SEARCH

SEARCH BY CITATION

Keywords:

  • Embryonic stem cells;
  • Differentiation;
  • LIF;
  • STAT3;
  • Self-renewal

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

A unique and complex signaling network allows ESCs to undergo extended proliferation in vitro, while maintaining their capacity for multilineage differentiation. Genuine ESC identity can only be maintained when both self-renewal and suppression of differentiation are active and balanced. Here, we identify Pramel7 (preferentially expressed antigen in melanoma-like 7) as a novel factor crucial for maintenance of pluripotency and leukemia inhibitory factor (LIF)-mediated self-renewal in ESCs. In vivo, Pramel7 expression was exclusively found in the pluripotent pools of cells, namely, the central part of the morula and the inner cell mass of the blastocyst. Ablation of Pramel7 induced ESC differentiation, whereas its overexpression was sufficient to support long-term self-renewal in the absence of exogenous LIF. Furthermore, Pramel7 overexpression suppressed differentiation in ESCs in vitro and in vivo. This process was reversible, as on transgene excision cells reverted to a LIF-dependent state and regained their capacity to participate in the formation of chimeric mice. Molecularly, LIF directly controls Pramel7 expression, involving both STAT3-dependent transcriptional regulation and PI3K-dependent phosphorylation of glycogen synthase kinase 3β. Pramel7 expression in turn confers constitutive self-renewal and prevents differentiation through inactivation of extracellular signal-regulated kinase phosphorylation. Accordingly, knockdown of Pramel7 promotes ESC differentiation in presence of LIF and even on forced STAT3-activation. Thus, Pramel7 represents a central and essential factor in the signaling network regulating pluripotency and self-renewal in ESCs. STEM CELLS 2011;29:474–485


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

ESCs are derived from the inner cell mass (ICM) of blastocyst [1, 2] and represent an important tool for the study of early embryonic development and the pluripotent state, mostly because of their two distinctive properties, that is, their ability to undergo indefinite mitotic self-renewal and to differentiate into a range of specialized cell types. A tightly balanced interplay between different pathways is necessary to promote self-renewal in ESCs. Some of these pathways act through the repression of factors that initiate differentiation programs [3–7]. STAT3, OCT3/4, SOX2, and Nanog are transcription factors that regulate various aspects of ESC fate and safeguard the maintenance of the pluripotent state [3, 5].

Derivation and maintenance of murine ESCs were originally achieved by using feeders or the cytokine leukemia inhibitory factor (LIF) in combination with fetal calf serum or the growth factor bone morphogenetic protein. LIF acts through the leukemia inhibitory factor receptor/gp130 complex to maintain pluripotency [8, 9]. Cultivation of ESCs on Lif-deficient fibroblasts leads to differentiation, indicating that they mostly provide LIF [10]. LIF-independent maintenance of mouse ESCs with retention of pluripotency (adult chimerism) has been previously described for cell lines, which overexpressed Nanog or KLF2 [11, 12]. In vitro, the overexpression of PEM/RHOX5 also maintains pluripotency without LIF, even though contribution to chimera has not yet been proven [13, 14]. Nowadays, it is possible to bypass LIF, feeders, and serum requirement by using two inhibitors (2i conditions) which block mitogen-activated protein kinase (extracellular signal-regulated kinases [ERKs]) and glycogen synthase kinase 3β (GSK3β) [15]. Interestingly, maximal self-renewal is obtained by combination of LIF and 2i confirming LIF/STAT3 signaling as an essential component of self-renewal in ESCs. Despite the importance of this pathway, STAT3 downstream target genes have remained elusive. In a recently performed microarray study, we found Pramel6 and Pramel7 (preferentially expressed antigen in ESCs upon conditional in melanoma like 6 and 7) strongly upregulated in ESCs upon conditional overexpression of STAT3 [14]. These findings indicate a potential role of these genes in the stabilization of ESCs. In this work, we aimed at the functional characterization of Pramel6 and particularly of Pramel7 in ESCs. Our results demonstrate that Pramel7 is a new direct STAT3 target gene, fundamental for the LIF-mediated maintenance of pluripotency and for the inhibition of differentiation.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Cell Culture

Medium for E14 129/Ola and transgenic ESCs (complete medium [CM]-medium): Glasgow minimal essential (Sigma, Buchs, Switzerland, www.sigmaaldrich.com/switzerland-schweiz.html), 100 mM sodium pyruvate (Sigma), 10% fetal bovine serum (Gibco, Invitrogen, Basel, Switzerland, www.invitrogen.com), 50 mM β-mercaptoethanol (Gibco), 1× minimal essential medium nonessential amino acids (Gibco), 200 mM L-glutamine. Medium supplemented or not with 1,000 U/ml ESGRO murine LIF (Millipore, Chemikon, Zug, Switzerland, www.millipore.com). N2B27-medium: DMEM/F-12 (Invitrogen), Neurobasal (Invitrogen), 50 mM β-mercaptoethanol (Gibco), 200 mM L-glutamine, N2-Supplement 100× (Invitrogen), B27-Supplement 50× (Invitrogen). Medium supplemented with two inhibitors: 3 μM CHIR99021 (Stemgent, Cambridge, MA, www.stemgent.com) and 1 μM PD0325901 (Stemgent). Lif knockout fibroblast: Lif +/− mice [16] were mated and at E14.5 the fetuses were isolated. Heads and placentas were used for genotyping, whereas the rest of the embryo was trypsinized and cultured. Lif −/− fibroblasts were expanded and treated with mitomycin-C (10 μg/μl). Neural-differentiation medium: DMEM/F-12 (Invitrogen), B27-Supplement 50× (Invitrogen), N2-Supplement 100× (Invitrogen). For PI3K inhibition, 5 μM LY294002 were added to the N2B27 medium not supplemented with CHIR99021 and PD0325901.

