Regulation of gene expression in the mouse oocyte and early preimplantation embryo: developmental changes in Sp1 and TATA box-binding protein, TBP

During the growth phase, oocytes synthesize and accumulate RNA (mRNA, rRNA and tRNA), proteins and organelles (e.g., mitochondria) that constitute the maternal contribution to early development. In the mouse, this maternal contribution sustains and directs the first cell cycle, which is about 18-20 hours long (Howlett and Bolton, 1985). Oocyte maturation initiates the degradation of maternal mRNA and this destruction is essentially complete by the end of the 2-cell stage (Bachvarova and DeLeon, 1980; Bachvarova et al., 1985, 1989; Paynton et al., 1988). Zygotic gene activation (ZGA), which has definitely occurred during the 2-cell stage (Flach et al., 1982; Sawicki et al., 1982; Bensaude et al., 1983; Latham et al., 1991; Manejwala et al., 1991; Schultz, 1993), replaces maternal transcripts that are lost during this time and are common to both the oocyte and preimplantation embryo, as well as generating novel ones that are required for embryogenesis. In this regard, it should be noted that whereas cleavage to the 2-cell stage does not require transcription, subsequent cleavages do require transcription (Golbus et al., 1973; Bolton et al., 1984; Poueymirou and Schultz, 1989). Thus, ZGA in the 2-cell embryo is requisite for further development.

The onset of ZGA in the mouse appears to be a function of the time following fertilization rather than on cell cycle progression, i.e., ZGA may be regulated by a zygotic clock rather than the number of cell cycles (Schultz, 1993). Although the regulation of this maternal to zygotic transition is poorly understood at the molecular level, post-translational modifications of maternally derived proteins are implicated. Numerous changes in the pattern of protein synthesis occur during the first cell cycle and these are due to the recruitment of maternal mRNAs (Cascio and Wassarman, 1982) and post-translational protein modification, mainly protein phosphorylation (van Blerkom, 1981). Inhibition of the cAMP-dependent protein kinase (PKA), but not protein kinase C or the calmodulin-modulated protein kinase, inhibits ZGA (Poueymirou and Schultz, 1987, 1989, 1990). Also consistent with a role for post-translational regulation regulating ZGA is that the increase in hsp70 mRNA that occurs between the 1-cell and 2-cell stages is not inhibited by treatment of 1-cell embryos with cycloheximide (Manejwala et al., 1991).

More recent work indicates that although the 1-cell embryo has previously been referred to as transcriptionally inert, a transcriptionally permissive state develops in the late 1-cell embryo (Latham et al., 1992) and that expression of an Sp1-dependent luciferase reporter gene is detected in G2 of the 1-cell embryo (Ram and Schultz, 1993). It should be stressed that regulation of reporter gene expression likely reflects the same mechanisms that underlie regulation of endogenous gene expression, since the expression of a reporter gene is also regulated by the ‘zygotic clock’ (Wiekowski et al., 1991, 1993) and manifests the same dependence of PKA activity as do endogenous genes (Schwartz and Schultz, 1992). Interestingly, expression of this reporter gene in the 1-cell embryo is only detected following microinjection of the male pronucleus; no expression is detected following microinjection of the female pronucleus (Ram and Schultz, 1993). Moreover, this difference in the abilities of male and female pronuclei to support expression of an exogenous reporter gene is also observed following culture to times that correspond to the late 2-cell/early 4-cell stage (Wiekowski et al., 1993). These data suggest that ZGA may begin late in the 1-cell mouse embryo rather than during 2-cell stage and that there is a difference between the male and female pronuclei in their ability to support transcription of a reporter gene and, most likely, therefore, endogenous genes. In addition, expression of the Sp1-dependent luciferase reporter gene is not detected following microinjection of the germinal vesicle of the fully grown oocyte (Ram and Schultz, 1993). This was unanticipated because the fully grown oocyte is transcriptionally active (Wassarman and Letourneau, 1976) and synthesizes all classes of RNA (Brower et al., 1981), albeit the rate of synthesis is likely less than that in growing oocytes (Moore and Lintern-Moore, 1978).

