SEARCH

SEARCH BY CITATION

Keywords:

  • Embryonic stem cell;
  • Hematopoiesis;
  • Cell differentiation;
  • Transcription factor;
  • Mesoderm;
  • Gene expression

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References

The Oct4 transcription factor is essential for the self-renewal and pluripotency of embryonic stem cells (ESCs). Oct4 level also controls the fate of ESCs. We analyzed the effects of Oct4 overproduction on the hematopoietic differentiation of ESCs. Oct4 was introduced into ESCs via a bicistronic retroviral vector, and cells were selected on the basis of Oct4 production, with Oct4+ and Oct42+ displaying twofold and three- to fourfold overproduction, respectively. Oct4 overproduction inhibited hematopoietic differentiation in a dose-dependent manner, after the induction of such differentiation by the formation of day 6 embryoid bodies (EB6). This effect resulted from defective EB6 formation rather than from defective hematopoietic differentiation. In contrast, when hematopoiesis was induced by the formation of blast colonies, the effects of Oct4 depended on the level of overproduction: twofold overproduction increased hematopoietic differentiation, whereas higher levels of overproduction markedly inhibited hematopoietic development. This increase or maintenance of Oct4 levels appears to alter the kinetics and pattern of mesoderm commitment, thereby modifying hemangioblast generation. These results demonstrate that Oct4 acts as a master regulator of ESC differentiation.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References

Self-renewal is one of the major properties of stem cells. This property is required for continuous tissue regeneration and for the persistence of long-lived cells capable of promoting tissue homeostasis and repair.

However, the different stem cells in the human body differ considerably in their self-renewal capacities. Hematopoietic stem cells (HSCs), the best characterized adult stem cells, have considerable but limited self-renewal capacities and cannot be grown as permanent cell lines. They therefore cannot be propagated and expanded in vitro, limiting their use in clinical practice. It remains unclear whether this limitation is an intrinsic feature of HSCs or whether it results from a lack of knowledge of the growth factors regulating self-renewal and differentiation potential. In contrast, embryonic stem cells (ESCs) display extensive self-renewal potential in vitro and can be propagated as cell lines, through symmetric mitosis. They can also be induced to differentiate into multiple lineages in vitro. ESCs are derived from the inner cell mass (ICM) and can generate all embryonic cell types. Our understanding of the molecular mechanisms governing ESC self-renewal and differentiation has recently increased substantially. Two endogenous transcription factors, Oct4 and Nanog, have been shown to play key roles in ESC self-renewal [1, [2]3]. However, self-renewal is also regulated by two major signaling pathways: the leukemia inhibitory factor (LIF)/Gp130/Stat3 pathway and the recently characterized bone morphogenic protein/Smad1/Id pathway. These two pathways directly affect ESC self-renewal capacity and also have a profound effect on the differentiation potential of these cells [4, [5]6].

Oct4, a POU transcription factor, is produced in ESCs and epiblast and primordial germline cells [7, 8]. It is critical for the establishment of ICM pluripotency. Abnormal Oct4 production in blastocyst-stage mouse clones was found to be correlated with a low rate of post-implantation survival because developmental potential was restricted to the trophectoderm lineage [9]. Oct4 acts synergistically with Stat3 to induce ESC self-renewal. This effect of Oct4 on self-renewal is independent of its effect on the synthesis of one of its major targets, fibroblast growth factor-4 (FGF4), a pre-implantation growth factor required for blastocyst viability [10]. The precise level of Oct4 also tightly regulates the differentiation capacity of ESCs. A lack of Oct4 production is associated with the loss of pluripotentiality and the induction of trophectodermic differentiation [11, 12]. Surprisingly, doubling in Oct4 levels also abolishes the pluripotency of ESCs, inducing differentiation toward two of the three major embryonic lineages (mesoderm and endoderm) [11]. Furthermore, sustained Oct4 production accelerates the neuroectoderm differentiation of ESCs in serum-free, LIF-deficient medium [13]. Thus, Oct4 is a key determinant of the properties of ESCs because it activates or represses target genes [14, [15], [16]17]. Oct4 is not present in adult somatic cell types and is considered to be a transcription factor specific for early stages of embryonic development, with the exception of the germline. However, Oct4 was detected, albeit at a much lower level than in ESCs, in multipotent adult progenitor cells, putative adult stem cells capable of extensive self-renewal. In contrast, it is absent from HSCs [18].

The mechanisms controlling the pluripotency and differentiation of ESCs are unclear, and Oct4 is just one of the molecules affecting the fate of ESCs. The presence of Oct4 in ESCs, but not in their differentiated derivatives, suggests that downregulation of Oct4 gene is a prerequisite for the induction of differentiation. A doubling in Oct4 production leads to the induction of mesodermal markers, suggesting that hematopoietic differentiation should also be modified [11]. We studied the effects of Oct4 overexpression in ESCs on their hematopoietic differentiation of these cells. We demonstrated that the effects of Oct4 overproduction in ESCs depended on Oct4 levels, as previously reported during primordial germ layer specification, and on the differentiation pathways used to induce hematopoiesis from ESCs in vitro.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References

ESC Lines

The 129/Sv-derived wild-type (WT) D3 ESC line was kindly provided by F. de Sauvage (Genentech, Inc., South San Francisco, CA, http://www.gene.com).

