SEARCH

SEARCH BY CITATION

Keywords:

  • Embryonic stem cells;
  • Nanog;
  • Gata6;
  • Heterogeneity;
  • Primitive endoderm

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

Nanog is a critical homeodomain factor responsible for maintaining embryonic stem (ES) cell self-renewal and pluripotency. Of interest, Nanog expression is not homogeneous in the conventional culture of murine ES cells. A Nanog-high population expresses markers for pluripotent ES cells, whereas a Nanog-low population expresses markers for primitive endoderm, such as Gata6. Since the inner cell mass of early blastocysts has recently been reported to be heterogeneous in terms of Nanog and Gata6 expression, ES cells appear to closely resemble the developing stage from which they originate. We further demonstrate that Nanog can directly repress Gata6 expression through its binding to the proximal promoter region of the Gata6 gene and that overexpression of Nanog reduces heterogeneity during ES cell maintenance. Interestingly, Nanog heterogeneity does not correlate with the heterogeneous expression of stage-specific embryonic antigen-1, suggesting that multiple but overlapping levels of heterogeneity may exist in ES cells. These findings provide insight into the factors that control ES cell self-renewal and the earliest lineage commitment to primitive endoderm while also suggesting methods to promote homogeneity during ES cell maintenance.

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


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

Embryonic stem (ES) cells are isolated from the inner cell mass of preimplantation blastocysts and have the capacity for indefinite self-renewal in cell culture [1, 2]. ES cells are pluripotent and can be differentiated in vitro into multiple cell lineages by the hanging drop method [3]. In this method, ES cells are aggregated into embryoid bodies in droplets on the lids of Petri dishes. Upon aggregation and differentiation, ES cells recapitulate early embryonic development by forming an outer primitive endoderm layer, whereas the inner cells stay more epiblast-like [4]. Extrinsic factors, such as leukemia inhibitory factor (LIF) and bone morphogenetic protein-4, have been shown to be required for maintenance of ES cells in an undifferentiated state [5, [6]7]. Furthermore, multiple intrinsic factors, including Nanog, Oct4, and Sox2, are important for maintaining ES cells and preventing the onset of differentiation [8, [9], [10]11]. Likewise, several intrinsic factors, such as Gata4, Gata6, and Brachyury, are critical for the early stages of differentiation to primitive endoderm and mesendoderm lineages [4].

The homeodomain protein Nanog is considered to be a master transcriptional regulator of self-renewal and pluripotency in ES cells [12]. Nanog is expressed in ES cells and is rapidly downregulated during in vitro differentiation. Upon aggregation of ES cells, even in the presence of LIF, Nanog downregulation occurs at the outer layer, along with primitive endoderm formation [13]. Interestingly, Nanog overexpression maintains ES cells in the absence of LIF and blocks primitive endoderm differentiation upon aggregation [8, 13]. Similarly, Nanog is highly expressed in the inner cell mass of preimplantation blastocysts and is downregulated during further lineage specification [8]. Nanog knockout embryos fail to form epiblasts, and are mostly composed of disorganized extraembryonic tissue [8]. These data suggest that a primary function for Nanog is the repression of differentiation to primitive endoderm. The mechanism by which Nanog inhibits primitive endoderm differentiation is likely multifaceted, since Nanog, in combination with Oct4 and Sox2, serves as both an activator and repressor of multiple target genes responsible for self-renewal and differentiation [14].

Heterogeneity during ES cell differentiation has posed a particular problem in the formation of lineage-specific cell populations potentially useful for cell-transplantation therapies [15]. Recently, both human and mouse ES cells have been described as being heterogeneous cell populations during normal ES cell maintenance. For example, mouse ES cells were shown to heterogeneously express Pecam-1/CD31, and pluripotency factors were more highly expressed in the Pecam-1 positive population [16]. Also, the epigenetic status of some genes was different between the Pecam-1 positive and negative populations. Furthermore, human ES cells have been separated based on the expression of stage-specific embryonic antigen-3 (SSEA3) [17]. Differences in these populations showed variable expression of pluripotency factors and also changes in clonogenicity and cell cycle. In the present study, we have explored mouse ES cell heterogeneity in terms of Nanog gene expression.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

Maintenance of Mouse ES Cells

The cell lines used for this study were Nanog-TRE ES cells as we previously described [13], R1 ES cells (a gift from Dr. A. Nagy, Toronto), Nanog β-geo ES cells (a gift from Dr. S. Yamanaka, Kyoto, Japan) [8], Rosa ES cells (a gift from Dr. S. P. Oh, University of Florida, Gainesville, FL), and J1 ES cells (American Type Culture Collection, Manassas, VA, http://www.atcc.org). All embryonic stem cell lines were maintained in an undifferentiated state as described [13]. Briefly, ES cells were maintained in an undifferentiated state on gelatin-coated dishes in knockout Dulbecco's modified Eagle's medium (Gibco-BRL, Gaithersburg, MD, http://www.gibcobrl.com) containing 10% knockout serum replacement (Gibco-BRL), 1% fetal bovine serum (Atlanta Biologicals, Lawrenceville, GA, http://www.atlantabio.com), 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 25 mM HEPES (Gibco-BRL), 300 μM monothioglycerol (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com), and 1,000 U/ml recombinant mouse LIF (ESGRO; Chemicon, Temecula, CA, http://www.chemicon.com).

