SEARCH

SEARCH BY CITATION

Keywords:

  • Human embryonic stem cells;
  • Pluripotency;
  • STAT3;
  • gp130;
  • LIF

Abstract

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

The preservation of “stemness” in mouse embryonic stem (mES) cells is maintained through a signal transduction pathway that requires the gp130 receptor, the interleukin-6 (IL-6) family of cytokines, and the Janus Kinase-signal transducer and activator (JAK/STAT) pathway. The factors and signaling pathways that regulate “stemness” in human embryonic stem (hES) cells remain to be elucidated. Here we report that STAT3 activation is not sufficient to block hES cell differentiation when the cells are grown on mouse feeder cells or when they are treated with conditioned media from feedercells. Human ES cells differentiate in the presence of members of the IL-6 family of cytokines including leukemia inhibitory factor (LIF) and IL-6 or in the presence of the designer cytokine hyper-IL-6, which is a complex of soluble interleukin-6 receptor (IL-6R) and IL-6 with greatly enhanced bio-activity. Human ES cells express LIF, IL-6, and gp130 receptors, as well as the downstream signaling molecules. Stimulation of human and mouse ES cells with gp130 cytokines resulted in a robust phosphorylation of downstream ERK1, ERK2, and Akt kinases, as well as the STAT3 transcription factor. Loss of the pluripotency markers Nanog, Oct-4, and TRA-1-60 was observed in hES cells during gp130-dependent signaling, indicating that signaling through this pathway is insufficient to prevent the onset of differentiation. These data underscore a fundamental difference in requirements of murine versus hES cells. Furthermore, the data demonstrate the existence of an as-yet-unidentified factor in the conditioned media of mouse feeder layer cells that acts to maintain hES cell renewal in a STAT3-independent manner.


Introduction

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

Maintenance of the undifferentiated state and pluripotency in mouse embryonic stem (mES) cells requires the presence of mouse embryonic fibroblast (mEF) feeder layers or leukemia inhibitory factor (LIF). LIF is a member of the interleukin-6 (IL-6) family of cytokines, which also includes Oncostatin M, ciliary neurotropic factor, IL-6, IL-11, cardiotrophin-1, and cardiotrophin-like cytokine (CLC) [1]. LIF is known to bind to its transmembrane receptor, LIFR, which heterodimerizes with the signal-transducing receptor gp130. The pluripotency of mES depends on the intracellular signaling events that follow, including phosphorylation by the Janus family of tyrosine kinases (JAK), which leads to activation of the signal transducer protein STAT3 [2]. The combination of IL-6 and soluble IL-6 receptor also interacts with and activates a homodimer of gp130 and has been used to maintain mES cells without involvement of LIFR [3,4]. STAT3 activation is sufficient to maintain mES in the undifferentiated state, and inhibition of the gp130-triggered mitogen-activated protein (MAP) kinase (Erk1/2) pathway enhanced this effect [2]. Activation of Erk1/2, in turn, leads to the loss of pluripotency and the onset of differentiation, such that a balance between these opposing signaling pathways determines the fate of the cells [2,5]. Recently it was demonstrated that gp130 signaling has a physiologic role in the process of diapause that occurs naturally in lactating female mice. Mouse embryos arrest at the late blastocyst stage when implantation is prevented and gp130−/−embryos are unable to resume development after the end of diapause. The responsiveness of embryonic stem cells to gp130 signaling most likely has its origin in this reaction [6].

Human embryonic stem (hES) cells require mEFs to maintain pluripotency. Neither LIF nor IL-6 secreted by the mEFs is responsible for this effect since neither murine LIF nor murine IL-6 acts on human receptors [7,8]. Additionally, human LIF was not sufficient to maintain the hES cells in the undifferentiated state [9,10]. One possible explanation of the inability of LIF to maintain pluripotency might relate to deficient cell surface expression of the appropriate cytokine receptors [11]. While gp130 is known to be present on all cells in the body, the LIFR protein is not ubiquitously expressed [12].

