Reprogramming of Trophoblast Stem Cells into Pluripotent Stem Cells by Oct4§


  • Author contributions: T.W.: conception and design, collection and/or assembly of data, data analysis and interpretation, manuscript writing; H.W., J.H., L.K., Y.J., J.L., Y.Z., Z.K., L.L., and X.Z.: provision of study material, data analysis and interpretation; S.G.: conception and design, financial support, data analysis and interpretation, manuscript writing, final approval of manuscript.

  • First published online in STEM CELLS EXPRESS February 8, 2011.

  • §

    Disclosure of potential conflicts of interest is found at the end of this article.


ESCs and trophoblast stem (TS) cells are both derived from early embryos, yet these cells have distinct differentiation properties. ESCs can differentiate into all three germ layer cell types, whereas TS cells can only differentiate into placental cells. It has not been determined whether TS cells can be converted into ES-like pluripotent stem (PS) cells. Here, we report that overexpression of a single transcription factor, Oct4, in TS cells is sufficient to reprogram TS cells into a pluripotent state. These Oct4-induced PS (OiPS) cells have the epigenetic characteristics of ESCs, including X chromosome reactivation, elevated H3K27 me3 modifications, and hypomethylation of promoter regions in Oct4 and Nanog genes. Meanwhile, methylation of promoter region in the Elf5 gene occurred during reprogramming of TS cells. The gene expression profile of OiPS cells was very similar to ESCs. Moreover, OiPS cells can differentiate into the three germ layer cell types in vitro and in vivo. More importantly, chimeric mice with germline transmission could be efficiently produced from OiPS cells. Our results demonstrate that one single transcription factor, Oct4, could reprogram the nonembryonic TS cells into PS cells. STEM CELLS 2011;29:755–763


The first cell fate determination occurs at the eight-cell embryo stage in mice. Consequently, blastomeres give rise to two completely different cell types [1, 2]. Some blastomeres develop into trophoblast cells, triggered by both intrinsic and extrinsic signals. Upregulation of Cdx2 expression, mediated through the Hippo-Tead4 pathway, and simultaneous allocation of such blastomeres to the outside of the embryo initiate the formation of the trophectoderm [1–8]. The rest of the blastomeres, which express high levels of Oct4 and Sox2, are allocated inside of the embryo and develop into the pluripotent inner cell mass (ICM) [2, 9, 10].

In vitro, two types of stem cells can be established from embryos at the very early blastocyst stage, that is, trophoblast stem (TS) cells and ESCs [11–13]. In distinct culture conditions, both types of stem cells undergo self-renewal in vitro. Although both types of stem cells are derived from early embryos, they stably retain the cell lineage restrictions, and in turn, TS cells can never make embryonic tissues. It has been recently discovered that downregulation of Oct4 in ESCs can convert ESCs into TS-like cells [10]. However, it is not yet known if TS cells can be converted into ES-like pluripotent stem (PS) cells.

Generation of induced PS (iPS) cells from differentiated somatic cells, through overexpression of transcription factors that are highly enriched in embryonic stem (ES) cells [14–24], not only provides us with a revolutionary approach to potentially generate histocompatible patient-specific iPS cells for regenerative medicine application but also prompts us to apply this approach to understand cell fate determination. Recently, it has been demonstrated that certain cell types (such as neural stem cells) can be converted into PS cells through expression of only one or two transcription factors [25, 26]. Moreover, conversion of certain differentiated somatic cells into another type of differentiated cell has recently been demonstrated both in vivo and in vitro [27, 28].

Although TS cells and ESCs originate from early embryos, it remains unclear whether nonembryonic-derived TS cells can be converted into PS cells with pluripotency, and if so, how many transcription factors are required for this cell fate transition.



The specific pathogen-free grade mice were housed in the animal facility of the National Institute of Biological Sciences. All studies adhered to procedures consistent with the National Institute of Biological Sciences Guide for the care and use of laboratory animals.

