SEARCH

SEARCH BY CITATION

Keywords:

  • Embryonic stem cells;
  • Endoderm;
  • Intestine differentiation;
  • In vitro differentiation

Abstract

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

The studies of differentiation of mouse or human embryonic stem cells (hESCs) into specific cell types of the intestinal cells would provide insights to the understanding of intestinal development and ultimately yield cells for the use in future regenerative medicine. Here, using an in vitro differentiation procedure of pluripotent stem cells into definitive endoderm (DE), inductive signal pathways' guiding differentiation into intestinal cells was investigated. We found that activation of Wnt/β-catenin and inhibition of Notch signaling pathways, by simultaneous application of 6-bromoindirubin-3′-oxime (BIO), a glycogen synthase kinase-3β inhibitor, and N-[(3,5-Difluorophenyl)acetyl]-L-alanyl-2-phenylglycine-1,1-dimethylethyl ester (DAPT), a known γ-secretase inhibitor, efficiently induced intestinal differentiation of ESCs cultured on feeder cell. BIO and DAPT patterned the DE at graded concentrations. Upon prolonged culture on feeder cells, all four intestinal differentiated cell types, the absorptive enterocytes and three types of secretory cells (goblet cells, enteroendocrine cells, and Paneth cells), were efficiently differentiated from mouse and hESC-derived intestinal epithelium cells. Further investigation revealed that in the mouse ESCs, fibroblast growth factor (FGF) and bone morphogenetic protein (BMP) signaling act synergistically with BIO and DAPT to potentiate differentiation into the intestinal epithelium. However, in hESCs, FGF signaling inhibited, and BMP signaling did not affect differentiation into the intestinal epithelium. We concluded that Wnt and Notch signaling function to pattern the anterior-posterior axis of the DE and control intestinal differentiation. STEM Cells 2013;31:1086–1096


INTRODUCTION

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

The intestinal epithelium is a robust system for the study of cell proliferation and differentiation. There are four differentiated cell types of nonproliferative epithelial cells: enterocytes, goblet cells, enteroendocrine cells, and Paneth cells. Intestinal stem cells (ISCs) and progenitor cells reside in crypts, proliferate vigorously, and function as the source of differentiated epithelial cells. In contrast to adult tissue stem cells, embryonic development and differentiation from embryonic stem cells (ESCs) have not been as well investigated in the intestine. The intestinal epithelium is derived from the definitive endoderm (DE) and is covered by a single layer of epithelial cells, where the cells turnover rapidly. Intestinal endoderm is marked by the expression of the homeobox gene of the caudal family, Cdx2, which start to be expressed from embryonic day 8.5 (E8.5). Cdx2 is shown to play a central role in the regionalization of the primitive gut tube. Conditional ablation of Cdx2 in the DE resulted in the formation of abnormal intestinal epithelia, which was transformed into an anterior identity of the esophageal epithelium [1]. Cdx2 is also shown to be directly regulated by canonical Wnt/β-catenin signaling molecules Tcf4 and β-catenin [2].

Recently, it was reported that at high concentration of fibroblast growth factor 2 (FGF2), the differentiation of human ESCs (hESCs) into small intestinal progenitors increased at the expense of Pdx1+ pancreatic progenitors [3], and that the combination of Wnt3a and FGF4 was reported to be required for hindgut specification, and that FGF4 alone was sufficient for promoting hindgut morphogenesis [4]. During normal development, bone morphogenetic protein (BMP) signals are reported to be absent in the Hensen's node, and to be high in the posterior region in the chick [5]. However, it is not clear how BMP antagonizing signals exerts its anteriorizing effect. Loss of BMP2b signaling in zebrafish was reported to expand the anterior gut endoderm and reduce the posterior gut endoderm, while loss of chordin signaling resulted in opposite effects [6]. Conversely, posterior endoderm seems to be specified through FGF signal emitted from the posterior ectomesoderm [7, 8]. In the chick, we reported that segregation of the intestinal fate occurs very early, at two to five somite stages (ss), before pancreatic fate decision takes place at six to eight ss [9]. In gut development, secreted factors such as FGFs, BMPs, and Wnt/β-catenin are expressed in the endoderm or adjacent mesoderm and play important roles in endoderm regionalization [10, 11].

Studies of ESCs and induced pluripotent stem (iPS) cells have demonstrated that ESCs and iPS cells can be guided into the DE and its derivative organs, such as the pancreas, liver [12–14], and gut-like organ in vitro using embryoid body culture systems [15]. Recently, it is reported that hESCs could be differentiated into endodermal cells expressing Cdx2 in monolayer cultures at high doses of FGF2 [3] or a combination of high doses of Wnt3a and FGF4 [4]. FGF4 is reported to promote mesoderm expansion and morphogenesis, whereas FGF4 and Wnt3a synergy are required for the specification of the hindgut lineage [4]. More recently, it is reported that dual inhibition of transforming growth factor-β and BMP signaling in hESCs resulted in a potentiation of the anterior foregut differentiation [16].

We previously established an in vitro procedure to guide ESCs sequentially differentiated into the mesoderm, DE, and finally regional-specific DE-derived organs (pancreas, liver, and intestine), using a mesoderm-derived feeder cell line, M15, and added with activin and FGF2 [14, 17]. In this study, we demonstrate that Wnt/β-catenin and Notch signaling systems play crucial roles in the intestinal differentiation of ESCs. We also found that there are differences between mouse and hESCs in their responsiveness to FGF and BMP signaling.