The study reported here was undertaken to explore the possibility that “stemness” could be maintained in hES cells using the designer cytokine hyper-IL-6 (a complex of IL-6 and its soluble receptor sIL-6R), which is a potent activator of gp130 that does not require the presence of LIFR or IL-6R to activate signal transduction [13]. Although hyper-IL-6 has recently been shown to block differentiation in mES cells [14], we demonstrate that this is not the case for hES cells. LIFR, IL-6R, and gp130 were expressed on hES cells and, when stimulated with either LIF or IL-6, activated STAT3. However, augmented levels of gp130-stimulated STAT3 activation failed to maintain hES cells in the undifferentiated state, with the progressive loss of TRA-1-60, Nanog, and Oct-4 expression. Culture of hES cells on mEF feeder layers or in the presence of conditioned media from mEF cells also failed to induce STAT3 phosphorylation, although high levels of TRA-1-60, Nanog, and Oct-4 expression were observed. This is the first evidence demonstrating that gp130 can activate STAT3 in hES cells and that this pathway is not activated in the undifferentiated state. Taken together, these results suggest that the maintenance of hES cell “stemness” is STAT3 independent.

Materials and Methods

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

Cell Culture

The human stem cell line HSF6 has been previously described [15] and was maintained undifferentiated by passaging on mitomycin-C–treated mEF from CF1 mice. Stem-cell colonies were cultured at 37°C, 5% CO2 in DMEM serum replacement (DSR) medium, which consisted of high-glucose Dulbecco's Modified Eagle Medium (DMEM) containing knockout serum replacer, glutamine, nonessential amino acids, 0.1 mM β-mercaptoethanol (all from Gibco, Carlsbad, CA; http://www.invitrogen.com), and FGF2 (10 ng/ml) (Peprotech Inc., Rocky Hill, NJ; http://www.pepro-tech.com). For experimental procedures, cells were grown as previously described [16] in the presence of DSR-conditioned medium (CM) from mEF, or in DSR with or without the cytokines to be tested, on tissue-culture dishes precoated for 1 hour at 37°C with mouse laminin (20 μg/ml) (Chemicon, Temecula, CA; http://www.chemicon.com).

Mouse ES cells (ES-D3, ATCC no. CRL-11632, American Type Culture Collection, Manassas, VA; http://www.atcc.org) were grown in DMEM containing 10% fetal bovine serum (FBS), 0.1 mM β-mercaptoethanol (Sigma Chemical Corp., St. Louis, MO; https://www.sigma-aldrich.com) and mouse LIF (10 ng/mL) (Chemicon).

Both hES and mES cells were also cultured in the presence or absence of members of the IL-6 family of cytokines, hLIF (R&D Systems, Minneapolis, MN; http://www.rndsystems.com) or mLIF (both at 10 ng/ml), hIL-6 (50 ng/ml) (Peprotech), and the designer cytokine hyper-IL-6 (a complex of soluble hIL-6R and hIL-6 used at 25, 50, and 200 ng/ml) [13,14].

Immunofluorescence

Stem cell cultures were grown on coverslips coated with laminin (Chemicon), fixed with 4% paraformaldehyde, and stained. Primary antibodies were mouse anti-TRA-1-60 (Chemicon), rabbit antihuman nestin (a generous gift from Dr. Conrad Messam, National Institutes of Health), and rabbit anti-Oct-4 antiserum (a generous gift from Dr. Hans Scholer, University of Pennsylvania). Control slides were incubated with mouse immunoglobulin M (IgM) and rabbit immunoglobulin G (IgG). Affinity-purified rhodamine red–conjugated donkey antimouse IgM and fluorescein-conjugated donkey antirabbit IgG (Jackson ImmunoResearch Laboratories, West Grove, PA; http://www.jacksonim-muno.com) were directed against primary antibodies. Sections and coverslips were mounted in antifade medium (Biomeda Corp., Foster City, CA; http://biomeda.com) and viewed on a Nikon eclipse E800 microscope (NikonUSA, Melville, NY; http://www.nikonusa.com) equipped with a fluorescent attachment. Images were captured with a SPOT digital camera (Diagnostic Instruments Inc., Sterling Heights, MI; http://www.diaginc.com) and acquired through Image Pro Plus 4.0 (Media Cybernetics, Silver Spring, MD; http://www.mediacy.com). Color composite pictures were processed using Adobe Photoshop 6.0 (Adobe Systems, Mountainview, CA; http://www.adobe.com).