Differentiation of ES Cells

Fluorescent-activated cell sorting (FACS) separated cells were differentiated in monolayer culture by plating 40,000 cells onto gelatin-coated dishes in Dulbecco's modified Eagle's medium, 20% fetal bovine serum (Atlanta Biologicals), 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin (Gibco-BRL).

Site-Directed Mutagenesis

Mutagenesis of the Gata6 β-galactosidase reporter vector (a gift from Dr. Molkentin [18]) was performed using QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, http://www.stratagene.com) according to the manufacturer's instructions. Briefly, the following forward (GTGTTACAGCGCTGGATGGGCCTGGGTCGCTGGCC) and reverse (GGCCAGCGACCCAGGCCCATCCAGCGCTGTAACAC) polyacrylamide gel electrophoresis purified primers were ordered from Integrated DNA Technologies, Inc. (Coralville, IA, http://www.idtdna.com/Home/Home.aspx; mutated sites are italicized). Reaction consisted of 5 μl of 10× reaction buffer, 50 ng of vector template, 125 ng of each primer, 1 μl of deoxynucleoside-5′-triphosphate mix, and 1 μl of pfuTurbo DNA polymerase up to a final volume of 50 μl. Reaction conditions consisted of one cycle at 95°C for 30 seconds and then 18 cycles of 95°C for 30 seconds, 55°C for 1 minute, and 68°C for 9 minutes. Template vector was next degraded by the addition of 1 μl of DpnI restriction endonuclease and incubated at 37°C for 1 hour. Reactions were transformed into Max Efficiency DH5α chemically competent cells (Invitrogen, Carlsbad, CA, http://www.invitrogen.com), analyzed by gel electrophoresis, and sequenced. All the recombinant DNA experiments here and below were performed under the National Institutes of Health guidelines.

Reverse Transcription-Polymerase Chain Reaction

Total RNA was extracted using the RNA aqueous kit according to the manufacturer's instructions (Ambion, Austin, TX, http://www.ambion.com), and reverse transcription-polymerase chain reaction (RT-PCR) was performed as previously described [13]. Briefly, cDNA was synthesized using SuperScript II first-strand synthesis system with oligo(dT) (Invitrogen). PCR was performed using Taq DNA polymerase kit (Eppendorf AG, Hamburg, Germany, http://www.eppendorf.de). For each gene, primers were designed from different exons, avoiding pseudogenes, to make sure that the PCR product would represent the mRNA target and not background genomic DNA. Primer sequences are available upon request.

Chromatin Immunoprecipitation

Chromatin immunoprecipitation (ChIP) was performed using the ChIP assay kit from Upstate (Charlottesville, VA, http://www.upstate.com) according to the manufacturer's instructions. Briefly, a 100-mm plate of ES cells was fixed in formaldehyde to a final concentration of 1% and incubated for 10 minutes at 37°C. Cells were collected in cold phosphate-buffered saline (PBS) with protease inhibitors. Cells were sonicated by three 20-second bursts with 2-minute rests on ice using setting 7 on a Fisher Scientific Sonicator Dismembrator Model 100 (Fisher Scientific International, Hampton, NH, http://www.fisherscientific.com). Lysates were centrifuged for 10 minutes, and supernatants were collected, diluted, and a sample was kept as input DNA. The supernatant was next precleared and collected again following centrifugation. Approximately 5 μl of each antibody (control IgG, Sigma; H3-acetylated (general) and H3-K9-dimethylated, Upstate; Nanog, Chemicon) was added to the supernatant and incubated overnight at 4°C. The following day, 60 μl of Salmon Sperm DNA/Protein A Agarose-50% slurry was added to the reaction and incubated for 1-hour rocking at 4°C. The agarose complex was then collected and washed, and the DNA was eluted off in 1% sodium dodecyl sulfate and 0.1 M NaHCO3, and cross-linking was reversed by incubation at 65°C for 4 hours. Proteins were degraded by proteinase K for 1 hour at 45°C. Finally, DNA was recovered using the Qiagen PCR Purification Kit (Qiagen, Hilden, Germany, http://www1.qiagen.com), and PCR was performed with the Taq DNA polymerase kit (Eppendorf).

