Maintenance of pluripotency in mouse ES cells without Trp53

Tumor suppressor Trp53 works as a guardian of the genome in somatic cells. In mouse embryonic stem (ES) cells, it was reported that Trp53 represses pluripotency-associated transcription factor Nanog to induce differentiation. However, since Trp53-null mice develop to term, Trp53 is dispensable for both the maintenance and differentiation of the pluripotent stem cell population in vivo, suggesting the differential functions of Trp53 in ES cells and embryos. To reveal the basis of this discrepancy, here we established a new line of Trp53-null ES cells by sequential gene targeting and evaluated their ability to differentiate in vitro and in vivo. We found that Trp53-null ES cells had defects in differentiation in vitro as reported previously, whereas they were able to contribute to normal development in chimeric embryos. These data indicated that the requirement of Trp53 for maintaining and executing the ES pluripotency is not absolute.

M aintenance of the genome integrity in cells is important for keeping homeostasis of multi-cellular organisms. Tumor suppressor Trp53 is one of the most important components to protect the genome from the oncogenic mutations. It controls cell-cycle arrest, apoptosis and stem cell differentiation by activating and repressing its downstream targets 1,2 . Trp53 mainly acts as a transcription factor to activate and repress the target gene expressions. It is expressed ubiquitously in somatic cells and normally its protein product Trp53 is in rapid turnover by active degradation mediated by the E3 ubiquitin ligase Mdm2 or Mdmx. Induction of the DNA damage induces inactivation of Mdm2 that results in accumulation of Trp53 and its nuclear localization. Nuclear localized Trp53 causes arrest of cell-cycle progression and apoptosis to eliminate the cells with damaged genome from the organisms 3 .
Mouse embryonic stem (ES) cells are pluripotent stem cells derived from the inner cell mass of the blastocyststagte embryos 4,5 . They continue self-renewal in the optimal culture condition in vitro, which commonly contain the cytokine leukemia inhibitory factor (LIF) as a repressor of differentiation 6 . Even after a prolonged culture, their pluripotency is maintained as confirmed by injection of these ES cells into blastocyst that give rise to chimeric embryos in which ES cell-derived cells contribute to all germ layers including germ cells 7 . It was reported that Trp53 functions in a unique mode in mouse ES cells 2 . Trp53 is expressed in mouse ES cells, localized in cytoplasm and degradated in a Mdm2/Mdmx-dependent manner as found in other somatic cell types 8,9 . Induction of differentiation activates Trp53, which represses the pluripotency-associated transcription factor Nanog, suggesting its function to drive differentiation program properly 10 . This process could be regulated by a Trp53 deacetylase Sirt1 by controlling Trp53 subcellular localization 11 as well as by the expression of a specific isoform of Trp53, delta40p53, in ES cells 12 . Recently, Aurora kinase A was identified as a repressor of Trp53 by phospholyrating it directly, which could also be one of the mechanisms to maintain self-renewal by repressing the differentiation program induced by Trp53 13 .
In contrast to these suggested functions of Trp53 in mouse ES cells, it was known that although Trp53 is activated by DNA damage in mouse ES cells, it has no ability to activate Trp53-mediated DNA damage response such as cell-cycle arrest, apoptosis or senescense as found in somatic cells 14 . This might be due to the unique cellcycle regulation in mouse ES cells lacking the check point in transition from G1 to S phase 15 . Moreover, Trp53null mice develop normally although they showed high incidence of tumor formation, indicating that the function of Trp53 is dispensable for self-renewal and differentiation of pluripotent stem cells transiently appeared in the developmental process 16 . Why does the requirement of Trp53 in differentiation of pluripotent stem cells look different between embryos and ES cells? The distinct role of the LIF signaling in ES cells and embryo has been well analyzed: ES cells require the activation of Stat3 by LIF for continuous self-renewal in serum-containing culture condition 17 while Stat3-null embryos keep pluripotent cell population in pre-and early post-implantation development 18 . Interestingly, Stat3-null ES cells are able to be established in a defined serum-free culture condition without LIF 19 , indicating the context-dependent requirement of Stat3 function in ES cells. How about in the case of Trp53? It may be absolutely required in ES cells to execute the differentiation program. Alternatively, the requirement of Trp53 may be contextdependent and thus dispensable for differentiation of ES cells in the context of embryonic development, i.e. the context in which chimeric embryos from Trp53-null ES cells are developed as Trp53-null embryos. Since the ability of Trp53-null ES cells to contribute chimeric embryos has never been assessed, here we established a new line of Trp53-null ES cells and tested their character in vitro and in vivo. We revealed that Trp53-null ES cells retain the ability to contribute to chimeric embryos although they showed abnormality in differentiation in vitro.

