p53-Dependent S-Phase Damage Checkpoint and Pronuclear Cross Talk in Mouse Zygotes with X-Irradiated Sperm

Abstract

One difficulty in analyzing the damage response is that the effect of damage itself and that of cellular response are hard to distinguish in irradiated cells. In mouse zygotes, damage can be introduced by irradiated sperm, while damage response can be studied in the unirradiated maternal pronucleus. We have analyzed the p53-dependent damage responses in irradiated-sperm mouse zygotes and found that a p53-responsive reporter was efficiently activated in the female pronucleus. [³H]thymidine labeling experiments indicated that irradiated-sperm zygotes were devoid of G₁/S arrest, but pronuclear DNA synthesis was suppressed equally in male and female pronuclei. p53^−/− zygotes lacked this suppression, which was corrected by microinjection of glutathione S-transferase-p53 fusion protein. In contrast, p21^−/− zygotes exhibited the same level of suppression upon fertilization by irradiated sperm. About a half of the 6-Gy-irradiated-sperm zygotes managed to synthesize a full DNA content by prolonging S phase, while the other half failed to do so. Regardless of the DNA content, all the zygotes cleaved to become two-cell-stage embryos. These results revealed the presence of p53-dependent pronuclear cross talk and a novel function of p53 in the S-phase DNA damage checkpoint of mouse zygotes.

Ionizing radiation induces DNA double-strand breaks (DSB) and inflicts a variety of damage responses which include induction of cell cycle checkpoint and apoptosis. The p53 pathway plays a prominent role in the damage responses of higher eukaryotes (26). The signal generated by DSB is transduced by ATM and related sensor kinases to phosphorylate p53, which activates various functions of the protein (5, 7, 20).

It has been well documented that embryonic development in mouse is quite sensitive to radiation. A high level of p53 was correlated to radiation induction of apoptosis without cell cycle arrest in post-implantation stage mouse embryos (19, 23). This p53-dependent apoptosis was shown to be the major mechanism to suppress radiation-induced teratogenesis in mice (33). Earlier stages of mouse development, especially from fertilization to implantation, were shown to be highly radiosensitive (14). However, the mechanisms of the high radiosensitivity have been poorly elucidated. Mouse zygotes are known to possess an extremely high level of p53 (21). This indicates that p53 might play an important role in damage response of mouse zygotes.

One difficulty in studying the damage response is that the effect of damage itself and the effect of cellular responses are hard to distinguish, because these two always come together in irradiated cells. In a mammalian zygote, genomes of the incoming sperm and residing oocyte form two separate pronuclei. DNA synthesis occurs in two pronuclei, which thereafter merge to enter M phase. This gives a unique opportunity to analyze the mechanism of damage responses; damage can be delivered through irradiated sperm while the responses can be analyzed in damage-free female pronuclei. Indeed, our previous study clearly demonstrated that introduction of DNA damage by irradiated sperm induces genomic instability which destabilizes a maternally derived minisatellite allele in F₁ mice (32).

In the present study, we have investigated the p53-dependent radioresponse in mouse zygotes fertilized by X-irradiated sperm. The results definitely demonstrate p53-dependent pronuclear cross talk in zygotes. In addition, we have found a novel function of p53 in the S-phase DNA damage checkpoint, which suppresses the rate of pronuclear DNA synthesis in mouse zygotes.

MATERIALS AND METHODS

Mice.

ICR strains of mice 8 to 10 weeks old were used in the present experiment. The p53^−/− C57BL/6 × CBA mice were kindly provided by S. Aizawa (42). The p21^−/− 129 mice were purchased from The Jackson Laboratory (Bar Harbor, Maine). The p53⁻ allele was then introduced into ICR strains by repeated backcross. All mice were maintained in our animal facility on a 12 h of light and 12 h of dark cycle starting at 7:00 a.m. under standard temperature conditions (23 ± 2°C).

X-irradiation.

Partial irradiation was performed on the testicular area of male mice, with a lead shield covering other parts of the body, at a dose rate of 1 Gy/min (250 keV; 15 mA with 1-mm-thick Al filter; Rigaku Radioflex X-ray generator). Irradiation of isolated sperm and zygotes was carried out at room temperature in microdrops of the medium submerged in liquid paraffin. Plasmid DNA was irradiated in a buffer under the X-ray generator without the filter at a dose rate of 30 Gy/min.

In vivo and in vitro fertilization.