TS Cell Derivation and Culture

TS cells were derived and maintained as previously described by Dr. Janet Rossant [13, 29]. Briefly, day 3.5 blastocysts were flushed from the uterus of 129/sv and OG2 female mice, which were mated with Rosa26-M2rtTA male mice. Blastocysts were then transferred on to mitomycin C (MMC)-treated ICR mouse embryonic fibroblasts (MEFs) with TSC medium containing RPMI1640 (Gibco Invitrogen, USA,, 20% fetal bovine serum (FBS; Hyclone, Logan, UT,, 1 mM sodium pyruvate (Sigma, USA,, 100 μM β-mercaptoethanol (Chemicon, Millipore, Temecula, CA,, 2 mM L-glutamine (Chemicon), and F4H (25 ng/ml fibroblast growth factor 4 (FGF4); Invitrogen and 1.0 μg/ml heparin; Sigma). After disaggregation at 5–7 days later, culture medium was changed to 70% FCM (feeder condition medium) + 1.5× F4H and refreshed every other day. At the time of first passage, the medium was changed to 70% FCM +F4H.

In Vitro Differentiation of TS Cells

A total of 5 × 105 TS cells were plated onto gelatin-coated tissue culture dishes with TSC medium for up to 6 days and total RNA were extracted for analyzing the expression of differentiation related genes by real-time polymerase chain reaction (PCR).

iPS Cells Generation

The plasmids used and the procedure of iPS cells derivation was described previously [18, 30, 31]. Briefly, 293T cells was transfected with TetO-FUW-Oct4, Sox2, Klf4, and c-Myc plasmids separately with lentivirus packaging plasmids ps-PAX-2 and pMD2G. Medium containing virus was collected twice at 24 and 48 hours after transfection. About 30 ml virus medium was centrifuged at 25,000 rpm for 1.5 hours at 4°C and resuspended in 4 ml 70% FCM +F4H with 5 μg/ml polybrene. A total of 105 TS cells was plated on gelatin-coated dish and infected for 10 hours. TS cells were recovered for 12 hours in 70% FCM +F4H and then medium replaced with ES medium containing Dulbecco's modified Eagle's medium (Chemicon), 15% FBS, 1 mM L-glutamine, 0.1 mM mercaptoethanol, 1% nonessential amino acid (Chemicon), and 1,000 U/ml leukemia inhibitory factor (LIF) (Chemicon) supplemented with 1 μg/ml doxcycline (Sigma) and refreshed everyday. After the ESC-like colonies appeared, the medium were replaced with ES medium for 2–3 days. Then the colonies were mechanically picked up and digested to culture on MMC-treated MEF cells as ESCs.


Karyotyping of TS cells and iPS cells was preformed as previously described [18]. Chromosomal G-band analyses were performed at Lanzhou University First Hospital IVF Center.

Immunofluorescent Staining and Confocal Microscopy

Cells growing on slides were fixed by 4% paraformaldehyde overnight at 4°C and then permeabilized for 15 minutes with 0.5% Triton X-100. Slides were blocked in 2.5% bovine serum albumin for 1 hour at room temperature then incubated with the first antibodies to anti-Oct4 (1:500; Santa Cruz), Nanog (1:1,000; Cosmo BioCo), SSEA1 (1:50; Millipore), Cdx2 (1:50; Santa Cruz), H3K27me3 (1:250; Millipore), and EZH2 (1:50; Cell Signaling) for 1.5 hour at room temperature separately. After three times of washing, slides were incubated with Alexa-Fluor goat anti-rabbit or goat anti-mouse IgG secondary antibody (1:1,000; Invitrogen) for 1 hour at room temperature. DNA was stained with 1 μg/ml DAPI and the slides were mounted in antifade solution.

Stained cells on slides were observed on a LSM 510 META microscope (Zeiss, Oberkochen, Germany, using a Plan Neofluar ×40 DIC or ×63 Oil DIC objective and excitation wavelengths of 543 and 405 nm. All collected images were assembled using Adobe Photoshop software (Adobe Systems, San Jose, CA, without any adjustment of contrast and brightness to the images.

Reverse Transcription PCR and Quantitative PCR

Total RNA was extracted from TS cells or iPS cells after removal of feeder cells by TRIzol reagent (Invitrogen) and reverse transcribed by MMLV system (Promega, Madison, WI, Reverse transcription PCR (RT-PCR) was performed as following with 2× PCR Supermix Solution (Vigorous, Beijing, China, 94°C for 4 minutes; 30 cycle of 94°C for 30 seconds, 58°C for 30 seconds; and 72°C for 40 seconds; 72°C for 5 minutes. Quantitative PCR was performed under the instruction of manufacturer with SYBR PrimeScript RT-PCR kit (Takara, Dalian, China, on a ABI PRISM 7500 Real-time PCR System (Applied Biosystems, Foster City, CA, www.applied All the primer sets used were previously described [14].