Reverse Transcription Polymerase Chain Reaction (RT-PCR)

RNA was purified from mES and hES cells cultured in the absence of feeder layers using the RNeasy minikit including DNase treatment (Qiagen, Valencia, CA; http://www1.qiagen.com) and was then reverse transcribed using avian myeloblastosis virus (AMV) with 3.2 μg of random primer (both Roche, Indianapolis, IN; http://www.roche-applied-science.com) and 1 μg of total RNA in a reaction volume of 20 μL. A master-mix of the same cDNA from either mouse or human cells was made up of 1 μLof cDNA per PCR reaction. The master-mix was then added to the oligonucleotide primers shown in Table 1 to a final volume of 50 μL. The PCR product was loaded onto a 1.2% agarose gel and stained with ethidium bromide.

Table Table 1.. Reverse transcription polymerase chain reaction (RT-PCR) primers
  1. a

    -, previously unpublished primer sequence.

GenesForward primerReverse primerProduct (bp)CyclesAccession no.Refs
β-Actincgcaccactggcattgtcatttctccttgatgtcacgcac20322GI: 33878222[10]
Oct-4gagcaaaacccggaggagtttctctttcgggcctgcac30926GI: 22056734-
Nanoggcttgccttgctttgaagcattcttgactgggaccttgtc25526GI: 13376297-
h Lif Rggccgtggtactgattatgagctccagtcactccactct49730GI: 6042197-
h gp130ctggagtgaagaagcaagtggcgatgcacggtaccatct50230GI: 2174800-
h gp80gctccacgactctggaaaccgcacatggacactatgtag32030GI: 31317250-
m Lif Rcaaccaacaacatgcgagtgggtattgccgatctgtcctg67930GI: 7305234[26]
m gp130ccacatacgaagacagaccagcgttctctgacaacacaca49130GI: 2172014[26]
m gp80ccaaccacgaaggctgtgctgctccactggccaaggtcaa32030GI: 52692[26]

STAT3, ERK, and AKT Activation Assays

Mouse and human stem cells were plated on laminin and were kept undifferentiated in CM, or they were induced to differentiate in DSR for 48 hours. In addition, cells were stimulated acutely with LIF (10 ng/ml), IL-6 (50 ng/ml), and hyper-IL-6 (50 ng/ml)–supplemented DSR for 15 minutes.

Western Blotting

Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) immunoblotting was performed as previously described [17,18] with some alterations. Briefly, cells were lysed in SDS-urea lysis buffer (2% SDS, 6 M urea in 50 mM Tris) containing a cocktail of protease inhibitors (5 μg/ml leupeptin, 2 μg/ml pepstatin, 1 mM phenylmethylsulfonyl fluoride (PMSF), and phosphatase inhibitors (100 mM NaF, 10 mM Na pyrophosphate, 1 mM Na orthovanadate, 10 mM beta-glycerophosphate, and 1 μM microcystin). The lysates were sonicated, and protein was measured by the Micro-BCA method (Pierce, Rockford, IL; http://www.piercenet.com). Total cell lysates (20 μg) from mES and hES cells cultured in the absence of feeder layers were resuspended in reducing NuPage sample buffer (Invitrogen, Carlsbad, CA; http://www.invitrogen.com) and electrophoresed on a 10% SDS-polyacrylamide gel and transferred onto polyvinylidene fluoride (PVDF) membrane (Millipore, Billerica, MA; http://www.millipore.com). Membranes were blocked with tris-buffered saline (TBS) solution containing 5% nonfat milk and 0.2% Tween and incubated in mouse anti-phospho-Erk1/2 (1:2000), rabbit anti-phospho-Stat3(Tyr705) (1:1000), rabbit anti-phospho-Akt(Thr308) (1:1000) (all from Cell Signaling, Beverly, MA; http://www.cellsignal.com) or mouse anti-ß-actin (1:1000) (Sigma) overnight at 40°C. Blots were then incubated for 1 hour with peroxidase-conjugated antimouse/antirabbit antibodies (1:5000) (Jackson ImmunoResearch), followed by chemiluminescence detection (Amersham Biosciences, Picataway, NJ; http://www.amershambiosciences.com).