Immunofluorescence Staining and Confocal Microscopy Analysis

ES cells were grown on eight-chamber slides (BD Biosciences, San Diego, http://www.bdbiosciences.com). Immunostaining was performed by first fixing the differentiated cells in 3.7% formaldehyde. Cells were next permeabilized with 0.5% Triton X-100 in PBS. Cells were blocked for 1 hour in 5% bovine serum albumin/PBS and then incubated with the Nanog antibody (Calbiochem, San Diego, http://www.emdbiosciences.com), Oct4 antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com), and SSEA1 antibody (MC-480) (Developmental Studies Hybridoma Bank, Iowa City, IA, http://www.uiowa.edu/∼dshbwww) overnight in blocking solution. Cells were washed five times in PBS and incubated with secondary antibody, rhodamine-conjugated anti-Rabbit IgG for 45 minutes (Nanog), fluorescein isothiocyanate-conjugated anti-mouse IgG (Oct4), and fluorescein isothiocyanate-conjugated anti-mouse IgM (SSEA1). Cells were again washed five times in PBS. The slides were counterstained and mounted in antifade medium (Vectashield; Vector Laboratories, Burlingame, CA, http://www.vectorlabs.com) with 4′-6-diamidino-2-phenylindole. Fluorescence was documented using a laser scanning spectral confocal microscope (TCS SP2; Leica, Heerbrugg, Switzerland, http://www.leica.com).

5-Bromo-4-chloro-3-indolyl-β-d-galactoside Staining

We performed 5-bromo-4-chloro-3-indolyl-β-d-galactoside (X-gal) staining using the In Situ β-Galactosidase Staining Kit according to manufacturer's instructions (Stratagene). Medium from the cells was removed, and the cells were then fixed in 1× fixing solution for 10 minutes at room temperature. Cells were washed twice in PBS, and then 1× staining solution was added containing 1 mg/ml X-gal. Cells were incubated overnight at 37°C and analyzed by bright-field light microscopy.

β-Galactosidase Activity Assays

β-Galactosidase activity assays were performed using the β-Galactosidase Enzyme Assay System with Reporter Lysis Buffer kit (Promega, Madison, WI, http://www.promega.com) according to manufacturer's instructions. Briefly, cells were collected using a rubber scraper in 1× Reporter Lysis Buffer. Next, 150 μl of cell lysates were mixed with 150 μl of Assay 2× Buffer and incubated for 3 hours at 37°C. Reactions were stopped by adding 500 μl of 1 M sodium carbonate. The absorbance was read at 420 nm, and readings were compared with a standard curve using β-galactosidase enzyme. Activity is measured in units of β-galactosidase. All assays were performed in triplicate, and standard deviations and p values from Student's t test were determined.

Fluorescein di-β-d-Galactopyranoside Staining of ES Cells

Fluorescein di-β-d-galactopyranoside (FDG) (Sigma) was dissolved in dimethyl sulfoxide to make a 20 mM stock solution, and aliquots were frozen at −20°C. Staining was performed as previously described [19]. Cells were first washed twice in PBS and then incubated in PBS for 5 minutes at 37°C. Next, a 1 mM FDG solution in 50% PBS/H2O was prepared and warmed to 37°C. PBS was then removed from the cells, and FDG solution was added to the plate and incubated at 37°C for 1 minute. Cells were then placed on ice, and 1 ml of ice-cold PBS was added to the cells for 5 minutes. Immunofluorescence was detected using an Olympus (Tokyo, http://www.olympus-global.com) IX70 inverted fluorescent microscope with an Optronics (Muskogee, OK, http://www.optronicsinc.com) digital camera.

Fluorescent-Activated Cell Sorting

Cells were prepared in a single cell suspension by treatment with 0.05% trypsin/EDTA and incubated at 37°C for 5–10 minutes. A cell pellet was collected by centrifugation. To FDG stain cells for FACS, the following protocol was used: A 2-mM FDG solution in sterile warm deionized H2O was first prepared. Next, the cell pellet was resuspended in 100 μl of prewarmed reduced serum medium (Opti-MEM; Invitrogen). Then, 100 μl of FDG solution was added to the cells, and the mixture was incubated at 37°C for 1 minute. The cell suspension was next diluted in ice-cold growth medium and kept on ice for 30 minutes. Before FACS, 1 μg/ml propidium iodide (Sigma) was added to account for the dead cells resulting from the osmotic shock.