Immunohistochemical Analyses

E14, transgenic, and recombined ESCs were cultivated for 11 days with CM-medium with or without LIF on Lif knockout feeders. At day 11, all cell lines were analyzed for OCT3/4 and stage-specific embryonic antigen-1 (SSEA-1) expression. For the serum- and feeder-free experiment, the same cells were cultivated on gelatinized plates for 9 days. N2B27 medium was supplemented with either both or one inhibitor or only LIF. ESCs were fixed in 4% paraformaldehyde and incubated overnight at 4°C with primary antibodies against OCT3/4 (rabbit anti OCT3/4, Santa Cruz Biotechnology, Santa Cruz, CA, www.scbt.com) and SSEA-1 (Mouse mAb, Millipore). Secondary fluorescence-labeled antibodies were used for detection (anti rabbit Alexa Fluor 594 and anti mouse Alexa Fluor 488, Molecular Probes, Invitrogen). Nuclei of the cells were counterstained with DAPI (Roche, Basel, Switzerland, www.roche.ch).

Pramel7 Knockdown

shRNA vectors: four specific shRNA vectors against Pramel7 cloned in pGFP-V-RS vector (Origene, Rockville, MD, www.origene.com) and one negative control shRNA pGFP-V-RS vector. Sequences of shRNA are listed in Supporting Information Table 2. ESCs were transiently lipofected with either the shRNA constructs against Pramel7 or with the control vector by using FuGENE HD Transfection Reagent (Roche) and selected with puromycin. Transfection efficiency was monitored by EGFP fluorescence and Pramel7 knockdown was analyzed by real-time polymerase chain reaction (PCR). Both E14 wild-type (wt) and STAT3MER transgenic cells were cultivated on feeders with CM-medium either supplemented with LIF (E14 cells) or with hydroxy-tamoxifen (OHT; STAT3MER cells). STAT3MER cells were analyzed for alkaline phosphatase (AP) expression at day 8, respectively 4 of the knockdown. For STAT3MER cells, all the AP-positive colonies present in the 35-mm dishes were counted. For E14 cells, colonies with greater than 80% staining were classified as “undifferentiated,” 20%–80% staining as “mixed,” and less than 20% as “differentiated.”

LIF Induction Experiment

wt ESCs, established under feeder- and serum-free conditions, were cultivated on gelatinized 35-mm dishes in N2B27 + 2i medium. Cells were then incubated for 4 hours in N2B27 + 2i, N2B27 + PD0325901, N2B27 + CHIR99021, N2B27 + LY294002 or in N2B27 without inhibitors. After 4-hour incubation, LIF was added to the media. Total RNA was extracted after 30 minutes, 1 hour, 3 hours, and 5 hours of LIF incubation. Reverse transcription and real-time PCR were performed.

Part of Material and Methods is provided in the Supporting Information.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Characterization of Pramel6 and Pramel7 Genes

In the mouse, Pramel6 and Pramel7 cluster on chromosome 2(D) in opposite orientations (Supporting Information Fig. 1A). The amino acid composition of both proteins is very similar (Supporting Information Fig. 1B). Search for recognizable domains in the open reading frame of both genes using SMART (http://smart.embl-heidelberg.de) and myHits (http://myhits.isb-sib.ch/cgi-bin/motif_scan) revealed the presence of conserved leucine-rich repeats (LRRs), which usually participate in protein-protein interactions. The presence of these types of domains and the absence of conserved domains typical for transcriptional factors suggests that the Pramel family might not directly regulate gene transcription but rather act via protein-protein interaction.

Pramel6 and Pramel7 Expression Is Restricted to the Late Preimplantation and is Silenced in the Early Postimplantation Embryonic Stages

Whole mount in situ hybridization studies in the preimplantation embryos showed a homogeneous expression of Pramel6 in all cells of the morula and the blastocyst, whereas Pramel7 mRNA distribution was restricted to the interior part of the morula and the ICM of the blastocyst (Supporting Information Fig. 1C). Gene expression analyses of early postimplantation embryos (E6.5) revealed that at this developmental stage Pramel6 and Pramel7 genes are silenced (Fig. 1A). At this stage, the embryos expressed, as expected, both Nanog and DPPA3. The Nanog expression was restricted to the embryos, whereas DPPA3 was detected in the embryo as well as in the decidua (Fig. 1A).

thumbnail image

Figure 1. Endogenous expression of Pramel6 and Pramel7 and characterization of the transgenic clones. (A): Expression of Pramel6, Pramel7, Dppa3, and Nanog in early postimplantation embryos (E6.5). Pramel6 and Pramel7 expressions are absent in both decidual and embryonic tissue but present in ESCs. Negative control (−). (B): Real-time polymerase chain reaction (PCR) analysis of wild-type (wt) ESCs cultivated in presence of LIF and feeders (MEF) or in presence of the extracellular signal-regulated kinases and glycogen synthase kinase 3 inhibitors (2i). Expression levels of Pramel7, Pramel6, and Oct3/4 are compared with differentiated ESCs. (C, E): Representative morphology of Pramel7 and Pramel6 overexpressing cells and Cre-reverted (EGFP expressing) ESCs cultivated in presence of LIF and feeders. (D, F): Real-time PCR analyses of known pluripotency-related genes in Pramel7, Pramel6, and in Cre-reverted ESCs. Data were normalized to the expression level in the wt ESCs. Abbreviations: D, decidual tissue; E, embryonic tissue; ES, ESC; LIF, leukemia inhibitory factor; MEF, mouse embryonic fibroblasts.

