|Home | About | Journals | Submit | Contact Us | Français|
Nuclear transfer to eggs or oocytes provides a potential route for cell-replacement therapies  because oocytes directly reprogram transplanted mammalian somatic-cell nuclei such that they have an embryo-like pattern of gene expression. This includes a large increase in the mRNA level of the stem-cell marker gene oct4. We have developed a novel procedure to identify new proteins that greatly increase the level of oct4 mRNA upon nuclear transfer. We have isolated Xenopus oocyte proteins that bind to the regulatory region of the mouse oct4 gene and identified these by mass spectrometry. The proteins include the retinoic-acid-receptor γ, a known repressor of oct4 transcription, and Tpt1, a cancer-associated factor . The depletion of transcripts of retinoic-acid receptor γ from oocytes increases oct4 and nanog transcription as expected, and depletion of tpt1 transcripts in oocytes reduces oct4 and nanog transcription in injected HeLa nuclei. An elevation of tpt1 transcripts in oocytes results in an earlier activation of oct4 transcription. Therefore, we identify a novel role for tpt1 in activating pluripotency genes upon nuclear transfer. Our results help to elucidate the mechanism by which somatic-cell nuclei are reprogrammed to have an embryo-like pattern of gene expression.
Apart from global chromatin decondensation [4–7], the mechanisms and molecules involved in nuclear reprogramming and in establishing pluripotency are not well understood. oct4 is a gene that is normally expressed only in pluripotent, embryo-like cells. It is not expressed in somatic cells but is activated upon nuclear transfer into oocytes and in embryonic stem (ES) cells [2, 8]. When development and differentiation proceed [9–12], or when retinoic acid is present , oct4 transcription is downregulated. Pluripotency proteins such as Oct4 and Nanog are transcription factors with essential roles in permitting the proliferation of ES cells [14–16]. Even though several proteins, such as GCNF , SF1 , LRH1 , retinoic acid , and Sall4 , are already known to regulate the oct4 promoter region in cultured cells or in later development, it is still unknown what is responsible for the activation of oct4 transcription, which correlates with cloning efficiency [21, 22]. Xenopus oocytes provide abundant material and contain proteins that cause a large increase in the level of oct4 mRNA. The identification of such proteins is important for a better understanding of generating pluripotency and of the molecular events by which eggs or oocytes reprogram somatic-cell nuclei.
To identify proteins that bind to the oct4 regulatory region, we have used radioactively labeled DNA oligos corresponding to regions of the mouse oct4 promoter (Figures 1A and 1B). These oligos were incubated in oocyte extract and DNA-protein complexes analyzed on gels. Many of the oocyte-derived proteins showed sequence specificity for the oligos representing the oct4 regulatory region, as shown by the addition of increasing concentrations of unlabeled competitor DNA (Figure 1C). Competition experiments with CpG methylated versus unmethylated oligos revealed that the proteins prefer to bind to the unmethylated oligos (Figure 1D). To identify which proteins bind to the different oligos, we recovered the proteins from gels and identified them by mass spectrometry. We have identified proteins already known to have a function in nuclear reprogramming and have thereby validated our procedure. These proteins include nucleoplasmin [4, 7, 23] and the histone-binding protein N1/N2 . We identified FRGY2, which is involved in disassembling somatic nucleoli in egg cytoplasm; this disassembly process accompanies nuclear reprogramming . The retinoic-acid receptor γ was also identified. Retinoic acid induces retinoic-acid receptors in cultured cells, and an increase in both retinoic acid and its receptors causes differentiation of pluripotent cells [13, 18]. A retinoic-acid receptor has already been shown to bind to the oct4 regulatory region . Most intriguingly, we have identified Tpt1 by using mass spectrometry with oligo Sf1 (data not shown); Tpt1 has not so far been implicated in regulating oct4 transcription.
Tpt1 stands for tumor translationally controlled protein 1. It is widely expressed and conserved throughout vertebrates, and was first identified in a screen for genes involved in tumor reversion. tpt1 is the gene that shows the strongest differential expression between tumor and tumor-reversed states in human leukemia and breast-cancer cells . tpt1 is strongly expressed in tumor cells and is downregulated upon tumor reversal. Moreover, when Tpt1 is inhibited in tumor cells, suppression of the malignant phenotype is observed . oct4 is expressed in tumor cells , and we recovered Tpt1 with our oligos. Hence, we decided to focus on Tpt1.