In order to explore further the basis for these changes in expression of the Sp1-dependent luciferase reporter gene, we examined expression of this reporter gene during oocyte growth, as well as analyzing the temporal and spatial pattern of expression of Sp1 during oocyte growth, the first cell cycle, and in the 2-cell embryo by confocal microscopy and immunoblotting. We also conducted a similar set of experiments for the TATA-box-binding protein, TBP, since this protein is a central component of the basal transcription apparatus. We report that the expression of the reporter gene, which is readily detected in growing oocytes, decreases during oocyte growth and that this decrease correlates with a decrease in the nuclear concentration of Sp1, as well as the amount of Sp1. The nuclear concentration of TBP also decreases during oocyte growth. During the first cell cycle the nuclear concentration of both Sp1 and TBP increases and the concentration of each is greater in the male pronucleus; these increases are inhibited by either aphidicolin or cycloheximide. Last, the nuclear concentration of these transcription factors continues to increase between the 1-cell and 2-cell stages and these increases are independent of transcription and cell division.

Growing oocytes were obtained from prepubertal mice (CF-1, Harlan) 13, 15, 17 and 19 days of age as previously described (Eppig, 1977). Cumulus cell-free, fully grown oocytes were obtained from PMSG-primed CF-1 mice 6 weeks of age as previously described (Schultz et al., 1983). In all cases, 0.2 mM 3-isobutyl-1-methyl xanthine (IBMX) was included in the medium to inhibit germinal vesicle breakdown (Schultz et al., 1983). The collecting medium was bicarbonate-free minimal essential medium (Earle’s salts) supplemented with pyruvate (100 μg/ml), gentamicin (10 μg/ml), polyvinylpyrrolidone (3 mg/ml) and 25 mM Hepes, pH 7.2 (MEM/PVP).

Oocytes were maintained in culture at 37°C in an atmosphere containing 5% CO₂ in air and in MEM/PVP in which the Hepes was replaced by 25 mM sodium bicarbonate. 1-cell embryos were collected from superovulated mice mated to B6D2/J₁ males (Jackson Laboratory) and synchronized as previously described (Ram and Schultz, 1993). Briefly, following collection of the fertilized eggs, those containing a visible pronucleus were discarded. The remaining cells were examined every 30 minutes and those that formed a pronucleus were culled and used for the experiments described below. Embryo culture was conducted in CZB medium (Chatot et al., 1989) at 37°C in an humidified atmosphere containing 5% CO₂ in air.

The following drugs were used at the indicated final concentrations to inhibit specific events during preimplantation development: α-amanitin, 24 μg/ml; aphidicolin, 3 μg/ml; cycloheximide, 10 μg/ml; cytochalasin D, 1 μg/ml; H8 (N-[2-(methylamino)ethyl]-5-isoquino-linesulfonamide), 50 μM. 1-cell embryos were transferred into medium containing the indicated drugs immediately following pronucleus formation and incubated until S phase (6 hours), mid-S phase (9 hours), or G₂/M phase (12 hours). 2-cell embryos were obtained from 1-cell embryos cultured overnight in medium containing the indicated drugs. In these instances, the drugs were added at S phase (6 hours after pronucleus formation) and the embryos incubated overnight, except for cycloheximide, which was added at G₂/M. Cytochalasin D and aphidicolin-treated 1-cell embryos were cultured until they were chronologically at the 2-cell stage.

The germinal vesicle of oocytes at different stages of the growth phase was microinjected with 1 pl of 400 ng/μl of a luciferase reporter gene that is driven by the SV40 early promoter (pGL2-Promoter vector, Promega) in TE buffer (10 mM Tris-HCl, pH 7.6, 0.25 mM EDTA) as previously described (Ram and Schultz, 1993). The SV40 early promoter is utilized by the mouse embryo (Bonnerot et al., 1991; Vernet et al., 1992; Ram and Schultz, 1993) and the Sp1 transcription factor is required for expression of the SV40 early promoter. The oocytes were injected within a 60 minute period and about 80% survived the injection procedure. The microinjected oocytes were then cultured for 24 hours in MEM/PVP containing bicarbonate and 1 mM N⁶-monobutyryl cAMP. Following culture, the oocytes were removed and frozen in luciferase assay buffer. As a positive control for each experiment, the male pronucleus of 1-cell embryos at early S phase was also microinjected with the luciferase reporter gene, cultured for 24 hours in CZB medium containing aphidicolin and then frozen in luciferase assay buffer.

Luciferase assays were performed as previously described (Ram and Schultz, 1993). A standard curve was constructed using known amounts of firefly luciferase and the light emission present in the samples was converted to fg of luciferase. Under the assay conditions the background was 100–120 relative light units and the assay could readily detect 5 fg of luciferase, which typically resulted in ∼170–180 relative light units. The experiments were conducted on at least two, and usually three, independent occasions. Similar results were obtained in each case and the data were pooled.