Alkaline Phosphatase Staining

Alkaline phosphatase (AP) staining was performed under the instruction of manufacturer with Leukocyte Alkaline Phosphatase Kit (Sigma).

Teratoma Formation

A total of 2 × 106 iPS cells were suspended in 200 μl phosphate-buffered saline (PBS) and subcutaneously injected into the groin of severe combined immunodeficiency (SCID) mice. Four to six weeks later, the teratoma with a diameter of 2 cm was dissected for H&E staining performed at Center for Clinical Laboratory Development of Peking Union Medical College.

In Vitro Differentiation of iPS Cells

A total of 105 iPS cells were cultured in IMDM (Gibco) with 15% FBS and 1 mM L-glutamine on Ultra-low attachment six-well plate (Costar, Amsterdam, The Netherlands, After 4–6 days, embryoid bodies (EBs) were trypsinized and plated on gelatin-coated tissue culture dishes for an extra 7 days and then total RNA was extracted.

Chimeras Generation and Genotyping

Day 3.5 blastocysts of ICR mice were flushed and 10–15 cells were injected to the cavity of each blastocyst. Then the blastocysts were transplanted into the uterus of pseudopregnant ICR mice. The female chimeric mice with 8- to 10-week old produced from iPS cells were mated with male ICR mice to confirm the germline transmission of iPS cells. To test the multipotency of TS cells, days 12.5 and 18.5 fetus were obtained by caesarean section and genomic DNA of placenta and embryo were isolated by phenol:chloroform:isoamylalcohol (25:24:1). PCR for genotyping was performed using primers targeting rtTA sequence got from Mouse Genome Informatics Website.

Bisulphite Sequencing

Genomic DNA of cells was modified with the Methylamp DNA Modification kit (Epigentek, Brooklyn, NY, Two-round nested PCR was performed to amplify the promoter region of each gene and PCR products were purified from agarose gel by Gel Purification kit (Qiagen, Venlo, The Netherlands,

Western Blot Analysis

After removal of feeder cells, iPS cells were washed once by PBS and boiled to 100°C for 5 minutes in protein-loading buffer (BioRad) with 2% β-mercaptoethanol (Amersham, CT, Anti α-tubulin(1:2,000; Sigma) and Histone H3 (1:5,000; generated by Antibody Center at National Institute of Biological Sciences, Beijing) antibody were used as endogenous control and anti-H3K27me3 (1:2,000; Millipore), anti-H3K4me3 (1:1,000; Millipore), anti-H3K9me3 (1:1,000; Millipore), Ezh2 (1:1,000; Cell Signaling, Boston, MA,, Suz12 (1:1,000; Cell Signaling) were used. ECL peroxidase-labeled anti-mouse or anti-rabbit antibodies (Amersham) were used as the secondary antibodies.

Microarray Analysis

Total RNA was extracted from TS cells or iPS cells including three biological repeats by TRIzol after removal of feeder cells. Affymetrix Mouse Gene 1.0 ST Array (Affymetrix, Inc., CA, was used and all were performed at Beijing Capitalbio corporation. The microarray data had been submitted to GEO database (GSE25255).


Derivation and Characterization of TS Cells

First, we derived TS cell lines using day 3.5 blastocysts from 129/sv and OG2 female mice that were mated with Rosa26-M2rtTA mice [18]. Following disaggregation and the first passage of cells, cells cultured in fibroblast-conditioned medium (supplemented with FGF4 and heparin) formed colonies with distinct borders and tight epithelial morphology resembling typical TS cells (Fig. 1A). Expression of a TS-specific transcription factor, Cdx2, was detected in the established TS cells (Fig. 1B). Following karyotype analysis, we chose one TS cell line, 129R3, with the correct karyotype (40, XX) and typical morphology of TS cells for further study (Fig. 1C). Using RT-PCR, we found that the 129R3 TS cells expressed the majority of TS cell marker genes, including Cdx2, Eomes and Esrrb, Fgfr2, and Sox2. The expression levels of these genes were comparable with levels in the widely used Cdx2–green fluorescent protein (GFP) TS cell line (the Cdx2-GFP TS cell line was kindly provided by Dr. Janet Rossant; Fig. 1D). Quantitative RT-PCR was further performed and the results indicated that the expression levels of Sox2, c-Myc, and Esrrb in 129R3 TS cells were similar to the levels in ESCs, whereas Oct4 and Klf4 expression was silenced in TS cells. In addition, the other Klf-family members including Klf1 and Klf5, which can substitute Klf4 in iPS generation [32, 33], were expressed at moderate level in TS cells but lower than ESCs (Fig. 1E). The results presented above prompted us to test the possibility of reprogramming the TS cells with fewer transcription factors in subsequent experiments.