Growth and Differentiation of ESCs

Cells were cultured on a monolayer of mitomycin C-inactivated mouse embryonic fibroblasts (MEFs) derived from day-14 Swiss mouse embryos. ESCs were maintained in an undifferentiated state by culture in Dulbecco's modified Eagle's medium (DMEM; Invitrogen, Cergy Pontoise, France, http://www.invitrogen.com) supplemented with recombinant LIF (1,000 U/ml; ESGRO; Abcys, Paris, http://www.abcys.fr), 15% fetal bovine serum (FBS; Perbio Science, Brebieres, France, http://www.perbio.com), 150 μM monothioglycerol (MTG; Sigma, Saint Quentin Fallavier, France, http://www.sigma-aldrich.com), 1% l-glutamine (Invitrogen), and 1% penicillin-streptomycin (PS; Invitrogen). ESCs were passaged every 2 days, and the culture medium was changed daily. Cultures were maintained at 37°C in an atmosphere containing 5% CO2 and O2 (7% O2).

We used slightly modified versions of two methods initially described by Kennedy et al. [19] and Wiles et al. [20] to study hematopoietic differentiation from ESCs. Undifferentiated ESCs were treated with trypsin and cultured for 1 hour in DMEM without LIF in cell culture dishes to eliminate adherent embryonic fibroblasts. ESCs were then plated in Iscove's modified Dulbecco medium (IMDM; Invitrogen) supplemented with 15% FBS, 450 μM MTG, 50 μg/ml l-ascorbic acid (Sigma), and 200 μg/ml transferrin (Sigma). Under these conditions, ESCs give rise to embryoid bodies (EBs). EBs were generated on day 6 of differentiation (EB6) by plating 500 cells per milliliter in bacterial-grade dishes. ESC-derived blast cell colonies (BL) were generating by plating 4,500 cells per milliliter on these dishes. After 3 days of culture, EBs in their third day of differentiation (EB3) were rinsed twice with 1× phosphate-buffered saline (PBS; Invitrogen) and centrifuged. The pellet was disrupted by trypsin treatment, and the cells were replated in 1% methycellulose (Fluka, Buchs, Switzerland, http://www.emarketlabo.com) in IMDM supplemented with 1% l-glutamine, 1% PS, 10% FBS (Perbio Science), 450 μM MTG, 25 μg/ml ascorbic acid, 200 μg/ml transferrin, and 5 ng/ml vascular endothelial cell growth factor (VEGF). We plated 75,000 EB3 in a volume of 1.5 ml in 35-mm-diameter bacterial-grade dishes. In these conditions, BL were detected and scored on the fourth day of culture.

Hematopoietic Progenitor Assays

We quantified myeloid progenitors among BL or in EB6 as follows. Cells were collected, rinsed, centrifuged, and mechanically dissociated in IMDM and filtered through Millipore filters (Millipore Corporation, Billerica, MA, http://www.millipore.com) with 40-μm pores. Cells were plated in 1% methylcellulose in IMDM containing six growth factors (5 U/ml erythropoietin [EPO], 1,000 U/ml interleukin [IL]-1α, 100 U/ml IL-3, 10 ng/ml IL-6, 10 ng/ml granulocyte colony-stimulating factor [G-CSF], and 5% of baby hamster kidney/mouse c-Kit ligand [BHK/MKL] cell line conditioned medium containing stem cell factor [SCF]). Hematopoietic colonies were counted after 7 days.

Endothelial Cell Culture

On day 4, BL were collected, rinsed, centrifuged, dissociated, and filtered as described above. We determined the number of BL-derived endothelial foci by plating 0.5–4 × 103 cells in 12-well gelatin-coated tissue culture plates containing IMDM supplemented with 10% FBS, 10% horse serum (Roche Diagnostics, Mannheim, Germany, http://www.roche-diagnostics.com), 5 ng/ml VEGF, 10 ng/ml basic fibroblast growth factor, 100 μg/ml endothelial cell growth supplement (Sigma), 100 μg/ml sodium heparin (Sigma), 450 μM MTG, and 0.5% l-glutamine. Endothelial foci were counted after 3 days of culture.

Growth Factors

Murine recombinant IL-3, recombinant IL-1α, IL-6, and VEGF were purchased from Abcys. Human EPO was a gift from Cilag (Levallois-Perret, France, http://www.janssen-cilag.fr). Human G-CSF was obtained from Bellon Laboratory (Neuilly sur Seine, France). Murine BHK/MKL-SCF-producing cell lines were also used as a source of SCF.

Vector Construction

Mig-R1, the murine stem cell virus (MSCV) retroviral vector containing the encephalomyocarditis virus (ECMV) internal ribosomal entry sequence (IRES) and the coding sequence for enhanced green fluorescent protein (eGFP), were kindly provided by Warren S. Pear (University of Pennsylvania, Philadelphia). The Mig-R2 vector was constructed by replacing the ECMV-IRES cassette with the VEGF-IRES cassette. The murine Oct4 cDNA was generated from embryonic murine stem cell mRNA amplified and inserted into Mig-R2 (Mig-R2-Oct4).

For dual-luciferase assays (Dual-Luciferase Reporter Assay System; Promega, Madison, WI, http://www.promega.com), the reporter vectors were constructed by subcloning six repeats of the octamer sequence downstream from the thymidine kinase promoter in the pGL3 vector containing the firefly luciferase reporter gene (Promega) (i.e., :pGL3HSTK(6X)). The octamer sequence used was reported by Schöler et al. [21]. The pGL3 control plasmid, containing the ubiquitous SV40 (simian virus 40) promoter and enhancer sequences, and pGL3HSTK (plasmid without the octamer sequence) were used as positive and negative controls, respectively.