In vitro fertilization of mouse oocytes and subsequent culture were carried out as described (30, 44). Virgin ICR female mice were induced to superovulate with 5 U of pregnant mare's serum (Teikokuzoukiseiyaku Ltd., Tokyo, Japan), followed 46 to 48 h later by 5 U of human chorionic gonadotropin (Teikokuzoukiseiyaku Ltd.). Female mice were sacrificed at 15 to 16 h after human chorionic gonadotropin administration, and the oocytes were released from the ampullar portion of the oviducts into the medium submerged in equilibrated paraffin oil. Sperm was obtained from the cauda epididymides of male ICR mice and suspended in medium submerged in equilibrated paraffin oil. After capacitation by incubating for 2 h at 37°C under 5% CO₂ in an air environment, sperm was added to the oocytes. Zygotes were washed 6 h later and were transferred into fresh microdroplets of culture medium.

For in vivo fertilization, females of the proestrus stage were identified by examination of vaginal smears at 5:00 p.m. These females were caged at 6:00 a.m. the next morning with males for 1 h. Successful mating were identified by the presence of a vaginal plug. Females were sacrificed at 6 h after mating, and the zygotes were collected as described above. Zygotes were washed and transferred into microdroplets of culture medium, the cumulus cells were removed by treatment with hyaluronidase (10 mg/ml; Sigma), and the zygotes were cultured further for investigation.

Plasmid constructions, microinjection, analysis of lacZ expression, and protein purification.

A p53-responsive lacZ reporter plasmid was constructed. The HindIII fragment carrying 13 copies of the p53-responsive element (AGGCAAGTCCAGGCAGGCC) and the basic promoter of polyomavirus was isolated from PG13Py-lacZ plasmid (kindly provided by B. Vogelstein [22]). This was inserted in the HindIII site of plasmid pENL (kindly provided by Y. Nabeshima). The resulting plasmid, PG13-Py-NL-lacZ, carried the p53-responsive promoter upstream of the nuclear localizing lacZ.

The plasmid described above was microinjected into the female pronuclei of zygotes at 8 h postfertilization. Male and female pronuclei were distinguished by the size and the position with respect to the polar body. Expression of lacZ was examined at 20 and 24 h after fertilization by fixation with 1% glutaraldehyde and staining with 5-bromo-4-chloro-3-indolyl-β-d-galactoside. The success rate of microinjection into female pronuclei was around 50%. Some zygotes degenerated after microinjection, possibly due to harsh treatment, and some surviving zygotes did not stain. Therefore, we used at least 30 zygotes for each microinjection experiment. When all the zygotes were devoid of staining, as it was in the case of control zygotes and p53^−/− irradiated-sperm zygotes, then this was taken to represent a lack of response. When we saw staining in about 50% of the 30 zygotes, this was taken as a positive response. So, the data obtained here are qualitative rather than quantitative.

A construct was made in which human p53 cDNA was fused downstream of the glutathione S-transferase (GST) gene of pGEX plasmid (18). The plasmid was introduced into Escherichia coli, and the culture was induced with 0.1 mM isopropyl-β-d-thiogalactopyranoside. The GST-p53 fusion protein was purified with the standard procedure using glutathione agarose beads (39). The purified protein was prepared in a buffer (17) at a concentration of 1 mg/ml and microinjected into the cytoplasm of the zygotes at 4 h after fertilization.

Analysis of pronuclear DNA synthesis.

DNA synthesis and S-phase progression were monitored by pulse-labeling with ³H-labeled thymidine ([³H]TdR) (28). Briefly, zygotes were pulse-labeled for 1 h with [³H]TdR (1 μCi/ml) at 6 to 24 h after fertilization and chased with fresh medium containing 5 × 10⁻⁵ M cold TdR for another 1 h. The total amount of DNA synthesized was monitored by continuous labeling with [³H]TdR (0.1 μCi/ml) from 8 to 21 h after fertilization for in vitro-fertilized zygotes and from 8 to 18 h for in vivo-fertilized zygotes. After completion of labeling and chasing, zygotes were treated hypotonically in 0.9% sodium citrate and 3% fetal calf serum for 5 min at 37°C and fixed for 5 min in methanol-acetic acid-water (5:1:4, vol/vol/vol). The fixed zygotes were placed onto glass slides and washed with methanol-acetic acid (3:1, vol/vol). Slides were then treated with 5% trichloroacetic acid solution for 1 h at 0°C and were washed with ethanol for 5 min at room temperature. Slides were dipped in Kodak NTB2 emulsion and stored at 4°C for 2 weeks. After development, autoradiographs were stained with Giemsa stain, and the number of grains per nucleus was determined for male and female pronuclei.