Figure 1.

Derivation and characterization of TSCs. (A): Morphology of trophoblast stem cell line 129R3 derived from 129/sv ×Rosa26-MrtTA blastocysts. A picture of R1 ESC was shown. Left, ×40; right, ×10. (B): Immunofluorescent staining showing that the 129R3 TSC line is positive for Cdx2. (C): The karyotype of the 129R3 TSC was 40, XX. (D): Reverse transcription polymerase chain reaction results indicating that 129R3 TSCs express trophoblast stem cell marker genes including Cdx2, Sox2, and Esrrb. The Cdx2-GFP TSCs were used as a positive control. (E): Quantitative PCR demonstrates that 129R3 TSCs express transcription factors Sox2, c-Myc, and Esrrb at similar levels compared with R1 ESCs; but Klf4 expression level was much lower in TSCs. Oct4 expression was not detected in TSCs. Abbreviations: GFP, green fluorescent protein; MEF, mouse embryonic fibroblast; TSC, trophoblast stem cell.

To examine if the newly derived 129R3 TS cells possessed the unique TS cell differentiation properties, we cultured the 129R3 TS cells in the TS cell medium without FGF4 and heparin in vitro. Subsequently, 129R3 TS cells differentiated into giant cell-like cells, which are large and polyploid and belong to primary trophoblast cells (supporting information Fig. S1). The differentiated 129R3 TS cells expressed several genes of trophoblast cells, including Mash2 (specific for diploid trophoblast cells), Tpbpa (specific for spongiotrophoblast cells), Gcm1 (specific for labyrinthine trophoblast cells), and Pl1 (specific for trophoblast giant cells, supporting information Fig. S1) [13, 34]. We also examined the in vivo differentiation potential of 129R3 TS cells using chimera experiments. Ten to fifteen 129R3 TS cells were microinjected into ICR mouse blastocysts. Subsequently, rtTA-specific sequences were PCR amplified from placenta genomic DNA and genomic DNA from embryos collected at 12.5 dpc and 18.5 dpc. The rtTA sequence was only detected in genomic DNA from the placenta (supporting information Fig. S2). These results confirmed that the 129R3 TS cells we established are multipotent and can only commit to placenta cell lineages both in vitro and in vivo.

Reprogramming of TS Cells into iPS Cells

Reprogramming of 129R3 TS cells was further investigated using a doxcycline-inducible iPS system [30, 31]. As the 129R3 TS cells express Sox2 and c-Myc at moderate levels, and high expression of Esrrb is also detected in these cells, we hypothesized that fewer transcription factors, or even one factor (Oct4), would be sufficient to reprogram TS cells into PS cells. Evidence from recent studies has clearly shown that Oct4 alone could reprogram neural stem cells into PS cells because the other three transcription factors were also expressed at relatively high levels in neural stem cells [25, 26]. Although Klf4 is not expressed in TS cells, it has been recently demonstrated that Esrrb could substitute for Klf4 in reprogramming of fibroblast cells into PS cells [35]. On the basis of these studies and the gene expression characteristics of TS cells, we next examined the possibility of reprogramming TS cells with fewer transcription factors. Different transcription factor combinations were used, including all four factors, two factor combinations (which included Oct4 and one of the other three factors), and Oct4 alone. After lentiviral infection, 129R3 TS cells were allowed to recover for 12 hours in feeder condition medium, and then the medium was replaced with doxcycline-supplemented ES medium. All exogenous transcription factors were highly overexpressed in the TS cells within 48 hours after infection (supporting information Fig. S3). After 2–3 weeks, primary ES-like colonies appeared in all groups (supporting information Fig. S3). Variable reprogramming kinetics and efficiency were observed among the different transcription factor combinations (supporting information Fig. S3). More importantly, we succeeded in generating 10 iPS cell lines from Oct4-infected TS cells within 2.5 weeks (16 days), and the reprogramming kinetics were comparable with four factor-induced reprogramming (Fig. 2A and supporting information Fig. S3). Compared with the prolonged reprogramming kinetics (4–5 weeks) observed in neural stem cells [25], Oct4 conversion of TS cells into ES-like cells appeared relatively easier. Although the reprogramming efficiency of Oct4-infected TS cells was significantly lower than that of cells infected with Oct4 combined with either Klf4 or Sox2, it is comparable with four-factor MEF reprogramming [15]. Interestingly, the combination of Oct4 and c-Myc had no positive effects on reprogramming efficiency, and moreover, it prolonged the reprogramming kinetics (supporting information Fig. S3). Similarly, we could successfully reprogram the other TS cell line derived from OG2rtTA embryo into PS cells using one single transcription factor, Oct4 (supporting information Fig. S4).

