Breeding scheme to obtain mice with Orc4 deleted in oogenesis

We obtained frozen sperm from the C57BL/6NTac-Orc4 <tm1a(EUCOMM)Wtsi>/IcsOrl mouse strain from the European Mouse Mutant Archive (EMMA). This strain has a gene trap on exon 7 of the Orc4 gene. These sperms were injected into C57BL/6J mice, then bred to flippase mice to release the gene trap, then bred to homogeneity until a colony of Orc4^Flox/Flox mice were obtained. Deletion of both copies of the Orc4 gene is embryonically lethal (unpublished data from gene trap results of EMMA at  https://www.mousephenotype.org/data/genes/MGI:1347043), presumably because ORC4 is required for DNA synthesis ( Supplementary Figure 2 (supplementary_figures_ioab237.pdf)). Therefore, we used a well-established breeding scheme that allows for the production of homozygous Orc4^Flox/Flox mice that have at least one copy of the Zpe3-Cre gene to allow Cre to inactivate the Orc4 gene during oogenesis [13, 19]. To verify that the promoter to Zp3 activated Cre only in oocytes and as early as the primary follicle stage of oogenesis as has been reported [20], we first crossed a female mouse containing a tdTomato red fluorescent protein (RFP) reporter gene with a male ZP3-Cre⁺/⁺ mouse. Oocytes in primordial follicles from the female progeny did not express RFP but most oocytes in primary and secondary follicles did express RFP ( Supplementary Figure 3A–D (supplementary_figures_ioab237.pdf)). The surrounding somatic cells had no fluorescence. Germinal vesicle (GV) oocytes from adult females also showed tomato RFP expression only in the oocytes but not in the surrounding cumulus cells ( Supplementary Figure 3E and F (supplementary_figures_ioab237.pdf)). These data indicated that ZP3-Cre expression was limited to the oocytes and not the surrounding cumulus cells as expected.

We next bred female Orc4^Flox/Flox mice with male ZP3-Cre⁺/⁺ to create Orc4^Flox/wt; ZP3-Cre^+/- mice [13, 19]. In the final step, male Orc4^Flox/wt; ZP3-Cre^+/- mice were bred with female Orc4^Flox/Flox mice to create four different genotypes, only one of which, Orc4^FloxFlox; ZP3-Cre^+/- (referred to as Orc4-CKO), is a conditional knockout of Orc4 because it activates Cre during oogenesis to inactivate the full-length Orc4 gene by excising exon 7. Female Orc4-CKO mice were infertile as expected (explained next), so this breeding scheme had to be repeated for every experiment.

Preparation of GV oocytes, gametes, and embryo culture

Spermatozoa were obtained from B6 or transgenic male mice (8–10 weeks old). A caudal epididymis was dissected, sperm mass released into CZB [21] and incubated for 1 h at 37°C under 5% CO₂ in air to allow the spermatozoa to disperse in the medium. The spermatozoa suspension for intracytoplasmic sperm injection (ICSI) was mixed with polyvinylpyrrolidone (PVP)-saline (0.9% NaCl containing 10% w/v PVP [360 kDa; ICN Biochemicals]). To obtain mature MII oocytes, mature females, 8–10 weeks old, were induced to superovulate by injections of 5 IU eCG and 5 IU hCG, given 48 h apart. Oviducts were removed 14–15 h after the injections of hCG and placed in light mineral oil for embryo culture (Irvine, catalog number 9305). The cumulus cells were released from the oviducts into 0.1% bovine testicular hyaluronidase/CZB medium for 10 min to disperse cumulus cells. The cumulus-free oocytes were washed and kept in CZB at 37°C under 5% CO₂ in air. These oocytes were used for immunocytochemistry (ICC) and ICSI. To collect immature metaphase I (MI) and anaphase I (AI) oocytes, females were injected with 5 IU of eCG. Ovaries were removed 24 h thereafter and placed in HEPES-CZB in a Petri dish. The biggest follicles were broken to release GV-stage oocytes. GV oocytes with surrounding cumulus cells were placed in CZB drops under mineral oil and cultured for 2 h at 37°C, 5.0% CO₂. After 2 h, cumulus cells were removed by pipetting. The oocytes that underwent germinal vesicle breakdown were collected and cultured for further 3, 9, and 12 h to obtain MI, AI, and MII stages, respectively. Some of these oocytes were then used for ICC.