Measurement of DNA content.

Zygotes were washed with phosphate-buffered saline once and placed onto cover glasses. After fixation with 70% ethanol for 5 min, zygotes were stained with propidium iodide (PI) (0.1 mg/ml) in phosphate-buffered saline containing proteinase K (0.1 mg/ml) and RNase (1 mg/ml) at 37°C for 30 min. The relative intensity of PI fluorescence was measured by laser scanning cytometer (LSC 101; Olympus, Tokyo, Japan). Care was taken to compare the PI intensities of nuclei on the same slide so that the variation in the staining was minimized.

RESULTS

Transactivation of a p53-responsive reporter in female pronuclei of zygotes with irradiated sperm.

The p53-dependent cross talk between two pronuclei was examined in zygotes fertilized in vitro by unirradiated sperm (control zygotes) and irradiated sperm (irradiated-sperm zygotes). A p53-responsive reporter plasmid carrying the nucleus-localizing lacZ gene was microinjected into female pronuclei at 8 h after fertilization. These zygotes were cultured and then fixed at 20 h (zygote stage) or 24 h (two-cell stage) after fertilization. There was no staining for the lacZ activity in microinjected control zygotes (data not shown). In contrast, a strong staining was observed in both pronuclei when the control zygotes were irradiated with 3 Gy at 4 h after fertilization (Fig. 1a). This staining persisted in two-cell-stage embryos (data not shown). A strong staining was also observed in 6-Gy-irradiated-sperm zygotes (Fig. 1b). The polar body which does not share cytoplasm with the zygote was devoid of the staining (Fig. 1b). Activation of the reporter was not observed in p53^−/− zygotes regardless of irradiation of p53⁻ sperm (Fig. 1c).

FIG. 1.

Activation of p53-responsive reporter in irradiated-sperm zygotes. Construction of the p53-responsive reporter plasmid is described in Materials and Methods. This reporter plasmid was microinjected into the female pronucleus at 8 h after fertilization. Zygotes were fixed and stained with X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) at 20 h (zygote stage) or 24 h (two-cell stage) after fertilization. (a) Strong staining with X-Gal was observed both in male pronuclei (♂) and female pronuclei (♀) of control zygotes exposed to 3 Gy of X rays at 4 h after fertilization. (b) This staining was also observed in 6-Gy-irradiated-sperm zygotes, but the polar body (arrowhead) was not stained. (c) There was no staining in p53^−/− irradiated-sperm zygotes. (d) Weak staining was observed in control zygotes microinjected with 3,000-Gy-irradiated reporter plasmid.

Irradiation of the reporter plasmid with 3,000 Gy of X rays, which was expected to introduce about one DSB per plasmid, resulted in a weak but definite staining upon microinjection into control zygotes (Fig. 1d). In addition, a weak staining was also observed for the KpnI-linearized reporter (data not shown). These results indicate that introduction of DNA damage, possibly DSB, by irradiated sperm triggered the p53-dependent pathway, which then activated the reporter in the female pronucleus.

S-phase progression of irradiated-sperm zygotes.

S-phase progression in zygotes was examined by pulse-labeling with [³H]TdR of control zygotes and 6-Gy-irradiated-sperm zygotes fertilized in vitro and in vivo. Table 1 shows that pronuclear DNA synthesis was first detected at 8 h after fertilization, both in control and in irradiated-sperm zygotes in vitro. Control zygotes gave the labeling index of 23% at 8 h after fertilization, the value for irradiated sperm was 21%, and there was no arrest at the G₁/S border in irradiated-sperm zygotes, even through activation of p53 was evident from Fig. 1b. In order to avoid a possible artifact of in vitro fertilization, we also investigated in vivo-fertilized zygotes recovered from natural mating. Control and irradiated-sperm zygotes recovered this way both entered S phase at 8 h after mating. The labeling index of control zygotes was 44% at 8 h after in vivo fertilization, and the value for irradiated sperm was 45% (Table 1). Direct irradiation of these zygotes with 3 Gy of X rays at 4, 6, and 8 h after mating was not effective in induction of G₁/S arrest (data not shown).

TABLE 1.