MATERIALS AND METHODS

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

Cell Lines and Tissues

The R1 and NERTΔOP ESC lines were used. NERTΔOP is an ESC line introduced with an inducible short form of intracellular domain of murine Notch 1, and is shown to be functional, similarly as the full length Notch1 intracellular domain (NICD) [18]. The R1 and NERTΔOP ESC lines were maintained on mouse embryonic fibroblast (MEF) feeders in 4,500 mg/l glucose Dulbecco's modified Eagle's medium (DMEM) (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) supplemented with leukemia inhibitory factor (Wako, Chemical, Osaka, Japan, http://www.wako-chem.co.jp/english), 10% fetal bovine serum (FBS, Hyclone, Logan, UT, http://www.hyclone.com), 100 mM nonessential amino acids (NEAA, Invitrogen), 2 mM L-Gln (Nacalai Tesque, Kyoto, Japan, http://www.nacalai.co.jp), 50 units/ml penicillin, 50 mg/ml streptomycin (PS, Nacalai Tesque), and 100 μM β-mercaptoethanol (ME) (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich. com). MEF was isolated from mouse embryonic day (E) 12.5–14.5. The mesonephric cell line M15 [17] was graciously provided by Dr. T. Noce (Mitsubishi Kagaku Institute of Life Science, Tokyo, Japan) and Dr. M. Rassoulzadegan (University of Nice-Sophia Antipolis, Antipolis, France). We thank Dr. Andras Nagy for the R1 ESCs and Drs. Timm Schroeder and Jun Yamashita for the NERTΔOP ESCs. MEF and M15 cells were treated with mitomycin C (Sigma) and were used as previously described [14, 17].

Mouse fetal and adult small intestines were dissected from ICR mice at E14.5 and 6 weeks of age, respectively. Human fetal intestine (Biochain Institute, Inc., San Francisco, US, http://www.biochain.com/ #R1244226-50) and human adult intestine (Clonetech Laboratories, Inc., Missouri, US, http://www.clontech.com/ #636539) were purchased.

Intestinal Differentiation of Mouse ESCs

ESCs were cultured on M15 cells added with 20 ng/ml activin (R&D Systems, Inc., Minneapolis, http://www.rndsystems. com) and 50 ng/ml FGF2 (Peprotech, Rocky Hill, NJ, http://www.peprotech.com) in DMEM media containing 10% FBS and 4,500 mg/ml glucose for 5 days and were analyzed using flow cytometry for DE. For intestinal differentiation, ESCs were further cultured on M15 or MEF cells in the presence of 5 μM bromoindirubin-3′-oxime (BIO) (Calbiochem, San Diego, http://www.emdbiosciences.com) and 10 μM N-[(3,5-Difluorophenyl)acetyl]-L-alanyl-2-phenylglycine-1,1-dimethylethyl ester (DAPT) (Peptide Institute. Inc., Osaka, Japan, http://www.peptide.co.jp), or without BIO or DAPT but with FGFs, in media with 10% knockout serum replacement (KSR, Invitrogen) at a glucose concentration at 2,000 mg/ml.

Maintenance of hESCs

The hESCs (KhES-3) [19] were a gift from Dr. N. Nakatsuji and Dr. H. Suemori (Kyoto University, Kyoto, Japan) and were used following the hESC guidelines of the Japanese government. Undifferentiated hESCs were maintained on a feeder layer of MEF in Knockout DMEM/F12 (Invitrogen) supplemented with 5 ng/ml FGF2, 20% KSR, L-Gln, NEAA, and β-ME in 3% CO2. To passage hESCs, hESC colonies were detached from the feeder layer by treating with 0.25% trypsin (Invitrogen) and 0.1 mg/ml collagenase IV (Gibco, Grand Island, NY, http://www.invitrogen.com) in Phosphate Buffered Saline (PBS) containing 20% KSR and 1 mM CaCl2 at 37°C for 5 minutes, followed by adding culture medium and disaggregating ESC clumps into smaller pieces (5–20 cells) by gentle pipetting several times.

Intestinal Differentiation of hESCs

For differentiation studies, hESCs, pretreated with Y27632 (Wako) for 24 hours, were plated at 50,000 cells per well in 24-well plates that had been previously coated with M15 cells. ESCs were dissociated with 0.25% trypsin-EDTA and cultured in Y27632 containing ES maintenance medium for 1 day. Twenty-four hours after plating, cells were washed with PBS and medium was changed to differentiation media. The cells were cultured in first differentiation medium [RPMI1640 (Invitrogen) supplemented with 2% B27 (Invitrogen), NEAA, L-Gln, PS, and β-ME] on days 0–10, switched to second differentiation medium (DMEM supplemented with 10% KSR, glucose 2,000 mg/ml, NEAA, L-Gln, PS, and β-ME) at day 10, and cultured up to 35 days. Activin A (100 ng/ml) was added on days 0–10 of differentiation; BIO and DAPT were added on days 10–35. Media were replaced every 2 days with fresh medium supplemented with growth factors.