Parthenogenetic activation of MII oocytes

MII oocytes were incubated with CZB-free, Ca⁺²-supplemented 10 mM SrCl₂ for 2 h. The oocytes were incubated at 37°C in a humidified atmosphere of 5% CO₂ in air and cultured for up to 2 days to follow their progression to the stages up to the two-cell stage.

Intracytoplasmic sperm injection

ICSI was carried out as described by Ward and Yanagimachi [22]. ICSI was performed using Eppendorf micromanipulators (Micromanipulator TransferMan, Eppendorf, Germany) with a piezoelectric actuator (PMM Controller, model PMASCT150; Prime Tech, Tsukuba, Japan). A single sperm head was sucked, tail first, into the injection pipette and moved back and forth until the head-midpiece junction (the neck) was at the opening of the injection pipette. The head was separated from the midpiece by applying one or more piezo pulses. An oocyte was held to the holding pipette at the 9 o'clock position. After discarding the midpiece and tail, the sperm head was redrawn into the pipette and injected immediately into an oocyte. After ICSI, oocytes were cultured in CZB at 37°C under 5% CO₂ in air.

Isolation of primordial, primary, and secondary follicles

To collect these oocytes, 7-day-old pups from mating ZP3-Cre⁺/⁺ males with tdTomato females were used to collect ovaries. Isolated ovaries were placed in Accumax solution (catalog number A7089; Sigma Aldrich) in a Petri dish for 10 min. The follicles were manually isolated in HEPES-CZB using 30 G needles and categorized into primordial, primary, and secondary follicles according to their size and morphology [23]. These follicles were used to analyze the expression of RFP under fluorescent microscopy.

Antibodies

Polyclonal goat anti-ORC4 was obtained from Santa Cruz Biotechnology (C-15, catalog number sc-19,726; Santa Cruz Biotechnology, Santa Cruz, CA). The secondary antibodies included Alexa Fluor 488 rabbit antigoat (Invitrogen, Grand Island, NY). Click-it EdU Alexa Fluor 488 kit (catalog number C10350, Invitrogen) was used to detect DNA replication.

Immunocytochemistry

Oocytes and embryos were cultured in CZB until they reached the desired stage after ICSI or culture. The oocytes or embryos were fixed in 2% paraformaldehyde for 30 min at room temperature. After fixing, embryos were washed three times with 0.1% Tween/PBS (PBSw) for 10 min. Cells were permeabilized in 0.5% Triton X-100 for 15 min, then the cells were washed three times with PBSw containing 10% volume of 5% bovine serum albumin (BSA). Cells were blocked with 5% BSA for 1 h at room temperature, then incubated in primary antibody at 1:50 dilution overnight at 4°C. On the next day, the cells were again washed three times with PBSw containing 10% volume of 5% BSA, then incubated in secondary antibody at 1:1000 dilution at room temperature for 1 h. Cells were then washed three times with PBSw containing 10% volume of 5% BSA and finally the cells were mounted with ProLong Gold antifade reagent with DAPI (catalog number P-36931; Invitrogen). The staining oocytes or embryos were analyzed with an FV1000-IX81 confocal microscope from Olympus using the Fluoview v. 2.1 software. For each stage of oocyte or embryonic development, at least 20 embryos were examined by ICC, and the results were only reported if they were consistent in all embryos.

Aphidicolin treatment

MII oocytes collected from Orc4-CKO, and wild-type females were activated with 10 mM of SrCl₂ for 3 h. The activated oocytes were then incubated with Aphidicolin (3 µL of 1 mg Aphidicolin/mL) overnight at 37°C in a humidified atmosphere of 5% CO₂ in air to inhibit DNA synthesis. These oocytes were then used for DNA synthesis assays (catalog number C10350; Invitrogen) and then mounted with ProLong Gold antifade reagent with DAPI (catalog number P-36931; Invitrogen). DNA synthesis was analyzed with an FV1000-IX81 confocal microscope from Olympus using the Fluoview v. 2.1 software.

