SEARCH

SEARCH BY CITATION

Keywords:

  • Human embryonic stem cells;
  • Leukemia inhibitory factor;
  • STAT3;
  • Self-renewal

Abstract

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

Murine embryonic stem (mES) cells remain undifferentiated in the presence of leukemia inhibitory factor (LIF), and activation of signal transducer and activator of transcription 3 (STAT3) via LIF receptor (LIFR) signaling appears sufficient for maintenance of mES cell pluripotency. Anecdotal and contradictory accounts exist for the action of LIF in the culture of human embryonic stem cells, and the nature of LIF signaling and whether the LIF-STAT3 pathway is conserved in human embryonic stem cells (hESCs) has not been systematically explored. In this study, we show that the LIFRβ and the signaling subunit gp130 are expressed in hESCs and that human LIF can induce STAT3 phosphorylation and nuclear translocation in hESCs. Nevertheless, despite the functional activation of the LIF-STAT3 signaling pathway, human LIF is unable to maintain the pluripotent state of hESCs. Feeder-free culture conditions that maintain hESCs in an undifferentiated state do not show activation of STAT3, suggesting that distinct signaling mechanisms govern the self-renewal of hESCs.


Introduction

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

Human embryonic stem cells (hESCs) are pluripotent cells derived from the inner cell mass of blastocysts [1]. Previous studies have shown that hESCs can be stably maintained in culture for longer than 1 year and have the capacity to differentiate in vitro and in vivo into cell types from the three germ layers [2, 3]. These characteristics make them an invaluable resource for studies of cell differentiation and development. A major challenge is to determine the conditions that allow convenient and efficient large-scale culture of hESCs and their directed differentiation into therapeutic cells. hESCs are routinely cultured on feeder fibroblasts to maintain their undifferentiated state [1, 2], but feeder-free culture systems have recently been developed [4]. hESCs cultured on Matrigel supplemented by conditioned media from mouse embryonic fibroblasts (MEF-CM) remain positive for markers of the undifferentiated state (e.g., oct-4, hTERT, alkaline phosphatase, TRA-1-60, and SSEA-4), retain the morphology of undifferentiated cells, and retain the potential to differentiate into numerous cell types. This work suggests that one or more factors secreted by MEFs are required to maintain hESC self-renewal. Murine embryonic stem (mES) cells can be kept undifferentiated in culture under feeder-free conditions by adding leukemia inhibitor factor (LIF) to the medium [5, 6]. LIF is a cytokine that belongs to the interleukin-6 (IL-6) family, which includes IL-6, oncostatin M, cardiotrophin-1, IL-11, and ciliary neurotrophic factor. The LIF receptor (LIFR) consists of the following two subunits: gp130, which is common to all the cytokines from the IL-6 family, and LIFRβ (or gp190), specific for LIF [7]. Both signal transducer and activator of transcription (STAT) and Ras/mitogen-activated protein kinase pathways are activated downstream of gp130. The activation of STAT3 appears both necessary and sufficient for mES cell self-renewal [8]. Overexpression of a dominant-negative form of STAT3 in mouse embryonic stem (ES) cells inhibits their self-renewal and enhances their differentiation [9], whereas a conditionally active form of STAT3 is sufficient to maintain the undifferentiated state of mouse ES cells [10]. Although the central role of the LIF-STAT3 signaling has been well documented in mES, the function of this pathway has remained unclear in hESCs. While anecdotal reports suggest that LIF is not required for maintenance of hESCs [1, 2], others claim that “LIF helps retain the hES cells in an undifferentiated state” [11]. In this work, we investigated whether LIF was necessary and sufficient for maintaining hESCs in an undifferentiated state and whether pathways activated by LIF in hESCs are functionally conserved between mouse ES and hESCs.