ESC Infection

We infected 293T Phoenix-Eco cells stably transfected with the GAG-polymerase and ecotropic envelope sequence with an Oct4 retroviral supernatant containing vesicular somatitis virus glycoprotein particles. The supernatant was collected, filtered (0.45-μm pores), and concentrated by a factor of 100 by centrifugation through a Centricon 30 (Millipore Corporation). ESCs were washed, treated with trypsin, and plated in culture dishes for 1 hour in complete medium to remove adherent MEFs. They were then plated at a density 106 cells per milliliter of complete medium containing LIF and MTG, in bacterial-grade Petri dishes, in the presence of 4 μg of Polybrene per milliliter. Cells were infected by two incubations with retroviral supernatant for 4 hours each at a multiplicity of infection (MOI) of 10.

Real-Time Quantitative Polymerase Chain Reaction

Primers and internal probes for the amplification of Oct4 and Hprt sequences were designed using Primer Express Software (Perkin-Elmer Applied Biosystems, Foster City, CA, http://www.appliedbiosystems.com). Polymerase chain reaction (PCR) was carried out with the ABI Prism GeneAmp 5,700 Sequence Detection System (Perkin-Elmer Applied Biosystems) using TaqMan Universal PCR Master Mix (Perkin-Elmer Applied Biosystems) containing specific primers (1.2 μM) and the specific probe (0.1 μM). For each Oct4 fraction (Oct4-GFP+ and Oct4-GFP2+), Oct4 gene expression levels were expressed with respect to levels of expression for the Hprt housekeeping gene.

Luciferase Assays

For the dual-luciferase assay, 1 × 106 EBs were electroporated, at 220 V and 500 μF, with 10 μg pGL3HSTK(6X) or control vectors in 250 μl of PBS on day 4. Cells were then cultured in EB culture medium and harvested 48 hours after transfection by lysis in 100 μl buffer. Cell lysate (20 μl) was mixed with 100 μl of luciferase assay reagent, and firefly luciferase luminescence, reflecting the activity of the pGL3HSTK(6X) promoter, was immediately quantified in a luminometer (Lumax Industries, Inc., Altoona, PA, http://www.lumaxlighting.com). This reaction was then quenched, and the Renilla luciferase reaction was simultaneously initiated by adding 100 μl of Stop and Glo reagent (Promega) to the mixture. Firefly luciferase luminescence was used to normalize the amount of protein in each cell lysate.

Gene Expression Analysis

RNA was isolated with the SV total RNA isolation system (Promega), which includes DNaseI treatment to eliminate contaminating genomic DNA. Poly (A) RNA was reverse-transcribed with Superscript II RNase H reverse transcriptase (Invitrogen). The specific primers used were as follows [22, 23]: Hprt, forward, 5′CACAGGACTAGAACACCTGC3′; reverse, 5′GCTGGTGAAAAGGACCTCT3′, 249 bp; Zfp42, forward, 5′CAGGTTCTGGAAGCGAGTTC3′; reverse, 5′TTGAAATCCAGGGAGAAACG3′, 385 bp; Brachyury, forward, 5′TCCAGGTGCTATATATTGCC3′; reverse, 5′GCTGCCTGTGAGTCATAAC3′, 947 bp; Scl/tal1, forward, 5′ATTGCACACACGGGATTCTG3′; reverse, 5′GAATTCAGGGTCTTCCTTAG3′, 321 bp; Flk1, forward, 5′CACCTGGCACTCTCCACCTTC3′; reverse, 5′GATTTCATCCCACTACCGAAAG3′, 239 bp; Fgf5, forward, 5′AAAGTCAATGGCTCCCACGAA3′; reverse, 5′CTTCAGTCTGTACTTCACTGG3′, 376 bp; Gata4, forward, 5′CCGAGCAGGAATTTGAAGAGG3′; reverse, 5′GCCTGTATGTAATGCCTGCG3′, 496 bp; CoupTFI forward 5′AGCCATCGTGCTATTCACG3′; reverse, 5′TTCTCACCAGACACGAGGTC3′, 570 bp; Dab2 forward 5′GGCAACAGGCTGAACCATTAGT3′; reverse 5′TTGGTGTCGATTTCAGAGTTTAGAT3′, 283 bp; Gata6 forward 5′GCAATGCATGCGGTCTCTAC3′; reverse 5′CTCTTGGTAGCACCAGCTCA.

The specific primers used for quantitative PCR were the following: Hprt: forward primer 5′GCCCTTGACTATAATGAGTACTTCAGG3′, reverse primer 5′TCAACAGGACTCCTCGTATTTGC3′; TaqMan probe 5′TCATTAGTGAAACTGGAAAAGCCAAATACAAAGCC3′; OCT4: forward primer 5′CTCACCCTGGGCGTTCTCT3′, reverse primer 5′AGGCCTCGAAGCGACAGA 3′, TaqMan probe 5′TGGAAAGGTGTTCAGCCAGACCACC3′; Nanog: forward primer 5′TGCTACTGAGATGCTCTGCACA3′, reverse primer 5′TGCCTTGAAGAGGCAGGTCT3′, TaqMan probe 5′AGGCTGCCTCTCCTCGCCCTTC3′.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References