Download figure to PowerPoint

In vitro, Pramel6 and Pramel7 are constitutively expressed in ESCs (Fig. 1B) independent of whether the cells are cultivated with serum, feeders, and LIF, or with N2B27 + 2i (serum free, feeder free, and LIF free). Expression of both genes disappears as soon as the cells differentiate. In accordance with our previous results [14], the presence of LIF in the medium resulted in an increased expression of both genes, presumably because of a direct activation of STAT3 (Fig. 1B).

Taken together, these observations suggest a possible role of both Pramel6 and Pramel7 in the maintenance of pluripotency in vivo as well as in vitro.

Pramel7 but Not Pramel6 Overexpression in ESCs Induces General Upregulation of Known Pluripotency-Related Genes

We generated ESCs conditionally overexpressing LoxP-flanked open reading frames of Pramel6 or Pramel7, which can be excised by Cre-recombinase simultaneously bringing egfp under the CAG promoter (Supporting Information Fig. 2A). Overexpression experiments were performed in E14 129/Ola (E14) ESCs and, if not specified, they were cultivated in presence of feeders and serum (referred as CM). All clones showed a high expression of the transgene, which reverted to wt levels once recombined (Supporting Information Fig. 2B and 2C). All transgenic clones showed the classical morphology of pluripotent ESCs (Fig. 1C, 1E, and Supporting Information Fig. 2E) and expressed the pluripotency markers AP, SSEA-1, and OCT3/4 (data not shown). On the transcriptional level, as assessed by real-time PCR, the overexpression of Pramel7 induced a slight increase of most of the pluripotency genes (Fig. 1D). Accordingly, on Cre-recombination the expression of Nanog and OCT3/4 reverted to wt levels in combination with an increase in fibroblast growth factor 5 (FGF5) (Fig. 1D). Overexpression of Pramel6 did not induce significant changes on the expression of the investigated genes (Fig. 1F). Nanog overexpression did not influence the transcriptional levels of Pramel7, indicating that Pramel7 is not a target of Nanog (Supporting Information Fig. 2D).

Overexpression of Pramel7 Liberates ESCs from LIF Dependence and Maintains Pluripotency

We further analyzed the ability of ESCs overexpressing either Pramel6 or Pramel7 to be propagated clonally on Lif-knockout feeders [16] without LIF in the medium. After 11 days, E14 ESCs exhibited widespread differentiation and no longer expressed the pluripotency markers OCT3/4, SSEA-1, and AP (Fig. 2A, 2B and Supporting Information Fig. 3A). In contrast, Pramel7-overexpressing cells (referred as Pramel7 cells) were undistinguishable from Nanog-overexpressing cells (Supporting Information Fig. 3A) and only less than 10% of the colonies showed no AP activity (Fig. 2C). Accordingly, Cre-recombined Pramel7 ESCs (referred as Cre-Pramel7 cells) behaved like wt ESCs and exhibited more than 80% of differentiated colonies (Fig. 2C). In contrast, only about one-third of the colonies in Pramel6-overexpressing clones remained completely undifferentiated (Fig. 2C and Supporting Information Fig. 3A). Our data reveals that overexpression of Pramel7, but not of Pramel6, is sufficient to liberate ESCs from LIF dependence and to promote self-renewal.

thumbnail image

Figure 2. Characterization of Pramel7 and Pramel6 transgenic ESCs. (A): Immunostaining for OCT3/4 and SSEA-1 in E14 wt, Pramel6 and Pramel7 overexpressing ESCs cultivated for 11 days in absence of leukemia inhibitory factor (LIF) on a layer of Lif knockout feeders. Magnification: ×20. (B, C): Alkaline phosphatase (AP) staining and quantification of AP-positive colonies after 11 days in absence of LIF (100 colonies were counted). (D): Contribution of Pramel7 Cre-deleted cells to embryo and adult chimeric animals. Chimerism was assessed by green fluorescence. Wild-type pup (EGFP negative), EGFP-positive Cre-reverted Pramel7 pup and the adult chimera are shown. Abbreviations: DAPI, 4′,6-diamidino-2-phenylindole; SSEA-1, stage-specific embryonic antigen-1.

Download figure to PowerPoint

Finally, we investigated whether cells maintained in a self-renewing pluripotent state exclusively through overexpression of Pramel7 were retaining their embryo colonization capacity. Excision of Pramel7 restored LIF dependence and these cells contributed to the generation of healthy live chimeras upon morula aggregation (Fig. 2D and Supporting Information Table 3).