Growth Factors and Inhibitors

The following concentrations were used unless specifically indicated. BIO (Calbiochem) 5 μM; DAPT (Peptide Inst. Inc.) 10 μM, and recombinant human activin-A (R&D Systems), 20 ng/ml for mouse ESCs, 100 ng/ml for hESCs; recombinant human FGF2 (PeproTech) 50 or 256 ng/ml; SU5402 (Calbiochem) 10 μM; Noggin 500 ng/ml (R&D systems); Dorsomorphin 200 nM (Sigma-Aldrich); LE540 1 μM (Wako); AMD3100 10 μM (Sigma-Aldrich); BMP4 25 ng/ml (R&D systems); and 4-hydroxytamoxifen (4-OHT) 500 ng/ml (Sigma).

Flow Cytometry Analysis and Reculture of Sorted Cells

Cells were dissociated with Cell Dissociation Buffer (Invitrogen), adjusted to 1 × 106 cells/50 μl and stained with appropriate antibodies. The following antibodies were used: either biotin- or Alexa 488-conjugated anti-E-cadherin monoclonal antibody ECCD2, phycoerythrin (PE)-conjugated anti-mouse Cxcr4 mAb 2B11 (BD Pharmingen) Allophycocyanin-conjugated anti-human CXCR4 12G5 Ab (BioLegend), or PE-Cy7-conjugated anti-human CD117 104D2 Ab (BioLegend). The stained cells were recovered with fluorescence-activated cell sorting (FACS) Aria (BD Pharmingen, San Diego, http://www.bdbiosciences.com). Data were recorded using the BD FACS Diva Software program (BD Pharmingen) and analyzed using the Flowjo program (Tree Star, Star, Ashland, OR, http://www.treestar.com).

After sorting, E-cadherin+ Cxcr4+ cells were plated at 5,000 cells per well in 96-well plates precoated with either M15 or MEF cells. These cells were cultured in intestinal differentiation medium.

Reverse Transcription-Polymerase Chain Reaction Analysis

RNA was extracted from ESCs using TRI Reagent (Sigma-Aldrich) or the RNeasy micro-kit (Qiagen, Qiagen, Hilden, Germany, http://www1.qiagen.com) and then treated with DNase (Sigma-Aldrich). Three micrograms of RNA were reverse-transcribed using MMLV reverse transcriptase (Toyobo, Osaka, Japan, http://www.toyobo.co.jp/e) and oligo dT primers (Toyobo). The primer sequences and number of cycles are shown in supporting information Table S1. The polymerase chain reaction (PCR) conditions for each cycle were: denaturation at 96°C for 30 seconds, annealing at 60°C for 2 seconds, and extension at 72°C for 45 seconds. Reverse transcription-PCR (RT-PCR) products were separated by 5% nondenaturing polyacrylamide gel electrophoresis, stained with SYBR Green I (Molecular Probes, Eugene, OR, http://probes.invitrogen.com), and visualized using a Gel Logic 200 Imaging System (Kodak, Rochester, NY, http://www.kodak.com).

Antibodies

Antibodies used were as follows: mouse anti-Cdx2 (BioGenex, San Ramon, CA, http://www.biogenex.com), rat anti-mouse E-cadherin (TaKaRa BIO Inc., Shiga, Japan, http://www.takara-bio.co.jp), goat anti-HNF4a (Santa Cruz Biotechnology Inc., CA, http://www.scbt.com), and mouse anti-Villin (BD Transduction Laboratories, San Diego, http://www.bdbio sciences.com). Rabbit anti-Lysozyme (Diagnostic Biosystems, CA, US, http://dbiosys.com), rabbit anti-Chromogranin A (Epitomics Inc., CA, US, http://www.epitomics.com), biotin-conjugated Dolichos biflorus agglutinin (DBA) lectin (Sigma-Aldrich), goat anti-somatostatin (Santa Cruz Biotechnology Inc.), mouse anti-Muc2 (Visionbiosystems Novocastra, Benton Lane, UK, http://www.vision-bio.com), and rabbit anti-Claudin-7 (Abcam, Cambridge Science Park, UK, http://www. abcam.co.jp), anti-Nanog (Reprocell, Kanagawa, Japan, http://www.reprocell.com). Control IgG (mouse, rabbit, goat) (Beckman Coulter, Inc., CA, US, http://www.beckmancoulter.com), anti-activated β-catenin (Millipore, Billerica, MA, http://www.millipore.com), and anti-NICD (Abcam).

Alkaline Phosphatase Activity Measurement

The culture cells were fixed in 4% paraformaldehyde for 10 minutes. After washing with phosphate buffered saline containing 0.1% Tween 20 (TBST) for 20 minutes, a coloring reaction was carried out with 35 μg/ml nitroblue tetrazolium and 17.5 μg/ml 5-bromo-4-chloro-3-indolyl phosphate in Alkaline phosphatase buffer (100 mM Tris-HCl [pH 9.5], 100 mM NaCl, 50 mM MgCl2, 0.1% Tween 20, and 2 mM levamisole).

Periodic Acid-Schiff Staining

Cells were fixed in 4% paraformaldehyde for 10 minutes. Periodic acid-Schiff (PAS) staining solution (Muto Pure Chemicals, Tokyo, Japan, http://www.mutokagaku.com) was used according to the manufacturer's instructions.