Duration of S phase in oocytes fertilized with irradiated sperm in vitro and in vivo

Fertilization	Time after fertilization (h)	Sperm irradiated with:
		0 Gy		6 Gy
		No. of oocytes examined	No. of labeled oocytes (%)	No. of oocytes examined	No. of labeled oocytes (%)
In vitro	6	28	0 (0)	39	0 (0)
	8	31	7 (23)	29	6 (21)
	11	23	16 (70)	24	18 (75)
	14	15	15 (100)	11	11 (100)
	17	24	24 (100)	18	17 (95)
	20	36	9 (25)	22	4 (18)
	21	25	1 (4)	28	5 (18)
	22	26	0 (0)	25	0 (0)
In vivo	6	15	0 (0)	12	0 (0)
	8	18	8 (44)	31	14 (45)
	16	8	8 (100)	12	11 (92)
	17	8	3 (27)	15	9 (60)
	18	24	2 (8)	26	13 (50)
	19	28	0 (0)	39	14 (36)
	20			10	2 (20)

The labeling index increased to around 70% at 11 h and then to 100% for both control and irradiated-sperm zygotes fertilized in vitro. The index dropped at 21 h. Delay in cessation of DNA synthesis was noted in some of the irradiated-sperm zygotes, and for these, uptake of [³H]TdR ceased at 22 h after fertilization (Table 1).

A similar analysis was performed with control and 6-Gy-irradiated-sperm zygotes fertilized in vivo. The duration of S phase in control zygotes was from 8 to 18 h after fertilization in vivo (Table 1). Some of the 6-Gy-irradiated-sperm zygotes exhibited a delay in S phase, and on average the S phase of about a half of these embryos was longer by 2 h than that of control zygotes. These results indicate that irradiation of sperm did not trigger G₁/S checkpoint in mouse zygotes but retarded S phase in some.

Suppression of pronuclear DNA synthesis by irradiation to sperm.

The rate of DNA synthesized was analyzed by pulse-labeling experiments. The numbers of grains were counted in male and female pronuclei. The results indicated that the rate of DNA synthesis was lower throughout S phase in both of the pronuclei of 6-Gy-irradiated-sperm zygotes fertilized in vitro (Fig. 2). A small but definite delay in S phase was evident for irradiated-sperm zygotes.

FIG. 2.

S-phase progression in zygotes. DNA synthesis was measured in pronuclei of zygotes by pulse-labeling with [³H]TdR for 1 h at various time points after fertilization in vitro. The numbers of zygotes examined are described in Table 1. Average grain counts of control zygotes were plotted for male (—○—) and female (---○---) pronuclei. Similarly, average grain counts of irradiated-sperm zygotes were plotted for male (—•—) and female (---•---) pronuclei. Error bars, standard deviations.

The total amount of DNA synthesized in S phase was measured by continuous labeling of the zygotes. Zygotes were incubated with [³H]TdR from 8 to 21 h after in vitro fertilization and from 8 to 18 h after in vivo fertilization. Control zygotes had a large number of grains in both pronuclei, and incorporation of [³H]TdR was also noted in the polar body (Fig. 3a). The number of grains was suppressed drastically in both male and female pronuclei of 6-Gy-irradiated-sperm zygotes (Fig. 3b). The grain numbers of both pronuclei were scored for each of the in vitro-fertilized oocytes, and the values are plotted in Fig. 3c. The data indicated a strong reduction of grain numbers in both pronuclei by irradiation of sperm. The average of the grain counts is plotted in Fig. 3d, which demonstrates that the grain numbers in female and male pronuclei of irradiated-sperm zygotes were suppressed to 35.4 and 35.1%, respectively, of those of the control zygotes. Similar suppression was observed for irradiated-sperm zygotes recovered from natural mating, and the average grain counts decreased to 40.3% for female pronuclei and 44.1% for male pronuclei (Fig. 3e and f). Suppression of DNA synthesis was not specific for irradiated-sperm zygotes, and direct irradiation of zygotes to 3 Gy of X rays at 8 h after fertilization was also effective (data not shown).

FIG. 3.