Pramel7-Overexpressing ESCs Fail to Differentiate In Vitro

To investigate the role of Pramel7 in ESC differentiation, embryoid bodies were generated with Pramel7, Cre-Pramel7, and Nanog-overexpressing ESCs. As assessed by reverse transcription and PCR, expression of pluripotency markers, such as OCT3/4, Nanog, and REX1, was persisting after 10 days in Pramel7 cells as well as in the Nanog-overexpressing cells, whereas it was completely abolished in Cre-revertants (Fig. 3A). To further assess the capacity of Pramel7 cells to undergo defined differentiation, cells were cultivated with neural differentiation medium. E14 cells differentiated, whereas the majority of Pramel7 cells showed a large number of cells with pyknotic nuclei (Fig. 3B, 3C) indicating a high degree of cell death.

thumbnail image

Figure 3. Pramel7 overexpression results in differentiation defects in vitro and in vivo. (A): Expression levels of pluripotency genes (Pramel7, Nanog, Oct3/4, Pem/Rhox5, Stat3, Sox2, Rex1, Dppa3), endoderm (Gata-6), primitive ectoderm (Fgf5), and mesoderm (Brachyury-T) were measured by semiquantitative RT-PCR in ESCs (control) and in embryoid bodies generated from Pramel7-overexpressing cells, Pramel7-Cre revertants, and Nanog-overexpressing cells. Number of PCR cycles is annotated. (B, C): In vitro neural differentiation of Pramel7-overexpressing cells and parental E14 cells. Pramel7-overexpressing clones exhibited an extensive number of cells with pyknotic nuclei, as shown by DAPI staining. Magnification: ×20. Number of live and dead cells after 4 days of neural differentiation: Cells which showed condensed nuclei were defined as dead cells, whereas normal nuclei were considered as live cells. (D): Western blotting for phospho-Erk after LIF stimulation of wt and Pramel7-overexpressing ESCs for 10, 30, or 60 minutes reveals that Pramel7 maintains pluripotency through gradual suppression of pERK. (E): Morula aggregation of Cre-reverted Pramel7 cells (EGFP positive) and Pramel7-overexpressing cells (RFP labeled). Pramel7-overexpressing cells do not correctly take part to the development, whereas Cre-reverted Pramel7 cells integrate into the developing embryo and at E5.5 they are part of the epiblast. (F–H): Teratoma formation reveals impaired in vivo differentiation potential of Pramel7-overexpressing cells. Efficiency of teratoma formation in E14 wt ESCs, Nanog, Pem/Rhox5, Pramel6, and Pramel7-overexpressing cells was assessed by number of teratomas generated (F) and tumor volume (mm3) measurement (G), as well as by degree of tumor differentiation (H) analyzed by immunostaining for SMA, TuJ1, and GFAP. Magnification: ×10. Abbreviations: DAPI, 4′,6-diamidino-2-phenylindole; ERK, extracellular signal-regulated kinase; GFAP, glial fibrillary acid protein; LIF, leukemia inhibitory factor; MEF, mouse embryonic fibroblasts; pERK, phosphorylated ERK; SMA, smooth muscle actin; TuJ1, βIII-tubulin.

Download figure to PowerPoint

Pramel7 Maintains Pluripotency In Vitro Through Gradual Suppression of ERK Phosphorylation

ERK phosphorylation is known to be an early signaling event required for the differentiation of ESCs, and it has been proposed that the balance between LIF-induced activation of STAT3 and ERK may determine the efficiency of self-renewal and thereby influence stem cell fate [17, 18]. To further understand the mechanism underlying Pramel7-mediated self-renewal, we investigated by Western blotting activation of ERK in E14, Pramel6, Pramel7, and in Cre-recombined ESCs on LIF stimulation. Twenty-four hours LIF-depleted wt, Pramel6-, and Cre-Pramel7 cells showed progressive phosphorylation of ERK (pERK) on LIF induction (Supporting Information Fig. 3B). In Pramel7 cells, however, pERK was only detectable 10 minutes after LIF stimulation and drastically decreased after 30 minutes (Fig. 3D). Taken together, this data suggests that Pramel7 overexpression prevents ESCs from differentiation by promoting ERK dephosphorylation.

Pramel7 Cells Are Unable to Form Teratomas and to Contribute to Embryo Development

To test the capacity of Pramel7 cells to contribute to embryo development, we genetically labeled these cells with red fluorescent protein (RFP) and performed morula aggregations. After being transferred into the uterus of foster mothers, at E5.5 the embryos were isolated and analyzed. Embryos aggregated with Pramel7-RFP cells exhibited malformations (Fig. 3E and Supporting Information Table 3), indicating that Pramel7 cells do not enter normal development.

We further investigated the ability of Pramel6 and Pramel7 cells to form teratomas. We transplanted 106 ESCs subcutaneously into immunoincompetent nonobese diabetic/severe combined immunodeficiency mice (Fig. 3F). After 3 weeks, wt and Cre-Pramel7 ESCs produced teratomas of similar size (Fig. 3G and Supporting Information Fig. 4A) containing derivatives of all three germ layers (Supporting Information Fig. 4B). Pramel6 cells generated similar teratomas but with smaller volume than the ones generated by the control clones (Fig. 3G and Supporting Information Fig. 4A, 4B). Intriguingly, 14 independent injections with two different Pramel7 ESC-clones overexpressing similar amounts of Pramel7 never generated teratomas, even after 2 months incubation. Seeing that no previous publication describes a similar behavior of pluripotent cells, we tested the potential of teratoma formation of ESCs overexpressing Nanog or PEM/RHOX5. A total of 100% of the injections with Nanog-overexpressing ESCs produced teratomas undistinguishable in size (Fig. 3G) and histological composition from the ones generated from wt and Cre-revertant cells (Fig. 3H, Supporting Information Fig. 4A and 4B). This data indicates that Nanog overexpression is not sufficient to prevent differentiation in the context of teratoma.