Differentiation of Mouse and hESCs Under Feeder-Free Condition

For the differentiation study, mouse ESCs were plated at 6,900 cells per square centimeter in gelatin-coated dish. The cells were cultured in DMEM containing 4,500 mg/l glucose, supplemented with NEAA, L-Gln, PS, β-ME, Insulin-transferrin-sodium selenite media supplement (ITS) [10 μg/ml insulin (Sigma-Aldrich), 5.5 μg/ml transferrin (Sigma-Aldrich), and 6.7 pg/ml selenium (Sigma-Aldrich)], 0.25% Albmax (Invitrogen), and 10 ng/ml recombinant human activin-A (R&D Systems) for 7 days, and then, the media were changed to 10% KSR at a glucose concentration at 2,000 mg/l in the presence BIO and DAPT with growth factors.

For the differentiation study, hESCs were plated at 69,000 cells per square centimeter onto dishes precoated with gelatin. The cells were cultured for 7 days in RPMI 1640 medium (Invitrogen) supplemented with NEAA, L-Gln, PS, β-ME, 100 ng/ml recombinant human activin-A, and B27 SUPPLEMENT (Invitrogen), then the media were changed to 10% KSR at a glucose concentration at 2,000 mg/l in the presence BIO and DAPT with growth factors.

Measurement of Cell Proliferation

Cell proliferation was measured by the incorporation of EdU (Invitrogen) into genomic DNA during the S-phase of the cell cycle, and processed with a Click-iTtm EdU Kit, according to the manufacturer's protocol.

Caco2 Cell Culture and Differentiation

Caco2 cells were cultured by DMEM medium containing 10% FBS until confluent. After that, the medium was changed to DMEM medium containing 10% FBS or 10% KSR with or without FGF2.

RESULTS

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

Activation of the Canonical Wnt Signaling and Inhibition of Notch Signaling Potentiate Intestinal Differentiation of ESCs

To establish optimal conditions for differentiation into intestinal cell lineages, ESCs were differentiated into DE and were challenged with various growth factors or chemicals that affect certain signaling pathways from day 5, then assayed for intestinal marker expressions on day 12 (Fig. 1A). The growth factors and chemical inhibitors tested are based on our previous results on the expression profile of M15 cells (supporting information Table S2). Among these tested growth factors and chemicals, we found that the simultaneous application of 6-bromoindrirubin-3′-oxime (BIO), a glycogen synthase kinase-3β inhibitor, that activates canonical Wnt/β-catenin signaling, and DAPT, a γ-secretase inhibitor that functions as an inhibitor of Notch signaling, dramatically increased the expression of intestinal markers Cdx2 [20] and Intestinal fatty acid binding protein (Ifabp) [21], confirmed by real-time PCR on day 12 (Fig. 1B). Immunocytochemical analysis confirmed that a high proportion of ESCs are differentiated into Cdx2-expressing cells by adding BIO and DAPT simultaneously (day 12, Fig. 1C). RT-PCR revealed the expression of intestinal markers Villin1 [22], Cdx2, Ifabp, Lactase (Lct), and intestine-specific homeobox (Isx) [23] were sequentially turned on (Fig. 1D). The expressions of these markers, except Isx, were induced or much enhanced by these two chemicals (Fig. 1E). By contrast, BIO and DAPT did not induce Cdx2 expression in M15 cells alone (Fig. 1F). At this stage, the ESC-derived Cdx2+ cells were still highly proliferative (supporting information Fig. S1). Since Notch is reported to post-translationally regulate β-catenin [24], we then confirmed that the effects of BIO and DAPT were through activation of Wnt/β-catenin and inhibition of Notch signaling in our ESC differentiation system. Differentiated ESCs treated with BIO, DAPT, or both were harvested at day 12 and analyzed for activated β-catenin or NICD (Fig. 1G). NICD is detected as a 85 kDa fragment [25, 26]. BIO alone induced activated form of β-catenin without interfering with NICD. Decrease of NICD was observed by DAPT addition, which was further synergized by BIO (Fig. 1G). To further confirm that activated Notch is inhibitory for intestinal differentiation, we used a NERT ΔOP ESC line introduced with an inducible form of intracellular domain of murine Notch 1 (Fig. 1A) [27]. The NERTΔOP ESCs were used instead of R1 ESCs and were subjected to differentiation into the intestinal lineage at the presence of BIO and DAPT. The expression of Notch signaling effector Hairy and enhancer of split (Hes1) was still observed at the presence of BIO and DAPT, without tamoxifen induction. At the presence of BIO and DAPT, Notch signaling is activated upon tamoxifen (4-OHT) induction, as shown by the upregulation of a Hes1 (Fig. 1H). The elevated level of Notch signaling switched off Cdx2 expression (Fig. 1I, lower panel). Taken together, these results demonstrated that activation of Wnt/β-catenin and inhibition of Notch signaling resulted in differentiation into the intestinal epithelium of ESCs.