Materials and Methods

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

Culture of ES Cells

The hESC lines WA09 (H9) and UC06 (HSF-6) were cultured on feeder layers (primary MEFs from Specialty Media [Phillipsburg, NJ]) in the following medium: knockout-Dulbecco's modified Eagle's medium (Gibco, Grand Island, NY) supplemented with 20% serum replacement (Knockout SR [Gibco]), 1 mM L-glutamine, 1% nonessential amino acids, 0.1 mM beta-mercaptoethanol, and 4 ng/ml human basic fibroblast growth factor (bFGF; Peprotech, Rocky Hill, NJ). Cells were passaged every 3 days by dissociation with 0.25% trypsin/1 mM EDTA. hESCs were also maintained on Matrigel basement membrane matrix (BD Biosciences, Bedford, MA) with MEF-conditioned medium supplemented with bFGF (4 ng/ml). The mES cell lines CCE, J1, and E14 were cultured as described [12]. When elimination of the feeder layer was necessary, we harvested the cells with trypsin-EDTA and plated them in WA09 medium for 45 minutes to allow MEFs to adhere. Nonadherent cells were then collected.

Reverse Transcriptase Polymerase Chain Reaction Analysis

RNA was prepared using RNA STAT60 (Tel-test) according to the manufacturer's instructions. RNA (1.5 μg) was reverse transcribed with superscript reverse transcriptase (RT) (Invitrogen). For polymerase chain reaction (PCR), 1 μl of cDNA was used in a 25-μl final volume using 1.25 units of Taq polymerase (Promega, Madison, WI) with the following primers: gp130 forward: 5′-ggagtgctgttctgctttaa-3′, gp130 reverse: 5′-actgtgtaccacggtagaat-3′, LIFR forward: 5′-ccta acagatggtggagtg-3′, LIFR reverse: 5′-gctgatcgagtttccagaac-3′, actin forward: 5′-tggcaccacaccttctacaatgagc-3′, actin reverse: 5′-gcacagcttctccttaatgtcacgc-3′, Rex1 forward: 5′-cagatcctaaacagctcgcagaat-3′, Rex1 reverse: 5′-gcgtacgcaaat-taaagtccaga-3′, Nanog forward: 5′-actaacatgagtgtggatcc-3′, Nanog reverse: 5′-tcatcttcacacgtcttcag-3′. After 25 cycles,10μl of PCR product was separated on 1.5% agarose gel.

Immunoblot

hESCs or mES cells were plated on Matrigel or 0.2% gelatin, respectively, on 6-well plates (1 × 106 cells per well). One day later, the cells were washed with phosphate-buffered saline (PBS) and cultured with medium containing 0.1% serum replacement without cytokines. The next day, the cells were stimulated with 0.01 μg/ml murine LIF (mLIF) or human LIF (hLIF) (Chemicon, Temecula, CA) for 10, 20, 30, or 60 minutes, washed with PBS, and harvested with 0.25% trypsin-1 mM EDTA. Protein extracts were prepared by resuspending the cells in lysis buffer (150 mM NaCl, 20 mM Tris, pH 7.4, 10 mM NaF, 1 mM EDTA, 1 mM ZnCl2, 1 mM MgCl2, 1% Non-idet P-40 [NP-40], 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, 10 mM Na3VO4), and proteins were resolved on a 10% SDS polyacrylamide gel. The antibodies used were phospho STAT3 (Tyrosine 705), STAT3 antibody (Cell Signaling, Beverly, MA), phospho STAT3 (serine 727) antibody (Biosource, Camarillo, CA), oct 4 (BD Transduction Laboratories), Jak-2 (Upstate, Lake Placid, NY), and HDJ-2 (Lab Vision Corporation, Fremont, CA).

Assay for Differentiation

hESCs were plated at 1 × 105 on Matrigel or 0.2% gelatin and grown in the presence of MEF-CM, hLIF (10 ng/ml), or medium without hLIF. Expression of TRA-1-60 [13], a specific marker of undifferentiated hESCs, was checked after 8 and 16 days.

Surface-Antigen Expression

For antibody staining, hESCs were dissociated with Trypsin-EDTA. Cells were then stained with the primary antibody TRA-1-60 (provided by Dr. Peter Andrews). Phycoerythrin-conjugated goat anti-mouse IgM was used as secondary antibody. Samples were run through a FACScan cytometer (Becton, Dickinson, Franklin Lakes, NJ).