Retroviral Expression in ESCs

We generated an ecotropic supernatant for the transduction of ESCs with Oct4. A vector containing a VEGF IRES and the Oct4 gene under the control of an MSCV promoter was constructed, giving levels of Oct4 gene expression proportional to GFP fluorescence intensity. A vector containing the GFP gene alone was used as a control (Fig. 1A). ESCs were reproducibly efficiently infected, as shown by flow cytometry 48 hours after infection (Fig. 1B). With an MOI of approximately 10, we obtained more than 35% GFP+ cells. We characterized the effects of Oct4 by sorting ESCs into two fractions based on GFP fluorescence intensity: one with low to moderate levels of GFP (Oct4-GFP+ ESCs, 36%) and the other with high levels of GFP (Oct4-GFP2+ ESCs, 3.5%). Two ESC populations transduced with the control vector were also sorted in a similar manner on the basis of GFP fluorescence intensity, giving the GFP+-only (45%) and GFP2+-only (7%) fractions (Fig. 1B). We could not maintain ESC lines overproducing Oct4 in continuous culture in standard conditions (feeder, serum, and LIF), because they spontaneously differentiated and lost their self-renewal capacities. We therefore transduced and sorted new ESCs for each experiment investigating the hematopoietic potential of these cells.

thumbnail image

Figure Figure 1.. Retroviral construct and GFP detection in embryonic stem cells (ESCs). (A): Retroviral construct. The retroviral vector was derived from the Migr retrovirus in which the encephalomyocarditis virus (ECMV) IRES has been replaced by the IRES from vascular endothelial cell growth factor. (B): GFP levels in ESCs. ESCs were infected with either a control retrovirus or an Oct4-GFP retrovirus. GFP levels were assessed by flow cytometry 48 hours after infection. Approximately 36% of ESCs infected with the Oct4-GFP retrovirus were GFP+, and 3.5% were GFP2+. With the GFP control retrovirus, 45% of cells were GFP+ and 7% GFP2+. Solid peak (gray): control ESCs (wild-type); solid line: control virus containing GFP alone; dotted line: Oct4-GFP retrovirus. Abbreviations: GFP, green fluorescent protein; IRES, internal ribosomal entry sequence; LTR, long terminal repeat.

Download figure to PowerPoint

Oct4 Gene Expression in ESCs During the Induction of Differentiation

Oct4 plays an important role in self-renewal and cell fate determination in ESCs and the embryo. Oct4 gene expression is downregulated by retinoic acid in ESCs upon differentiation [7, 24, [25], [26]27]. In contrast, the pattern of Oct4 gene expression during the induction of hematopoietic differentiation from ESCs has not been defined. We used real-time quantitative PCR to track changes in Oct4 levels during the differentiation of ESCs into the hematopoietic lineage. We used two methods to induce differentiation: the formation of day-3 EBs (EB3) followed by BL (three-step method) and the formation of day-6 EBs (EB6) (two-step method). The three-step method was used to study primitive bipotent progenitors (hemangioblasts), whereas the two-step method targets more committed progenitors with hematopoietic potential only. In both cases, levels of the Oct4 transcript decreased and remained at the detection threshold in BL and EB6 (Fig. 2). Endogenous Oct4 gene expression decreased, the kinetics of this decrease depending on the method used to induce differentiation (two- or three-step methods).

thumbnail image

Figure Figure 2.. Schematic representation of two methods used to induce hematopoietic differentiation. Quantitative endogenous Oct4 mRNA produced during embryonic stem cell (ESC) hematopoietic differentiation. Three-step method: 0 day: wild-type ESCs on feeder cells; 3 days: EB3; 7 days: BL; 14 days: BL-derived hematopoietic colonies. Two-step method: 6 days: EB6; 13 days: EB6-derived hematopoietic colonies. Oct4 mRNA level was determined by quantitative reverse transcription-polymerase chain reaction, using TaqMan oligonucleotides, and was expressed with respect to mRNA levels for the Hprt housekeeping gene (mean of three independent experiments). Abbreviations: BL, blast colonies; EB3, embryoid bodies in their third day of differentiation; EB6, embryoid bodies in their sixth day of differentiation; LIF, leukemia inhibitory factor; N, mRNA not detectable; VEGF, vascular endothelial cell growth factor.

Download figure to PowerPoint

We evaluated transduction efficiency by comparing total Oct4 levels during cell differentiation in control ESCs and cell lines transduced with the retroviral vector, by real-time quantitative PCR (Fig. 3A). Oct4-GFP+ and Oct4-GFP2+ ESCs had twice as many and three to four times as many Oct4 transcripts, respectively, as did ESC control cells (GFP only). During differentiation, the level of Oct4 transcription decreased in all cases, but differentiated cells derived from Oct4-GFP+ and Oct4-GFP2+ ESCs had higher Oct4 transcript levels than did control cells. Indeed, Oct4-GFP+ and Oct4-GFP2+ BL had twice as many and three to four times as many transcripts, respectively, as did control (WT, GFP+, and GFP2+-only). However, in the two-step method, retroviral expression was strongly silenced at the EB6 stage and Oct4 transcript levels were similar in cells derived from Oct4-GFP+ and Oct4-GFP2+ ESCs. During hematopoietic differentiation, weak Oct4 levels were detected (approximately 10% that in ESCs) but retroviral transcripts were consistently higher than endogenous levels at the same stage. Endogenous Oct4 transcripts were undetectable in hematopoietic cells derived from control ESCs.

thumbnail image

Figure Figure 3.. Detection of Oct4 during ESC differentiation after transduction with retrovirus. (A): Total amount of Oct4 mRNA in infected cells during hematopoietic differentiation. (a) ESCs, (b) three-step hematopoietic differentiation method, and (c) two-step hematopoietic differentiation method. Control cells correspond to the mean of wild-type, GFP+, and GFP2+-only cells. (B): Expression of Zfp42/Rex1, an Oct4 target gene, during differentiation, as assessed by reverse transcription-polymerase chain reaction. One profile representative of three independent experiments is shown. (C): Luciferase activity: Embryoid bodies on day 4 were transiently electroporated with pGL3HSTK or pGL3HSTK(6X). After 48 hours, luciferase activities were evaluated in cells corresponding to day-6 embryoid bodies and expressed in arbitrary units. The amount of protein in each cell lysate was normalized based on Renilla luciferase activity, following co-transfection with the pRL-TK vector. pGL3HSTK(6X) was three times more active than pGL3HSTK. Three independent experiments were carried out in duplicate. Abbreviations: ESC, embryonic stem cell; GFP, green fluorescent protein; ND, not determined; BL, blast colonies.