For sorting of the SSEA1 stained cells, ES cell pellets were resuspended in PBS containing 1% bovine serum albumin and were stained with anti-SSEA1 antibody (MC-480) (Developmental Studies Hybridoma Bank) for 60 minutes on ice. The cells were then washed and resuspended in the buffer containing phycoerythrin-conjugated anti-mouse IgM (BioLegend, San Diego, http://www.biolegend.com). After 60 minutes of incubation on ice, the cells were washed. Sorted cells were collected by Vantage (Becton, Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com) using CellQuest Acquisition data analysis software (Becton, Dickinson).

Gene Expression Profiling

RNA was extracted from FDG sorted cells using RNAqueous Kit (Ambion) as already described. Gene expression profiling was performed by GenUs Biosystems (Northbrook, IL, http://www.genusbiosystems.com). Total RNA samples were quantitated by UV spectrophotometry (optical density 260/280). Quality of total RNA was assessed using an Agilent Bioanalyzer (Agilent Technologies, Palo Alto, CA, http://www.agilent.com). First and second strand cDNA were prepared from the total RNA samples. Biotinylated complementary RNA (cRNA) target was prepared from the DNA template and verified on the Bioanalyzer. cRNA was fragmented to uniform size and verified on the Bioanalyzer. CodeLink Mouse Whole Genome Bioarrays (GE Healthcare, Piscataway, NJ, http://www.gehealthcare.com) containing approximately 36,000 gene targets were hybridized with the cRNA target and stained with Cy5-streptavidin. Slides were washed and then scanned on an Axon GenePix 4000B scanner (Axon Instruments/Molecular Devices Corp., Union City, CA, http://www.moleculardevices.com). Data were analyzed with CodeLink and GeneSpring software packages. Intensity values were normalized to the median value from the array, and those genes that were differentially expressed by twofold or more were annotated.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

Nanog Expression Is Heterogeneous in ES Cells

We have previously shown that Nanog represses primitive endoderm formation [13]. However, we and others have also reported the expression of the primitive endoderm marker, Gata6, in ES cells [13, 20, 21]. To confirm the expression of Nanog and Gata6 in ES cells, we examined the RNA expression of these genes and other pluripotency genes, Sox2, Oct4, and Rex1, by RT-PCR (Fig. 1A). Indeed, we found that both the pluripotency genes and the primitive endoderm marker, Gata6, were expressed in ES cells. Nanog β-geo ES cells were originally generated by S. Yamanaka's group, in which the β-galactosidase-neomycin fusion gene (β-geo) was inserted under the endogenous Nanog gene promoter [8]. Using these Nanog β-geo ES cells, we have previously found the expression pattern of Nanog by X-gal staining during ES cell maintenance to be heterogeneous [20]. We confirmed these data and again found that both weakly and strongly expressing cells were evident after staining, compared with Rosa ES cells, which show ubiquitous staining (Fig. 1B). To further examine the expression of Nanog at the protein level, we performed immunostaining and confocal microscopy analysis of R1, J1, and Nanog β-geo ES cells (Fig. 1C). Again, we found that Nanog is heterogeneously expressed in all the murine ES cell lines we tested, whereas Oct4 is rather homogenously expressed. SSEA1 has been used as a marker for pluripotent murine ES cells and is also known to be expressed heterogeneously in ES cells [16, 22]. To investigate how the expression of Nanog correlates with the expression of SSEA1, we examined their expressions by confocal microscopy in R1 ES cells (Fig. 1D). Interestingly, SSEA1 expression did not correlate with Nanog. FACS separation of SSEA1 high and low cells shows that Nanog and Gata6 are equally expressed in the SSEA1 populations (Fig. 1E). Together, these data suggest that multiple heterogeneous, but overlapping, populations may exist within ES cells.

thumbnail image

Figure Figure 1.. Heterogeneous expression of Nanog in embryonic stem (ES) cells. (A): Expression of pluripotency markers Nanog, Oct4, Sox2, Rex1, and primitive endoderm marker Gata6 in R1 ES cells and Nanog β-geo ES cells by reverse transcription-polymerase chain reaction (RT-PCR). (B): 5-Bromo-4-chloro-3-indolyl-β-d-galactoside staining of R1 ES cells, Rosa ES cells, and NβG ES cells; bar: 50 μm. (C): Immunofluorescence staining and confocal microscopy analysis of Nanog and Oct4 in R1, J1, and NβG ES cells; bar: 50 μm. (D): Immunofluorescence staining and confocal microscopy analysis of Nanog and SSEA1 in R1 ES cells; bar: 50 μm. (E): Fluorescent-activated cell sorting separation of SSEA1 high and low expressing cells followed by RT-PCR for Nanog and Gata6; β-actin serves as loading control. Abbreviations: DAPI, 4,6-diamidino-2-phenylindole; NβG, Nanog β-geo; PE, phycoerythrin; SSEA, stage-specific embryonic antigen.