Cell Fractionation

Both mES and hES cells cultured in feeder-free conditions were resuspended in lysis buffer (20 mM 4-2-hydroxyethyl-1-piperazineethanesulfonic acid [HEPES], pH 7.5, 2 mM eth-ylenediaminetetraacetic acid [EDTA], 2 mM ethylenegly-coltetraacetic acid [EGTA], 5 mM MgCl2, 300 μM PMSF 1 mM vanadate, 40 μg/ml leupeptin, and 1 uM microcystin) and sonicated. Whole-cell lysates were centrifuged at 100,000 g for 30 minutes at 4°C, and the resulting supernatant (cytosol) was removed. The pellet containing the membrane-associated proteins was solubilized in lysis buffer containing 1% Triton X-100 [19]. Samples were separated on 8% SDS polyacrylamide gels and transferred onto PVDF membrane. Gp130 was detected using a rabbit anti-gp130 antibody (1:1000) (Upstate, Lake Placid, NY; http://www.upstate.com).

Results

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

HSF6 Cells Differentiate in the Presence of the IL-6 Family of Cytokines

To remain in an undifferentiated state, hES cells require either a feeder layer of mouse fibroblasts or plating on laminin and CM from mouse fibroblasts [16]. The morphology of the hES cell line HSF6 grown on feeder layers (Fig. 1A) or in CM on laminin-coated plates (Fig. 1B) was similar. Under both conditions, cells grew as undifferentiated colonies that stained uniformly for the stem cell markers TRA-1-60 (Fig. 1C–E) and Oct-4 (Fig. 1E), with rare cells staining for nestin, a marker of differentiation, at the edge of the colonies (Fig. 1C, D). In contrast, cells grown on laminin in DSR medium with or without hLIF (Fig. 1F), IL-6 (data not shown), or hyper-IL-6 (Fig. 1G, H) differentiated rapidly, as determined by the loss of TRA-1-60 and the appearance of nestin-positive cells. The rate of differentiation was similar at all concentrations of hyper-IL-6 (data not shown). After 24–48 hours, cytokine-treated cells increased in volume, flattened, and migrated from the colonies. After 1 week, the stem cell colonies had almost disappeared, leaving behind spreading monolayers of single cells. Immunofluorescence of these cells indicated that most cells were nestin positive (Fig. 1G), while TRA-1-60 (Fig. 1F, G) and Oct-4 staining was much reduced (Fig. 1F). After 2 weeks, TRA-1-60 expression was not detected (Fig. 1H). No difference was observed between cells grown with or without the cytokines, either morphologically or in the expression of differentiation and stem cell markers (data not shown).

thumbnail image

Figure Figure 1.. Differentiation of human embryonic stem (hES) cells in the presence of the interleukin-6 (IL-6) family of cytokines. Morphology of undifferentiated HSF6 cells cultured on mouse embryonic fibroblast (mEF) feeder layers (A) or on laminin in the presence of conditioned media (CM) from feeder layers (B) was similar when viewed under phase contrast microscopy. Immunohistochemical analysis of undifferentiated HSF6 cells cultured on mEF feeder layers (C) or on laminin in the presence of CM from feeder layers (D) was also similar. Uniform staining for the human stem cell marker TRA-1-60 was observed, with only rare cells at the edges of the colonies staining for nestin, a marker for differentiating cells. Nuclear staining for the stem cell marker Oct-4 was also observed in undifferentiated cells on feeder layers (not shown) and on laminin with CM (E). After 1 week of culture in the absence of CM and feeder layers and in the presence of either human leukemia inhibitory factor (hLIF) (F) or hyper-IL-6 (G), staining for TRA-1-60 was rarely seen (F, G), and Oct-4 staining was much reduced (G), while differentiating cells stained for nestin (G). After 2 weeks culture with hyper-IL-6 in the absence of CM, no stem cell markers were observed and only nestin-positive cells could be seen (H). Magnification, bar = 25 μMA–D and 12.5 μM in E–F.