Oocytes and embryos were collected and the zona pellucida was removed with acid Tyrode solution (Bornslaeger and Schultz, 1985). The zona pellucida-free oocytes/embryos were briefly washed in phosphate-buffered saline containing 3 mg/ml polyvinylpyrrolidone (PBS/PVP) and fixed in 3.7% paraformaldehyde overnight at 4°C. All subsequent steps were carried out at room temperature in a humidified chamber. Oocytes/embryos were washed in 2 drops of PBS/PVP and permeabilized in 0.1% Triton X-100/PBS for 15 minutes. The cells were then washed in PBS/PVP and placed in blocking solution (0.1% BSA, 0.01% Tween-20, PBS) for 15 minutes. All subsequent washes and antibody incubations were performed in blocking solution. The oocytes/embryos were incubated for 60 minutes in primary antibody. The cells were washed in 4 drops of blocking solution for 15 minutes each and incubated in 0.5 μg/ml of secondary antibody (goat anti-rabbit IgG conjugated to rhodamine, Cappel) for 60 minutes in the dark. The cells were washed again and mounted with coverslip compression in VectaShield antibleaching solution (Vector Labs). To determine specificity of the primary antibody, individual antibodies were neutralized with the appropriate antigen (see below) overnight at 4°C prior to incubation with the 2-cell embryos. Sp1 affinity-purified polyclonal antibody, which recognizes both the 95,000 and 105,000 M_r forms of Sp1, and its blocking peptide were purchased from Santa Cruz Biotechnology. The blocking peptide encodes amino acids 520–538 of the Sp1 protein and is located in the carboxy terminus just upstream of the zinc finger DNA-binding domain. The final concentration of the Sp1 antibody was 1.0 μg/ml while its blocking peptide was 20 μg/ml. TATA-binding protein (TBP) polyclonal antibody was a gift from Frank Pugh; a 1:100 dilution was used. Recombinant TBP (Promega) was used at a final concentration of 0.2 μg/ml. As a control, 2-cell embryos were incubated with affinity-purified normal rabbit IgG as a control 1° antibody or the 2° antibody alone.

Fluorescence was detected on a laser-scanning confocal micro-scope equipped with Bio-Rad MRC-600 software or by an Image I analysis system (Interactive Video Systems, Inc., Concord, MA) where specified. Confocal images were taken at 60× magnification and a zoom level equal to 1 to give 0.65 μm sections through an oocyte or embryo. All data were quantified using an Image I/AT image processor (Interactive Video Systems, Inc., Concord, MA). For each experimental series, which consisted of either oocytes or embryos, oocytes or embryos at the different stages were processed together under the same conditions and images were captured using the same number of filters. In addition, 2-cell embryos were included in each experiment to serve as an internal control (see below). The concentration of fluorescence reported for the confocal images was determined by averaging the pixel value of fluorescence observed within a constant area from 4 or 5 different regions from the nucleus and the cytoplasm. The average fluorescent pixel value of the cytoplasm was then subtracted from the average fluorescent pixel value of the nucleus to obtain the average concentration of nuclear fluorescence. Since the relative fluorescence was greatest in the nuclei of the 2-cell embryo, this value was arbitrarily set as 100% and the values obtained for the other stages are expressed relative to this value.

For the results of experiments shown in Figs 6 and 7, the images were collected and analyzed on an Image I/AT image processor. The concentration of fluorescence was determined and expressed relative to that in 2-cell embryos as described above.

Whole-cell protein extracts were prepared as follows. 300 oocytes or embryos were added to 30 μl of protein extraction buffer (100 mM NaCl, 20 mM Tris-HCl, pH 7.5, 0.5% Triton X-100, 0.5% NP-40) containing the protease inhibitors phenylmethylsulfonyl fluoride (1 mM) and leupeptin-pepstatin (1 μg/ml) and incubated on ice for 15 minutes. The extract was briefly vortexed, centrifuged and frozen at −80°C. Extracted proteins, in addition to pure Sp1 protein (Promega) that was used as a separate sample, were denatured by heating to 85°C for 10 minutes in the presence of Laemmli buffer (Laemmli, 1970) and 1% β-mercaptoethanol and then subjected to SDS-PAGE using a 7.5% gel. The fractionated proteins were transferred onto nitrocellulose (Schleicher and Schuell) according to Sambrook et al. (1989). The nitrocellulose was blocked with 5% skim milk in 10 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.2% Tween-20 (TBST) for 30 minutes. The blocked nitrocellulose membrane was incubated for 60 minutes with either 5.0 μg/ml of Sp1-specific polyclonal antibody in 5% skim milk/TBST or 5.0 μg/ml of Sp1-specific polyclonal antibody neutralized with 5.0 μg/ml of Sp1 peptide in 5% skim milk/TBST overnight at 4°C. The blot was washed 5 times with TBST for 5 minutes for each wash and then incubated with 5.0 μg/ml of biotinylated goat antirabbit secondary antibody (Vector Labs) in 5% skim milk/TBST for 60 minutes. The blot was again washed 5 times with TBST for 5 minutes for each wash and then incubated with the VectaStain ABC reagent for horseradish peroxidase detection prepared in 5% skim milk/TBST for 30 minutes. The blot was washed 5 times with TBST for 5 minutes for each wash and proteins were detected using the Enhanced Chemiluminescence detection system following the manufacturer’s instructions (Amersham).