Quantitative reverse transcriptase-polymerase chain reaction

Total RNA from 100 Orc4-CKO MII oocytes was extracted and purified using the TRizol and Trace kit (Thermo Fisher Scientific, catalog number A33250). cDNA was then synthesized from this total RNA by reverse transcription of polyadenylated RNA using superscript reverse transcriptase IV following the manufacturer's protocols (Thermo Fisher Scientific, catalog number 18090050). Then, reverse transcriptase-polymerase chain reaction (RT-PCR) was performed using SYBR Green PCR Master Mix on an ABI Step-OnePlus machine (Applied Biosystems, Carlsbad, CA). RT-PCR reactions were performed at 95°C for 10 min followed by 35 PCR cycles (10 s at 95°C and 60 s at 60°C). Three types of PCR reactions were performed: (1) Primers designed at exons 7 and 9 were used to amplify ORC4 variant 1 (v1-Orc4) from wild-type and Orc4^Flox/Flox oocytes; (2) primers designed at exons 6 and 8 were used to amplify v1-Orc4 mRNA from wild-type, Orc4^Flox/Flox, and Orc4CKO oocytes; and (3) primers designed at exons 6 and 7* (7* is a variant exon) were sued to amplify ORC4 variant 2 (v2-Orc4) ( Supplementary Table 1 (supplementary_tables_ioab237.pdf)). All of the reactions were performed in triplicate per assay, and β-actin was included in every PCR reaction as a loading control. The different values in PCR cycles for β-actin for a given experimental sample were subtracted from the mean ΔCt of the reference samples (Orc4^Flox/Flox) (ΔΔCt) [24, 25]. The quantification of the ORC4 knockdown values were further normalized to ΔΔCt values of β-actin. Primers are listed in  Supplementary Table 1 (supplementary_tables_ioab237.pdf).

Single oocyte PCR

A single Orc4-CKO MII oocyte or an Orc4^Flox/Flox oocyte was added to a PCR tube containing 4 µL of GNTK buffer (50 mM KCl, 1.5 mM MgCl₂, 10 mM Tris, pH 8.5, 0.45% Triton X100, 0.45% Tween 20) supplemented with 100 µg/mL proteinase K. These tubes were incubated at 55°C for 2.5 h with a Bio-RAD C100 Touch thermo cycler to extract total DNA. These DNA tubes were directly used to perform PCR using a KOD hot start DNA polymerase kit (Millipore/Sigma, catalog number 71086). The primers used were CKO-F and CKO-R ( Supplementary Table 1 (supplementary_tables_ioab237.pdf)). The PCR amplification reactions were performed using a PCR Bio-RAD C100 Touch thermocycler. The PCR cycle protocol consisted of 2 min hold at 95°C, followed by 35 cycles of 10 s at 98°C, 30 s at 60°C, and 2 min at 72°C, then with a final step for 5 min at 72°C. The PCR products were used to run with 1% agarose supplement with ethyl bromide. The gel was imaged using a Fujifilm LAS-3000 imager.

Image presentation

All micrograph images were constructed by placing the raw images on a single figure. Immunocytochemical images were all then enhanced together so that relative comparisons can be seen. Phase-contrast images were enhanced separate together but separately from the fluorescent images.

Statistical analysis

For statistical analysis comparing the development of Orc4CKO and Orc4^Flox/Flox GV oocytes to MII oocytes, percentages were transformed to arcsine values ( Supplementary Table 3 (supplementary_tables_ioab237.pdf)), then compared by the Student t-test [26, 27].