Results
Transient nuclear localization of Trp53 during differentiation in mouse ES cells. It was reported that Trp53 is localized in the cytoplasm and translocates into the nuclei by the induction of DNA damage 8 . However, when we stained mouse ES cells with anti-Trp53 mouse monoclonal antibody 1C12, we found a few cells possessing strong signal of nuclear-localized Trp53 in conventional culture condition, which were 14% of total cells, whereas others had weak signal of nuclear Trp53 (Fig. 1a, 1e). We obtained similar observation with other anti-Trp53 antibody (data not shown). These strong nuclear localized Trp53 population merged with the Oct3/4-positive/Rex1-negative population in OCRG9 ES cell line that carries Oct3/4-Ecfp fusion gene and Rex1-promoterdriven Egfp gene 20 . Since the Oct3/4-positive/Rex1-negative population represents the pluripotent stem cells in the late developmental stage that are ready for undergoing differentiation 21 , these data suggested that the nuclear localization of Trp53 was induced at the initiation of the differentiation event.
To confirm the regulation of Trp53 localization in differentiation process, we tested the localization of Trp53 in ES cells undergoing differentiation by withdrawal of LIF from the culture medium. The mesoderm marker T (also known as Brachyury) 22 was transcriptionally up-regulated by day 4 after the withdrawal of LIF (data not shown) and its nuclear staining was detectable from day 3 heterogeneously by immunostaining (Fig. 1b). In parallel, the pluripotency-associated transcription factor Oct3/4 was transcriptionally down-regulated after day 2 (data not shown). Trp53 started to accumulate in the nuclei on day 2 (Fig. 1d) and its nuclear localization reached to the maximal level on day 3 (Fig. 1b), which was 53% of total cells (Fig. 1e), although no obvious change was observed in its transcription level during this period (data not shown). Interestingly, Oct3/4 signal, which was retained only in few cells on day 3 after withdrawal of LIF, never merged with T during the differentiation, and Trp53 signal always merged with Oct3/4 but not with T ( Fig. 1b), suggesting that the nuclear Trp53 might mark the pluripotent stem cells that are ready to exit the pluripotency and enter into the differentiated state.
To evaluate the transcriptional activity of nuclear localized Trp53 during differentiation, we tested the expression of Pml in self-renewing mouse ES cells since it was reported that Pml is a direct target of Trp53 23 . Pml is a component of the macromolecular nuclear structure, PML body. As shown in Fig. 1c, large PML bodies were detected in the Oct3/4-positive/Rex1-negative population as found in the case of the nuclear Trp53 (Fig. 1a), suggesting that the nuclear Trp53 is active in these cells to direct the expression of the target genes. These data indicated that Trp53 is transiently localized in the nuclei in the pluripotent stem cell population during differentiation and is functionally regulating the expression of the target genes.
How about the relationship between Trp53 and Nanog? We tested nuclear localization of Trp53 and Nanog in ES cells carrying the Oct3/4-Ecfp fusion gene by co-immunostaining. As shown in Fig. 1d, Nanog was exclusively expressed in Trp53-negative cells in undifferentiated ES cells (1LIF; middle line), which was consistent with our previous observation that Nanog is predominantly expresseed in Rex1-positive population 20 and with the reciprocal expression of Rex1 and Trp53 shown above. When differentiation was induced by withdrawal of LIF, the expression of Nanog most likely disappeared after day 2 (2LIF; top line) and also the transcriptional level was down-regulated (data not shown) coinciding with the predominant accumulation of Trp53 in the nuclei. When DNA damage was induced in undifferentiated ES cells by the treatment with Doxorubicin for 3 hours, strong nuclear accumulation of Trp53 was observed with loss of Nanog staining although Oct3/4-Ecfp fusion protein remained in the nuclei (1Doxorubicin; bottom line). These data suggest the negative regulation of Nanog by nuclear localized Trp53 as reported previously 10 . Since it was known that Trp53 is important as a guardian of the genome stability, we first assessed the karyotypes of the Trp53-null ES cells independently isolated from the heterozygous ES cells in which the second Trp53 allele was correctly disrupted. As a result, we found that many of the Trp53-null ES cell lines possessed aneuploidy, but a few of them retained the normal karyotype (Fig. 2d). Interestingly, once the Trp53-null ES cell lines with normal karyotype were established, they never showed the genomic instability as reported previously. Therefore, the loss of Trp53 has a functional impact for keeping normal karyotype but it is able to be compensated, possibly by the function of the other Trp53-related genes such as Trp73.