Subcellular Fractionation

Nuclear and cytoplasmic fractions were prepared as previously described [14]. Briefly, cells were harvested in CE buffer (10 mM HEPES [pH 7.6], 60 mM potassium chloride, 1 mM EDTA, 1 mM dithiothreitol) and supplemented with 0.275% NP-40 and protease inhibitors. After centrifugation at 2,000 rpm for 5 minutes at 4°C, the supernatant (cytoplasmic fraction) was stored on ice and the nuclear pellet was washed twice with CE buffer, resuspended in 60 μl of norepinephrine buffer (20 mM Tris [pH 8.0], 420 mM NaCl, 1.5 mM MCl2, 0.2 mM EDTA, 25% glycerol), and supplemented with protease inhibitors. After 10 minutes on ice, both fractions were centrifuged at 13,000 rpm for 10 minutes at 4°C, and supernatants were transferred to fresh tubes.

Retroviral Transduction of hES with a Constitutively Active Form of STAT3

Murine STAT3C cDNA (provided by Daniel C. Link) was cloned as an EcoR1 fragment into the MSCViresGFP (MIG) retroviral vector [15]. Retroviruses were produced by transient cotransfection of 293T cells with the following three plasmids: the viral vector MIG-STAT3C or MIG-empty, the gag-pol genes encoding retroviral packaging plasmid pUMVC3, and a plasmid encoding the vesicular stromatitus virus G protein [16]. Viruses containing supernatants were collected 48 hours after transfection and concentrated by ultracentrifugation. The mES cells and hESCs were transduced with retroviruses by a single-round infection (for 24 hours) at a multiplicity of infection (MOI) of 1, 10, and 50.

Results

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

Effect of LIF on the Maintenance of hESCs in an Undifferentiated State

We first assessed the potential of hLIF to maintain hESCs in an undifferentiated state. WA09 (H9) cells were grown on Matrigel or gelatin in the presence or absence of hLIF or with a combination of hLIF and bFGF. To assess the state of differentiation, we measured the level of expression of the TRA-1-60 antigen that selectively recognizes undifferentiated hESCs [1, 2]. WA09 cells grown on MEFs or on Matrigel supplemented with MEF-CM maintained consistent and high-level expression of the TRA-1-60 antigen (Fig. 1A). In contrast, when WA09 cells were cultured on gelatin or Matrigel supplemented with hLIF, marked downregulation of TRA-1-60 expression was seen after 7 days and was almost complete by 15 days (Fig. 1A). The same pattern of TRA-1-60 downregulation was found for UC06 (HSF-6) cells grown under comparable conditions (data not shown). Downregulation of TRA-1-60 expression in both hESC lines correlated with overt signs of morphological differentiation (not shown). We counted the number of cells for each condition (Fig. 1B) and showed that hESCs grown on Matrigel with MEF-CM expand in culture whereas cells grown on Matrigel and gelatin (with or without LIF) stopped proliferating after 3 to 5 days and assumed a differentiated morphology. We also assayed by RT-PCR the expression of two markers of ES cell pluripotency, Nanog and Rex1 [1719]. Both Nanog and Rex-1 are expressed at high levels in WA09 and UC06 cells grown for 7 days on Matrigel and MEF-CM (Fig. 1C). Corroborating the downregulation of TRA-1-60 expression and morphologic differentiation, Nanog and Rex-1 expression likewise decreased when the cells were cultured on Matrigel alone, with or without hLIF. Unlike WA09 cells, UC06 still expressed low levels of Rex-1 after 7 days with or without hLIF. This difference is likely attributable to the lower growth rate of UC06 compared with WA09 and therefore slower kinetics of differentiation. These data confirmed that hLIF is not sufficient to maintain hES in the pluripotent, undifferentiated state.

thumbnail imagethumbnail image

Figure Figure 1.. hESC differentiation assay. (A): hESCs were plated on MEF, Matrigel + MEF-CM, gelatin, gelatin + hLIF, Matrigel + hLIF, or Matrigel + hLIF + bFGF. At days 7 and 14, the state of differentiation of these cells was evaluated by quantifying TRA-1-60 expression by fluorescence-activated cell sorter analysis. (B): Total cell number was counted at days 3, 7, and 14 for each condition. (C): WA09 and UC06 hESC lines were plated on Matrigel (M) + MEF-CM, + hLIF, or – hLIF. After 7 days in culture, total RNA was prepared from these cells and the expression of Nanog and Rex-1, two markers of pluripotency, was assessed by reverse transcriptase–polymerase chain reaction. Abbreviations: CM, conditioned medium; FGF, fibroblast growth factor; hESC, human embryonic stem cell; hLIF, human leukemia inhibitory factor; MEF, mouse embryonic fibroblast.