Deletion of Orc4 in Orc4-CKO mice occurs by the MII stage

Inactivating a floxed gene using Zp3-Cre has been reported to be capable of being 100% effective [15]. We tested the efficiency of the Zp3-Cre in our hands by single-oocyte PCR. We used a set of primers for two loci just outside the loxP sites flanking exon 7 that are present in all genotypes (CKO-F and CKO-R,  Supplementary Figure 2 (supplementary_figures_ioab237.pdf)). We performed single oocyte PCR on 27 MII oocytes from 4 Orc4^Flox/Flox mice, which have exon 7, and 92 oocytes from 6 Orc4-CKO mice, in which we expected exon 7 to be removed by Zp3-Cre. We did not obtain PCR products from all oocytes, but every PCR product was the expected size ( Supplementary Figure 4 (supplementary_figures_ioab237.pdf),  Supplementary Table 2 (supplementary_tables_ioab237.pdf)). We examined 92 Orc4-CKO oocytes and 51 had PCR products that were consistent with exon 7 being excised, as expected. The other 41 had no PCR products, indicating that our technique was not sensitive enough to amplify DNA from every oocyte. However, none of the positive PCR products contained exon 7. Conversely, 19 of the 27 Orc4^Flox/Flox oocytes that had PCR products were all positive for exon 7 as expected. None of Orc4CKO oocytes retained exon 7, and all of the Orc4^Flox/Flox oocytes were positive for exon 7. Fewer Orc4-CKO oocytes had PCR products than Orc4^Flox/Flox oocytes (55.4% vs. 70.4%, respectively), but this was not surprising given that the Orc4-CKO MII oocytes were more poorly developed and may have had degenerated DNA. We conclude that Zp3-Cre was remarkably effective at removing Orc4 exon 7 during oogenesis.

Depletion of Orc4 transcripts in Orc4-CKO mice

The PCR data described above demonstrated that the exon 7 of the Orc4 genes were undetectable in MII Orc4-CKO oocytes. This deletion causes a frame-shift in the remainder of the Orc4 gene resulting in an inactive Orc4 transcript. However, it is possible that some complete Orc4 mRNA could have been transcribed before the gene was inactivated during oogenesis and stored for use in the MII stage or later. We therefore performed RT-PCR on MII oocytes from Orc4^Flox/Flox and Orc4-CKO mice to determine the levels of detectable transcripts. The major Orc4 mRNA variant (v1-Orc4, NCBI reference sequence: NM_011956.3) contains all 14 exons and is translated to the full-length ORC4 protein ( Supplementary Figure 5 (supplementary_figures_ioab237.pdf)). There is another splice variant of Orc4 known as variant 2 (v2-Orc4) that contains the first 6 exons of Orc4 and a different exon 7 (NCBI reference sequence: NM_001177213.1) [28]. We expected that Orc4CKO oocytes would not produce v1-Orc4 transcripts but might produce a transcript that did not contain exon 7 and exons 8–14 in a frameshift ( Supplementary Figure 5 (supplementary_figures_ioab237.pdf)). We also expected that Orc4-CKO oocytes could still produce v2-Orc4 ( Supplementary Figure 5 (supplementary_figures_ioab237.pdf)). We therefore tested for the presence of mRNAs using forward and reverse primers located in exons 6 and 8, respectively, which would amplify both wild-type v1-Orc4 and Orc4-CKO mRNAs, and using forward and reverse primers located in exons 7 and 9, respectively, which would only amplify wild-type v1-Orc4 transcript. In Orc4-CKO oocytes, we observed the expression of transcripts amplified with both primer pairs, but transcripts amplified with the primers located in exons 7 and 9 were at very low levels (0.15-fold) whereas transcripts amplified with the primers located in exons 6 and 8 were approximately half lower (0.58-fold) when compared to the control ( Supplementary Figure 6 (supplementary_figures_ioab237.pdf)). Interestingly, the v2-Orc4 transcript increased by 5.21 + 3.25-fold in Orc4-CKO mice as compared to the controls. These data suggested that the production of normal, full-length Orc mRNA was severely reduced in Orc4-CKO mice but increased in v2-Orc4.

Table 1.

Female Orc4-CKO mice are infertile, but males are fertile. Female Orc4-CKO mice were mated with normal B6 males (wild-type), and no pups were obtained. Male Orc4^{Flox/Flox;Cre/}- mice were able to sire pups when mated with normal females.

Fertility of Orc4-CKO mice

When Orc4-CKO mice were mated with wild-type mice, female Orc4-CKO mice were completely infertile. No pups were born from four mating attempts (Table 1). Male Orc4^{Flox/Flox:Cre}+^/- mice were fertile when mated with wild-type females.