Trp53-null ES cells show defects in the induction of differentiation markers after withdrawal of LIF. Previous reports indicated that
Trp53-null ES cells show defects in up-regulation of the differentiation marker genes and down-regulation of the pluripotencyassociated genes 10 . We tested the ability of differentiation in the Trp53-null ES cells with normal karyotype that we established. When cultured without LIF, they underwent a morphological differentiation as same as the wild-type and the Trp53-heterozygous ES cells (see phase contrast images in Supplemental Fig. 1a) with the change of the cell-cycle profile from ES-specific type to the somatic cell type 25 monitored by Fucci system 26 (Supplemental Fig.  1a, 1b). However, when the relative mRNA expression level of the differentiation marker genes were examined by RT-qPCR, we found that the up-regulation of the mesoderm markers T, Mixl1 27 , Gsc 28 and Lhx1 29 , and the extra-embryonic endoderm markers Gata4 and Gata6 30 were impaired in the Trp53-null ES cells (Fig. 3a). This attenuation was also consistent with the negative immunostaining of T and Gata4 at day 4 ( Fig. 3b and data not shown). Impaired upregulation of T is consistent with a previous report 10    In contrast to the drastic change in the induction of the differentiation markers, the kinetics of down-regulation of the pluripotencyassociated genes looked normal (Fig. 3a). Although a previous report mentioned the persistence of Nanog expression after induction of differentiation 10 , we observed that Nanog was transcriptionally down-regulated properly from day 2 (Fig. 3a), which was also confirmed by immunostaining (Fig. 3c). These data indicated that Trp53-null ES cells have deficiencies in the induction of differentiation markers during the differentiation process triggered by withdrawal of LIF but the repression of Nanog occurs normaly in the absence of Trp53. Trp53-null ES cells contribute to chimeric embryos. We confirmed that the new Trp53-null ES cell lines with normal karyotype show defects in differentiation as reported previously. To test their ability to adopt the normal developmental process, we injected them into the blastocysts to form the chimeric embryos. By the injection of a single Trp53-null ES cell into the blastocyst cavity, we succeeded to obtain chimeric embryos in which the ES-derived cells contributed to the three germ layers as efficiently as wild-type ES cells (Fig. 4 and Supplemental Fig. 4). These data clearly demonstrated that Trp53null ES cells retain pluripotency.

ES cells differentiate normally in embryoid bodies without Trp53.
The differentiation event induced by the withdrawal of LIF on twodimensional dish surface is somewhat artificial. In contrast, the differentiation of mouse ES cells in floating culture to form the embryoid bodies (EB) is regarded as mimicking the event in normal development 32 , encouraging us to test the ability of differentiation of the Trp53-null ES cells in this context. The Trp53-null ES cells were cultured in a hanging drop to form EB and the regulation of the marker gene expression was examined. Surprisingly, the up-regulation of T and Mixl1 occurred in the similar kinetics in EB derived from the Trp53-null ES cells to that of the wild-type ES cells (Fig. 5a). Both Gata4 and Gata6 expressions were also considerably recovered in the Trp53-null EB differentiation to the same level as in the EB derived from wild-type cells. The kinetics of the down-regulation of Nanog was also indistinguishable between the EB derived from the Trp53-null and wild-type ES cells (Fig. 5b). Therefore, the Trp53-null ES cells retained the normal ability to form EB as the wild-type ES cells and its defect in differentiation was context-dependent.

Discussion
The process of mouse development is largely divided into two stages by the event to attach maternal uteri, implantation. At the pre-implantation development, the zygote develops into the blastocyst by generating two extraembryonic cell lineages, trophectoderm and extra-embryonic endoderm, which contribute to implantation. After implantation, the epiblast that consists of pluripotent cells form primitive ectoderm then undergo embryonic development 33 . Mouse ES cells derived from the ICM of blastocysts mimic the character of epiblast cells at early post-implantation embryos 34 . Since they undergo differentiation into the three germ layers and germ cells in vitro, they are regarded as a good model system of peri-implantation development that is hard to be analyzed in vivo due to the small size embryos with inaccessibility 35,36 . Here we demonstrated that Trp53-null ES cells showed defects in two-dimensional differentiation in vitro but differentiate normally in three-dimensional differentiation in vitro as well as in chimeric embryos after injection into blastocysts, indicating the difference of these contexts on the requirement of autonomous gene function.