ESCs-overexpressing PEM/RHOX5 were previously shown to form teratomas containing undifferentiated cells [13]. Interestingly, 75% of the injections with PEM/RHOX5-overexpressing cells failed to generate teratomas, in a similar way to the Pramel7 cells (Fig. 3G). Nevertheless, the only teratoma isolated contained differentiated tissue similar to the wt and Nanog cells (Fig. 3H). In summary, overexpression of Pramel7 impairs the capacity of ESCs to generate teratomas and the inability to differentiate probably causes the death of the cells.

Pramel7 Is Necessary for LIF- and STAT3- Dependent Maintenance of Pluripotency in ESCs

To investigate whether Pramel7 is required for maintenance of pluripotency in ESCs, we performed knockdown experiments in E14 wt ESCs cultivated in the presence of LIF. Knockdown was achieved by transient ESCs transfection with a cocktail of four shRNA-containing vectors specific for Pramel7. Four days after transfection, Pramel7 mRNA was completely silenced, although the efficiency decreased 4–5 times in comparison with the control cells at day 6 (data not shown). Despite the presence of LIF, knockdown of Pramel7 induced loss of AP activity and differentiation (Fig. 4A, 4B), whereas control cells did not. The data indicates that LIF-mediated self-renewal in ESCs depends on Pramel7 expression.

thumbnail image

Figure 4. Pramel7 expression is indispensable for LIF/STAT3-dependent maintenance of pluripotency and is a direct STAT3 target gene. (A, B): Representative appearance of AP-stained colonies (A) after Pramel7 knockdown in E14 ESCs in presence of LIF. Control (E14 mock), Pramel7 knockdown (E14 shPramel7). Quantification of AP-positive colonies (B) after Pramel7 knockdown in E14 wt cells. Colonies that showed more than 80% positive cells were identified as “undifferentiated,” 20%–80% as “mixed,” and less than 20% as “differentiated”. (C): Pramel7 knockdown in STAT3MER cells cultivated in presence of 4-hydroxy-tamoxifen. Total number of AP-positive colonies is presented. (D): Time-pulse assay with LIF in E14 wt ESCs. Phosphorylation levels of STAT3 (pSTAT3) were monitored by Western blot, whereas the amount of Pramel7 mRNA was measured by real time-polymerase chain reaction (PCR). (E): Real-time PCR analysis of Pramel7 mRNA expression in wt and Stat3 null ESCs cells after LIF stimulation for 1 or 24 hours. (F): STAT3 directly transcribes Pramel7 gene. Real-time PCR analysis of Pramel7 mRNA expression in E14 wt ESCs deprived from LIF for 24 hours followed by LIF stimulation for 1 or 3 hours in the presence or absence of cycloheximide (50 μg/ml). Abbreviations: AP, alkaline phosphatase; LIF, leukemia inhibitory factor.

Download figure to PowerPoint

To further assess whether self-renewal induced by STAT3 overexpression also depends on Pramel7, we exploited STAT3MER ESCs, which conditionally express a STAT3 fusion protein with a mutated estrogen receptor. These cells maintain pluripotency in the absence of LIF by the sole activation of STAT3MER with 4-OHT [19] and were shown to upregulate Pramel7 [14]. Surprisingly, STAT3MER activation failed to promote stem cell maintenance in Pramel7 knockdown cells, resulting in a drastic decrease in the total number of AP-expressing colonies (Fig. 4C). These results clearly indicate that the expression of Pramel7 is essential for both LIF- and STAT3-mediated maintenance of pluripotency.

Pramel7 Is a Novel Direct Downstream Target of STAT3 in the LIF/STAT3 Pathway

On the basis of our previous [14] and recent results, we assessed whether Pramel7 might represent a novel effector of the LIF/STAT3-pathway. We therefore first monitored STAT3 phosphorylation and Pramel7 mRNA expression in response to three cycles of 10 minutes of LIF incubation. Depletion of LIF for 24 hours (t0) and successive incubation with LIF-containing medium for 10 minutes (t1) induced in E14 cells a fast phosphorylation of STAT3 (pSTAT3) as assessed by Western blotting (Fig. 4D). Concomitantly, mRNA levels of Pramel7 strongly increased during this period (Fig. 4D). The same behavior was observed also after the third round of LIF incubation, suggesting that Pramel7 is a downstream target of LIF/STAT3.

To test the robustness of this hypothesis, we monitored Pramel7 expression after LIF induction in Stat3 null cells [12] cultivated in N2B27 + 2i medium without LIF. In the wt cells, LIF induced an upregulation of Pramel7 mRNA, whereas Stat3 knockout cells failed to regulate Pramel7 expression (Fig. 4E). These results clearly indicate that Pramel7 transcription is STAT3-dependent. Furthermore, LIF stimulation in wt ESCs, in the presence of the protein synthesis inhibitor cycloheximide, resulted in strong activation of Pramel7 transcription (Fig. 4F) confirming that STAT3 directly drives Pramel7 transcription.