Download figure to PowerPoint

Nanog and Gata6 Exist Within Separate Subpopulations During ES Cell Maintenance

To confirm the heterogeneous expression pattern of Nanog and to isolate the Nanog high- and low-expressing populations during ES cell maintenance, we used FDG to stain Nanog β-geo ES cells (Fig. 2A). This substrate is cleaved by β-galactosidase to release fluorescein, and live cells may be visualized by fluorescent microscopy or isolated by FACS. Like X-gal staining, staining with the FDG substrate showed a heterogeneous expression pattern for Nanog. To isolate the FDG positive and negative populations, we performed FACS separation (Fig. 2B). We found that approximately 6% of ES cells were negative for FDG. To further understand the expression profiles of these two subpopulations, we analyzed the RNA expression of the FDG negative and positive populations by RT-PCR (Fig. 2B). Although Nanog, Rex1, and Sox2 were more highly expressed in the FDG positive population, we found that Gata6 was more highly expressed in the FDG negative population. Furthermore, Oct4 was found to be expressed in both the negative and positive populations. These data suggest that Nanog and Gata6 are expressed heterogeneously during ES cell maintenance, and their expression is specific to separate, distinct subpopulations.

thumbnail image

Figure Figure 2.. Staining and isolation of Nanog high and low expressing cells using fluorescein di-β-d-galactopyranoside. (A): FDG staining of Nanog β-geo ES cells confirming a heterogeneous expression pattern for Nanog; bar: 100 μm. (B): Fluorescent-activated cell sorting separation of FDG treated NβG ES cells and RNA expression analysis by reverse transcription-polymerase chain reaction. (C): Gene expression profiling of FDG sorted Nanog β-geo ES cells from Nanoghigh and Nanoglow populations. Abbreviations: ECM, extracellular matrix; ES, embryonic stem; FDG, fluorescein di-β-d-galactopyranoside; NβG, Nanog β-geo; Prim. Endo., primitive endoderm.

Download figure to PowerPoint

Gene Expression Profiling of the Nanoghigh and Nanoglow ES Cells

To further understand the expression profiles of the Nanoghigh and Nanoglow cell populations, we performed a microarray analysis (Fig. 2C, supplemental online Table 1). We found that 917 genes showed higher levels of expression by twofold or more in the Nanoghigh cells, whereas 542 genes showed more than twofold higher expression in the Nanoglow cells. More than 1,200 genes were expressed in common, above background, between the two populations. Similar to our RT-PCR data, we found that the Nanoghigh cells had increased expression of pluripotency markers, whereas the Nanoglow cells showed increased expression of primitive endoderm genes (Table 1). Other interesting findings from the gene expression profiling showed that there was increased expression of genes involved in cell cycle (supplemental online Table 2) and mitochondrial function (supplemental online Table 3) in the Nanoghigh population. In the Nanoglow population, however, we found increased expression of genes that were involved in cell cycle inhibition and extracellular matrix formation (supplemental online Tables 2, 4). These data together suggest that the Nanoghigh and Nanoglow cell populations are quite distinct and may represent different lineages.

Table Table 1.. Gene expression profile of pluripotency and primitive endoderm genes by microarray analysis
Thumbnail image of

Differentiation and Self-Renewal Characteristics of Nanoghigh and Nanoglow Cells

To evaluate the differentiation potential of the Nanog subpopulations, we separated the Nanoglow and Nanoghigh cells using FDG staining and FACS and then differentiated the cells in monolayer culture by the removal of LIF. Morphologically, there was a clear difference in the potential of the Nanog subpopulations, as the Nanoglow cells remained as a homogenous population of flat cells after 4 and 8 days of differentiation, whereas the Nanoghigh cells randomly differentiated into multiple morphological cells types (Fig. 3A). RNA expression analysis by RT-PCR showed that the Nanoglow cells seemed to specifically form extraembryonic endoderm populations that expressed Gata6 and Sparc (Fig. 3B). Nanoghigh cells, however, expressed a mesendoderm marker, Brachyury, a primitive ectoderm marker, Fgf5, and extraembryonic endoderm markers Gata6 and Sparc after 4 days of differentiation. After 8 days of differentiation, the Nanoghigh cells also expressed a cardiac marker, Nkx2.5. These data suggest that the Nanoghigh cells represent a pluripotent cell type capable of differentiating into multiple lineages, whereas the Nanoglow cells may only differentiate into extraembryonic endoderm.