thumbnail image

Figure 1. Simultaneous BIO and DAPT addition potentiates intestinal differentiation of mouse embryonic stem cells (ESCs). A serial of growth factors and inhibitors were tested. Addition of BIO and DAPT revealed to be effective in potentiating ESC differentiation into intestinal epithelia on feeders. (A): A schematic drawing of the experimental design. Mouse ESCs were differentiated into definitive endoderm for 5 days, then switched to BIO and DAPT containing medium on M15. (B): Real-time PCR and (C) immunocytochemical analyses of ESCs treated with different combinations of BIO and DAPT. (D): RT-PCR analyses of time-dependent expression of various intestinal markers in the presence of BIO and DAPT. (E): RT-PCR analyses of marker expression at day 20 differentiation, at the absence or presence of BIO and DAPT (BD). (F): Immunocytochemical analysis of M15 treated with BIO and DAPT. M15 cells did not express Cdx2. (G): Western blot analysis of day 12 differentiated ESCs for activated β-catenin and NICD by BIO and DAPT treatments. (H, I): 4-OHT induced NICD and initiated Hes1expression, detected by real-time PCR (H). This resulted in a downregulation of Cdx2 expression, detected by immunohistochemistry (I). Values represent means ± SEM (n = 3). *, p < .05 versus control by Student's t test. DW, a negative control without cDNA. Scale bars = 100 μm. 4500FBS, glucose at 4,500 mg/ml, with FBS. 2000KSR, glucose at 2,000 mg/ml with KSR. Abbreviations: AI, adult intestine; DAPI, 4′,6-diamidino-2-phenylindole; FBS, fetal bovine serum; FI, fetal intestine; FGF2, fibroblast growth factor 2; NICD, Notch1 intracellular domain; 4-OHT, 4-hydroxytamoxifen.

Download figure to PowerPoint

BIO and DAPT Induced Posterior Patterning of the DE on M15 Cells

To confirm the DE origin of the differentiated cells, Cxcr4- and E-cadherin (ECD)-positive fraction were isolated on day 4 by flow cytometry and recultured on M15 feeders with BIO and DAPT until day 15 (Fig. 2A, 2B). After replating, the DE cells differentiated into Cdx2 or Villin-expressing intestinal cells (Fig. 2C), thus indicating the DE origin of the Cdx2 or Villin-expressing intestinal cells.

thumbnail image

Figure 2. Addition of BIO and DAPT posteriorizes the DE. Embryonic stem cell-derived DE cells were isolated and tested for its responsiveness to graded concentrations of BIO and DAPT. (A): A schematic drawing of the experimental design. On day 4, DE (Cxcr4+/ECD+) cells were sorted and replated on M15 until day 15. (B): DE (Cxcr4+/ECD+) cells (square) sorted on day 4 were then replated. (C): This yielded Cdx2 (left) and Villin (right)-positive cells. (D, E): Simultaneous BIO and DAPT addition at graded concentrations induced anterior-to-posterior patterning of DE. Scale bar = 100 μm. Abbreviations: AI, adult intestine; DAPI, 4′,6-diamidino-2-phenylindole; DE, definitive endoderm; FBS, fetal bovine serum; FI, fetal intestine; FGF2, fibroblast growth factor 2; KSR, knockout serum replacement.

Download figure to PowerPoint

We next applied different concentrations of BIO and DAPT and examined the expression of a panel of markers indicative of anterior-to-posterior region specificity. Paired box gene 8 (Pax8) [28], an anterior marker for thyroid differentiation, was maximally expressed at the lowest BIO and DAPT concentration tested. Cdx2 and Ifabp were induced at higher concentrations, up to 5 μM BIO and 10 μM DAPT. Homeobox C8 (Hoxc8) [29], a posterior marker, was induced maximally at 15 μM BIO and 30 μM DAPT (Fig. 2D). When BIO was applied at 5 μM, increasing DAPT concentration inhibited the expression of Pax8 (Fig. 2E), without affecting much on Hoxc8 expression. When DAPT was applied at 10 μM, increasing BIO concentration induced Hoxc8 expression. Therefore, activation of canonical signaling by BIO and blockade of Notch signaling prevented the anterior fate of the thyroid gland endoderm to turn on and promoted the posterior fate of the intestinal endoderm. Conversely, the expression of tissue-specific markers such as Pdx1 or Alb was not induced by BIO and DAPT (S.O., unpublished observations). Taken together, activation of Wnt/β-catenin and inhibition of Notch signaling pathways play a role in anterior-posterior specification of the intestinal epithelium, by antagonizing the expression of anterior marker Pax8 and potentiating the expression of posterior markers Ifabp, Cdx2, and Hoxc8.

Mouse ESCs Differentiate into All Lineages of the Absorptive and Secretory Cells of the Intestine

We then examined the cell types of the intestinal cells differentiated from ESCs more closely by immunocytochemistry. DE cells were sorted by flow cytometry, replated onto MEF with BIO and DAPT, and assayed at day 20. MEF was also potent for supporting ESC differentiation into intestinal cells, and thus was used instead of M15. 87.8% ± 3.0% of cells turned into Cdx2-positive cells (Fig. 3A). These Cdx2-expressing cells also expressed E-cadherin, an epithelial marker [30] and Hepatic nuclear factor 4 alpha (HNF4a) [31], a key factor for intestinal epithelial architecture (Fig. 3B). They are also positive for Claudin7, an intestinal epithelial tight junction marker [32] (Fig. 3C). Mature cell type-specific markers, such as Lyz1 as a Paneth cell marker (2.4% ± 0.5%), ChromograninA (Chga) (3.0% ± 0.9%), Somatostatin (Sst) as enteroendocrine markers (Fig. 3D), Mucin2 (Muc2) [33], and lectin DBA [34] (10.4% ± 3.5%) as goblet cell markers, were expressed (Fig. 3E). MEF were confirmed not to express Cdx2 under this differentiation condition (supporting information Fig. S2A). No stainings were observed with specific isotype negative controls of the primary antibodies (supporting information Fig. S2B). These results indicate that ESCs are induced to differentiate into all four mature intestinal cell types, the absorptive enterocytes and three secretory cells, namely, Paneth cells, goblet cells, and enteroendocrine cells of the intestinal cell lineages.