We tested whether Tpt1 binding to the Sf1 oligo could be confirmed. We repeated the gel shift, but this time in the presence of a Tpt1 antibody (Figure 2A). This resulted in a supershift, indicating that Tpt1 in fact binds to the Sf1 oligo and is not an artifact of the mass-spectrometry results. In addition, Tpt1 was also found to bind in vivo, as determined by chromatin immunoprecipitation (ChIP) analysis with Xenopus oocytes (Figure 2B).
To determine the developmental function of Tpt1, we carried out a test that depends on the downregulation of the tpt1 mRNA. We wanted to find out whether oct4 activation is inhibited in reduced abundance of tpt1. Antisense DNA oligos (ODN, oligo deoxynucleotides) that successfully degrade endogenous tpt1 mRNA in living oocytes were designed (Figures 2C and 2D). Antisense oligos were also prepared for the identified retinoic-acid receptor γ (RAR), which serves as a control because of its known function in downregulating oct4 transcription  (Figures 2C and 2E). After downregulation of both tpt1 and RAR mRNAs, the levels of other oocyte mRNAs were unchanged (Figure 2F). As a control, antisense oligos against gfp did not cause downregulation of any other genes tested (Figures 2C–2F). This suggests that the antisense oligos against RAR and tpt1 act specifically. Moreover, antisense oligos against Xenopus tpt1 mRNA resulted in specific downregulation of Xenopus Tpt1 protein but not of mouse Tpt1 protein, which is not targeted by these antisense oligos (Figure 2G).
To reinforce our evidence for the importance of tpt1 for oct4 activation, we have asked whether tpt1 is required for an increase in oct4 mRNA in nuclear-transfer experiments. Forty hours after oocyte injection of antisense oligos, approximately 100 HeLa nuclei were injected into the germinal vesicles of oocytes, which were cultured for up to 24 hr. We took different time points and analyzed them by RT-PCR in order to observe the activation of oct4 (Figure 3A). Because our method is very sensitive, we found that oct4 and nanog were expressed in nonpluripotent cells, such as human HeLa cells, but at a very low background level. However, if a large number (several thousand) of nonpluripotent HeLa cells was analyzed, the signal was strong and was used as a positive control for the RT-PCR. The time course revealed a gradual increase of oct4 mRNA (Figure 3C). In control samples that were only injected with HeLa nuclei and not antisense oligos, we observed a 100-fold increase in oct4 mRNA within several hours of nuclear transfer (Yen-Hsi Kuo, personal communication). We find that oocytes depleted of RAR mRNA activate human oct4 transcription earlier and more strongly than do oocytes with a normal endogenous level of RAR (Figure 3C). This increased level of oct4 transcription due to RAR mRNA downregulation agrees with previous findings that RAR is an inhibitor of pluripotency and induces differentiation [13, 18] and thus validates our functional test. When antisense oligos against tpt1 were injected into oocytes, the opposite result was obtained. We found that a reduction in the oocyte content of tpt1 greatly reduced oct4 transcription (Figure 3D). This suggests that tpt1 is involved in oct4 transcriptional activation.
To test the specificity of our antisense oligo against tpt1 mRNA, we have rescued Xenopus tpt1 depletion by injecting mouse tpt1 mRNA into oocytes previously injected with antisense oligos against tpt1 (Figure 3E). Importantly, mouse tpt1 mRNA does not contain the sequence that is targeted by the Xenopus antisense oligos. Hence, mouse tpt1 mRNA was not degraded and rescued the depletion of Xenopus tpt1 mRNA and, subsequently, oct4 expression (Figure 3E). This result confirms that the downregulation of tpt1 mRNA is specific.