Download figure to PowerPoint

We investigated whether Oct4 mRNA was efficiently translated into protein, by investigating the expression of Zfp42, a direct target gene of Oct4, encoding the transcription factor Rex1, during hematopoietic differentiation. Rex-1 levels decrease rapidly during ESC-induced differentiation [28, 29]. Consistent with these results, we did not detect the Rex-1 transcript in EB3 control cells. After the induction of differentiation in Oct4+ ESCs, Rex-1 mRNA was detected in EBs on day 3 but not later. In contrast, Rex-1 mRNA was detected in EB6 in most experiments and in BL when hematopoietic differentiation was induced in Oct4-GFP2+ ESCs (Fig. 3B, left). We then analyzed the pattern of expression of two other target genes of Oct4: FGF4 and Nanog. These genes are regulated by a composite element containing an octamer and a sox-binding site. Transcription is regulated by a sox2-oct4 complex [10, 30, 31]. Reverse transcription-PCR showed that FGF4 mRNA was present in ESCs, its level gradually decreasing until the formation of day-4 EBs in GFP+-only control cells. In contrast, FGF4 mRNA continued to be detected in Oct4-GFP+ cells until the formation of day-5 EBs (Fig. 3B, right). Real-time quantitative PCR showed that Nanog mRNA was present in ESCs, its level decreasing during hematopoietic differentiation in GFP-only control cells. Oct4-GFP+ and Oct4-GFP2+ cells displayed a kinetic pattern similar to GFP-only control cells but with lower mRNA levels in ESCs and EB3 cells. The reason for this difference is unclear but may involve a negative effect of Oct4 on the Nanog promoter as a function of the ratio between Sox2, Oct4, and/or perhaps another transcription factor. These data suggest that in our conditions Oct4 protein is overproduced at least until EB3 formation, and this overproduction modifies the expression of target genes. We then tried to detect Oct4 transcriptional activity directly, using an artificial promoter containing the octamer binding site (pGL3HSTK(6X)) upstream from the luciferase gene. We used pGL3HSTK as a negative control (background). In EB6 derived from Oct4-GFP+ ESCs, pGL3HSTK(6X) activity was three times higher than pGL3HSTK plasmid, with a p value of .008. In contrast, no specific activity was detected in control EB6 (Fig. 3C). These observations strongly suggest that the Oct4 mRNA was efficiently translated into a functional protein during differentiation, until the formation of EB6, derived from ESCs overexpressing Oct4.

Consequences of Enforced Oct4 Production for ESC Hematopoietic Differentiation

Previous studies showed that artificially increasing Oct4 levels in ESCs led to an increase in the expression of endodermal and mesodermal markers [11]. We were interested in the effects of Oct4 on hematopoietic differentiation, so we first determined whether the overproduction of Oct4 in ESCs influenced the formation of EBs, the first step in the mesodermal differentiation of ESCs, and the content of EBs in hematopoietic progenitor cells. The cloning efficiency of ESCs in EB6 decreased with increasing Oct4 expression (Fig. 4A). Oct4-GFP+ and Oct4-GFP2+ ESCs were associated with 25% and 50% fewer EB6, respectively, than the controls. In contrast, the production of GFP alone had no effect on EB formation (data not shown). We determined the progenitor cell content of these EB6 in methylcellulose in the presence of six growth factors. Similar numbers of GFP-only and Oct4-GFP overproducing EBs were replated for hematopoietic differentiation. Myeloid progenitor content was identical in EB6 derived from control and Oct4+ ESCs but was significantly lower (40% lower) in Oct4-GFP2+ ESCs (Fig. 4B). However, this decrease was not associated with qualitative changes in the type of hematopoietic progenitor (data not shown). Thus, Oct4 overproduction modifies the fate of ESCs, because high Oct4 levels (associated with prolonged Oct4 production) slightly inhibited hematopoietic differentiation.

thumbnail image

Figure Figure 4.. Consequence of enforced Oct4 production for hematopoietic differentiation. (A): Effect of Oct4 overproduction on EB6 cloning efficiency. Oct4+ and Oct42+ ESCs displayed 25% and 50% decreases, respectively, in EB6 cloning efficiency. (B): Oct4 overproduction modified the hematopoietic potential of EB6 (expressed as a percentage of control including GFP+ and GFP2+-only, GFP = 100% ± SEM). The number of myeloid progenitors in Oct4-GFP+ cells was identical to that of the control, whereas that in Oct4-GFP2+ cells was 25% lower. (C): A doubling of Oct4 levels in ESCs resulted in an increase in the cloning efficiency of BL-CFC, whereas higher levels of Oct4 expression (increase by a factor of three to four) led to 50% of BL-CFC colony formation. (D): The hematopoietic potential of BL-CFC was affected only by the larger increase in Oct4 levels. All results are expressed with respect to control cells (mean of wild-type, GFP+-only, and GFP2+-only cells). *, significant difference compared with control. Abbreviations: BL-CFC, blast colony-forming cell; EB6, embryoid bodies in their sixth day of differentiation; ESC, embryonic stem cell; GFP, green fluorescent protein.