thumbnail image

Figure 3. Mouse embryonic stem cells (ESCs) differentiated to form all intestinal cell types. Mouse ESCs differentiated on feeders and addition of BIO and DAPT were assayed for expression of intestinal markers. (A–E): Mouse ESCs differentiated on mouse embryonic fibroblast were examined on differentiation day 20, with their expression of Cdx2, E-Cadherin (ECD), Hnf4a, Villin, Claudin7, along with various differentiated markers, Lyz1, Chga, Sst, Muc2, and DBA. Numbers in parentheses are the proportions of positively stained cells within the replated cells. Scale bars = 100 μm. Abbreviation: ECD, E-cadherin.

Download figure to PowerPoint

Differentiation of hESCs into All Mature Intestinal Cells

We next examined if hESCs could also be directed into intestinal cells using the equivalent procedure. hESCs, KhES-3 [19], were differentiated into DE by culturing on M15 and added with activin A for 10 days, then were continuously cultured on M15 or replated onto MEF, with BIO and DAPT (Fig. 4A). CDX2 was detectable from day 25 and reached a maximum on day 30 by immunohistochemistry (Fig. 4B). An approximately 20-fold upregulation in CDX2 expression by BIO and DAPT was confirmed by real-time PCR (Fig. 4C). Molecular markers for enterocytes (IFABP and ISX), goblet cells (TFF3) [35], Paneth cells (LYZ), enteroendocrine cells (GAST, SYP, and SST), and an ISCs marker gene, human Leucine-rich repeat-containing G protein-coupled receptor-5 (LGR5) [36], were induced, as revealed by RT-PCR (Fig. 4D). Differentiated cells were positive for alkaline phosphatase (ALP) (79%) and PAS stainings, indicating that functional enterocytes were induced at a high efficiency and goblet cells are also derived (Fig. 4E, 4F). Immunocytochemical analysis showed that KhES-3-derived CDX2-expressing intestinal cells (87.6% ± 3.9%, within total cells) were positive for HNF4a (Fig. 4F). Among VILLIN-positive intestinal cells, LYZ-positive Paneth cells were 5.5% ± 1.3%, CHGA-expressing enteroendocrine cells were 5.8% ± 1.0%, and MUC2/DBA expressing cells were 11.1% ± 1.5% (Fig. 4G). Thus, hESCs can be differentiated into all intestinal cell lineages, similarly to mouse ESCs, through activation of Wnt/β-catenin and inhibition of Notch signaling, and the induced Cdx2-expressing cells then differentiate into all differentiate cell types of the intestine.

thumbnail image

Figure 4. Human embryonic stem cells (ESCs) differentiated into all intestinal cell types. Human ESCs differentiated on feeders and addition of BIO and DAPT were assayed for expression of intestinal markers. (A): A schematic presentation of the experiment. KhES-3 were differentiated into definitive endoderm on M15 for 10 days, then switched to intestinal differentiation medium (with BIO and DAPT) on M15 (B–E) or MEF (F). Time course of the appearance of CDX2-expressing intestinal cells detected by immunohistochemistry (B) or CDX2 transcripts quantified by real-time PCR on day 30 (C). **, p < .01 versus control by Student's t test (n = 3). Time course of the expression of various intestinal markers is shown (D). Differentiated cells showed ALP activities (ALP+/CDX2+ = 79%) (E) and were positive for PAS staining (F). (G): Expression of markers for differentiated intestinal cell types. Numbers in parentheses are the proportions of positively stained cells within differentiated KhES-3 cells (CDX2) or CDX2-positive intestinal cells (LYZ, CHGA, and DBA). Scale bars = 100 μm. Abbreviations: AI, adult intestine; ALP, alkaline phosphatase; DAPI, 4′,6-diamidino-2-phenylindole; FI, fetal intestine; FGF2, fibroblast growth factor 2; KSR, knockout serum replacement; MEF, mouse embryonic fibroblast; PAS, periodic acid-Schiff.

Download figure to PowerPoint

The Induction of Cdx2-Expressing Intestinal Endoderm is Mediated Through FGF, and Potentiated by BMP Signaling Pathway, Concurrently with Wnt and Notch in the Mouse ESCs

Wnt are known to be expressed in gradients and acts in patterning the neuronal identity along the anterior-posterior (A-P) axis, in which Wnt acts with FGF or other signaling molecules [37]. The above experiments were performed on feeder cells. To reveal the role of the feeders, we then explored the underlying signaling pathways that work concurrently with Wnt/β-catenin and Notch signaling in our present differentiation procedure, using chemical inhibitors. DE were isolated by flow cytometry and replated onto MEF, and were treated with BIO and DAPT, together with various chemical inhibitors to screen for signaling pathways that interfere Cdx2 expression (Fig. 5A). SU5402 showed inhibitory effect for DE cells to differentiate into Cdx2-expressing cells, detected by immunohistochemistry and RT-PCR on day 15 (Fig. 5B, 5C). Noggin and dorsomorphin (inhibitors for BMP) but not LE540 (a retinoic acid inhibitor) or AMD3100 (stromal cell-derived factor inhibitor) showed inhibitory effects for inducing Cdx2-expressing cells (Fig. 5D). We previously reported that M15 cells expressed FGF2 and BMP4 [17]. Our above results therefore indicated that FGF and BMP signals that potentiate differentiation might be emitted from feeder cells, and concurrently function with Wnt/β-catenin and Notch signalings to induce Cdx2 expression in ESC-derived DE cells.