Download figure to PowerPoint

We next wanted to examine the ability of hyper-IL-6 to prevent hES cell differentiation. Robust expression of the stem cell markers Oct-4 and Nanog was observed in undifferentiated cells grown on laminin with CM (Fig. 2, lane 1). Expression of these markers was decreased in hES cells grown on laminin without CM for 1 week (Fig. 2, lane 2) and absent after 2 weeks (Fig. 2, lane 4). Treatment of the cells with hyper-IL-6 did not prevent differentiation over the same 2-week time period (Fig. 2, lanes 3 and 5).

thumbnail image

Figure Figure 2.. Semiquantitative reverse transcription polymerase chain reaction (RT-PCR) of hES cells for Oct-4 and Nanog cultured in the presence of conditioned media (CM) (lane 1), without CM (lanes 2, 4), or hyper-IL-6 (lanes 3, 5). Cells were maintained in CM for 2 weeks in parallel with cells grown without CM or in hyper-IL-6. Samples grown without CM or treated with hyper-IL-6 cells were collected at 1 week (lanes 2, 3) and at 2 weeks (lanes 4, 5). Both Nanog and Oct-4 expression decreased after 1 week, both in the absence of CM and in the presence of hyper-IL-6, and was absent at 2 weeks (lanes 4, 5). Abbreviation: RT, reverse transcriptase.

Download figure to PowerPoint

HSF6 Cells Express the Receptors for LIF, gp130, and IL-6

Mouse ES cells require LIF to maintain an undifferentiated state. Binding of the cytokine to the LIFR results in heterodimerization of LIFR with the signaling receptor gp130 and activation of the JAK/STAT pathway. The self-renewal program has been attributed to STAT3-mediated transcription [2,12]. In contrast to mES cells, LIF does not maintain hES cells in an undifferentiated state, and this has been attributed to the lack of the LIFR on hES cells [11]. RT-PCR experiments showed that a high level of LIFR mRNA was expressed in both the hES and mES cells (Fig. 3A and B respectively), suggesting that LIFR alone does not explain the difference between mouse and human ES cells.

thumbnail image

Figure Figure 3.. Expression of leukemia inhibitory factor receptor (LIFR), gp130, and IL-6R (gp80). (A, B): RT-PCR was performaed on a master-mix of the same cDNA from either hES or mES cells. LIFR and gp130 were expressed at high levels in the mouse and human cells. In addition, the relative level of mouse gp130 compared with mouse LIFR was the same as the relative level of human gp130 compared with human LIFR. IL-6 (gp80) was present at lower levels than LIFR and gp130 expression in mouse and human cells. (C): Western blot for gp130 in the membrane fractions of both mES and hES cells. Abbreviations: cyt, cytoplasmic fraction; Mb, membrane fraction; +/−, reverse transcriptase.

Download figure to PowerPoint

LIFR is not the only receptor to dimerize with gp130. The IL-6 receptor (gp80) also binds gp130 to signal to the JAK/STAT pathway [1]. Although gp80 was expressed in both hES and mES cells, it was at a lower relative level than LIFR and gp130 (Fig. 3A, B).

We next wanted to determine whether hES cells expressed gp130 mRNA. RT-PCR analysis (Fig. 3A, B) and western blotting (Fig. 3C) indicate the presence of the receptor, suggesting that the signaling machinery required for the activation of the JAK/STAT pathway are present in both the human and mouse stem cells.