Pramel7 Expression Is Regulated Through the Parallel Circuitry of the LIF Signal Pathways

We observed that within 10 minutes LIF induces a rapid upregulation of Pramel7 transcripts under standard culture conditions (Fig. 4D). Interestingly, the presence of 2i causes a significant delay in the LIF-dependent Pramel7 upregulation (Fig. 4E). In the presence of 2i, LIF supply is dispensable and the cells can be propagated in the absence of active LIF/STAT3 pathway [15]. We therefore considered whether the retarded responsiveness of Pramel7 transcription to LIF was due to the absence of active STAT3 or to the presence of the ERK- and/or GSK3β-inhibitors.

We examined induction of Pramel7 on addition of LIF in absence of 2i or in presence of either PD0325901 (ERK-inhibitor) or CHIR99021 (GSK3β-inhibitor). ESCs were then incubated for 30 minutes, 1 hour, 3 hours, or 5 hours with LIF. Intriguingly, the presence of CHIR99021 completely blocked Pramel7 transcription (Fig. 5A) indicating that the inhibition of GSK3β activity impairs STAT3-mediated transcription of Pramel7. We therefore analyzed the effect of LY294002, a phosphoinositide-3-kinase (PI3K) inhibitor on Pramel7 regulation. Because of the known apoptotic effect of PI3K inhibition, cells were incubated with low inhibitor concentrations and for a short time. STAT3 mRNA expression was not influenced by the presence of the PI3K inhibitor and as expected, increased during the LIF incubation times. Whereas the level of Pramel7 transcripts remained unaltered even after 5 hours of LIF stimulation (Fig. 5B). We can therefore conclude that Pramel7 transcription is mediated by the parallel activity of the LIF/STAT3 and the PI3K/GSK3β pathways.

thumbnail image

Figure 5. Transcriptional regulation of Pramel7. (A): Quantitative real-time polymerase chain reaction (PCR) analysis of Pramel7 gene expression in E14 wild-type (wt) ESCs stimulated with LIF in presence of 2i, PD0325901 (PD), or CHIR99021 (CHIR) at different time points. (B): Quantitative real-time PCR analysis of Pramel7 and Stat3 expression in E14 wt ESCs stimulated with LIF in presence of LY294002. (C): Gene expression analyses in E14 wt cells cultivated for 5 or 9 days in the presence of CHIR99021. (D): Immunostaining for SSEA-1 and OCT3/4 of E14 and Pramel7 cells after 5 days in N2B27 medium supplemented only with CHIR99021. Transgenic Pramel7 cells retained the expression of OCT3/4 and SSEA-1 markers, whereas E14 wt cells differentiated. (E): Quantification of alkaline phosphatase-positive colonies after 9 days of culture in N2B27 medium supplemented only with CHIR99021. Abbreviation: LIF, leukemia inhibitory factor.

Download figure to PowerPoint

It is known that the sole inhibition of ERK or GSK3β is not sufficient for maintaining ESCs undifferentiated. To test whether Pramel7 overexpression can overcome this situation, we cultivated wt and Pramel7 cells in N2B27 medium supplemented only with CHIR99021. After 5 days, wt cells showed significant upregulation of the differentiation markers Brachyury-T, GATA6, and GATA4, and after 9 days, the expression of REX1, Nanog, OCT3/4, and SOX2 was reduced (Fig. 5C). Immunohistochemistry for OCT3/4 and SSEA-1 showed a reduced expression in wt cells and colonies started to spread out loosing their compact shape (Fig. 5D). In contrast, Pramel7 cells formed high compact colonies, which were homogeneously positive for both pluripotency markers (Fig. 5D). Moreover, they formed more than 70% undifferentiated, homogenously AP-positive colonies, whereas E14 cells were only about 25% (Fig. 5E). This demonstrates that overexpression of Pramel7 in combination with the GSK3β-inhibitor is sufficient to maintain ESCs undifferentiated, indicating that the presence of Pramel7 can compensate for the need of MEK/ERK inhibition.

Pramel7 Overexpression Under Feeder- and Serum-Free Conditions Is Only Partially Able to Maintain Pluripotency

We asked next if Pramel7 is able to maintain ESCs undifferentiated in N2B27 medium in the absence of feeders. E14 and Pramel7 ESCs were adapted to the N2B27 + 2i conditions (without LIF) by extensive passaging and both cell lines after 11 passages showed homogenous AP staining (Fig. 6A). We therefore used these cells for our further experiments. We first investigated if overexpression of Pramel7 is sufficient for prolonged cell expansion in N2B27 medium without 2i and LIF, a condition that normally does not support self-renewal. After two passages at clonal density, wt cells died, whereas Pramel7 cells could be split for one more passage, but finally they also died (Fig. 6B). This indicates that the sole overexpression of Pramel7 is not sufficient for prolonged cells expansion under these conditions. We therefore investigated if Pramel7 cells were able to self-renew in N2B27 medium supplemented with LIF alone, a condition that is also normally not sufficient for maintaining ESCs. After passaging, almost all the E14 cells either died or stopped to self-renew. However, Pramel7 cells showed higher self-renewing rate and formed round, compact colonies (Fig. 6C). OCT3/4 and SSEA-1 expression was detected only in the transgenic cells, but not in the wt E14 cells (Fig. 6D). Quantification for AP-positive colonies showed that Pramel7 cells exhibited 5% of completely undifferentiated colonies and about 50% of partially differentiated colonies when stained for AP (Fig. 6E). In contrast, more than 80% of the wt cells were completely differentiated and showed no AP activity at all (Fig. 6E). This was also confirmed by real-time PCR (Fig. 6F). In summary, the combination of Pramel7 overexpression and LIF increases self-renewal capacity of ESCs, facilitating the maintenance of the undifferentiated state. Nevertheless, this synergistic effect was only partial, as complete elimination of differentiation was not observed.