Pure human Sp1 protein (Promega) was used to obtain an estimate of the amount of Sp1 present in oocytes or embryos, as well as to determine which molecular weight form predominates at a particular stage during development. The purified Sp1 exists predominantly as the 95,000 M_r unphosphorylated form, with the 105,000 M_r phosphorylated form running as a smear above it (Jackson and Tjian, 1988).

We previously noted that the Sp1-dependent luciferase reporter gene was not expressed following microinjection of germinal vesicle-intact fully grown oocytes, whereas it was expressed following microinjection of 1-cell embryos (Ram and Schultz, 1993). Other investigators, however, using a similar type of Sp1-dependent reporter gene have detected expression following microinjection of oocytes obtained from mice 13–14 days of age (Bonnerot et al., 1991). Accordingly, we examined expression of this luciferase reporter gene during oocyte growth to ascertain if there was a growth-associated decrease in reporter gene expression.

As previously reported (Ram and Schultz, 1993), no luciferase was detected in the fully grown oocyte (Fig. 1). In contrast, luciferase activity was readily detected in growing oocytes and this activity markedly decreased as the oocytes grew. This expression of luciferase activity required the SV40 early promoter, since microinjection of a vector not containing the promoter (pGL-2 Basic, Promega) into the germinal vesicle of growing oocytes did not result in any detectable luciferase activity (data not shown). In addition to the decrease in the amount of luciferase present in these oocytes, the fraction of microinjected oocytes that displayed detectable luciferase activity also decreased with oocyte growth. Whereas luciferase activity was detected in 80% of the oocytes about 40-49% of the final oocyte volume, luciferase activity was only detected in 39% of the oocytes about 51-80% of the final oocyte volume, and no activity was ever detected following injection of fully grown oocytes (Fig. 1).

It should also be noted that these experiments were conducted following culture of the oocytes in medium containing 1 mM N⁶-monobutyryl cAMP. The reason for this is that the fraction of oocytes capable of resuming spontaneous maturation in vitro increases with oocyte size (Sorensen and Wassarman, 1976; Wickramasinghe et al., 1991) and transcription terminates following germinal vesicle breakdown. Spontaneous maturation is inhibited by membrane-permeable activators of PKA, such as N⁶-monobutyryl cAMP and IBMX (Schultz, 1990). Similar levels of luciferase expression were observed for meiotically incompetent oocytes obtained from mice 13 days of age (none of these oocytes resume spontaneous maturation in vitro) that were cultured in the presence or absence of N⁶-monobutyryl cAMP (control, 443±56; N⁶-monobutyryl cAMP, 348±80; P>0.05, ANOVA). In contrast, when these oocytes were cultured in medium containing 0.2 mM IBMX an 80% inhibition in luciferase activity was observed (control, 443±56; IBMX, 93±15; P<0.01, ANOVA). The basis for this inhibition is not known.

As described above, the level of expression of the Sp1-dependent luciferase reporter gene decreases during oocyte growth and then increases following fertilization (Ram and Schultz, 1993). To determine if these changes in reporter gene expression correlate with changes in the nuclear concentration of Sp1, we examined the temporal changes in nuclear Sp1 con-centration by immunofluorescent confocal microscopy¹. The ability to quantify the Sp1-specific immunofluorescence in optical sections permitted calculation of changes in the relative nuclear Sp1 concentration. This is because, although the nuclear volume increases during oocyte growth, the same volume in each optical section was analyzed at each developmental stage. It should be noted that it is important to determine changes in the nuclear concentration of a transcription factor, rather than changes in its amount, since the activity of these proteins depends on their nuclear concentration.