thumbnail image

Figure Figure 3.. Differentiation and self-renewal properties of Nanog subpopulations. (A): Morphology of FDG− and FDG+ populations after 4 or 8 days of growth in differentiation medium. (B): Reverse transcription-polymerase chain reaction (RT-PCR) of FDG− and FDG+ populations after 4 or 8 days of differentiation; Brachyury, mesendoderm marker; Fgf5, primitive ectoderm marker; Nkx2.5, cardiac myocyte marker; Gata6, primitive endoderm marker; Sparc, parietal endoderm marker. (C): RT-PCR analysis before and after replating FDG sorted Nanog β-geo embryonic stem cells for 24 hours in leukemia inhibitory factor (LIF) medium. (D): 5-Bromo-4-chloro-3-indolyl-β-d-galactoside staining of FDG sorted populations after 24 hours in LIF medium; bar: 100 μm. Abbreviations: FDG, fluorescein di-β-d-galactopyranoside; Hrs, hours.

Download figure to PowerPoint

To examine whether the FDG negative (Nanoglow) and positive (Nanoghigh) populations are convertible to each other, we again isolated each population by FACS. Cells were collected (0-hour time point) or replated for 24 hours in ES cell maintenance medium with LIF. Initially, we analyzed the expression of Nanog and Gata6 in these cells by RT-PCR (Fig. 3C). We found that, after replating the Nanoglow cells for 24 hours in LIF medium, there was a significant reduction in Gata6 expression to near background levels and also an increase in Nanog expression. Indeed, when we evaluated the expression of Nanog by X-gal staining after replating the Nanoglow cells (Fig. 3D), we detected X-gal positive cells within 24 hours. The Nanoghigh-sorted population also showed heterogeneous X-gal staining within 24 hours, although there was no detectable change in the expression of Nanog or Gata6 mRNA. In a long-term culture, both Nanoglow and Nanoghigh-sorted populations became similar to their parental ES cells in terms of both X-gal staining and RNA expression (data not shown).

Nanog Reduces Heterogeneity and Directly Represses Gata6

We have previously developed a Nanog-overexpressing tetracycline-inducible (Nanog-TRE) ES cell line [13]. When doxycycline is removed from the medium, Nanog is overexpressed, and the cells can be maintained in the absence of LIF. To examine how Nanog overexpression would affect heterogeneity, Nanog-TRE ES cells and parental cells (TopES) were maintained in LIF medium for 3 days with or without the addition of doxycycline. RNA was extracted from these cells, and the expression of pluripotency genes, primitive endoderm genes, extracellular matrix (ECM) genes, and cell cycle inhibitor genes was analyzed by RT-PCR (Fig. 4A). We found that when Nanog was overexpressed by the removal of doxycycline, there was a decrease in the expression of primitive endoderm genes such as Gata4, Gata6, and Dab2; ECM genes such as Fibrillin1, Fibronectin, and Laminin; and cell cycle inhibitor genes such as Ink4b and Kip2.

thumbnail image

Figure Figure 4.. Nanog reduces heterogeneity and represses Gata6. (A): RNA expression analysis of parental (TopES) and Nanog inducible (Nanog-TRE) cells with or without the addition of doxycycline. Markers used for pluripotency, Oct4 and Nanog (total Nanog, endogenous Nanog, and ectopically expressed Nanog) [20]; extraembryonic endoderm, Gata4, Gata6, and Dab2; extracellular matrix, Fibrillin1, Fibronectin, and Laminin α1; and cell cycle inhibitors, p15/Ink4b and p57/Kip2; β-actin was used as a loading control. (B): β-Galactosidase activity assays with transfections of a Gata6 β-galactosidase reporter or a control EF2 β-galactosidase reporter into Nanog-TRE embryonic stem cells in the presence or absence of doxycycline; * p < .05. (C): Chromatin immunoprecipitation with various antibodies, including Nanog, using primer sets from five independent regions of the Gata6 promoter containing one or more putative Nanog binding sites. (D): Site directed mutagenesis of the putative Nanog binding site determined by chromatin immunoprecipitation in the Gata6 β-galactosidase reporter and subsequent β-galactosidase activity assays in the Nanog-TRE cell line; * p < .05. Abbreviations: β-gal, β-galactosidase; Dox, doxycycline; Mut, mutated.

Download figure to PowerPoint

As Gata6 is considered to be a master regulator for primitive endoderm formation [21], the repression of Gata6 by Nanog overexpression was of particular interest. To determine whether Gata6 repression by Nanog overexpression occurs on the transcriptional level, we measured the activity of the Gata6 promoter using a β-galactosidase reporter (Fig. 4B). Nanog-TRE ES cells were transfected with the Gata6 β-galactosidase reporter vector or a β-galactosidase control vector and maintained in the presence or absence of doxycycline for 3 days. Nanog overexpression significantly reduced β-galactosidase activity from the Gata6 reporter by almost twofold but had no significant effect on the control reporter.