thumbnail image

Figure 6. The sole Pramel7 overexpression in serum- and feeder-free condition is not completely sufficient for maintaining the undifferentiated state of ESCs. (A): Representative appearance of alkaline phosphatase (AP)-positive colonies in E14 wt and Pramel7-overexpressing cells after 11 passages in N2B27 + 2i medium (feeders free and LIF free). Both cell lines show AP activity. (B): Representative morphology of wt and Pramel7 ESCs after three passages in N2B27 medium in absence of 2i and LIF. (C): E14 and Pramel7 cells cultivated for four passages in N2B27 medium supplemented with LIF only. (D): Immunostaining for OCT3/4 and SSEA-1 expression in wt and Pramel7-overexpressing cells cultivated in N2B27+LIF. E14 cells lost OCT3/4 and SSEA-1 expression, whereas Pramel7 cells remained undifferentiated. (E): Quantification of AP-positive colonies after 9 days of culture in N2B27 medium supplemented only with LIF. (F): Gene expression analysis of ESCs cultivated 5 and 9 days in N2B27+LIF. E14 cells showed progressively downregulation of pluripotency genes and upregulation of differentiation marker genes. Abbreviations: LIF, leukemia inhibitory factor; SSEA-1, stage-specific embryonic antigen-1.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

The JAK/STAT3-pathway was shown to be essential and sufficient in mouse ESCs to mediate LIF signals thereby contributing to the maintenance of pluripotency [19, 20]. Even though a complete bypass of LIF signaling is possible under certain circumstances [15], optimal self-renewal is obtained by combination of LIF and 2i, confirming LIF/STAT3-pathway as an essential component in ESCs. Moreover, STAT3-activation was recently described to be a limiting factor for reprogramming to ground state pluripotency [21]. Expression of Pramel6 and Pramel7 was increased upon overexpression of STAT3 in ESCs [14] and therefore represent potential candidates involved in regulating the homeostasis of ESCs [14, 22, 23].

Here, we demonstrate that Pramel7 is not only a direct target of STAT3 but also that STAT3-mediated maintenance of pluripotency strongly depends on Pramel7 expression. During the preimplantation embryo stages the LIF/gp130 pathway is dispensable for early development without diapause [24]. Nevertheless, even in the absence of LIF stimuli, Pramel7 is expressed in the inner part of the morula and in the ICM of the embryo. Intriguingly, in Stat3-knockout cells, where the LIF/STAT3-pathway is not active, Pramel7 expression was upregulated compared with the wt cells. These findings suggest that there might be LIF/STAT3 compensatory mechanisms or factors, which drive and regulate Pramel7 expression. We exclude the possibility of an autoregulatory activity of Pramel7, as the protein lacks the characteristic domains typical for transcriptional factors but contains LRR domains mediating protein-protein interactions [25].

Pramel7 regulation occurs through a parallel circuit involving both the STAT3 and the PI3K-pathway. The regulation of GSK3β by PI3K is involved in the transcription of Pramel7 upon LIF stimulation and thus for the maintenance of LIF-mediated control of ESC self-renewal. We suggest that GSK3β is directly involved in the control of STAT3-mediated Pramel7-transcription, but the exact nature of this regulation remains to be elucidated. Interestingly, recent data suggests that STAT3-activation is dependent on GSK3β [26]. Beurel et al. reported that GSK3β does not function upstream of STAT3-activating tyrosine kinases but instead is required for the recruitment of STAT3 to the receptor and for its tyrosine phosphorylation-mediated activation. Even though the authors did not directly analyze ESCs, they demonstrated that the dependence of STAT3-activation on GSK3β was a widespread regulatory interaction observed with multiple stimuli and in several types of cells [26], suggesting that the same molecular process likely occurs in ESCs. This is interesting because it was previously shown that in ESCs, STAT3 is not a target of PI3K action and that the loss of self-renewal and the consequent differentiation of the cells after inhibition of PI3K was due to an increase in pERK upon LIF-stimulation [27]. The ERK-pathway is continuously activated in undifferentiated ESCs predominantly by signaling through FGF receptor [28]. It is widely accepted that in self-renewing ESC cultures, the provision of LIF and the following activation of STAT3 acts downstream of pERK to override the auto inductive capacity of FGF4. In this study, we identify for the first time Pramel7 as a protein that links the three LIF/gp130-induced pathways (LIF/STAT3, LIF/MEK/ERK, and LIF/PI3K/GSK3β). We suggest that the concerted activity of STAT3 and GSK3β controls Pramel7 transcription, which in turn regulates the phosphorylation of ERK leading to an abrogation of ESC differentiation (Fig. 7A).