Further, because RT-PCR for gp130, LIFR, and gp80 was performed on a master-mix of the same cDNA from either the human or mouse cells, it is possible to compare the relative levels of human gp130 to human LIFR, as it is also possible to compare the relative levels of mouse gp130 to mouse LIFR. This is an important observation, since efficient activation of STAT3 requires dimerization of gp130 and LIFR. Mouse gp130 and LIFR were expressed at the same relative level, which is sufficient to activate STAT3 (Fig. 4). Human gp130 was also expressed at a similar level to human LIFR. Hence, the ratio of mouse gp130 to LIFR required for STAT3 activation was the same as the ratio of human gp130 to LIFR.

thumbnail image

Figure Figure 4.. Chronic and acute stimulation of gp130 signal transduction in mES and hES cells. pSTAT3 (Tyr705), pErk1/2 (Thr202/Tyr204), pAkt (Thr308), and β-actin were detected on the same blot that contained positive and negative controls for pSTAT3 only. Acute treatment was performed on undifferentiated ES cells for 15 minutes and included no treatment: ES cell media (mES: DMEM + FBS; hES: DSR + FGF); or either LIF, IL-6, or hyper-IL-6 supplemented ES cell media. Chronic treatment was performed for 24 hours and included the following: Undiff, undifferentiated mouse and human cells stimulated with ES cell media supplemented with LIF or CM, respectively; Diff, mouse and human cells stimulated with their respective ES cell media. Western blots are representative of three separate experiments done in duplicate.

Download figure to PowerPoint

LIF and IL-6 Induce STAT3 Activation in HSF6 Cells

The ability of the IL-6 family of receptors to activate JAK/STAT signal transduction in response to their ligands was investigated in the hES cells using the mES cells as a positive control (Fig. 4). Consistent with the role of STAT3-mediated transcriptional activation of “self-renewal,” undifferentiated mES cells were characterized by a robust level of phosphorylated STAT3, which was lost upon differentiation. Activation was observed 15 minutes after acute stimulation with LIF and was sustained for 24 hours. Hyper-IL-6 also activated STAT3 in the mES cells. In contrast, phospho-STAT3 was not observed in either undifferentiated or differentiated hES cells. Acute stimulation with CM did not activate STAT3, and phosphorylation was not observed in undifferentiated hES cells grown on feeder layers (data not shown). The lack of activated STAT3 in the hES cells was not due to missing or inactive receptors, since the addition of the hLIF resulted in a robust activation of STAT3. Hyper-IL-6, which served as a positive control, activated STAT3 to a higher level than hLIF, while IL-6 alone induced a lower level of STAT3 phosphorylation. The lower levels may relate to the lower level of IL-6R expression relative to LIFR and gp130. Furthermore, there was a larger relative difference between the expression levels of both mouse LIFR and gp130 compared with IL-6R. This may explain the lack of STAT3 activation in mES cells when stimulated with IL6 alone.

The IL-6 family of receptors also activates other signal transduction pathways such as the MAP kinase Erk1/2 and Akt (Fig. 4). The presence or absence of LIF in the media for either mES (DMEM + FBS) or hES (DSR + FGF) had no effect on the acute activation of Erk1/2. However, when compared with differentiated cells, undifferentiated cells in the presence of LIF (mES) or CM (hES) maintained a higher level of pErk1/2 expression 24 hours after stimulation. This difference upon differentiation was more apparent in the mES cells and may be a result of the FGF-supplemented hES cell media. Similarly, the level of Akt activation in the presence of LIF over 24 hours was maintained at a higher level in the undifferentiated mES cells than in the differentiated cells. In contrast, Akt phosphorylation remained constant in both undifferentiated and differentiated hES cells, and following treatment with LIF, IL-6, or hyper-IL-6.

These results are the first demonstration of a functional gp130 signal transduction pathway in hES cells. Furthermore, STAT3 activation via gp130 is not observed in undifferentiated hES, suggesting the self-renewal program is regulated independently of gp130 signal transduction.