Since Nanog expression could block Gata6 activity, we decided to test whether Nanog could directly interact with the Gata6 promoter. The Nanog binding sequence has been previously determined using Systematic Evolution of Ligands by Exponential Enrichment (SELEX) to be “(C/G)(G/A)(C/G)C(G/C)ATTAN(G/C)” [8]. Using this sequence, we manually analyzed the previously characterized Gata6 promoter/enhancer region of approximately 5.5 kilobases to look for the highly conserved “ATTA” domain, characteristic of many homeodomain-containing factors. We identified at least 12 of these regions as putative Nanog binding sites (Table 2).

Table Table 2.. Potential Nanog binding sites in the 5.5-kilobase Gata6 promoter
Thumbnail image of

To perform a chromatin immunoprecipitation assay based on these putative binding sites, we first confirmed the ability of the antibody to immunoprecipitate Nanog (data not shown). We next developed five primer sets surrounding these putative binding sites, and analyzed Nanog binding to these regions by chromatin immunoprecipitation (Fig. 4C). We found that Nanog only immunoprecipitated with the most proximal binding site, suggesting this is a likely binding site on the Gata6 promoter for Nanog in vivo. To further confirm that this site is required for Nanog repression of Gata6, we mutated the conserved ATTA domain of this most proximal binding site to a “GGCC” on the Gata6 β-galactosidase reporter and repeated the activity assays (Fig. 4D). Using the mutated reporter with the Nanog-inducible cells, we surprisingly found that mutation of this site led to a loss of activity for the reporter. These data suggest that this site on the Gata6 promoter is not only critical for Nanog binding and potential repression but may also be required for its activation.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

It has been reported that Nanog and Gata6 are expressed heterogeneously in the inner cell mass of E3.5 preimplantation blastocysts in a “salt and pepper” configuration, each expressed in a random mosaic pattern but exclusively from the other [23]. Our data suggest that ES cells are strikingly similar to their inner cell mass counterparts. Nanog and Gata6 appear to be expressed heterogeneously in ES cells in separate subpopulations. Interestingly, despite the long-term passaging of ES cells, they still do not form a uniform, homogeneous population, but rather recapitulate the developing stage from which they originate. This supports the idea that ES cells are a useful in vitro tool to study early embryonic development.

Gata6 is considered to be a potential master regulator for primitive endoderm formation. For example, the overexpression of Gata6 was found to induce primitive endoderm differentiation of ES cells [21]. Thus, the Nanoglow/Gata6high cells in ES cell culture may be a subpopulation that is already differentiated into primitive endoderm. Indeed, after analyzing the heterogeneous cell populations by gene expression profiling, we found that the Nanoglow/Gata6high population was representative of a primitive endoderm population, showing expression of other primitive endoderm markers, many ECM genes, and also cell cycle inhibitors. However, expression of parietal or visceral endoderm markers, such as Sparc or α-fetoprotein, respectively, was not detected at significant levels. The Nanoglow/Gata6high population may then be more representative of primitive endoderm cells not yet fated to parietal or visceral endoderm lineages. This was in contrast to the outermost layer cells of ES cell aggregates, where they expressed visceral endoderm markers including α-fetoprotein [13, 20]. Primitive endoderm-like cells existing in the ES cell maintenance culture should further differentiate into visceral endoderm-like cells when they are sorted out to the outermost layer by cell aggregation. In fact, the Nanoglow/Gata6high cells started to express some parietal endoderm markers when cultured in a differentiation medium.

On the other hand, the Nanoghigh/Gata6low population, likely representing the “true” stem cell population capable of self-renewal, maintained high expression of pluripotency markers, cell cycle genes, and genes important for mitochondrial functioning. We could imagine that those cells undergoing higher levels of proliferation and cell division may require increased mitochondrial respiration and have higher energy demands. Furthermore, the Nanoghigh/Gata6low cells were also capable of multilineage differentiation, whereas the Nanoglow/Gata6high cells seemed to only form extraembryonic endoderm, again suggesting the Nanoghigh/Gata6low cells to be the stem cell population.