Figure 2.

Conversion of 129R3 trophoblast stem (TS) cells into PS cells by overexpressing Oct4 or the combination of four factors. (A): Morphology of 129R3 TS cells after 7 days of induction, indicating that the cell morphology has changed (top), forming primary colonies at 14 or 16 days (middle), and established iPS cell lines (bottom). Left, induced by Oct4; right, induced by Oct4, Sox2, Klf4, and c-Myc. (B): Reverse transcription polymerase chain reaction results showing that OiPS and 4F iPS cells expressed ESC markers Oct4, Nanog, Sox2, etc. (C): Immunofluorescent staining results, indicate that OiPS and 4F iPS cells are positive for Oct4, Nanog, and SSEA1 but negative for Cdx2. Abbreviations: iPS, induced pluripotent stem cell; OiPS, Oct4-induced PS cell; SSEA1, stage-specific embryonic antigen 1.

Oct4-induced iPS cells (OiPS) and four factor-induced iPS cells (4F iPS) were indistinguishable from ESCs in their morphology and were AP-positive (Fig. 2A and supporting information Fig. S5). After doxcycline withdraw, the expression of exogenous genes was silenced in OiPS and 4F iPS cells (supporting information Fig. S6). After karyotype analysis, OiPS-8 and 4F iPS-4 cell lines derived from 129R3 TS cells were used for molecular characterization (supporting information Fig. S7). Both OiPS cells and 4F iPS cells expressed pluripotent genes including Oct4, Sox2, Nanog, Ecat, and Gdf3 at levels comparable with that of normal R1 ESCs (Fig. 2B). In contrast, the expression of TS-specific Cdx2 was silenced in both types of iPS cells. Moreover, endogenous Oct4, Sox2, Nanog, Klf4, and c-Myc were expressed at similar levels in OiPS, 4F iPS, and ESCs (supporting information Fig. S8). Expression of Oct4, Nanog, and SSEA-1 at protein level was detected in OiPS and 4F iPS cells by immunocytochemistry staining, but Cdx2 was no longer expressed in the Oct4-induced iPS cells (Fig. 2C).

Epigenetic Characteristics of Oct4-Induced PS Cells

DNA methylation-mediated gene regulation plays an important role in determining the expression of transcription factors, which are specifically expressed in either ES or TS cells. Oct4 and Nanog are two master pluripotency transcription factors expressed in pluripotent cells, we next analyzed DNA methylation of the Oct4 and Nanog promoters in 129R3 TS, R1, OiPS and 4F iPS cells, and bisulfite results illustrated that demethylation of Oct4 and Nanog promoters occurred in both OiPS and 4F iPS cells, whereas Oct4 and Nanog promoters were highly methylated in 129R3 TS cells (Fig. 3). Similarly, Elf5 is one of the master transcription factors specifically expressed in TS cells, and moreover, its expression is determined by DNA methylation [36]. Therefore, we further analyzed the DNA methylation status of the Elf5 promoter region in all the different cell lines. The bilsufite results revealed that the Elf5 promoter was unmethylated in TS cells but highly methylated in R1 ESCs (Fig. 3). Interestingly, methylation of the Elf5 promoter was observed in the TS-reprogrammed OiPS and 4F iPS cells (Fig. 3). These results indicated that DNA methylation can be reprogrammed in the process of cell fate transition from TS cells to iPS cells.

Figure 3.

DNA methylation status in the promoter regions of Oct4, Nanog, and Elf5. Bisulfite sequencing results showed that the highly methylated Oct4 and Nanog promoter regions in 129R3 trophoblast stem (TS) cells were demethylated in OiPS, 4F iPS cells. In the contrast, unmethylated promoter region of Elf5 in the 129R3 TS cells was methylated in OiPS and 4F iPS cells. R1 ESCs were used as a positive control. Open and closed circles indicated the unmethylated and methylated CpGs. For each gene, 10 samples were sequenced and each row indicated a repeat. Abbreviations: iPS, induced pluripotent stem cell; OiPS, Oct4-induced PS cell.