Download figure to PowerPoint

We then determined the consequences of Oct4 overproduction for blast colony-forming cells (BL-CFC). BL-CFC-derived colonies were obtained from EB3 in methylcellulose cultures containing VEGF. Cultures from Oct4-GFP+ ESCs yielded 2.5 times as many BL as did control cells (GFP-only) (Fig. 4C). Surprisingly, higher levels of Oct4 production (Oct4-GFP2+ ESCs) inhibited BL-CFC colony formation by 35% with respect to control GFP-only cells (Fig. 4C). The number of BL-CFC therefore differed by a factor of 3.5 between Oct4-GFP+ and Oct4-GFP2+ ESCs, suggesting that the development of this progenitor is strongly regulated by Oct4 levels. The hematopoietic progenitor content of BL did not differ markedly between control ESCs, Oct4-GFP+, and Oct4-GFP2+ ESCs, but slightly fewer (25% fewer) hematopoietic progenitors were observed for Oct4-GFP2+ ESCs (Fig. 4D). In contrast, EBs from Oct4 and control ESCs had similar secondary EB formation capacities (data not shown). BL derived from Oct4-GFP+ and Oct4-GFP2+ ESCs were unable to generate secondary colonies.

Thus, prolonged Oct4 overproduction in ESCs primarily disrupts the development of BL-CFC, having only a moderate effect on later stages of hematopoietic differentiation. High levels inhibited hematopoietic differentiation (25%; Fig. 4D), whereas moderate overproduction had no effect. We then analyzed the endothelial potential of BL. We compared the endothelial potential of Oct4/GFP BL cells with that of control cells with similar levels of GFP fluorescence. In three independent experiments, GFP-only and Oct4-GFP BL gave rise to endothelial cells in limiting dilution assays, in the same way (Fig. 5). Similar numbers of endothelial cells foci for cells with moderate and high levels of Oct4 were counted. High levels of GFP production in the control also seemed to inhibit endothelial cell differentiation, especially if plates were seeded with 4,000 BL cells. Thus, increases in Oct4 levels influenced hematopoietic progenitors in the same way (20%–25% inhibition) regardless of the method used to induce cell differentiation (three- or two-step method). No effect on endothelial differentiation from hemangioblasts was observed.

thumbnail image

Figure Figure 5.. Consequences of enforced Oct4 production for the endothelial differentiation potential of BL-CFC. No difference was observed between Oct4 and GFP control. Error bars indicate SE of three experiments performed in triplicate. Abbreviations: BL-CFC, blast colony-forming cell; GFP, green fluorescent protein.

Download figure to PowerPoint

Gene Expression Analysis During Hematopoietic Differentiation

We examined the expression profiles of two genes involved in the earliest stages of endothelial and hematopoietic development (Flk1 [VEGF-R2] and SCL/tal-1) to confirm that overproduction primarily affected early stages of differentiation. Three independent experiments were carried out, one of which is illustrated in Figure 6. We detected SCL and Flk1 expression by day 3 of differentiation (Fig. 6), consistent with previous reports [19, 32, 33]. During the differentiation of Oct4-GFP+ ESCs, Flk1 gene expression levels were higher in day-3 EBs than in the control whereas SCL expression was not affected, suggesting that Oct4 overproduction initially favors mesodermal differentiation (Fig. 6). We were unable to study Oct42+ ESCs before day 6, because too few such cells were available after retroviral infection. In EB6 and in hematopoietic colonies derived later from Oct4-GFP2+ ESCs, we observed no change in SCL or Flk1 gene expression.

thumbnail image

Figure Figure 6.. Analysis of gene expression by reverse transcription-polymerase chain reaction, during the differentiation of ES cells overproducing Oct4. Two controls (GFP+-only and GFP2+-only) were used corresponding to the GFP levels similar to those of Oct4+ and Oct42+ cells. We obtained polymerase chain reaction products of the following sizes: Hprt, 249 bp; SCL/tal-1, 321 bp; Flk1, 239 bp; Gata4, 496 bp; Fgf5, 376 bp; Brachyury, 947 bp; and Gata6, coupTfI, 570 bp. Results from one of three independent experiments are shown. Abbreviations: BL, blast colonies; EB, embryoid body; ES, embryonic stem; GFP, green fluorescent protein.

Download figure to PowerPoint

We analyzed the expression profile of marker genes for the various germ layers to confirm this effect on mesodermal differentiation. We detected no change in the expression pattern of genes restricted to the primitive ectoderm and parietal/visceral endoderm germ layers (Fig. 6) [34, 35]. Fgf5, Gata4, Gata6, and Coup-tfI were induced during hematopoietic differentiation, with no major difference between GFP control and Oct4-GFP cells. In contrast, Brachyury, the mesodermal marker, persisted later in differentiation in Oct4-overproducing ESCs [36, 37]. Brachyury mRNA was more abundant in EB3 derived from Oct4-GFP+ ESCs than in EB3 derived from the control. However, the kinetic profile of Brachyury expression did not differ markedly between the two types of cell; this mRNA was not detected in day-6 EBs derived from either type of ESC. In contrast, during the differentiation of Oct4-GFP2+ ESCs, Brachyury gene expression persisted in EB6 and BL (Fig. 6). These expression kinetics for brachyury suggest that Oct4 can slow down the differentiation of mesoderm cells.