As expected, Sp1 was localized to the nucleus in 2-cell embryos (Fig. 2A) and this staining was specific, since an initial incubation of the 1° antibody with the blocking peptide abolished the nuclear fluorescence (Fig. 2B). In addition, no specific nuclear staining was observed when either affinity-purified normal rabbit IgG was used as the 1° antibody or when the 2° antibody was used alone (data not shown). The nuclear concentration of Sp1 decreased during oocyte growth (Fig.ability to quantify the Sp1-specific immunofluorescence in optical sections permitted calculation of changes in the relative nuclear Sp1 concentration. This is because, although the nuclear volume increases during oocyte growth, the same volume in each optical section was analyzed at each developmental stage. It should be noted that it is important to determine changes in the nuclear concentration of a transcription factor, rather than changes in its amount, since the activity of these proteins depends on their nuclear concentration.

As expected, Sp1 was localized to the nucleus in 2-cell embryos (Fig. 2A) and this staining was specific, since an initial incubation of the 1° antibody with the blocking peptide abolished the nuclear fluorescence (Fig. 2B). In addition, no specific nuclear staining was observed when either affinitypurified normal rabbit IgG was used as the 1° antibody or when the 2° antibody was used alone (data not shown). The nuclear concentration of Sp1 decreased during oocyte growth (Fig. 3A,a-e) and quantification of these data is shown in Fig. 4A. It should be noted that the relative Sp1 concentration was significantly less than that observed in the nuclei of 2-cell embryos. In all cases, Sp1 was localized to the germinal vesicle but was excluded from the nucleolus (Fig. 3A,a-e).

We also examined changes in nuclear concentration of TBP during oocyte growth. TBP is an essential component of the basal transcription machinery and is required for the successful activation of genes transcribed by RNA polymerase II (Weis and Reinberg, 1992 and references therein). Nuclear staining of TBP was observed (Fig. 2C) and this staining was specific, since it was abolished following incubation of the anti-TBP antibody with recombinant TBP (Fig. 2D); the reason for the apparent membrane staining is not known. Similar to Sp1, a decrease in the relative nuclear TBP concentration was observed during oocyte growth (Fig. 4B), although the kinetics of this decrease were different from those for Sp1. The decrease in the nuclear concentration of TBP was greatest in oocytes obtained from juveniles 13 and 15 days old, whereas the decrease in Sp1 occurred between days 17 and 19. In addition, the relative nuclear concentration of TBP in the oocyte was significantly less than that in the 2-cell embryo.

The decrease in Sp1 nuclear concentration observed during oocyte growth could result from (1) a decrease in the total cellular amount of Sp1, (2) a translocation of Sp1 from the nucleus to the cytoplasm, or (3) a masking of the epitope recognized by the antibody. To determine if a decrease in the total cellular amount of Sp1 occurred during oocyte growth, we undertook an immunoblot analysis.

Immunoblot analysis with the Sp1 antibody of an equal number of growing and fully grown oocytes revealed the presence of a band of M_r=95,000 in both growing and fully grown oocytes and incubating the antibody with the peptide resulted in a >97% decrease in the intensity of the signal (Fig. 5, lanes 2-5). Image I analysis of these bands indicated that the amount of Sp1 present in the fully grown oocytes was 71% that in the growing oocytes. When the difference in oocyte volume was taken into account – the volume of the fully grown oocytes was 2.5-fold greater than that of the growing oocytes as determined by measuring their diameters at 200-fold magnification with an ocular micrometer – the relative concentration of Sp1 in the fully grown oocyte was about 28% that in the growing oocytes. Results of another experiment indicated that the relative concentration of Sp1 in the fully grown oocyte was about 21% that in the growing oocytes. The average decrease in the relative concentration of Sp1 (about 76%) was very similar to that observed for the decrease in Sp1 nuclear concentration determined (81%) (Fig. 4A).