Discussion

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

In mouse ES cells, activation of the JAK/STAT pathway via the gp130 receptor for the IL-6 family of cytokines triggers a signaling cascade that preserves expression of the genetic program for “stemness.” Here we report that, in contrast to mouse ES cells, STAT3 phosphorylation was not observed in undifferentiated hES cells maintained in either CM or on mEF. Both mouse and human ES cells express LIFR, IL-6R, and gp130 receptors; furthermore, stimulation of human and mouse ES cells with LIF and IL-6 resulted in a robust phosphorylation of the STAT3 transcription factor. However, the cytokine-treated hES cells proceeded to differentiate, even in the presence of gp130-dependent signaling, with the progressive loss of markers of pluripotency, including Nanog, Oct-4, and TRA-1-60. Differentiation also occurs in the presence of the designer cytokine hyper-IL-6, a complex of soluble IL-6R and IL-6, with greatly enhanced bioactivity. These results show that gp130-mediated STAT3 activation, even with augmented levels of the bioactive molecule hyper-IL-6, is insufficient to prevent the onset of differentiation. Therefore, the presence of other as-yet-unknown factors and signaling events in maintaining the pluripotency of hES cells are evident.

Self-renewal requires a threshold level of STAT3 activation to be maintained in order for mES cells to remain in an undifferentiated state [20,21]. The level of STAT3 activation in the hES in response to LIF and IL-6 observed in this study may not be sufficient for the maintenance of pluripotency. This could explain the lack of responsiveness of the hES cell lines HES1/2 and H9 to LIF. Stimulating gp130 and STAT3 with the bioactive hyper-IL-6 has previously been shown to activate STAT3 to the threshold required to maintain pluripotency [14]. Indeed, culturing the hES cells in the presence of hyper-IL-6 induced robust phosphorylation of STAT3, but it failed to prevent the onset of differentiation. These results, taken in conjunction with the lack of STAT3 phosphorylation in the presence of CM or mEFs, suggests hES self-renewal is independent of STAT3. Alternatively, the threshold levels of STAT3 activation may differ in human and mouse ES cells. This level may be undetectable using phosphospecific antibodies on western blots. STAT3 activation beyond this threshold level, which is detectable on western blot, may, in fact, induce differentiation. Although this situation is unlikely it cannot be ruled out.

Gp130 signal transduction also activates MAP kinase (Erk1/2) and Akt. Acute stimulation either with or without LIF/IL-6 cytokines resulted in a similar level of Erk1/2 activation, but 24 hours after stimulation, Erk1/2 phosphorylation was sustained at higher levels in the undifferentiated cells—that is, when the mES cells were in the presence of LIF and hES cells were in the presence of CM. Indeed, Erk1/2 phosphorylation has been associated with the onset of mES differentiation [5]. However, the small difference observed in the sustained Erk1/2 phosphorylation upon differentiation is unlikely to be directly responsible for the loss of hES pluripotency. Akt phosphorylation was also sustained at a slightly higher level in the undifferentiated mES cells, while no difference was observed following acute stimulation. It is unlikely that Akt plays a direct role in hES cell pluripotency, since expression levels remained unchanged upon differentiation and following acute stimulation.

The inability of hES cells to maintain pluripotency in response to gp130-dependant JAK/STAT signaling is not exclusive to human cells. Indeed, the LIF-responsive clones from the 129 strain of mouse ES cells are themselves unique when compared with LIF-resistant clones from the 129 strain [2,22] and other murine ES cells [2], rat stem cells [23], and nonhuman primate ES cells [24]. Pluripotent human embryonic germ (EG) cells derived from the primordial gonadal ridge are partially dependant on LIF for their propagation; however, differentiated embryoid bodies are also formed and collected in the presence of LIF [25]. The inability of LIF to stimulate self-renewal has also been observed in the human embryonic carcinoma (hEC) cell line Ntera/D1, even though all receptor components were expressed [26]. In contrast to the hES cells, LIF did not activate STAT3 in the hEC cells, possibly due to the constitutive expression of the negative feedback protein suppressor of cytokine signaling-1 (SOCS-1). This suggests that if gp130-dependant signaling could promote hES cell pluripotency, the pathway is blocked downstream of STAT3 activation.