Grain counts for total DNA synthesized in S phase. DNA synthesis was monitored by continuous labeling with [³H]TdR of zygotes obtained from in vitro fertilization and in vivo natural mating. (a) A large number of grains was observed in control zygotes fertilized in vivo. (b) Grain count was suppressed in irradiated-sperm zygotes fertilized in vivo, and this suppression was observed both in male pronuclei (♂) and female pronuclei (♀). The polar body is indicated by an arrowhead. Grain count was plotted for female pronuclei (abscissa) and male pronuclei (ordinate) for in vitro-fertilized zygotes (c) and for in vivo-fertilized zygotes (e). The average of grain counts was plotted for in vitro-fertilized zygotes (d) and for in vivo-fertilized zygotes (f). Average grain count was suppressed in irradiated-sperm zygotes (solid bars) compared with that of control zygotes (open bars). Error bars, standard deviations.

p53-dependent suppression of pronuclear DNA synthesis.

Involvement of p53 in suppression of pronuclear DNA synthesis was tested. p53^−/− males were irradiated with 6 Gy of X rays and mated immediately with p53^−/− females. Zygotes were recovered at 6 h after mating and labeled with [³H]TdR throughout S phase. The onset and the duration of S phase were not affected by the p53-null state of the zygotes (data not shown). Grain count analysis (Fig. 4) indicated that sperm irradiation had no effect on DNA synthesis in p53^−/− zygotes. However, the suppression was observed in p53 heterozygous zygotes in which the p53-null allele was derived either from the father or from the mother (data not shown). These results demonstrate that the suppression of pronuclear DNA synthesis in irradiated-sperm zygotes is p53 dependent. Direct irradiation of zygotes was also tested, and suppression of grain counts was not observed in p53^−/− zygotes (data not shown).

FIG. 4.

Grain count analysis of p53^−/− and p21^−/− zygotes. The average grain count was plotted for both pronuclei of control and 6-Gy-irradiated-sperm wild-type (n = 18 and 36), p53^−/− (n = 12 and 18), p53^−/−-GST-p53 (n = 20 and 10), and p21^−/− (n = 11 and 12) zygotes. Grain count was not suppressed in p53^−/− zygotes by sperm irradiation. Average grain count was not suppressed in p53^−/− control zygotes by microinjection of GST-p53 fusion protein. In contrast, grain count was clearly suppressed by microinjection of GST-p53 fusion protein in p53^−/− irradiated-sperm zygotes. In p21^−/− irradiated-sperm zygotes, grain count was suppressed as much as that in wild-type irradiated-sperm zygotes. Error bars, standard deviations.

In order to confirm the direct involvement of p53 in the suppression, we microinjected GST-p53 fusion protein into the cytoplasm of p53^−/− irradiated-sperm zygotes. Microinjection of the fusion protein had no effect on pronuclear DNA synthesis of p53^−/− control zygotes (Fig. 4). In contrast, microinjection of the protein into p53^−/− irradiated-sperm zygotes restored the suppression, and the average grain count in female and male pronuclei decreased to 48 and 38% of the control values, respectively (Fig. 4).

Similar analysis was performed with p21^−/− zygotes. The results show that pronuclear DNA synthesis in p21^−/− irradiated-sperm zygotes was suppressed as that in wild-type irradiated-sperm zygotes was (Fig. 4). Therefore, suppression of pronuclear DNA synthesis was not dependent on p21.

Dose response of suppression of pronuclear DNA synthesis.

The dose response of suppression of pronuclear DNA synthesis was analyzed. Grain count analysis was done after continuous labeling of in vivo-fertilized zygotes from 8 to 18 h after mating. Figure 5 shows the average grain count of female and male pronuclei of p53^+/+ and p53^−/− zygotes where the male mice were exposed to 1, 2, 3, and 6 Gy of X rays and mated with females immediately. In p53^+/+ zygotes, X-irradiation of sperm at doses higher than 1 Gy suppressed grain count in both pronuclei and the grain count decreased dose dependently up to 3 Gy. In contrast, DNA synthesis was not suppressed by sperm irradiation in p53^−/− zygotes. Rather, the grain count even increased in these zygotes when spermatozoa were irradiated at doses of 1, 2, and 3 Gy. This increase in p53^−/− zygotes was not due to the enhanced DNA synthesis, because DNA content as assessed by the PI value was not affected by sperm irradiation (data not shown). Therefore, the increase was likely due to a decreased nucleotide pool size in irradiated-sperm p53^−/− zygotes.

FIG. 5.

Dose response of suppression of grain count in zygotes. (a) Average grain count of pronuclei was plotted against dose at which spermatozoa were irradiated. Each data point is the average of values for more than 10 zygotes. In p53^+/+ zygotes, grain counts of both female pronuclei (---○---) and male pronuclei (—○—) decreased in a dose-dependent manner up to 3 Gy of irradiation. In contrast, grain counts of p53^−/− zygotes did not decrease in both female pronuclei (---▵---) and male pronuclei (—▵—). Average grain counts even increased in p53^−/− zygotes at doses of 1, 2, and 3 Gy. Error bars, standard deviations.