It should also be noted that the apparent molecular weight of Sp1 in both the growing and fully grown oocytes was less than that of the human Sp1 used as a standard (Fig. 5, compare lanes 1, 4 and 5) and that the apparent molecular weight of Sp1 in fully grown oocytes was slightly greater than that in the growing oocytes. Sp1 bears multiple O-linked N-acetylglucosamine residues that appear to enhance the ability of Sp1 to activate transcription in vitro (Jackson and Tjian, 1988). In addition, Sp1 can be phosphorylated (Jackson et al., 1990); the phosphorylated species has an M_r=105,000 whereas the nonphosphorylated species has an M_r=95,000. Although the differences in apparent molecular weight that are observed during oocyte growth may be attributed to differences in the extent of these post-translational modifications, the molecular basis for this change in apparent molecular weight is not known. Last, for the experiment shown in Fig. 5, the signals obtained from 30 pg and 300 pg were also determined and a linear relationship was observed, i.e., a constant number of pixels/pg of protein. From this, we calculated that 300 fully grown oocytes contained 58 pg of Sp1 and this corresponds to 0.0008% of total oocyte protein, assuming there are 25 ng of protein/oocyte (Schultz and Wassarman, 1977). This concentration of Sp1 is in the range (0.0017–0.0025%) to that observed in Hela cells (Briggs et al., 1986).

Immunoblot blot analysis of TBP in oocytes and embryos was unsuccessful due to the insensitivity of the TBP antibody.

As described above the level of expression of the Sp1-dependent reporter gene increases during the first cell cycle and preferential expression is observed from the male pronucleus (Ram and Schultz, 1993). To determine if these changes in reporter gene expression correlate with changes in the pronuclear concentration of Sp1, we examined the temporal changes in nuclear concentration of Sp1, as well as TBP, by immunofluorescent confocal microscopy.

For both Sp1 and TBP there was a dramatic increase in the pronuclear concentration following fertilization (Figs 3A,B,f-h; 4A,B); the relative change in Sp1 concentration was greater than that observed for TBP. This increase, which was present 6 hours following pronuclear formation, continued during the course of the first cell cycle and by G2 the pronuclear concentration of Sp1 and TBP was very similar to that observed in the 2-cell embryo. In addition, the concentration of Sp1 and TBP was greater in the male pronucleus at all times examined and the difference was more pronounced for TBP. This difference was unlikely due to a sampling artifact, since image analysis of each nucleus in the 2-cell embryo for either Sp1 (n=12) or TBP (n=11) indicated that the concentration difference between these two nuclei in the same embryo was <1% (data not shown).

The timing of ZGA is regulated apparently by anzygotic clock and is independent of cell cycle progression (Schultz, 1993). To ascertain if the observed changes in the nuclear concentration of Sp1 and TBP that occur during the first cell cycle were also independent of cell cycle progression, 1-cell embryos were treated with a variety of inhibitors and their effects analyzed.

Culture of 1-cell embryos that had recently formed a pronucleus in either aphidicolin, which inhibits DNA synthesis, or cycloheximide, which inhibits protein synthesis, inhibited the time-dependent increase in nuclear concentration of both Sp1 and TBP (Fig. 6A,B). Note again, that the concentration of Sp1 and TBP was greater in the male pronucleus than in the female pronucleus.

Culture of 1-cell embryos in the presence of either α-amanitin or the inhibitor of PKA, H8, each of which inhibits ZGA in the mouse embryo (Poueymirou and Schultz, 1989), did not substantially inhibit the increase in nuclear concentration of either Sp1 or TBP that occurred between the 1-and 2-cell stages (Fig. 7). These inhibitors were added during S of the first cell cycle, and thus, the additional increase in nuclear concentration was not due to ZGA. Consistent with this was that the similar amounts of Sp1 were detected by immunoblotting following culture of 1-cell embryos to the 2-cell stage in either the presence or absence of α-amanitin (Fig. 5, lanes 6 and 7). Immunoblot analysis also revealed that cleavage was associated with the conversion of the higher molecular weight species to a lower molecular weight one.

Cycloheximide inhibited the increase and the effect was more pronounced for TBP than for Sp1 (Fig. 7); this difference could reflect differences in the stabilities of these proteins. The increase in nuclear concentration was affected very little by inhibiting cytokinesis with cytochalasin D, whereas inhibiting DNA synthesis resulted in a marked inhibition. It should be noted that an increase in nuclear concentration was still observed for both Sp1 and TBP when compared to values obtained for 1-cell embryos analyzed 12 hours post-pronuclear formation, and that the difference in concentration between the pronuclei was still apparent.

We report here that developmental changes in the expression of an Sp1-dependent luciferase reporter gene during oocyte growth and early embryogenesis correlate positively with changes in the nuclear concentration of Sp1, as well as TBP. In addition, the preferential reporter gene expression observed following microinjection of the male pronucleus, when compared to the expression observed following microinjection of the female pronucleus, is also positively correlated with a higher concentration of Sp1 and TBP in the male pronucleus. The decrease in nuclear Sp1 concentration during oocyte growth is likely due to a decrease in the amount of Sp1 present in these oocytes. The increase in nuclear concentration of Sp1 and TBP that occurs during the first and second cell cycles requires DNA synthesis, but is independent of cytokinesis or activation of the embryonic genome.