As the 129R3 TS cell line karyotype is 40, XX, we examined whether other epigenetic modification events, including X chromosome inactivation in TS cells, could be precisely reprogrammed during reprogramming. Female TS cell has one active and one inactive X chromosome, which is in accordance with its in vivo counterpart, that is, day 3.5 of the trophoblast embryo [37, 38]. However, both X chromosomes are active in female ESCs [39, 40]. As polycomb repressive complex (PRC2) complex components Eed and Ezh2-mediated H3K27me3 have an important role in the initiation of X chromosome inactivation [38], we used immunofluorescence to detect nuclear H3K27me3 foci, a hallmark of X chromosome inactivation. Apparently, one X chromosome was inactivated in 129R3 TS cells, and after reprogramming no H3K27 me3 foci were observed in either OiPS or 4F iPS cells, suggesting that the inactivated X chromosome was reactivated (Fig. 4A).

Figure 4.

Epigenetic characterization of inducible PS cells. (A): Immunofluorescent staining results showing that 129R3 trophoblast stem (TS) cells have H3K27me3 foci (a hallmark of X inactivation), whereas both OiPS and 4F iPS cells possess two active X chromosomes. The right panel is an enlarged region of the left panel. (B): Quantitative PCR demonstrated that 129R3 TS cells express lower levels of Ezh1, Ezh2, and Eed, and similar level of Suz12 and Rnf2 compared with OiPS, 4F iPS cells, and R1 ESCs. (C): Western blot indicating that H3K27me3 modification in TS cells is lower than in OiPS and 4F iPS cells, whereas H3K4me3 and H3K9me3 levels are similar in all cell lines tested. The PRC2 component Ezh2 has a much lower expression level in TS cells compared with both OiPS and 4F iPS cells. Abbreviations: iPS, induced pluripotent stem cell; OiPS, Oct4-induced PS cell.

Recently, it has been reported that there are distinct histone modifications in three stem cell lines derived from early embryos: ES, TS, and extra-embryonic endoderm (XEN) (derived from primitive endoderm) cells [41]. The H3K27me3 level in TS and XEN cells was much lower than in ESCs due to the reduced expression of the PRC2 components Ezh2, Eed, and Ezh1 [41]. Therefore, we examined the histone modifications and the expression level of PRC2 and PRC1 components in our TS cells and in the iPS cells using both Western blot and quantitative RT-PCR assays. The mRNA levels of Ezh2, Ezh1, and Eed in 129R3 TS cells were much lower than in either R1 ES or iPS cells, whereas the expression level of Suz12 and the PRC1 component Rnf2 was similar in all cell lines (Fig. 4B). In addition, H3K27me3 modifications in 129R3 and Cdx2-GFP TS cells were lower than in OiPS and 4F iPS cells as detected by Western blot. Similarly, the amount of Ezh2 protein in 129R3 TS cells was much lower than in either R1 ES or iPS cells (Fig. 4C). These results demonstrate that both OiPS and 4F iPS cells have the same epigenetic modifications, including X chromosome reactivation and elevated H3K27 me3 modifications, as normal ESCs and are distinct from the parental 129R3 TS cell line.