DNA content analyses of zygotes.

[³H]TdR grain count can be influenced by the size of the nucleotide pool in the cells. Therefore, decreased grain count in irradiated-sperm zygotes may not represent a real suppression of DNA synthesis. It may be due to the increase in the nucleotide pool size and the resulting decrease in the specific activity of [³H]TdR. In order to rule out this possibility, we measured the PI values (in arbitrary units) of the pronuclei. In Fig. 6a, the PI value of both pronuclei was measured in control zygotes at 8 h after mating (pre-S phase) and 18 h after mating (post-S phase). The PI values of both pronuclei increased 2.1 times from pre-S phase to post-S phase, demonstrating that the method was reliable in estimating the relative DNA content. In contrast, the PI values of irradiated-sperm zygotes increased only 1.5-fold in female pronuclei and 1.4-fold in male pronuclei from 8 to 18 h after mating (Fig. 6b). Thus, suppression of S phase was confirmed by grain count analysis and by PI staining of the irradiated-sperm zygotes.

FIG. 6.

DNA content analyses of zygotes. DNA content was estimated by PI staining of zygotes at 8 h (pre-S-phase stage [open bars]) and 18 h (post-S-phase stage [striped bars]) after mating. PI value was measured in male pronuclei (♂) and female pronuclei (♀). (a) The PI value of control zygotes is plotted. The value of both pronuclei increased 2.1-fold from 8 h (n = 12) to 18 h (n = 13). (b) PI value of zygotes produced with sperm irradiated with 6 Gy of X rays. The value increased only 1.5-fold in female pronuclei and 1.4-fold in male pronuclei from 8 h (n = 17) to 18 h (n = 15). (c) The same analysis was performed in a p53^−/− background. The PI value of p53^−/− control zygotes was plotted. The value increased 2.1-fold in female pronuclei and 1.9-fold in male pronuclei from 8 h (n = 11) to 18 h (n = 11). (d) PI value of p53^−/− zygotes produced with sperm irradiated with 6 Gy of X rays. The value increased 2.1-fold in female pronuclei and 1.9-fold in male pronuclei from 8 h (n = 13) to 18 h (n = 13). Error bars, standard deviations.

Similar analyses were made in p53^−/− zygotes. PI values of both pronuclei in p53^−/− control zygotes increased about 2 times from 8 to 18 h after mating (Fig. 6c). PI values of both pronuclei in p53^−/− irradiated-sperm zygotes also increased about 2 times from 8 h to 18 h (Fig. 6d). These results are consistent with those of labeling experiments (Fig. 4) and once again demonstrate that the suppression of DNA synthesis is p53 dependent.

M phase progression of irradiated-sperm zygotes.

M phase progression was examined for in vivo fertilized zygotes by scoring two-cell-stage embryos (Fig. 7a). Control zygotes started to cleave at 17 h and all of them completed the division by 22 h. M phase progression was delayed by about 2 h in irradiated-sperm zygotes. The cleavage was little delayed in irradiated-sperm p53^−/− zygotes.

FIG. 7.

Progression of M phase in zygotes. (a) Zygotes were recovered from in vivo fertilization and analyzed for cleavage. The frequency of two-cell-stage embryos was scored from 16 to 23 h after mating. Cleavage started at 17 h and was completed at 22 h in p53^+/+ control zygotes (—○—) (n = 112) and in p53^−/− control zygotes (—▵—) (n = 28). In p53^+/+ irradiated-sperm zygotes (---•---) (n = 186), cleavage was delayed for 2 h, while the delay was marginal in p53^−/− irradiated-sperm zygotes (---▴---) (n = 44). (b) The PI fluorescence value in two-cell-stage embryos was measured and plotted for control embryos (open bars) and irradiated-sperm embryos (solid bars). DNA contents in irradiated-sperm early-cleaving embryos (cleaved before 19 h [n = 14]) were less than those in control early-cleaving embryos (n = 14). In contrast, irradiated-sperm late-cleaving embryos (cleaved at 21 h [n = 10]) had the same DNA content as control late-cleaving embryos (n = 14). Error bars, standard deviations.

