johnson
,
M. H.
,
Braude
,
P. R.
,
Taylor
,
A. S.
and
Bolton
,
V. N.
(
1982
).
The transition from maternal to embryonic control in the 2-cell mouse embryo
.
EM BO J
.
1
,
681
686
.
Gilbert
,
S. F.
and
Solter
,
D.
(
1985
).
Onset of paternal and maternal Gpi-1 expression in preimplantation mouse embryos
.
Devi Biol
.
109
,
515
517
.
Harper
,
M. I.
and
Monk
,
M.
(
1983
).
Evidence for translation of HPRT enzyme on maternal mRNA in early mouse embryos
.
J. Embryol. exp. Morph
.
74
,
15
28
.
Monk
,
M.
and
Petzoldt
,
U.
(
1977
).
Control of inner cell mass development in cultured mouse blastocysts
.
Nature, Lond
.
265
,
338
339
.
Padua
,
R. A.
,
Bulfield
,
G.
and
Peters
,
J.
(
1978
).
Biochemical genetics of a new glucosephosphate isomerase allele (Gpi-I) from wild mice
.
Biochem. Genet
.
16
,
127
143
.
Peters
,
J.
and
Ball
,
S. T.
(
1986
).
Induced mutations of Gpi-ls
.
Mouse News Letter
74
,
92
.
Peterson
,
A.
,
Choy
,
F.
,
Wong
,
G.
,
Clapoff
,
S.
and
Friar
,
P.
(
1985
).
Glucosephosphate isomerase (GPI-1) expression in mouse ova: cis regulation of monomer realisation
.
Biochem. Genet
.
23
,
827
846
.
Peterson
,
A. C.
and
Wong
,
G. G.
(
1978
).
Genetic regulation of glucose phosphate isomerase in mouse oocytes
.
Nature, Lond
.
276
,
267
269
.
Quinn
,
P.
,
Barros
,
C.
and
Whittingham
,
D. G.
(
1982
).
Preservation of hamster oocytes to assay the fertilizing capacity of human spermatozoa
.
J. Reprod. Fert
.
66
,
161
168
.
Rosenfteld
,
M. G.
,
Nelson
,
C.
,
Crenshaw
,
E. B.
, III
,
Elsholtz
,
H. P.
,
Lira
,
S. A.
,
Mangalam
,
H. J.
,
Franco
,
R.
,
Waterman
,
M.
,
Weinberger
,
C.
,
Hollenberg
,
S. H.
,
Gigere
,
V.
,
Ong
,
E.
and
Evans
,
R. M.
(
1987
).
Developmental and hormonal regulation of neuroendocrine gene transcription
.
Recent Prog, tn Hormone Res
.
43
,
499
534
.
Serfling
,
E.
(
1989
).
Autoregulation - a common property of eukaryotic transcription factors?
Trends in Genetics
5
,
131
133
.
Spielmann
,
H.
,
Eibs
,
H.-G.
,
Jacob-Muller
,
U.
and
Bischoff
,
R.
(
1978
).
Expression of lactate dehydrogenase isozyme 5 (LDH-5) in cultured mouse blastocysts in the absence of implantation and outgrowth Biochem
.
Genet
.
16
,
191
202
.
Takagi
,
N.
and
Sasakj
,
M.
(
1976
).
Digynic triploidy after superovulation in mice
.
Nature, Lond
.
264
,
278
281
.
West
,
J. D.
and
Fisher
,
G.
(
1984
).
A new allele of the Gpi-lt temporal gene that regulates the expression of glucose phosphate isomerase in mouse oocytes
.
Genet. Res
.
44
,
169
181
.
West
,
J. D.
and
Flockhart
,
J. H.
(
1989
).
Non-additive inheritance of glucose phosphate isomerase activity in mice heterozygous at the Gpi-ls structural locus
.
Genet. Res
.
54
,
27
35
.
West
,
J. D.
,
Flockhart
,
J. H.
,
Angell
,
R. R.
,
Hillier
,
S. G.
,
Thatcher
,
S. S.
,
Glasier
,
A. F.
,
Rodger
,
M. W.
and
Baird
,
D. T.
(
1989
).
Glucose phosphate isomerase activity in mouse and human eggs and pre-embryos
.
Human Reprod
.
4
,
82
85
.
West
,
J. D.
and
Green
,
J. F.
(
1983
).
The transition from oocyte-coded to embryo-coded glucose phosphate isomerase in the early mouse embryo
.
J. Embryol. exp. Morph
.
78
,
127
140
.
West
,
J. D.
,
Leask
,
R.
,
Flockhart
,
J. H.
and
Fisher
,
G.
(
1987
).
High activity of an unstable form of glucose phosphate isomerase in the mouse
.
Biochem. Genet
25
,
543
561
.
West
,
J. D.
,
Leask
,
R.
and
Green
,
J. F.
(
1986
).
Quantification of the transition from oocyte-coded to embryo-coded glucose phosphate isomerase in mouse embryos
.
J. Embryol. exp. Morph
.
97
,
225
237
.