Orc4-CKO GV oocytes have reduced polar body extrusion

We isolated 221 GV oocytes from 10 adult Orc4-CKO mice and compared their progression in vitro with 84 oocytes from 3 Orc4^Flox/Flox control mice. These mice had the Orc4^Flox/Flox genotype ( Supplementary Figure 2B (supplementary_figures_ioab237.pdf)) in all cells, except for the oogenic cells, which had the Orc4^-/phenotype ( Supplementary Figure 2C (supplementary_figures_ioab237.pdf)). GV oocytes have 4C DNA, before the first polar body extrusion releases half the DNA. Normal GV oocytes progress to the MII oocyte stage within 12 h of culture, which includes the extrusion of the first polar body. However, Orc4-CKO GV oocytes did not progress normally in culture. We found that only 25% of Orc4-CKO GV oocytes progressed to the MII stage (Figures 1A–D and 2) whereas almost half (48%) were arrested at the GV, MI, or AI stages (Figures 1A–C and 2,  Supplementary Figures 7 and 8 (supplementary_figures_ioab237.pdf)). The remaining Orc4CKO GV oocytes progressed to aberrant stages including oocytes with two or more pronuclei, or oocytes with one pronucleus and one polar body (Figure 1E and F). Orc4-CKO oocytes had some ORC4 staining (Figure 1C and H), suggesting that some ORC4 that was already present before Zp3-Cre is activated might have remained in the oocyte during development. To test this, we stained primary follicles, the stage at which Zp3 is first activated, for cytoplasmic ORC4 and found that it was present (Figure 3). Nevertheless, only little ORC4 was retained and our immunofluorescence data demonstrated that deletion of exon 7 of the Orc4 gene caused severe disruption in normal PBE.

Figure 1.

Development of Orc4-CKO GV oocytes in vitro. GV oocytes from Orc4-CKO females were isolated and allowed to mature in vitro for 24 h. This resulted in some normal and abnormal developmental stages which were stained for ORC4 using ICC (green) and counterstained with DAPI for DNA visualization (blue). (A–C) Oocytes that were arrested at normal stages of maturation in vitro, but did not progress. (A) An oocyte that was arrested at the GV stage. Some ORC4 staining is present just below the oolemma. (B) An oocyte that was arrested at MI. (C) An oocyte that was arrested at anaphase I, with a strong ORC4 staining pattern. (D) Some oocytes progressed to the MII stage, which was the normal point of progression. (E–H) Abnormal oocytes that represent aberrant stages of the first meiotic division. (E) An oocyte in which the second metaphase plate failed to develop and resolved into a pronucleus (larger nucleus) and a polar body (G1-PB). (F) An oocyte where the metaphase I plate resolved into two pronuclei with no polar body (2-PN). (G) An oocyte in which the metaphase plate resolved into multiple pronuclei with no polar body (MPN). (H) An oocyte in which the first meiotic division resulted in two cells (2-Cell) (Note: One nucleus was in a different focal plane from the rest of the image and is shown in a white box, inset). All images are shown at the same magnification (bar = 20 µm). For phase-contrast and single-stained images, see  Supplementary Figure 7 (supplementary_figures_ioab237.pdf).

Figure 2.

Maturation of Orc4-CKO GV oocytes in vitro. A total of 221 GV oocytes were isolated from 10 Orc4-CKO females and 84 GV oocytes from Orc4^Flox/Flox controls and cultured for 24 h. The developmental stage that each GV oocyte reached after 24 h was recorded by immunocytochemical analysis. None of the aberrant stages was noted in the controls. All values are means of oocytes from different mice, and error bars are standard deviations. The difference in the development of Orc4-CKO and Orc4^Flox/Flox GV oocytes against MII oocytes was highly significant (P = 0.0001, noted with an a above the bars). The numbers of oocytes in each category for each mouse are shown in  Supplementary Table 4 (supplementary_tables_ioab237.pdf). (MI, metaphase I; MII, metaphase II; GV, germinal vesicle; AI, anaphase I; G1-PB, an oocyte with one pronucleus and one polar body; 2PN, an oocytes with two pronuclei; MPN, an oocytes with more than two pronuclei; 2C, two-cell embryo).