DNA content in the nucleus was monitored in irradiated-sperm zygotes from 16 to 23 h after mating (data not shown). At 20 h, the PI values were low in some of the irradiated-sperm zygotes, but others showed values similar to those of the control. This suggested that some irradiated-sperm zygotes, especially those with late cleavage, may have managed to synthesize a normal amount of DNA during the delay. We have confirmed this possibility by measuring the PI values of two-cell-stage embryos cleaved by 19 h (early cleavage) or those cleaved at 21 h (late cleavage) after mating. PI values of irradiated-sperm two-cell-stage embryos of early cleavage were lower than normal, but they were in the normal range for those of late cleavage (Fig. 7b). This suggested that late-cleaving zygotes were those which spent more time in S phase to catch up to the normal DNA content even under the lower rate of DNA synthesis. The PI values of the two-cell-stage irradiated-sperm embryos with the p53^−/− genotype had a normal DNA content (data not shown).

DISCUSSION

p53-dependent cross talk between male and female pronuclei in irradiated-sperm zygotes.

The present study has shown that irradiated sperm activated a p53-responsive reporter microinjected into the female pronucleus of the zygotes. Although DNA is packaged tightly by protamine in the sperm head (4), this does not exempt DNA from being damaged by ionizing radiation (2). The number of DSB induced per unit dose of X rays was calculated to be around 30 in somatic cells (27), and a similar number might be expected in the sperm head. High-dose X-irradiation and KpnI digestion of the reporter also resulted in activation. These indicate that DSB might be responsible for induction of the p53-dependent pronuclear cross talk observed in this study.

The p53-dependent pronuclear cross talk occurs across the cytoplasm. This is in contrast to the nuclear limited activation of the RAD9-dependent damage checkpoint in Saccharomyces cerevisiae (9). Activation of p53 by radiation involves phosphorylation at the serine 15 and serine 20 residues (31). Phosphorylation of the serine 15 site is carried out by ATM, while that of the serine 20 site is carried out by Chk2 (5, 7, 20). It is easily envisaged that DNA damage is detected during decondensation of the sperm head and activates these protein kinases. Phosphorylated p53 would then be distributed in the male pronucleus in cis and in the female pronucleus in trans during pronuclear formation.

It is interesting that the activation was also observed in both p53^+/− and p53^−/+ irradiated-sperm zygotes, and the latter indicates that the p53 gene product is made from a paternally inherited allele. One-cell-stage transcription of the p53 gene is consistent with the previous report (21).

Lack of G₁/S arrest in irradiated-sperm zygotes.

One of the major functions of the p53-dependent damage response is to arrest the cells at the G₁/S border. In our experiment, however, the onset of DNA synthesis was not affected by 6-Gy X-irradiation of sperm, even though activation of p53 was evident by the reporter experiment. p53 is known to transactivate p21, which inhibits cyclin D-CDK4/6 complexes to arrest the cells at the G₁/S border (31). It was reported that embryonal stem cells and embryonal carcinoma cells lack the G₁/S checkpoint even though p53 is highly activated after X-irradiation (37, 40). In a recent study on γ-irradiated F9 embryonal carcinoma cells, p53 was shown to activate transcription of the p21 gene, but a rapid degradation blocked accumulation of p21 protein, leading to elimination of the G₁/S checkpoint (29). A similar mechanism may also be responsible for the lack of G₁/S checkpoint in mouse zygotes.

p53 dependence of S-phase damage checkpoint in mouse zygotes.

Suppression of DNA synthesis is attained by depletion of the pyrimidine nucleotide pool by treatment of the cells with N-(phosphonacetyl)-l-aspartate, and this suppression was reported to be p53 dependent (1). In this study we observed suppression of DNA synthesis in both pronuclei of irradiated-sperm mouse zygotes. The suppression was dependent on the presence of functional p53. Microinjection of the GST-p53 fusion protein restored the suppression in irradiated-sperm p53^−/− zygotes. The same treatment did not affect DNA synthesis in control zygotes, showing that the fusion protein by itself had no ability to suppress DNA synthesis in the absence of radiation damage. Thus, the present report is the first to implicate p53 in the S-phase DNA damage checkpoint of mouse zygotes.

It is well known that p21 is transactivated by p53 and arrests somatic cells at the G₁/S and G₂/M border (6, 10, 11). Lack of G₁/S arrest in zygotes may be due to the failure of accumulation of p21, as discussed in the previous section, or p21 may have a different function in mouse zygotes. In our study, suppression of DNA synthesis was observed in p21^−/− irradiated-sperm zygotes, indicating that p21 is not involved in the S-phase damage checkpoint. Indeed, S-phase block by hydroxyurea results in accumulation of p53 but not accumulation of p21 (15). Detailed analysis of a series of mutant p53 proteins is in progress to elucidate the domain required for the S-phase damage checkpoint.