Expression of LIFR and gp130 in hESCs

Given the lack of response of hESCs to hLIF, we investigated whether the components of LIF receptor signaling were expressed on the cells. The receptor for the cytokine LIF consists of two subunits, the LIF receptor LIFRβ (gp190) and the signal transducer gp130. After depleting the WA09 and UC06 hESCs of contaminating MEFs, we prepared total RNA, generated cDNA by RT, and performed PCR amplification using primers for LIFR or gp130. Both LIFRβ and gp130 were detected in WA09 and UC06 cells (Figs. 2A, B). No amplification was obtained from MEF cDNA, which confirms that the primers were specific for the human sequence. As positive controls, cDNA from K562 and TF1, two human cell lines known to express the LIF receptor, were amplified. Furthermore, expression of gp130 protein was confirmed by immunoblotting WA09 cells, mES, and MEFs with an antibody that recognizes both mouse and human gp130 (Fig. 2C). These data indicate that the components of the LIF receptor complex are expressed in hESCs.

thumbnail image

Figure Figure 2.. Expression of gp130 and LIFRβ in hESCs. (A): Total RNA was reverse transcribed, and the cDNA was subjected to PCR using primers for gp130, LIFRβ, and actin (as a loading control). (B): Total RNA from WA09 and UC06 human embryonic stem cell lines was reverse transcribed, and gp130 and LIFR were amplified by PCR. (C): Total cell extracts from hESCs, MEF, and mES cells were analyzed by immunoblotting using an antibody directed against murine and human gp130. Abbreviations: hESCs, human embryonic stem cells; M, Matrigel; MEF, mouse embryonic fibroblast; mES, murine embryonic stem; PCR, polymerase chain reaction.

Download figure to PowerPoint

Activation of STAT3 Phosphorylation by LIF

In mES cells, the binding of LIF to its receptor triggers the activation of associated Jak tyrosine kinases, which in turn phosphorylate STAT3 [20]. Therefore, we asked whether LIF could induce STAT3 phosphorylation in hESCs. Both mES cells (on gelatin) and hESCs (on Matrigel) were cultured for 12 hours without LIF in 0.1% serum replacement before stimulation with either mLIF or hLIF. We prepared whole-cell extracts from these cells and assayed for STAT3 phosphorylation using an antibody recognizing phosphotyrosine at position Y705, a residue critical for STAT3 dimerization and nuclear translocation [21]. mLIF induced Y705 STAT3 phosphorylation in mES cells (Fig. 3A). The signal can be detected after 10 minutes, reaches a peak at 20 minutes, and decreases after 30 minutes, in concordance with previous reports [22, 23]. In contrast, no STAT3 Y705 phosphorylation was observed in hESCs in response to mLIF, although STAT3 was expressed (Fig. 3A). We repeated the same experiment with hLIF, and phosphorylation of STAT3 on Y705 was detected in both mES and hESCs (Fig. 3B). These results show that hLIF is able to induce STAT3 phosphorylation in hESCs, whereas mLIF fails to activate hLIF receptor signaling.

thumbnail image

Figure Figure 3.. STAT3 phosphorylation after stimulation of mES and hES with murine or human LIF. mES and hESCs were factor-deprived for 12 hours and then stimulated with mLIF (10 ng/ml) (A) or human LIF (10 ng/ml) (B) for indicated times. Immunoblots of whole-cell extracts were then probed with an antibody directed against phosphotyrosine 705 on STAT3 (Y705 P-STAT3) or total STAT3. (C): Immunoblots of whole-cell extracts prepared from mES and hESCs stimulated with mLIF and hLIF, respectively, for indicated times were probed with an antibody directed against phosphoserine 727 on STAT3 (S727 P-STAT). Abbreviations: hESC, human embryonic stem cell; hLIF, human leukemia inhibitory factor; LIF, leukemia inhibitory factor; mES, murine embryonic stem; mLIF, murine leukemia inhibitory factor; STAT3, signal transducer and activator of transcription 3.

Download figure to PowerPoint

STAT3 contains a second phosphorylation site on Serine 727 that is critical for optimal induction of STAT3 transcriptional activity [24]. Therefore, we analyzed STAT3 phosphorylation at S727 upon LIF addition to mES and hESCs. As anticipated, S727 was phosphorylated upon addition of mLIF to mES cells (Fig. 3C). In hESCs, we could detect a basal level of S727 phosphorylation in the absence of LIF and a slight increase of S727 phosphorylation after addition of hLIF Fig. 3C). Taken together, these results suggest that the transcription factor STAT3 can be phosphorylated at both Y705 and S727 by LIF receptor occupancy in hESCs.