Gene Expression Profile of Oct4-Induced PS Cells

To investigate the global gene expression profile, we conducted microarray analysis on OiPS, 4F iPS, R1, and 129R3 TS cells using Affymetrix Mouse Gene 1.0 ST Array (Affymetrix, Inc.). Pearson correlation analysis was used to cluster the cells, and the results demonstrate that the gene expression profile of both OiPS and 4F iPS cells is highly similar to that of R1 ESCs, but distinct from the parental 129R3 TS cells (Fig. 5A). The volcano plot analysis was used to compare gene expression profiles between two different cell lines and the results illustrate that iPS cells are similar to each other as well as to R1 ESCs (Fig. 5B). We also found that the expression level of pluripotent genes in both OiPS and 4F iPS cells is comparable with that of ESCs (supporting information Fig. S9A). However, the expression of many X-chromosome genes was significantly upregulated in both OiPS and 4F iPS cells compared with R1 ESCs. Among them, 93 genes in OiPS and 35 genes in 4F iPS cells expressed a greater than twofold difference compared with R1 ESCs (selected representative genes are shown in supporting information Fig. S10A). On the other hand, expression of several Y-chromosome genes was lower in both OiPS and 4F iPS cells than in R1 ESCs (supporting information Fig. S10B). Apparently, upregulation of X-chromosome genes and low expression of Y-chromosome genes in iPS cells was due to the fact that iPS cells contain two active chromosomes. Several X-chromosome genes were expressed at much higher levels in OiPS cells than in 4F iPS cells (supporting information Fig. S9B). Results from a recent study have shown that expression of pluripotent genes (including Oct4, Nanog, and Sox2) is coordinated with Xist repression during embryogenesis [42]. It has also been reported that Oct4 functions upstream of Xist and depletion of Oct4 leads to inactivation of both X chromosomes in female cells [43]. TS cells inherit the imprinted X chromosome inactivation of the trophoblast [44], and reprogramming of TS cells by pluripotent genes cause repression of Xist expression, which in turn induces expression of other X-chromosome genes. However, whether reprogramming using different combinations of transcription factors causes different states of X chromosome “activation” needs to be further investigated. We analyzed the expression of imprinted genes and found that the reciprocal IGF2/H19 genes had higher expression level in iPS cells compared with R1 ESCs, whereas Peg5 and Peg10 had lower expression in iPS cells than R1 ESCs (supporting information Fig. S9C). Through an analysis of published microarray data [45], we found that expression of the H19 gene varied within different iPS cell lines. Moreover, biallelic expression of H19 in human iPS cells has been reported [46]. Expression of Gtl2 imprinted gene clusters has been suggested to be important for iPS cells pluripotency [45, 47]. Our array data showed that expression of Gtl2 and Rian was downregulated in OiPS and 4F iPS cells, whereas Dlk1 expressed higher and Dio3 expressed at similar level as compared with R1 ESCs (supporting information Fig. S10C). Taken together, the microarray data suggest that successful reprogramming occurred in both Oct4 and four factor-induced iPS cells, whereas disruption in the expression of some imprinted genes might also occur during the reprogramming process.

Figure 5.

Global gene expression profile of iPS cells. (A): Hierarchical clustering of global transcriptional profiles generated from microarray results from 129R3, OiPS-8, 4F iPS-4, and R1 cells, indicating greater similarities among OiPS, 4F iPS, and R1. (B): Volcano plot analysis was used to compare gene expression profiles from two different cell lines. The results show that OiPS and 4F iPS cells are similar to each other as well as to R1 but distinct from 129R3. Abbreviations: iPS, induced pluripotent stem cell; OiPS, Oct4-induced PS cell.

Pluripotency of Oct4-Induced PS Cells

To investigate the differentiation potential of both OiPS and 4F iPS cells in vitro, we first determined whether both one factor and four factor-induced iPS cells could differentiate into the three germ layers using the EB differentiation assay. When iPS cells were cultured in low-adherent dishes without feeder cells or LIF, EBs could be efficiently generated from iPS cells (supporting information Fig. S11). After being plated onto tissue culture dishes for an extra 7 days, differentiated cells expressed marker genes including Gata6, Brachyury, and Map2 (from the three germ layers, endoderm, mesoderm, and ectoderm, respectively) as determined by quantitative RT-PCR (supporting information Fig. S11).

To further investigate the differentiation potential of both OiPS and 4F iPS cells in vivo, we used teratoma formation and chimera-contribution assays. Teratomas were formed 4–5 weeks after subcutaneous transplantation of both types of iPS cells into SCID mice. The teratomas contained tissues from all three germ layers as revealed by H&E staining (Fig. 6A). Finally, we used the chimera contribution assay to assess the developmental potential of both OiPS and 4F iPS cells. Both OiPS and 4F iPS cells could efficiently generate chimeric mice as determined by the coat color of the mice (Fig. 6B). We further assessed the germline transmission capability of these chimeric mice. After mating female chimeric mice with ICR male mice, germline transmission was observed (Fig. 6B). Our data demonstrate that OiPS cells have a developmental potential that is comparable with that of normal ESCs.

Figure 6.

Pluripotency characterization of Oct4-induced PS cells. (A): H&E staining shows that teratomas produced from OiPS and 4F iPS cells contain tissues from the three germ layers. (B): Chimeric mice of OiPS-8 cells with high chimerism (left) and germ line transmission (right) from the chimeric mice. Abbreviations: iPS, induced pluripotent stem cell; OiPS, Oct4-induced PS cell.