The failure to detect expression of the Sp1-dependent luciferase reporter gene in fully grown oocytes is surprising, since the fully grown oocyte is transcriptionally active (Wassarman and Letourneau, 1976; Brower et al., 1981). Nevertheless, several lines of evidence suggest that the rate of transcription decreases substantially during growth. First, a dramatic decrease in both RNA polymerase I and II activity is observed (Moore and Lintern-Moore, 1978) when an in situ assay that detects RNA polymerases actively engaged in transcription is used (Moore, 1975). When accounting for the increase in nuclear volume that occurs during oocyte growth, an 8-fold decrease in RNA polymerase activity occurs between follicle stages 5b and 6/7. Consistent with a decrease in the rate of transcription during oocyte growth is that both rRNA (Jahn et al., 1976; Brower et al., 1981) and mRNA (Bachvarova, 1981; Brower et al., 1981) are very stable in the growing oocyte and that oocytes accumulate essentially all of their rRNA (Kaplan et al., 1982) and mRNA (Sternlicht and Schultz, 1981) when they are about 70% the volume of the fully grown oocyte. Last, chromatin condensation occurs during oogenesis (Wickramasinghe et al., 1991; Debey et al., 1993) and this most likely contributes to a reduced rate of expression. It should be noted that a decrease in reporter gene expression during oocyte growth has also been noted by others (Bevilacqua et al., 1992).

Consistent with the decreased reporter gene expression observed in this study is the decrease in nuclear concentration of both Sp1 and TBP. The excellent agreement between the extent of the decrease in Sp1 nuclear concentration and the decrease in Sp1 detected by immunoblotting minimizes the likelihood that the decrease in nuclear Sp1 concentration is due to masking of the epitope recognized by the antibody or partitioning of Sp1 to the cytoplasm. A similar conclusion cannot be reached for TBP since the antibody used in our studies was not sensitive enough for immunoblotting. The decrease in nuclear concentration of these two transcription factors does provide an explanation at the molecular level for the decrease in expression of the Sp1-dependent luciferase reporter gene and reporter genes used by others.

As with oocytes of other species, proteins accumulate during the growth phase (Schultz and Wassarman, 1977), and the synthesis and accumulation of these proteins constitute part of the maternal contribution to early development. The decrease in the amount of Sp1, and the likely decrease in the amount of TBP that occurs during oocyte growth, is therefore surprising since Sp1 is required for transcription of a large number of genes (Dynan, 1986; Kadonaga et al., 1986) and TBP is an essential component of the transcription apparatus for virtually all genes (Sharp, 1992). For RNA polymerase II transcripts, TBP is essential for transcription of both TATA box-containing and TATA box-free promoters. For TATA box-free promoters, Sp1 may play a pivotal role in recruiting TBP to the initiation complex (Pugh and Tjian, 1990, 1991). The decrease in nuclear concentration for each of these two central transcription factors suggests that the transcription of many genes may decrease during oocyte growth and this is consistent with the apparent decrease in RNA polymerase II activity. The transcription of genes can be exquisitely sensitive to the concentration of a transcription factor due to cooperative interactions. For example, relatively small changes in the concentration of bicoid result in very dramatic changes in the transcription of the target gene, hunchback (Driever and Nüsslein-Volhard, 1988, 1989). The cooperative nature of protein-protein interactions involved in the formation and activation of the transcription complex on a promoter may account for the observation that the relatively modest changes in the nuclear concentration of Sp1 and TBP that occur during oocyte growth correlate with dramatic changes in expression of the Sp1-dependent reporter gene. Moreover, the changes in nuclear concentration that we observe for these two transcription factors may result in even more pronounced changes in their activity due to post-translational modifications that alter either their ability to bind DNA or activate, directly or indirectly, RNA polymerase (e.g., Gottesfeld et al., 1994).