Activation of STAT3 by MEF-CM

Because MEF-conditioned medium supplemented with FGF (4 ng/ml) can sustain hESC self-renewal on Matrigel, we asked whether this condition was accompanied by STAT3 phosphorylation. After 12 hours on Matrigel with 0.1% serum replacement, hESCs were stimulated with MEF-CM + bFGF (4 ng/ml) for 15, 30, and 60 minutes. Total protein extracts were prepared from these cells, and STAT3 phosphorylation was assessed by immunoblotting with antibodies directed against phosphotyrosine 705 and phosphoserine 727. The combination of MEF-CM + bFGF did not induce Y705 STAT3 phosphorylation in hESCs, whereas phosphorylation is induced by hLIF (Fig. 4). In contrast, S727 STAT3 phosphorylation is modestly increased after addition of MEF-CM + bFGF (Fig. 4). These data demonstrate that conditions critical for maintaining hESCs in an undifferentiated state (MEF-CM + bFGF) are not associated with STAT3 phosphorylation on the critical residue Y705.

thumbnail image

Figure Figure 4.. STAT3 phosphorylation after stimulation of hESCs with MEF-CM + bFGF. hESCs were plated on Matrigel. After 12 hours of factor deprivation, cells were stimulated with MEF-CM + bFGF (4 ng/ml). Immunoblots of whole-cell extracts were probed with antibodies directed against phosphotyrosine 705 and phosphoserine 727 on STAT3 or total STAT3. Abbreviations: bFGF, basic fibroblast growth factor; CM, conditioned medium; hESC, human embryonic stem cell; MEF, mouse embryonic fibroblast; STAT3, signal transducer and activator of transcription 3.

Download figure to PowerPoint

STAT3 Subcellular Localization

Phosphorylation of Y705 is known to promote dimerization of STAT3 and translocation to the nucleus, where it functions in transcriptional activation [21]. We examined STAT3 sub-cellular localization in mES cells (CCE) and hESCs (WA09) after acute addition of LIF, as well as in mES and hESCs cultured under conditions that maintain the undifferentiated state. Cytosolic and nuclear fractions were prepared from WA09 and CCE cells treated for 10 minutes with hLIF (10 ng/μl) and mLIF (10 ng/μl), respectively. LIF induced the nuclear translocation of STAT3 in both WA09 and CCE cells (Fig. 5A). Moreover, using an antibody against phosphotyrosine 705, we confirmed that the phosphorylated form of STAT3 is found exclusively in the nucleus (Fig. 5A). Effective subcellular fractionation was verified by the detection of the transcription factor Oct-4 protein within the nuclear fractions and the cytoplasmic Jak-2 or HDJ-2 proteins within the cytosolic fractions (Fig. 5A). We determined the localization of STAT3 in undifferentiated mES cells maintained in log-phase growth in the presence of LIF and undifferentiated hESCs grown on Matrigel with MEF-CM. Using an antibody to detect total protein, STAT3 was found predominantly in the cytoplasm of mES cells; however, using the more specific antibody directed against phosphotyrosine 705, the modification critical for nuclear translocation [21], we indeed detected phosphorylated STAT3 in the nuclear fraction (Fig. 5B). In contrast, phosphorylated STAT3 was not found in the nuclear extracts from hESCs cultured under conditions that maintain self-renewal. Taken together, these results indicate that nuclear localization of STAT3 is not associated with conditions that maintain hESCs self-renewal and apparently is not required for maintenance of hESCs in the undifferentiated state.