To determine more precisely how the level of tpt1 influences oct4 transcription, we injected different concentrations of mouse tpt1 mRNA into oocyte cytoplasm. An incubation period of 16 hr allowed for the mRNA to be translated, and HeLa nuclei were then injected into the oocyte germinal vesicle. After 8 hr of incubation, the oct4 signal was already enhanced in comparison to control oocytes (Figure 3F). As a result, mouse tpt1 mRNA injected into Xenopus oocytes was effective at enhancing oct4 transcription. These observations were independent of the mRNA concentrations used (Figure 3E). Hence, in our nuclear-transfer experiments, tpt1 overexpression enhanced oct4. This suggests that Tpt1 might have an important role in regulating oct4; namely, it might ensure a level required for pluripotency and normal development.
We were interested in whether Tpt1 is a transcription factor specific for oct4 or whether it is a more generally important component that might act on other pluripotency genes. Using the same up- and down-regulation procedure just described for oct4, we have tested the transcriptional activation of the human pluripotency gene nanog by using nuclear transfer to oocytes (Figure 3G) [27, 28]. nanog seems to be an important pluripotency gene because its overexpression increases the efficiency of nuclear reprogramming by 200-fold in cell-fusion experiments . Overall, we found that a reduction in the RAR did not significantly affect nanog expression. However, we had expected some effect on nanog expression because retinoic acid is known to downregulate pluripotency genes [13, 18]. One explanation for our finding is that Nanog or Nanog orthologs do not exist in amphibians (NCBI blast). As a result, some aspects of transcriptional activation upon nuclear transfer might not be conserved between mammals and amphibians. However, we saw a reduction in nanog transcription after downregulation of tpt1 in oocytes (Figure 3G). To test for binding of Tpt1 to nanog, we carried out ChiP analysis in mouse ES cells and found that Tpt1 bound to oct4, but not significantly to the immediate promoter of nanog (Figure 3B). This suggests that Tpt1 regulates oct4 by binding to its DNA directly, whereas Tpt1 might regulate nanog in an indirect way or by binding to a distant site.
Overall, Tpt1 seems to represent a global regulator of pluripotency. We suggest that Tpt1 might have a somewhat general role in establishing an embryonic pattern of gene expression because it controls transcription of oct4 and nanog. In support of this view, we have also found that the transfection of tpt1 antisense RNA into mouse ES cells can reduce the transcription from the oct4 promoter (Figures 4A–4C), suggesting a conserved mechanism between mammals and amphibians.
The molecular events occurring after nuclear transfer and the initiation of pluripotency are not well understood. We describe here the first protein that has been discovered to be required for activating transcription of pluriptoency genes such as oct4 and nanog after nuclear transfer into oocytes. In the future, this type of work might provide a generally applicable experimental design, which in combination with the advantages of Xenopus oocytes, might lead to the identification of other proteins that activate pluripotency genes and thereby further elucidate mechanisms involved in nuclear reprogramming.
Complementary single-stranded 30 bp oligos that represent DNA fragments of the oct4 promoter  were annealed to double-stranded (DS) DNA. Longer DS oligo fragments were made with PCR amplification. Oligos were end labeled with γ-P32 ATP.
The phosphate binding buffer consisted of 20 mM phosphate (pH 7.4), 10% glycerol, 3 mM MgCl2, 17 mM KCl, 2.5 μM DTT, and 50 mM–2 M NaCl. We used 8 μl oocyte extract per sample and prepared it by mixing freshly obtained oocytes in presence of protease inhibitors. Centrifuging the mixture overnight eliminated debris and lipids. For the gel shift, 8 μl oocyte extract, 2 μg BSA, and 10 μl phosphate binding buffer were mixed with approximately 0.1 pmol radioactively labeled oligos and run on a native gel. To confirm binding of Tpt1 to any one oligo, we added 1 μl of Tpt1 antibody (monoclonal antibody anti-HRF/TCTP, Medical & Biological Laboratories) to the sample before was running it on a gel.
Proteins were recovered directly from the gels. The samples were washed, concentrated, and analyzed with MALDI-TOF mass spectrometry.
We thank S. Maslen and E. Stephens for help with the mass spectrometry, A. Surani for Oct4-GFP ES cells, Stina Simonsson for initial guidance, and A. Bannister for discussion and advice. This work was supported by the Wellcome Trust.