Changes in the nuclear concentration of these two transcription factors could also have pronounced effects on the expression of endogenous genes. During oocyte growth, the transcription of genes containing weak promoters will likely be more affected than genes containing stronger promoters and it is interesting to note that, during mouse oocyte growth, the synthesis of a number of proteins, as assessed by two-dimensional gel electrophoresis, decreases (Schultz et al., 1979). It should also be noted that although Sp1 is frequently involved in transcription of house-keeping genes, changes in Sp1 activity can profoundly affect the transcription of ‘differentiated’ genes. For example, a decrease in Sp1 activity occurs during chicken embryogenesis and this decrease can account for 90% of the developmental switch in embryonic ρ-globin expression in primitive erythroid lineage cells (Minie et al., 1992).

The nuclear concentration of both Sp1 and TBP increases following fertilization and pronuclear formation, and the nuclear concentration of each of these proteins is greater in the male pronucleus. This difference correlates well with the enhanced expression of reporter genes from the male pronucleus (Ram and Schultz, 1993; Wiekowski et al., 1993) and provides an explanation at the molecular level for this difference in transcription. Although the molecular basis for this difference is not known, it could reflect chromatin remodeling that occurs in the male pronucleus as a consequence of the replacement of the sperm-derived protamines with the maternally derived histones (Clarke, 1992; Perreault, 1992); there is no evidence for chromatin remodeling of the female pronucleus during the first cell cycle. This remodeling must entail formation of nucleosomes, which inherently repress basal transcription by inhibiting the binding of the basal transcription apparatus to the TATA box (Wolffe, 1991; Felsenfeld, 1992; Svaren and Hörz, 1993; Struhl, 1993). Transcription factors can compete with histones for DNA sequences (Laybourn and Kadonaga, 1991; Felsenfeld, 1992), and thus chromatin remodeling may make promoter sequences more accessible to transcription factors such as Sp1 and TBP.

The increase in the male and female pronuclear concentrations of both Sp1 and TBP that occurs during the first cell cycle requires DNA synthesis, since aphidicolin inhibits this increase. A possible explanation for this result is that DNA replication transiently disrupts chromatin structure and therefore could make more accessible DNA sequences recognized by transcription factors. Consistent with this is that the transcription factor Gal4-VP16, which is a fusion protein containing the DNA-binding domain of yeast Gal4 and the transcriptional activation domain of herpes simplex virus protein VP16, can potentiate transcription in vitro during, but not after, DNA replication and nucleosome formation (Kamakaka et al., 1993). The inhibitory effect of cycloheximide on the increase in pronuclear concentration of Sp1 and TBP may be attributed to its ability to inhibit DNA synthesis (Burhans et al., 1991). The cycloheximide effect could also result from inhibiting translation of maternal mRNAs that encode for these transcription factors. Last, the decrease in nuclear concentration of these transcription factors may reflect their turnover. It also should be noted that the male pronucleus forms before the female pronucleus and that DNA synthesis may initiate in the male pronucleus prior to initiating in the female pronucleus (Luthardt and Donahue, 1973; Abramczuk and Sawicki, 1975). If DNA replication does facilitate the binding of transcription factors to DNA, this temporal advantage may result in an initially greater concentration of transcription factors in the male pronucleus when compared to their concentration in the female pronucleus. The biological consequence(s), if any, of this preferential transcription from the male pronucleus is unknown.

Inhibiting ZGA with either α-amanitin or H8, each of which inhibits ZGA by a different mechanism (Poueymirou and Schultz, 1989), has little, if any, inhibitory effect on the increase in nuclear concentration of either Sp1 or TBP that occurs between the 1-cell and 2-cell stages. In fact, similar amounts of Sp1 are detected in 2-cell embryos cultured from the 1-cell stage in the presence or absence of α-amanitin. These results are consistent with maternally derived transcription factors providing the components of the transcription machinery required for ZGA. The finding that inhibiting cytokinesis has little inhibitory effect on the increases in nuclear concentration of these transcription factors was anticipated, since ZGA is independent of cytokinesis. Inhibiting protein synthesis with cycloheximide has a more pronounced effect on inhibiting the increase in nuclear concentration of TBP than it has for Sp1. This difference may reflect differences in the half-lives of these two proteins, as well as when transcription of these genes initiates following ZGA.

In summary, the nuclear concentration of Sp1 and TBP correlate positively with changes in reporter gene expression during oocyte growth and early embryogenesis. Future studies will address the molecular basis for these changes, as well as analyzing genes that are regulated by these and other transcription factors.

This research was supported by a grant from the NIH (HD 22681) to R. M. S; D. M. W. was supported in part by an NRSA Fellowship (F32 HD 07803). The authors would like to thank Mel DePamphilis for stimulating discussions concerning regulation of zygotic gene activation.