As the first differentiated cell lineage during embryogenesis, trophoblast cells only contribute to the placenta in vivo. In vitro, TS cells were established and have been widely used to investigate the mechanism of embryo implantation. Interestingly, it was discovered that ESCs derived from the other early blastocyst cell lineage (ICM cells) could be converted into TS-like cells through either forced repression of Oct4 or overexpression of Cdx2 [10]. Therefore, it became imperative to understand whether TS cell fate could be converted into an ES-like cell fate. Our present study demonstrates that a single transcription factor, Oct4, is sufficient to allow this cell fate conversion.

In early embryos, Oct4 and Cdx2 form a complex and the expression level of either of these transcription factors determines the earliest cell fate commitment [10]. Subsequently, Cdx2 is predominantly expressed in TS cells, and we speculate that overexpression of Oct4 in TS cells may repress expression of Cdx2, thus, re-establishing a PS cell transcription regulatory network where Sox2 is moderately expressed in TS cells. Indeed, our present study proved this hypothesis. We propose that the principle in TS cell reprogramming resembles neural stem cells reprogramming but with slight differences [25, 26]. All the rest three reprogramming factors including Sox2, Klf4, and c-Myc, are expressed in neuron stem (NS) cells but the Klf4 was found silenced in TS cells even though expression of Sox2 and c-Myc was detected. However, expression of another transcription factor, Esrrb, in TS cells was detected. It has been previously shown that Esrrb is capable of substituting for Klf4 in reprogramming of fibroblasts to pluripotency [35]. Moreover, moderate expression of Klf-family members including Klf1 and Klf5 in TS cells might also play certain role in replacing Klf4 during reprogramming. Therefore, the unique gene expression characteristics of TS cells make it another suitable cell type to be reprogrammed by a single transcription factor Oct4. Different from the slow reprogramming kinetics observed in NS cells reprogramming, the successful conversion of TS cells into PS cells only need 16 days. This phenomenon might indicate that molecular events occurred in the process of TS cell reprogramming might be distinct from NS cell reprogramming. More in-depth investigations need to be performed in the future to explore the precise molecular mechanism leading to this cell fate conversion.

Interestingly, in the initial experiments, we presumed that knock down of Cdx2 in TS cells might be beneficial for the conversion of TS cells into PS cells. However, we found that downregulation of Cdx2 expression has no positive effects in this process (data not shown), which may be due to the fact that downregulation of Cdx2 in TS cells causes TS cell differentiation instead of boosting the reprogramming process.

Epigenetic reprogramming plays a central role in nuclear reprogramming mediated by either somatic cell nuclear transfer or defined transcription factors [48, 49]. Demethylation of the promoter regions in the pluripotency transcription factors has been persistently used to evaluate if the reprogramming is complete. In the process of cell fate transition from TS cells to PS cells, demethylation of the promoter regions in the Oct4 and Nanog genes indeed occurred. Moreover, methylation of the promoter region of Elf5 gene, which was unmethylated in TS cells, occurred in the reprogrammed PS cells. To our knowledge, this is the first evidence showing that the promoter region of a lineage-specific gene has to be methylated in the process of reprogramming.

Recent investigations have demonstrated that distinct epigenetic configurations are persisted in ES and TS cells [41]. H3K4me3/H3K27me3 bivalent marks are found in a specific set of genes encoding developmental regulators in ESCs, whereas this bivalent mark is not observed in TS cells. It remains to be determined in the future to address how this epigenetic configuration re-established in the reprogrammed iPS cells derived from TS cells. Our system provides some advantages in answering such questions because only one transcription factor Oct4 is capable of reprogramming the TS cells into iPS cells, which excludes the effects of other transcription factors that are normally used in fibroblasts reprogramming.


In conclusion, this study demonstrated that a single transcription factor Oct4 is sufficient to convert nonembryonic-derived TS cells into PS cells with pluripotency. Our study not only provides an invaluable tool to further decipher the molecular mechanism underlying cell fate determination in early embryos but also suggests that TS cells from the placenta might be a better cell type for obtaining iPS cells.


We are grateful to our colleagues in our laboratory for their assistance with experiments and in the preparation of this article. We also thank Dr. Janet Rossant for giving us great suggestions during the preparation of the article. This project was supported by the Ministry of Science and Technology (grants 2008AA022311, 2010CB944900, and 2011CB964800).


The authors indicate no potential conflicts of interest.