Figure 3.

Primary follicles contain cytoplasmic ORC4. Primary follicles were isolated from 7-day-old pups and stained for the presence of ORC4. (A–C) DAPI, ORC4 stained, and merged fluorescent images, respectively, of a primary follicle stained with ORC4 antibody. (D–F) As in (A–C) except that no primary antibody was included. Bar = 10 µm.

Orc4-CKO MII oocytes progress to the two-cell stage without DNA synthesis

When Orc4-CKO females were stimulated with eCG and hCG to superovulate, they produced fewer MII oocytes than controls (Table 2). Most of the Orc4-CKO MII oocytes that were stimulated with SrCl₂ to activate parthenogenesis progressed to the two-cell stage, but none of these two-cell embryos developed any further (Table 3). These embryos had extremely condensed nuclei that appeared to have degraded DNA (Figure 4A and B; for an example of a normal parthenogenetic embryo's two-cell nuclei, see Figure 5A). Examining 58 oocytes from three different Orc4-CKO mice, we found that an average of 72.4% had either unequal distribution of DNA or no DNA in one or both cells ( Supplementary Table 5 (supplementary_tables_ioab237.pdf)). In some cases of these aberrant embryos, there was no DNA in one of the two cells (Figure 4C and D) or no DNA detectable in either embryonic cell, even when DNA was visible in the residual polar body (Figure 4E and F). This suggests that parthenogenetically activated Orc4-CKO two-cell embryos had degraded DNA.

Table 2.

Average number of MII oocytes from control and Orc4-CKO mice. The number of MII oocytes that were obtained from control (Orc4^Flox/Flox) and Orc4-CKO mice were counted. The differences between the total number of oocytes per mouse between the Orc4^Flox/Flox and Orc4-CKO was statistically significant (P = 0.0001, Student t-test).

The progression to the two-cell stage in Orc4-CKO oocytes was inconsistent with the second of the two ORC4 functions we studied, its role in DNA synthesis. ORC4 is thought to be required for DNA synthesis but progression from G1 through mitosis usually requires DNA replication. We therefore tested whether the Orc4-CKO MII oocytes that progressed to the two-cell stage did so without replicating their DNA, or if, as has been shown in some cancer cells [29], they could replicate DNA in the absence of the ORC proteins. Control parthenogenetic zygotes incubated for 24 h in EdU had robust evidence of DNA replication in both nuclei at the two-cell stage (Figure 5A–C). However, Orc4-CKO two-cell embryos did not have evidence for any EdU incorporation during this time period even in those embryos that retained visible levels of DNA (Figure 5G–I). The progression of Orc4-CKO embryos to the two-cell stage in the absence of DNA synthesis was in contrast to previous work in which zygotes that were cultured in the presence of the DNA polymerase inhibitor aphidicolin were arrested at the one-cell stage [30]. We repeated this previously published experiment as a control for this study, illustrating that cytokinesis of normal mouse zygotes is inhibited by the inhibition of DNA synthesis with aphidicolin (Figure 5D–F). We next tested the remote possibility that aphidicolin might also arrest cytokinesis in parthenogenetically activated Orc4-CKO MII oocytes as a secondary effect unrelated to its direct inhibition of DNA polymerase. We activated Orc4-CKO MII oocytes in the presence of aphidicolin and assessed them after 24 h. We found that parthenogenetically activated Orc4-CKO MII oocytes still progressed to the two-cell stage in the presence of aphidicolin (Figure 4J–L). These data suggest that inhibition of DNA synthesis by ORC4 depletion occurs by a different mechanism than that of aphidicolin treatment. Note that in aphidicolin-treated Orc4-CKO two-cell embryos, the nuclei are not condensed and the DNA complement appears to be normal (compare Figure 5G with Figure 5J). These results have implications for cell cycle control in the zygote as discussed next.