Thus, the marked differences in BL-CFC colony formation between Oct4-GFP+ and Oct4-GFP2+ cells and controls may result from differences in differentiation kinetics. We therefore investigated the pattern of BL-CFC development from EBs. It has been shown that the potential to form BL-CFC is limited to a very short period of EB development. Kinetic analysis of the developmental potential of BL-CFC confirmed that EBs derived from Oct4-GFP+ ESCs gave rise to two to three times as many BL as the control, whereas Oct4-GFP2+ ESCs gave rise to only half as many BL as the control (Fig. 7). No marked differences in the kinetics of BL-CFC development were found between the three types of ESCs, although EBs began to form slightly later (up to day 3.5) for Oct4-GFP+ ESCs. In contrast, BL-CFC developed from Oct4-GFP2+ ESCs with kinetics similar to the control. Thus, low levels of Oct4 overproduction slightly prolonged the window of development of BL-CFC (day 3.8, 1 GFP-only control BL against 11 Oct4-GFP+ BL), whereas higher levels of overproduction had no effect. Finally, we showed that the number of secondary EBs was strictly similar in the different conditions (Oct4-GFP+ or Oct4-GFP2+ vs. GFP-only control).

thumbnail image

Figure Figure 7.. Analysis of the kinetics of blast colony-forming cell development. EBs were cultured for various times (2–4 days), dissociated, and replated on semisolid medium containing vascular endothelial cell growth factor. Blast colony (BL) and secondary EBs were counted. The number of BL was expressed per 105 EBs (three independent experiments performed in triplicate). Oct4-GFP+ and Oct4-GFP2+ were compared with GFP control (mixture of GFP+-only and GFP2+-only cells). Abbreviations: EB, embryoid body; GFP, green fluorescent protein.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References

In this report, we studied the effects of Oct4 overproduction on hematopoietic differentiation from ESCs. Oct4 overproduction was induced using a bicistronic Oct4 and eGFP-containing retroviral vector with high infection efficiency in ESCs. This made it possible to select cells with two different eGFP levels, corresponding to a doubling and tripling of Oct4 production. As previously reported, ESCs overproducing Oct4 lost their self-renewal capacity in the presence of LIF and underwent spontaneous differentiation. We had to infect new ESCs for each differentiation experiment because no system for inducing Oct4 overproduction was available. The retroviral vector used gave high levels of Oct4 production in ESCs, but Oct4 levels fell continually during differentiation. This decrease may be due to silencing of the MSCV promoter during differentiation, by methylation or by more profound changes to chromatin [38, 39]. However, Oct4 transcripts were detected in BL, in EB6, and in hematopoietic colonies during differentiation but at only 30%, 10%, and 5%, respectively, of the level in ESCs. In these conditions, retroviral transcripts were not constitutively produced, but they were consistently produced in larger amounts than endogenous Oct4 transcripts at the same stage. Retroviral Oct4 production during differentiation seemed to decrease with kinetics similar to those reported for the endogenous.

In vitro experiments have shown that increasing endogenous Oct4 levels in ESCs by more than 50% leads to the expression of endodermal and mesodermal markers. These data were obtained with ESCs maintained in culture with LIF [11]. In our in vitro conditions, the overproduction of exogenous Oct4 modified mesodermal marker expression but had no effect on endodermal marker expression. We focused on mesodermal differentiation and, more precisely, on hematopoietic differentiation. We observed that the effects on progenitors of Oct4 overproduction in ESCs and of the maintenance of Oct4 production in ESC-derived hematopoietic cells differed according to Oct4 levels and the approach used to induce hematopoietic differentiation. With the two-step method, which allowed the development of EB6-derived hematopoietic progenitor colonies, Oct4 overproduction led to dose-dependent inhibition of progenitors. This inhibition was due mainly to a decline in the cloning efficiency of ESCs in EB formation. A decrease in number of day-6 EBs was observed up to 50% in Oct4-GFP2+ ESCs. In contrast, the hematopoietic progenitor content of EB6-derived colonies was modified only in Oct4-GFP2+ ESCs (40%), decreasing ESC Oct4-GFP2+ hematopoietic potential by 70%. The main effects of this enforced expression on hematopoiesis relate to the impact of Oct4 on the differentiation fate of ESCs, as demonstrated by the loss of EB6 colony formation. Decreases in the hematopoietic progenitor content of EB6 may have resulted from the initially high level of Oct4 or the sustained production of Oct4 during differentiation. These data suggest that a threshold was necessary to affect hematopoietic progenitors potency and hematopoietic colonies. Moderate Oct4 overproduction in ESCs inhibited EB formation. Day-3 to day-6 EBs were affected in the same way during EB development (a 20%–30% decrease). We hypothesized that defects observed in EB formation could be a result of dysregulation of survival signals regulated by Oct4 and perhaps combined with other transcription factors [40]. We could not exclude a maintenance of ESCs in an undifferentiated cell fate [41].