The S-phase damage checkpoint operates by two mechanisms: suppression of replication origin firing and suppression of replication fork movement (35). Many of the works on the S-phase DNA damage checkpoint come from the study of S. cerevisiae (34). Suppression of origin firing in hydroxyurea-treated budding yeast was shown to be dependent on Mec1 and Rad53 (36). Involvement of Rad53 in the S-phase damage checkpoint was also reported (38). In higher eukaryotes, the S-phase checkpoint and suppression of replication origin are dependent on the gene product ATM, a mammalian homologue of Mec1, and this has explained the long-known radioresistant DNA synthesis in AT cells (24). The p53 gene product was reported to have no relevance in the ATM-dependent S-phase damage response in many of the cell types studied (25, 43).

Sperm irradiation resulted in suppression of DNA synthesis throughout S phase. This suggested that replication fork movement was likely to be the target of p53-dependent suppression of DNA replication in irradiated-sperm zygotes. Replication machinery of eukaryotes is composed of DNA polymerase A and primase. Replication protein A (RPA) is also required. Mutation of any of these genes results in a defect of the S-phase damage checkpoint (13). DNA polymerase, primase, and RPA form foci in the nucleus during S phase. p53 is known to bind to RPA and suppress its ability to associate with single-stranded DNA (12). It is possible that p53 might suppress replication fork movement by binding to RPA. Indeed, colocalization of RPA and p53 was found in the S-phase nucleus of Xenopus laevis embryos after UV and γ-irradiation (41).

It has long been assumed that replication fork movement is retarded in irradiated cells because the progression of the replication machinery is hindered at the site of DNA damage. Our present data demonstrated that DNA synthesis was suppressed not only in the damaged male pronucleus but also in the female pronucleus which had no radiation damage. Therefore, suppression of DNA synthesis in mouse zygotes is truly a checkpoint rather than a mechanical block of replication machinery at the damaged site.

The biological function of p53-dependent S-phase damage checkpoint.

In vivo fertilization gives better-synchronized zygotes than in vitro fertilization, and this is the reason for the shorter S phase in the former zygotes (Table 1). Delay in S phase can be better demonstrated for in vivo-fertilized zygotes, and about half of the in vivo-fertilized irradiated-sperm zygotes exhibited retardation of S phase by 2 h (Table 1). These delayed zygotes managed to synthesize the normal amount of DNA, while those without delay had less than the normal DNA content (Fig. 7b). Nevertheless, all zygotes went through M phase (Fig. 7a).

Mouse zygotes irradiated prior to pronuclear formation were reported to exhibit no G₂/M arrest, while irradiation at later stages led to cleavage block (16). This arrest has recently been shown to correlate with the decreased activity of p34^cdc2 (3). Irradiation of sperm is equivalent to irradiation of zygotes prior to pronuclear formation. DNA damage at the pre-pronuclear stage, either from sperm irradiation or from direct irradiation of zygotes, led to activation of the p53-dependent S-phase damage checkpoint in mouse zygotes.

We propose that during the p53-dependent suppression of DNA synthesis, DNA damage is modified in such a way that it is no longer capable of activating the G₂/M checkpoint. In contrast to somatic cells, the irradiated-sperm zygotes which fail to synthesis full DNA content still can proceed to G₂ and M (Fig. 7b). p53^−/− zygotes had no S-phase damage checkpoint, and they also did not show delay in M-phase entry, which is consistent with the requirement of p53 in the G₂/M checkpoint (8).

The biological function of the p53-dependent S-phase checkpoint is likely to be one of protecting early-stage mouse development. Our unpublished observation demonstrated that all zygotes produced from 6-Gy-irradiated sperm progressed at least to the eight-cell stage, and half of the fetuses were born normally. This suggests the lack of p53-dependent apoptosis in pre-implantation stage embryos. In contrast, 6-Gy-irradiated-sperm p53^−/− zygotes lacking S-phase suppression had a higher frequency of micronucleus occurrence during subsequent cleavages and did not implant. Thus, the p53-dependent S-phase damage checkpoint in mouse zygotes is yet another function of p53 as a guardian of the mammalian genome.