thumbnail image

Figure 5. FGF and bone morphogenetic protein (BMP) signalings potentiated intestinal differentiation. Inhibitors for FGF and BMP signaling pathways were tested and revealed to inhibit intestinal differentiation of mouse embryonic stem cells. (A): A schematic drawing of the experimental design. (B, C): Intestinal differentiation (on MEF) by BIO and DAPT (BD) was partially inhibited by SU5402 (BD+SU5402) and analyzed by real-time PCR (n = 3) (B) or immunocytochemistry with anti-Cdx2 antibody (C) on day 12. (D):Cdx2 expression was inhibited by BMP inhibitors, such as Noggin and Dorsomorphin, but not affected by retinoic acid inhibitor, LE540, or Cxcr4 inhibitor, AMD3100 (n = 4). Scale bars = 100 μm. *, p < .05; **, p < .01 versus control (white bars) by Student's t test. Abbreviations: AI, adult intestine; DAPI, 4′,6-diamidino-2-phenylindole; DE, definitive endoderm; FBS, fetal bovine serum; FGF2, fibroblast growth factor 2; KSR, knockout serum replacement; MEF, mouse embryonic fibroblast.

Download figure to PowerPoint

We then try to perform gain of function experiments by differentiating ESCs under feeder-free condition, and asked if the activation of FGF and BMP signaling instead of using feeder cells could induce ESC differentiation into Cdx2-expressing cells. In the feeder-free differentiation system, ESCs were added with activin A and ITS to induce DE (days 0–7), then added with BIO and DAPT and tested with various growth factors until day 15 and assayed for Cdx2 expression by immunohistochemistry and quantified by flow cytometry (Fig. 6A). Under this feeder-free system, 52.4% of ESCs differentiated into ECD/Cxcr4-positive DE cells, detected on day 7 (Fig. 6B, without [upper] or with [lower] primary antibodies), and then differentiated into Cdx2-expressing cells (Fig. 6C, day 15). Addition of FGF2 increased the proportion of Cdx2-positive cells, assayed on day 12 or day 15 (Fig. 6D). BMP4 alone had no effects, but exerted potentiating effects when added simultaneously with FGF2, on day 15, but not day 12. These results indicated that BMP4 worked at the presence of FGF2 to potentiate differentiation at late stages (Fig. 6D). We examined the differentiated cell types and confirmed that cells positive for ALP or PAS activity and ChgA-expressing cells were induced, albeit at a lower efficiency compared to those cultured on M15 or MEF feeders (Fig. 6E). In these experiments, no Nanog-positive cells were observed in day 15 differentiated ESCs treated with BIO and DAPT, as compared to undifferentiated ESCs (supporting information Fig. S3), indicating that the PAS, Cdx2-positive cells were intestinal differentiated cells but not pluripotent cells.

thumbnail image

Figure 6. FGF2 and BMP4 upregulated intestinal differentiation from mouse ESCs under feeder-free conditions. Mouse ESCs were differentiated into intestinal fate under feeder-free condition. Addition of FGF2 and BMP4 synergistically acted with BIO and DAPT to potentiate intestinal differentiation. (A): A schematic drawing of the experiment. Mouse ESCs were differentiated into definitive endoderm (DE) for 7 days, then switched to intestinal differentiation medium (with BIO and DAPT) containing FGF2, or BMP4. (B, C): Mouse ESCs were assayed for DE (Cxcr4+/ECD+) cells by flow cytometry on day 7; without (upper) or with (lower) primary antibodies (B), or Cdx2-positive intestinal cells on day 12 (C). (D): The proportions of Cdx2-positive cells, determined by flow cytometry, was upregulated by FGF2, this was further potentiated by BMP4, assayed on day 12 (upper) or day 15 (lower). *, p < .05; **, p < .01 versus control (white bars) by Student's t test (n = 4). (E): Enterocytes (ALP) and goblet cells (PAS) were observed on day 15, and Chga-positive enteroendocrine cells were observed on day 20. (F): A schema of intestinal differentiation from mouse ESCs. Scale bars = 100 μm. Abbreviations: ALP, alkaline phosphatase; BMP, bone morphogenetic protein; DAPI, 4′,6-diamidino-2-phenylindole; ESC, embryonic stem cell; FGF2, fibroblast growth factor 2; KSR, knockout serum replacement; PAS, periodic acid-Schiff.

Download figure to PowerPoint

Taken together, in mouse ESCs, FGF signaling cooperatively acts with Wnt/β-catenin and Notch pathways, and later on, BMP signaling then potentiate differentiation into Cdx2-positive intestinal epithelial. These cells then generated differentiated cell types of enteroendocrine, enterocytes, and goblet cells (Fig. 6F).