thumbnail image

Figure Figure 5.. Subcellular localization of STAT3. (A): Nuclear (N) and cytoplasmic (C) fractions were prepared from mES cells and hESCs cultured without LIF (control) or 10 minutes after addition of mLIF or hLIF (10 ng/ml), respectively. (B): Nuclear and cytoplasmic fractions were prepared from undifferentiated mES cells maintained in log-phase growth in the presence of mLIF and undifferentiated hESCs grown on Matrigel with MEF-CM. STAT3 localization was determined by immunoblotting with antibodies directed against phosphotyrosine 705 on STAT3 and total STAT3. Immunoblotting to detect the nuclear transcription factor Oct-4 was used to validate the nuclear fractionation, whereas detection of the cytoplasmic proteins Jak-2 and HDJ-2 was used as controls for the cytosolic fractionation. Abbreviations: CM, conditioned medium; hESC, human embryonic stem cell; hLIF, human leukemia inhibitory factor; LIF, leukemia inhibitory factor; MEF, mouse embryonic fibroblast; mES, murine embryonic stem; mLIF, murine leukemia inhibitory factor; STAT3, signal transducer and activator of transcription 3.

Download figure to PowerPoint

Expression of STAT3C in ES Cells

Although our results show that STAT3 activation by LIF appears insufficient to maintain self-renewal of ES cells, we asked whether a constitutively activated allele of STAT3, which has been shown to maintain self-renewal of mES cells, might likewise sustain hESC self-renewal. STAT3C is a mutant form of STAT3 with a substitution of two cysteine residues in the SH2 domain that spontaneously dimerizes, translocates to the nucleus, and activates transcription [25]. We transduced mES and hESCs with MIG-empty or MIG-STAT3C retroviruses at increasing MOIs. Both vectors, MIG-empty and MIG-STAT3C, confer green florescent protein (GFP) fluorescence in infected cells. Although most mES cells remain GFP+ 15 days after infection with MIG-STAT3C at MOI 100 (>75% of infected cells; Fig. 6), the percentage of WA09 cells infected with MIG-STAT3C decreased dramatically from 48% at day 2 to 4% at day 15. We attempted to isolate the low numbers of GFP+ cells by flow cytometry to select for clones that maintain STAT3C expression and self-renewal. However, these attempts failed repeatedly, because growth of STAT3C-expressing hESCs apparently could not be sustained. The level of GFP+ cells infected with MIG-empty remained constant for both mES cells and hESCs. We likewise attempted to express a conditionally activated fusion protein of STAT3 and the estrogen receptor that had been shown previously to support the self-renewal of mES cells [10] but again failed to obtain stable lines of hESCs that remained undifferentiated. Our experience suggests that constitutive activation of STAT3 is not sufficient to maintain hESCs in the undifferentiated state and may be associated with enhanced differentiation or apoptosis.

thumbnail image

Figure Figure 6.. Maintenance or loss of mES and hES cells after retroviral transduction with a constitutively activated form of STAT3. mES and hES cells were infected with MIG-STAT3C at a multiplicity of infection of 10, 50, or 100 or with a control vector (MIG-empty). The percentage of GFP-positive cells was quantified by fluorescence-activated cell sorter after 2, 5, and 15 days for mES (top) and after 2, 7, and 13 days for hES (bottom). Abbreviations: GFP, green fluorescent protein; hES, human embryonic stem; mES, murine embryonic stem; MIG, MSCViresGFP; STAT3, signal transducer and activator of transcription 3.

Download figure to PowerPoint

Discussion

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

Several studies have shown that LIF-induced STAT3 signaling plays a central role in the maintenance of mES cell pluripotency and self-renewal [5, 6], but the consequences of LIF treatment of hESCs have remained largely undocumented. Some publications have concluded a lack of response of hESCs to LIF [1, 2], whereas others have reported a beneficial effect from adding LIF to the culture medium [11]. Given these anecdotal accounts, significant uncertainty remains regarding whether the LIF-STAT3 signaling pathway is expressed and functionally conserved in hESCs. In this study we show that the receptor and signaling components of the pathway (LIFRβ and gp130) are indeed expressed in two hES cell lines (WA09/H9 and UC06/HSF6) and that exposure of hLIF to hESCs induces characteristic tyrosine and serine phosphorylation of STAT3 and translocation to the nucleus. However, despite the apparently intact nature of the signaling pathway in hESCs, LIF cannot sustain self-renewal in the absence of differentiation and therefore does not play a comparable role in mouse ES cells and hESCs.