thumbnail image

Figure 7. Hypothetical transcriptional mechanisms, which drive Pramel7 expression. (A): Schematic representation of Pramel7 transcription directly controlled by the LIF/STAT3-pathway and also by the LIF/PI3K-pathway. LIF/gp130 receptor leads to the activation of three different pathways, that is, the LIF/STAT3, the LIF/PI3K/GSKβ3, and the LIF/MEK/ERK-pathway. Pramel7 transcription is directly controlled by the transcription factor STAT3, whereas its phosphorylation is probably regulated by the GSK3β kinase. The combinatorial effect of LIF/STAT3 and GSK3β drives and controls Pramel7 transcription, which in turn blocks phosphorylation of ERK and therefore ESCs differentiation. (B): Schematic representation of Pramel7 expression and effects in vivo and in vitro. In the late preimplantation embryo, Pramel7 expression is only detectable in the inner part of the morula and in the ICM of the blastocyst. Immediately after implantation Pramel7 expression disappears, when differentiation processes occur in the embryo. In vitro wild-type (wt) ESCs do express Pramel7 but need LIF for self-renewing. After LIF depletion wt ESCs differentiate into different cell types and also form EBs. In Pramel7-overexpressing cells, addition of LIF in the medium is not necessary anymore for maintaining the undifferentiated and self-renewing state of the cells. These cells are not able to differentiate and either form undifferentiated embryoid bodies or die. Once the overexpression of Pramel7 is reverted to the wt level, these cells are able to differentiate and take part to embryo development. Abbreviations: EBs, embryoid bodies; ERK, extracellular signal-regulated kinase; ICM, inner cell mass; LIF, leukemia inhibitory factor; MEK, MAPK/ERK Kinase; pERK, phosphorylated ERK.

Download figure to PowerPoint

Importantly, Pramel7 blocks teratoma formation capacity and differentiation potential in vitro and in vivo, indicating that Pramel7 silencing is an absolute requirement for differentiation (Fig. 7B). Similar to Pramel7 overexpression, forced expression of Nanog confers LIF-independent self-renewal and prevents differentiation of ESCs [11]. Unlike Pramel7-overexpressing cells, though, Nanog-overexpressing clones were able to generate teratomas-containing derivatives of all three germ layers. Even though forced expression of both genes in absence of LIF leads to self-renewal of ESCs, our data highlights a different reaction to differentiation stimuli and therefore a different function of these genes in maintaining pluripotency. Nanog is expressed at similar levels in both ICM and ESCs, whereas Pramel7 expression is higher in the ICM than in ESCs [22]. Moreover, it was reported that the essential function of Nanog is to define the pluripotent pool of cells of the ICM and the germ cells, and once the cells are established, its function is dispensable, so that Nanog null ESCs are able to self-renew [29]. Our data also confirms the idea that Nanog is important for the pluripotent identity of the ICM and of ESCs, whereas Pramel7 is more probably involved in allowing/blocking the start of differentiation rather than actively taking part in the processes maintaining pluripotency. In agreement with this hypothesis, elimination of Pramel7 expression by knockdown induces ESC differentiation independently if LIF or even STAT3 overexpression is present. Taken together, this suggests that Nanog is priming a cell to become pluripotent, whereas Pramel7 inhibits a pluripotent ESC from commitment and suggests that Pramel7 acts as the judge in the trial between pluripotency and differentiation. Its presence maintains the cells in a self-renewing state by retarding commitment.

Assuming that Pramel7 is not directly acting as a transcriptional factor but solely through binding to other proteins, further experiments aimed at the identification of potential Pramel7 binding partners are necessary to reveal the mechanisms underlying maintenance of pluripotency through Pramel7.

CONCLUSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Our data show that the combined activity of STAT3 and GSK3 controls Pramel7 transcription, which in turn regulates the phosphorylation of ERK leading to the inhibition of ESC differentiation. Accordingly, Pramel7 ablation causes ESC differentiation, whereas its overexpression sustains long-term self-renewal in the absence of LIF. These observations prove Pramel7 as an essential factor of the signaling network regulating pluripotency and self-renewal in ESCs.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

We thank Zsuzsanna Pataki, Dimitri Goriounov, and Leonardo Mamaril for competent technical assistance; Peter Richards for critical reading of the article; Austin Smith and Jason Wray for the Stat3 null cells, for advice, discussions, and comments on the article; Ian Chambers for providing the Nanog expression vector; and Christian Grimm for sharing the Lif-knockout mice. This work was supported by Swiss National Science Foundation (Grant 31003A-118361 to P.C. and K.B.).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Additional supporting information available online.

FilenameFormatSizeDescription
STEM_588_sm_suppinfo.doc35KSupporting Information
STEM_588_sm_suppinfofig1.tif1966KSupporting Information Figure 1
STEM_588_sm_suppinfofig2.tif745KSupporting Information Figure 2
STEM_588_sm_suppinfofig3.tif2933KSupporting Information Figure 3
STEM_588_sm_suppinfofig4.tif3216KSupporting Information Figure 4
STEM_588_sm_suppinfotable1.doc47KTABLE 1: primer used
STEM_588_sm_suppinfotable2.doc29KTABLE 2: shRNA against Pramel7 sequences
STEM_588_sm_suppinfotable3.doc31KTABLE 3: List of Morula Aggregations

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.