Differentiation of ES cells on flat surface has several advantages to analyze the cell-autonomous gene function and the effect of extrinsic factors on particular differentiation process as well as to obtain synchronous differentiation into particular cell types 37 . When ES cells were cultured on gelatin-coated surface without LIF, the primitive ectoderm marker Fgf5 31 was up-regulated at day 1 followed by the upregulation of the mesoderm marker T at day 2, suggesting the sequential differentiation of ES cells as found in embryonic development, which mimic the sequence of differentiation events in vivo. It was reported that ES cells undergo selective differentiation to neuroectoderm in serum-free culture 38 and lateral plate mesoderm on the surface coated by type IV collagen 37 , suggesting the similarity between in vitro differentiation and the developmental process in vivo can be extended to the later developmental stages. However, in our observation the requirement of Trp53 for the differentiation event in a two-dimensional culture was clearly distinct from that in a  three-dimensional culture in vitro or in a chimera formation in vivo. In the absence of Trp53, the induction of T was impaired in a twodimensional culture but not in a three-dimensional culture; it should also be normal in vivo since Trp53-null ES cells contributed to the mesoderm lineage in chimeric embryos. Therefore, the differentiation of ES cells in a two-dimensional culture is somewhat artificial although it remains unclear what the differences from embryonic development are. Indeed, differentiation of cardiomyocytes is easily induced by three-dimensional culture but not in two-dimensional condition 39 , suggesting their difference in mesoderm differentiation. Adequate attention should be paid for the interpretion of the data obtained using these systems.
Here we showed an unequivocal result about the function of Trp53 in mouse ES cells saying that Trp53 is dispensable for the maintenance of pluripotency and the induction of differentiation. Moreover, the role of Trp53 for the repression of Nanog should be a minor effect except the DNA damage response as found in Fig. 1d because Nanog was expressed heterogeneously in undifferentiated ES cells and down-regulated in differentiating ES cells without Trp53. It was recently reported that the auto-repression of Nanog is the major mechanism to generate the heterogeneity in undifferentiated ES cells 40 , suggesting the minor role of Trp53 on the repression of Nanog. However, it does not mean that Trp53 has no function in ES cells. Indeed, many Trp53-null ES cell lines we established showed abnormal karyotype, indicating the role of Trp53 as a major guardian of the genomic stability. A few Trp53-null ES cell lines kept normal karyotypes, suggesting that a functional compensation for the loss of Trp53 occurred in these cells. A previous report indicated that Trp73 partly share the function with Trp53 since Trp73 was essential for suppressing polyploidy and aneuploidy when Trp53 was inactivated 41 . To reveal the overlapaped function of Trp53 and Trp73 in ES cells, it will be required to generate double knockout lines of both genes probably by an inducible-knockout strategy to avoid the genetic instability during serial gene manipulation.

Methods
Cell culture and differentiation. ES cell lines (EB5) were cultured without feeder cells in Glasgow minimal essential medium (GMEM; Sigma) supplemented with 10% fetal calf serum, 1 mM sodium pyruvate, 10 24 M 2-mercaptoethanol, 13 nonessential amino acids, and 1,000 U of leukemia inhibitory factor (LIF) per ml on gelatin-coated dishes. For induction of differentiation in two-dimensional culture, ES cells were seeded (1,000 cells/cm 2 ) without LIF on gelatin-coated dishes. Embryoid bodies were generated by the culture of 300 ES cells in a 15 ml drop of the culture medium without LIF.
Quantitative PCR analysis. To quantify the levels of mRNA transcripts, cDNA were synthesized from 1 mg of total RNA using ReverTra (Toyobo), and quantified by realtime PCR using a CFX384 system (BioRad). Utilized primers were listed on supplemental Figure 3. All samples were tested in triplicate, and the mean relative amounts of each transcript were calculated by normalization to an endogenous control Gapdh. Results are given as the mean with 6SD. Statistical analysis was conducted using Student t-test. P , 0.05 was considered significant.