Why hES and mouse ES cells should differ in their mechanisms of self-renewal remains an interesting question. One principal function of LIF in murine development is to enable embryonic diapause [26], the temporary arrest of blastocyst development in lactating female mice that affords optimal timing of multiple and repetitive pregnancies in this highly fecund species. LIF acts as an antidifferentiation factor for cells of the inner cell mass, which can then be propagated as ES cells in what is arguably a fortuitous artifact of cell culture. Because human embryos are not susceptible to diapause, it is perhaps not surprising that blastocyst-derived ES cell lines from the human fail to respond similarly to human LIF, which nonetheless plays critical functions in the hematopoietic, reproductive, endocrine, and central nervous systems [27]. Although by nomenclature, hESCs and mouse ES cells represent a comparable in vitro facsimile of the inner cell mass, the differential response to LIF together with the fact that hESCs but not mouse ES cells show trophoblastic potential [28, 29] argue that hESCs may derive from a more developmentally primitive embryonic cell type that responds to different soluble and cell-associated factors. Because conditioned media from several murine and human cells can indeed sustain hESCs in an undifferentiated state, important self-renewal or antidifferentiation factors remain to be identified.

Our results confirm previous observations about the lack of cross-reactivity of murine and human forms of LIF for the LIF receptor [30]. Indeed, despite high sequence homology between hLIF and mouse LIF (78% amino acid identity) and between the human and mouse LIF receptors (76% identity), mLIF shows a species-restricted binding activity. In contrast, hLIF binds both mouse LIF and hLIF receptor and even shows a higher affinity to the mouse receptor than does mouse LIF itself [31]. In response to LIF cytokine stimulation, the LIF receptor common signaling subunit gp130 activates janus kinases to phosphorylate the transcription factor STAT3 on tyrosine residue 705. Y705 phosphorylation induces STAT3 dimerization, STAT3 translocation to the nucleus, and subsequent DNA binding [21]. The transcriptional activity of STAT3 is also regulated through phosphorylation of serine 727, because an alanine-727 mutant shows reduced transcription factor activity [24]. Both sites must be phosphorylated to achieve the full transcriptional activity of STAT3. In this report, we demonstrated that hLIF could stimulate phosphorylation of both STAT3 Y705 and S727 in hESCs. We also demonstrated that STAT3 translocates to the nucleus after addition of hLIF, and we have detected activation of STAT3 target genes after LIF stimulation (not shown). However, despite apparent functional activation of the STAT3 pathway, hESCs will differentiate despite the presence of hLIF. These results suggest that STAT3 activation is not sufficient to maintain hESCs in an undifferentiated state. This conclusion is additionally strengthened by our failure to isolate self-renewing hESC lines after transduction with either a conditional or a constitutive allele of STAT3. In using a conditionally active form of STAT3 to demonstrate the sufficiency of STAT3 signaling, Matsuda et al. [10] reported that there was a threshold of STAT3 activity required to maintain self-renewal of mES cells, and thus one might argue that hLIF does not achieve the required threshold of STAT3 activation in hESCs. However, we think this is unlikely, because even with methods that yield high gene transduction efficiency, our attempts to express a constitutively activated form of STAT3 in hESCs failed. Indeed, our collective experience suggests increased differentiation or induction of apoptosis in STAT3-expressing hESCs. Finally, we investigated the status of STAT3 activation in hESCs grown on Matrigel with MEF-CM and bFGF, conditions known to maintain these cells in a pristine and undifferentiated state. We found that STAT3 was not phosphorylated on Y705 and could detect only a modest modification at S727. Because Y705 phosphorylation provides the essential prerequisite for dimerization, nuclear translocation, and biological activity, these results suggest that STAT3 is not activated in hESCs cultured under maintenance conditions. Corroborating these observations, we showed that STAT3 was not found in the nucleus of hESCs grown in maintenance media, whereas it is transiently localized in the nucleus of mES cells cultured with LIF.

The work presented here documents the lack of participation of the LIF-STAT3 signaling pathway in the maintenance mechanisms of self-renewal of hESCs. Extensive efforts are underway to unveil the pathways and factors required for propagating ES cells without differentiation, because such agents will be profoundly useful for enhanced ES cell culture.

Acknowledgements

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

We thank Daniel C. Link for providing the STAT3C construct. This research was supported by grants from the National Institutes of Health (DK59279 and HL71265) and the National Science Foundation Biotechnology Process Engineering Center. G.Q.D. is the Birnbaum Scholar of the Leukemia and Lymphoma Society of America. M.W.L. is a fellow of the Leukemia and Lymphoma Society of America.

References

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