ICSI does not rescue Orc4-depleted oocytes

There was an unlikely possibility that the paternal genome could rescue the first round of DNA synthesis by either supplying ORC4 itself, or by supplying a functional copy of the Orc4 gene that could be transcribed in the zygote [31, 32]. We injected Orc4-CKO MII oocytes with wild-type sperm and incubated the zygotes with EdU for 8 h to detect DNA synthesis. Orc4^Flox/Flox showed robust DNA synthesis (Figure 6A–C) whereas Orc4-CKO zygotes had no evidence of replication (Figure 6D–F). These data suggest that the zygote depends on ORC4 supplied by the oocyte for the first round of DNA synthesis and that a normal Orc4 gene supplied by the sperm cell cannot rescue zygotic DNA synthesis in the absence of ORC4. The data also demonstrated that the infertility of the female Orc4-CKO mice was not due to the inability to activate when fertilized or to decondense the sperm nucleus because these zygotes formed normal pronuclei (Figure 6D).

Figure 4.

Distribution of DNA in two-cell embryos from Orc4-CKO parthenogenesis. Orc4-CKO MII oocytes were collected and activated with SrCl₂ and imaged after 24 h. This figure shows three examples with DAPI staining to reveal the DNA (A, C, and E) and the corresponding phase-contrast images (B, D, and F, respectively). (A, B) An embryo that had roughly equal distribution of DNA in both blastomeres even though the DNA appears degraded or dispersed. (C, D) An embryo which had visible DNA in one cell but none in the other. Note that the DNA in the polar body is visible. (E, F) An embryo in which there was no DNA visible in either blastomere. Normal parthenogenotes should have clearly visible nuclei in both blastomeres (Figure 5A). Bar = 20 µm.

Fertility of Orc4-CKO mice

Orc4-CKO female mice were infertile. This was expected because the original–gene-trapped mice, Orc4^{tm1a(EUCOMM)Wtsi}, ( Supplementary Figure 2A (supplementary_figures_ioab237.pdf)) were embryonically lethal when the deletion was homozygous. In our Orc4-CKO mice ( Supplementary Figure 2B (supplementary_figures_ioab237.pdf)), exon 7 was removed only in oocytes, after the last oocyte DNA synthesis, by ZP3-Cre ( Supplementary Figure 2C (supplementary_figures_ioab237.pdf)). Thus, we expected normal development of both male and female first-generation Orc4-CKO because the ZP3-Cre would not be activated in zygotes created by mating female Orc4^Flox/Flox mice with male ZP3-Cre mice. Male Orc4-CKO mice were fertile, as expected, because ZP3-Cre was never activated. Female Orc4-CKO mice were expected to be infertile because DNA synthesis was inhibited in the fertilized zygote due to the lack of ORC4. But even if DNA synthesis was restored, it is also probable that the impaired PBE due to the reduction in ORC4 would have also reduced fertility.

ORC4 and polar body extrusion

One major goal of this work was to test whether deletion of Orc4 during oogenesis would inhibit PBE as we have predicted from previous experiments [4, 5]. We found that when Orc4 was deleted during oogenesis, normal PBE was reduced by 75% in GV oocytes cultured in vitro (Figure 2). Although 25% of GV oocytes progressed to normal MII oocytes in vitro, there was only a 54% reduction in the number of MII oocytes obtained from Orc4-CKO ovaries (Table 2). This difference in the number of MII oocytes actually produced in Orc4-CKO females and the progression of GV oocytes in culture may be related to the stresses of in vitro culture that made the GV oocytes more susceptible to the partial loss of ORC4 available for PBE [33, 34]. Thus, these data support our hypothesis that ORC4 is required for normal PBE. There are three nonmutually exclusive possible explanations for the fact that Orc4 deletion did not fully inhibit PBE. First, although we demonstrated that the Orc4 gene was successfully targeted by the MII stage ( Supplementary Figure 4 (supplementary_figures_ioab237.pdf)), it is possible that intact Orc4 mRNA was still being transcribed through the GV stage. This would have allowed the oocyte to produce full-length cytoplasmic ORC4 after the primary follicle stage, and this protein could have contributed to PBE. Second, it is possible that enough cytoplasmic ORC4 had already been produced by the follicle stage and the protein retained during oogenesis to allow for some PBE. This is supported by continued visible staining of ORC4 in GV oocytes. Third, we have previously shown that the v2-ORC4 variant of ORC4 protein is present in the ORC4 cage along with v1-ORC4 during PBE [4] and it is possible that contribution of v2-ORC4 was enough to allow for the inefficient PBE we observed. Our RT-PCR data show that the level v2-Orc4 mRNA is increased in Orc4-CKO oocytes suggesting that the oocyte might be compensating for the loss of full length v1-ORC4. If any of these possibilities is true, then it is unlikely that a transgenic mouse model that deletes Orc4 will allow for the complete inhibition of PBE. We are currently testing whether deletion of an earlier exon that would abrogate both variants of Orc4 will inhibit PBE more clearly.