With the technique by which ESCs were induced to undergo differentiation into hematopoietic colonies via the VEGF-driven generation of BL from EB3, marked differences were observed, depending on the level of Oct4 overproduction. A doubling of Oct4 production markedly increased BL formation, whereas BL formation was inhibited at higher levels of Oct4 production. BL-CFC, which give rise to BL, are mesenchymatous progenitors with a bipotent endothelial and hematopoietic potential theoretically corresponding to a hemangioblast [42]. Oct4 overproduction had a profound effect on the generation of BL-CFC but not on the hematopoietic progenitor of BL-CFC, which was only 20% lower in BL derived from ESCs with high Oct4 levels (Fig. 4D). The mechanisms by which Oct4 modifies this differentiation process are unknown. Our data suggest that Oct4 overproduction affects early mesodermal precursors in a dose-dependent manner but that the final effect on hematopoietic progenitors is the same in all cases. A model of hematopoietic and endothelial development has suggested that a common precursor gives rise to hematopoietic progenitors and hemangioblasts [43, 44]. This model may explain the increase in number of BL and the decrease in number of EB6 when Oct4 is produced at low levels. In early progenitors, Oct4 acted as a transcriptional regulator of the expression/repression of key genes. We suggest that low levels of Oct4 inhibit genes associated with the hematopoietic lineage whereas the expression of genes associated with both the endothelial and hematopoietic lineages was maintained. Thus, monopotent progenitors (EB6) decreased in number and bipotent hemangioblasts increased in number. In these conditions, hematopoietic potency was maintained. In contrast, high levels of Oct4 seem to inhibit both specific hematopoietic genes and genes involved in the development of both lineages (endothelial/hematopoietic). The number of EB6 decreased strongly with decreases in hematopoietic potency (40%). However, the number of BL-CFC and their hematopoietic potency also decreased (25%), with no effect on the endothelial lineage. Nanog is probably one of the genes regulated in this context. The Nanog promoter contains an octamer-binding sequence that can bind Oct4 protein in cooperation with Sox2. Little is known about the regulation of Nanog, but in our study increases in Oct4 production led to the inhibition of Nanog transcription in ESCs and EB3 and the induction of differentiation. We suggest that Oct4 induces differentiation (mesodermal and endodermal) initially by inhibiting the transcription of Nanog. Nanog may regulate differentiation by repressing the transcription of genes promoting differentiation (Gata6), and the level of this protein is crucial. Oct4 has been reported to affect proliferation. In liquid culture, we observed similar numbers of EB3 control, Oct4-GFP+, and Oct4-GFP2+ cells. We concluded that changes in proliferation were not responsible for the observed differences in BL number. No real difference was observed in the size of colonies. However, an increase in the amount of FLK1 mRNA, encoding the VEGF receptor, was observed in EB3 derived from Oct4-GFP+ ESCs (Fig. 6). This increase may improve the response to VEGF, accounting for the higher cloning efficiency of BL-CFC. Alternatively, the increase in FLK1 mRNA levels in EB3 may result from greater specification of ESCs toward a mesodermal fate. Furthermore, FLK1+ cells isolated from EB3 had a BL potential similar to those from the entire EB population (data not shown). In contrast, as previously reported, FLK1 GFP or Oct4 cells were unable to generate BL [45]. No modification of FLK1 cells was found that could lead to the generation of BL or the secretion of factors with the capacity to modulate FLK1+ potential (data not shown). In addition, although EB3 secreted VEGF into the blast cell culture medium, the overproduction of Oct4 in these cells did not lead to BL formation unless exogenous VEGF was provided.

We found that the window of BL Oct4+ generation was extended (lasting until day 3.5), suggesting a change in embryonic lineage determination (Fig. 7). These data suggest that BL-CFC in Oct4-GFP+ cell culture should be more than bipotent. Higher levels of Oct4 overproduction may inhibit BL-CFC generation by profoundly affecting mesodermal differentiation. During the differentiation of ESCs, Brachyury, an early mesoderm marker absent from ESCs, was detected in EB3 but not in BL [22]. In contrast, when ESCs producing large amounts of Oct4 were induced to differentiate, Brachyury continued to be detected in BL, which also expressed FLK1 and SCL/Tal 1. Thus, the gene expression profile of BL is identical to that of the transitional colony, defined as an early EB-derived colony [33]. Transitional colonies contain cells that are already committed to the erythroid and endothelial differentiation pathways but also include mesodermal cells not yet committed. It is believed that, during mesoderm differentiation, there is a continuum between transitional cells which may represent a mesodermal stage of development and BL-CFC which correspond to a later stage with commitment to the hematopoietic and endothelial lineages. This suggests that high levels of Oct4 production markedly slow down mesodermal differentiation, possibly even impairing this differentiation; we did not observe several waves of colonies in the presence of VEGF. Thus, BL derived from Oct4-GFP2+ ESCs may contain both mesodermal and post-mesodermal cells, whereas normal BL are purely post-mesodermal. This may also explain the lower hematopoietic progenitor content of these BL. If we wish to understand the direct effects of Oct4 on hematopoietic differentiation, we will need to produce Oct4 directly in HSCs and progenitors. An approach based on a system for the inducible expression of the Oct4 gene at specific stages of differentiation would be extremely useful [46].

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References

Our study showed that the effects of Oct4 overproduction on mesodermal progenitors from ESCs differ according to the amount of Oct4 produced. High levels of overproduction markedly inhibit hematopoietic differentiation, whereas low levels of Oct4 production may increase the number of bipotent progenitors (BL-CFC) without affecting potency. In cells producing small amounts of Oct4, prolongation of the period of BL-CFC development should be seen as indicating a much larger change in potency. It would be interesting to determine which type of potency is affected.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References

This paper is dedicated to F. Le Pesteur who died in 2003. We thank Philippe Rameau and Yann Lecluse for cell sorting experiments, Isabelle Godin for helpful criticism, and Lorna Saint Ange for editing. This work was supported by grants from the INSERM and la Ligue Nationale contre le Cancer (équipe labelisée 2004). V.C.-C. received a grant from the Fondation sur la Leucémie.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References