Immunohistochemistry and Western blot. Cells were fixed by 4% paraformaldehyde in phosphate-buffered saline (PBS) for 30 min at 4uC and then permeabilized by 0.1% Triton X-100 in PBS for 15 minutes at room temperature (RT). After brief washing with PBS followed by blocking with PBS containing 2% FCS, the cells were incubated with following primary antibodies: anti-Trp53 mouse monoclonal antibody (1C12, Cell Signaling Technology), mouse anti-Pml mouse monoclonal antibody (Upstate Biotechnology), anti-T goat polyclonal anitibody (Santa Cruz Bitotechnology), anti-Gata4 goat polyclonal anitibody (Santa Cruz Biotechnology), and anti-Nanog rat monoclonal antibody (R&D) for overnight at 4uC. After washing with PBS, The cells were incubated with Alexa Fluor 488-or 555or 633-conjugated donkey antibodies (Invitrogen) were used in a proper combination of species specificity as indicated in Figure legends. The fluorescent images were captured with an IX51 microscope with DP70 digital camera (Olympus) or a Leica SP2 confocal microscope (Leica).
Western blot was performed with anti-Trp53 mouse monoclonal antibody (1C12, Cell Signaling Technology) for the total cell lysates of wild-type and Trp53-null ES cells.
Gene targeting of Trp53. For generation of Trp53-KO vector, genomic DNA fragment for 59 and 39 homology arms (Chr:11, 69583689-69587696 and 69591432-69595365 in GRCm38) were amplified from the EB5 genomic DNA using the primer pairs 59aatgtcgaccggt (Sal I, AgeI) AACAGTCTTAAACCAGATGTGGTGGCTC-39 and 59aatctagaagatct (Xba I, Bgl II) CAGAGAAAAAGAGGCATTGAAAGGTC for 59 homology arms, and 59-aaatctagaagat (XbaI, Bgl II) CTGCCTCTGCATCCCGTC CCCATCACC-39 and 59-aatgcggccg (Not I) CTTTCTCTTTTGGTGATAGTACT GGGTGG-39 for 39 homology arms, respectively. The PCR product of 59 homology arm (4.0 kb) was digested by Sal I and Xba I. That of 39 homology arm (3.9 kb) was digested by Xba I and Not I. These DNA fragments were sub-cloned into the Sal I and Not I site of pBR-Blue II (pBR322 -derived plasmid carrying multicloning sites from pBluescript), resulting in pBR Blue II 59 1 39. For drug selection, the Frt flanked SA-IRES-bgeo selection cassette 24 was inserted into Bgl II sites of pBR Blue II 59 1 39, resulting in a promoter-trap Trp53 KO vector. The 100 mg of plasmid DNA was linealized by NotI and introduced into 10 7 ES cells in 0.4-cm cuvette by electroporation with a Gene Pulser (Bio-Rad, 800 V, 3 mF) followed by the culture with selection drug, G418 (160 mg/ml, Sigma) for 7 to 10 days. G418-resistent colonies were picked up and the first allele targeting was confirmed by genomic PCR. The inserted FRT cassette was removed by subsequent transfection with the FLPe expression vector. This strategy was repeated for the targeting of the second allele to create homozygous Trp53 KO ES cells.
Karyotyping. Trypsin/EDTA-dissociated ES cell pellet was resuspended in 0.075 M KCl hypotonic solution and incubated at RT for 10 min. After that, the equal volume of Carnoy's fixative (acetic acid5methanol 5 153) was added and then centrifuged (1,000 rpm 3 5 min). After removal of supernatant, Carnoy's fixative was added and then centrifuged (1000 rpm 3 5 min). This was repeated once and resuspended in Carnoy's fixative. Cell suspensions were dropped onto 50% ethanol-washed slides and dried overnight at RT. After chromosome staining with Hoechst33342, the slides were mounted with coverslips. The karyotype was evaluated by counting unscattered metaphase spreads with microscopic imaging with a 100x/NA1.4 oil immersion lens (UPLSAPO, Olympus).
Production of chimeric embryos. For assessing in vivo contribution, Trp53 (1/2) and Trp53 (2/2) ES cells were transfected with pPB-CAG-Egfp-IZ using the piggy-Bac transposon system 42 and the clones expressing Egfp homogeneously were selected. To obtain chimeric embryos, these cells were injected into C57/BL6J blastocysts. Embryos were dissected at 14.5 dpc and Egfp signals were detected under fluorescent dissecting microscope. All animal experiments were performed according to the guidelines for animal experiments of RIKEN Center for Developmental Biology and approved by the Animal Experiment Committee of the RIKEN Kobe Institute.