This fundamental difference in the maintenance of pluripotency may represent a difference in the developmental stage of each of the stem cell lines. Certainly, gp130 signaling is required for implantation in the mouse, since female Lif mutants are infertile due to implantation failure [27]. However, loss of pluripotency is not observed in either Lif mutants, which survive into adulthood, or Lifr mutants, which die close to partuition due to deficits in the cells derived from the neurectoderm [28,29], and gp130 mutants, which die mid-to-late gestation due to placental, cardiac, haematopoietic, and neuronal malformations [30,31]. Gp130 signaling does maintain the preimplantation blastocyst during diapause, however, a situation in which implantation is delayed due to physiological state of the mother and the blastocyst remains viable in the uterus for several weeks [6]. It is possible that LIF-responsive ES cells resemble the developmental stage at which the late blastocyst enters diapause; ES cells derived from earlier or later stages may not require gp130 signaling for self-renewal.

The factor that is secreted from the mEFs and which stimulates hES cell pluripotency remains elusive. This study demonstrates that, in contrast to mouse ES cells, LIF, IL-6, or IL-6 in the presence of soluble IL-6R–induced STAT3 activation via gp130 does not contribute to pluripotency. Other gp130 cytokines like IL-11 and oncostatin M (OSM) also activate STAT3 via gp130. The lack of STAT3 activation in the undifferentiated hES cells suggests that, in contrast with mES cells, gp130 activation is not involved in the block of differentiation of hES cells. It is likely that not one single lig-and is responsible for the maintenance of self-renewal, which may involve the interaction of several signaling events that influence the balance of differentiation over self-renewal. One such factor might be the partially purified, soluble, secreted molecule from a differentiated, LIF-deficient parietal endodermal cell line that supports stem cell pluripotency without STAT3 activation [2,22].

Recent studies show that a newly described gene Nanog, which codes for a homeobox protein–promoting mES cell self-renewal, pluripotency, and epiblast formation, may have a leading role in “stemness” [32,33]. Nanog is crucial for preimplantation, and although Oct-4 is not required for Nanog expression, Nanog cannot act in its absence. Parallel pathways between Oct-4, Nanog, and STAT3 have been suggested in mES cells; however, Oct-4 and Nanog are obligatory and can maintain pluripotency in the absence and independently of STAT3 activation [2, 3234]. These results highlight the possible redundancy of the STAT2 pathway in mES cells, and the current study demonstrates hES cells maintain pluripotency in the absence of STAT3 activation. However, the role of Nanog and Oct-4 may be conserved in mES and hES cells. Indeed, the independence of Nanog and Oct-4 expression from the STAT3 pathway was demonstrated in both undifferentiated hES cells, which expressed Nanog and Oct-4 in the absence of STAT3 activation, and differentiating hES cells, which lost Nanog and Oct-4 expression in the presence of hyper-IL-6. Therefore, the interplay between Nanog and Oct-4 may be conserved between the hES and mES cells, while STAT3 is possibly a redundant pathway in mES cells and not required for hES cell self-renewal.

One immediate consequence of our study is to define the STAT3-independent pathway triggered by the contact of hES cells to murine feeder layers and/or by a soluble factor present in the CM. It can be anticipated that this new pathway will act to activate the transcription factors Oct-4 and Nanog. The molecular definition of stimulators of this new pathway will be extremely helpful in the elucidation of optimal culture conditions for primate and nonprimate ES cells.

Acknowledgements

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

The authors wish to thank Dr. Ulupi Jhala for her input during the preparation of the manuscript and Liora Newfield for her technical assistance. This work was funded by the Larry L. Hilblom Foundation.

References

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