As shown in Figure 3D and also supported by a differential expression profile in cell cycle regulatory genes, the Nanoghigh/Gata6low population appears to proliferate much faster than the Nanoglow/Gata6high. How, then, can such heterogeneity be maintained in the ES cell cultures? We consider that the Nanoghigh/Gata6low ES cells constantly differentiate into the Nanoglow/Gata6high primitive endoderm cells at a certain rate in conventional ES cell maintenance cultures. Indeed, the isolated Nanoghigh/Gata6low cells became heterogeneous again when maintained in the ES cell medium. Furthermore, this spontaneous differentiation was blocked by overexpression of Nanog, as shown here, or by addition of an inhibitor of fibroblast growth factor receptor (FGFR) to the culture [20]. In these conditions, Nanog-positive ES cells can be maintained for a long term in the absence of Gata6-positive primitive endoderm cells, indicating that the existence of primitive endoderm cells is not a requirement for ES cell self-renewal. On the other hand, we were unable to maintain cells that sustain the high expression of Gata6 under ES cell maintenance conditions. Taken together, a most reasonable explanation for the Nanog/Gata6 heterogeneity would be that self-renewing Nanog positive stem cells are constantly producing Gata6-positive primitive endoderm cells in ES cell cultures. Furthermore, Gata6-positive cells are also constantly lost from the culture by a failure to thrive in the culture conditions and a lower growth rate. It should be noted, however, that the isolated Nanoglow/Gata6high cells started to express Nanog, too. This unexpected phenomenon may be explained by Nanoghigh expressing cells contaminating the Nanoglow/Gata6high cells and quickly outgrowing this population. However, the present study does not exclude the intriguing possibility for the Nanoglow/Gata6high cells to dedifferentiate to a more Nanoghigh expression status. To answer these questions, careful lineage-tracing analysis experiments must be performed in the future, such as by the use of a Cre/lox system under the control of the Gata6 promoter.

The overexpression of Nanog was able to reduce the heterogeneity of ES cells and the expression of primitive endoderm marker Gata6. Nanog was predicted to function as a repressor of Gata6, as the Nanog binding motif was identified in the Gata6 promoter [8]. Indeed, Nanog was found to bind to the Gata6 promoter from ChIP-based genome-wide screening using microarrays in human ES cells [14]. Furthermore, Orkin and colleagues identified two new proteins, Nac1 and Zfp281, which physically associate with Nanog and bind to the Gata6 promoter. Additionally, they found that the knockdown of Nac1 and Zfp281 led to an increase in Gata6 expression, suggesting that these proteins cooperate with Nanog to repress Gata6 [24]. We found that Nanog does indeed directly repress Gata6 expression through its binding to a motif in the proximal promoter region of the Gata6 gene. Mutation of this site, surprisingly, led to an inactivity of the reporter. We predict a mechanism where Nanog protein complexes may compete with other factors that may bind to and activate Gata6 expression.

Based on the data provided here, we find that during the growth of an ES cell colony, a reduction in Nanog expression occurs in a subpopulation of cells. This leads to a relief in the direct transcriptional repression on Gata6 imposed by Nanog. In the presence of LIF medium, these cells are unable to grow or differentiate further. However, if they are plated in differentiation medium without LIF, these cells will only expand into extraembryonic endoderm. Ectopic expression of Nanog maintains Nanog protein levels in all cells, thus preventing the formation of Gata6-positive cells and subsequent primitive endoderm. How the initial heterogeneity of Nanog expression first develops remains unclear, albeit we demonstrated that FGFR was essential in the process [20]. It will be interesting to see if it is related to changes in the cell cycle.

Heterogeneity of ES cells during differentiation poses a significant obstacle in obtaining lineage-specific cell populations useful for cell-based transplantation therapies [15]. The present study demonstrates that ES cells are heterogeneous from the beginning, representing the origin from which they are derived, the inner cell mass of early blastocysts. Our data also suggest that heterogeneity of ES cells may exist at multiple levels, as SSEA1 also appeared to be expressed heterogeneously in ES cells, but this heterogeneity did not correlate with Nanog. The findings presented here not only provide insight into the early mechanisms regulating the loss of pluripotency and formation of primitive endoderm, but may also lead to advances in the technologies to control the quality of ES cells for a future therapeutic purpose.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

We thank Dr. S. Yamanaka and Dr. J. D. Molkentin for providing the Nanog-β-geo ES cells and Gata6 β-galactosidase reporter construct, respectively. We also thank Doug Smith for technical assistance with FACS and confocal microscopy. This work was supported in part by National Institutes of Health Grants DK059699 to N.T. and AG000196 (a T32 Training Grant in Aging) to A.M.S. and by the University of Florida McKnight Brain Institute.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure of Potential Conflicts of Interest
  8. Acknowledgements
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
sc-07-0126-SupTable1.pdf125KSupplemental Table 1
sc-07-0126-SupTable2.pdf16KSupplemental Table 2
Sup_Table3_pdf.pdf16KSupplemental Table 3
sc-07-0126-SupTable4.pdf14KSupplemental Table 4

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.