ORC4 is required for the first zygotic DNA synthesis

The second function of ORC4 that we examined was its more established role in DNA synthesis. This function was completely inhibited in embryos resulting from oocytes from Orc4-CKO females. Our data clearly demonstrate that ORC4 is required for the first round of DNA synthesis in the mouse embryo. Although this was expected from the known function of the ORC in the licensing of DNA replication origins [7, 35, 36], recent data on some mammalian systems have demonstrated that it is possible for cells to replicate DNA in the absence of core components of this protein complex. Shibata et al. [29] demonstrated that certain human cancer cell lines can replicate DNA when ORC1 and ORC2 are deleted. This suggests that mammalian cells do have alternate mechanisms for replicating DNA. Okano-Uchida et al. [37] created a mouse that had a floxed ORC1 gene and crossed this mouse with Sox2-Cre that deactivated the gene during early development. These authors showed that ORC1 was required for embryonic development, but in this case zygotic embryos replicated efficiently, most likely due to the fact that SOX2 becomes activated after the two-cell stage. These Orc1^-/- embryos developed to the blastocyst stage but not much further. Trophoblast lineages, however, continued to replicate DNA. When these mice were crossed with Albumin-Cre, adult livers retained the ability to regenerate, replicating by endoreduplication. Together, these studies demonstrated that mammalian cells do not always require the intact ORC for DNA replication. The ZP3-Cre model to inactivate Orc4 allowed demonstrating that the intact ORC is required in the first zygotic DNA synthesis.

ORC4 is part of the complex that recruits CDC6 to the licensing site [8], so the lack of ORC4 would be expected to inhibit licensing, and, thereby, the initiation of DNA synthesis. The progression of activated oocytes to the two-cell stage in the absence of ORC4 was not prevented by the inability of these embryos to license DNA for replication. This has two potential implications. The first is that mouse zygotes do not have a functional DNA licensing checkpoint. The p53-dependent checkpoint arrests cells in either S-phase or G1, depending on the cell type, until the chromatin is fully licensed [38–40]. The fact that Orc4-CKO zygotes were not arrested in the absence of DNA licensing suggests that zygotes do not have a licensing checkpoint. Secondly, the data support a “brake” model for the progression through the cell cycle. Recent studies have demonstrated that the transition from S-phase to M-phase is probably due to an inhibition of PLK1 during the S-phase with a gradual release of inhibition during G2 that culminates in mitosis (reviewed in [41]). Lemmens et al. [42] demonstrated that in mammalian cells, when licensing was inhibited by specifically degrading CDC6, cells progressed directly to mitosis without DNA synthesis (these authors used a cell line in which p53 was deleted to prevent activation of the licensing checkpoint so that they could analyze the S/G2/M transitions). CDK-controlled progression through the cell cycle is temporarily halted during S-phase and renewed when S-phase is completed [42]. The Orc4CKO zygotes were not arrested in G1, presumably because of the absence of the licensing checkpoint, then continued to progress through to mitosis because of the lack of the brake in S-phase that normally temporarily halts cell cycle progression. Normal oocytes that were treated with aphidicolin would have had normal licensing as aphidicolin directly inhibits DNA polymerase [43], which would not have disrupted DNA replication licensing. The brake for cell cycle progression is activated but never released because DNA synthesis is never completed. In the Orc4-CKO zygotes, aphidicolin did not arrest the cell cycle because DNA synthesis was never initiated. These results therefore support the interpretation that the lack of intact ORC4 prevented licensing in the oocytes, and this was the cause of the inhibition of DNA synthesis.