FGF Signaling Is Inhibitory for Induction of Cdx2-Positive Cells in KhES-3 Cells

We then examined whether in hESCs, FGF and BMP signalings also act similarly in regulating intestinal differentiation (Fig. 7A). KhES-3 cells were cultured under feeder-free condition, differentiated into DE cells, expressing SOX17 and FOXA2 (Fig. 7B, 7C). Flow cytometry analysis confirmed that 83% of cells in culture was CXCR4- and CD117- double positive DE cells on day 7 (Fig. 7D). Then, when added with BIO and DAPT, khES-3 cells differentiated into CDX2-positive intestinal cells (Fig. 7E).

thumbnail image

Figure 7. FGF2 inhibited intestinal differentiation from KhES-3 embryonic stem cells (ESCs). Human ESCs were differentiated into intestinal fate under feeder-free condition. Addition of FGF2 inhibited and BMP did not affected BIO and DAPT-mediated intestinal differentiation. (A): A schematic presentation of the experiment. KhES-3 cells were seeded onto gelatin-coated dish, and differentiated into definitive endoderm (DE) for 7 days, then switched to intestinal differentiation medium (with BIO and DAPT) containing FGF2 (50 or 250 ng/ml). (B–D): DE differentiation of KhES-3 was analyzed by immunocytochemistry (B), RT-PCR (C), and flow cytometry (D) on day 7. (E–G): CDX2-expressing cells detected by immunocytochemistry on day 20 (E). The effects of FGF2, BMP4, or both on CDX2+ cells are shown (F). FGF2 (50 and 250 ng/ml) inhibited differentiation of CDX2-positive cells assayed on day 9 (left panel) or day 20 (right panel) are shown (G). *, p < .05; **, p < .01 versus control (white bars) by Student's t test (n = 5). Images in (B) and (E) are counterstained with DAPI for nuclei. Scale bars = 100 μm. Abbreviations: BMP, bone morphogenetic protein; DAPI, 4′,6-diamidino-2-phenylindole; FGF2, fibroblast growth factor 2; KSR, knockout serum replacement.

Download figure to PowerPoint

Next, we added FGF2 and BMP4 with BIO and DAPT to test whether they enhance intestinal differentiation. FGF2 turned out to be inhibitory for intestinal differentiation, 50 and 250 ng/ml FGF2 inhibited intestinal differentiation (Fig. 7G). BMP4 did not affect intestinal differentiation. We tested the effects of SU5402 on khES-3 cells differentiated on feeders (supporting information Fig. S4). In the presence of SU54502, increase in CDX2 transcript expression was observed in the initial stage of differentiation at day 15, which was then caught up later on day 20 in control SU5402 untreated samples (supporting information Fig. S4). Taken together, unlike mouse ESCs, BMP signaling has no effects, while FGF2 inhibits intestinal differentiation of khES-3 cells.

DISCUSSION

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

We have reported that M15 potentiated differentiation into DE-derived tissues of the pancreas, liver, and intestine [17], through growth factors and extracellular matrices synthesized by M15 or ESCs themselves [38]. Here, we found that mouse or hESCs could be directed into the intestinal epithelium by activation of Wnt and inhibition of Notch signalings. Moreover, when cultured on M15 or MEF cells, the ESCs further differentiate into all intestinal cell types. It is reported that laminin α1 is expressed in the embryonic intestinal cells, therefore, the extracellular matrix in M15 or MEF seems to be important to support differentiation into mature cell types of the intestine [39].

The Wnt signaling cascade is implicated in the differentiation of intestinal cells. Tcf1 and Tcf4, which are targets of Wnt signaling, expressed in the hindgut region at E8.5 mouse embryo. Both Tcf1 and Tcf4 knockout mouse showed severe deficient of hindgut [40]. Here, we found that Notch signaling is inhibitory for differentiation into the intestinal fate from ESCs/iPS cells. Activation of Notch signaling inhibited differentiation into Cdx2-expressing intestinal epithelium. Inhibition of activated Notch by BIO and DAPT is crucial for intestinal differentiation. In our ESC differentiation system, DAPT alone did not completely inhibit activated Notch. However, BIO and DAPT synergistically inhibited Notch signaling. Wnt and Notch signaling activities are closely interdependent with each other, and that activation of Wnt signaling could cause a reduction of Notch signaling [41, 42]. Therefore, BIO and DAPT synergistically inhibited Notch signaling and released the inhibitory effect of Notch signaling for intestinal differentiation. At the presence of BIO and DAPT, the Notch activation was suppressed, but there were still some residual Hes1 expression. It is reported that the Hes1-Notch signaling promotes ISC proliferation and inhibited secretory cell development [43, 44]. The residual Hes1 expression might be permissive for enterocyte differentiation and the proliferation of the ISC in our present ESC system.

Conversely, a high concentration of DAPT inhibited the expression of Pax8, and only marginally triggered the expression of posterior marker Hoxc8. By contrast, increasing concentration of Wnt/β-catenin signal triggers the expression of posterior markers (Fig. 2D, 2E). Therefore, these results indicate that DAPT inhibits the expression of Pax8 while Wnt β-catenin posteriorizes the endoderm.

M15 and MEF cells secrete FGF2 and BMP4, revealed by expression profiling analysis [17], thereby exerting the potentiating effects. However, in the mouse ESCs, FGF2 and BMP4 signaling work cooperatively with Wnt/β-catenin activator and Notch inhibitor to drive ESCs toward Cdx2-expressing intestinal epithelium, which was not observed in a hESC line, khES-3 cells. The BMP signaling is reported to regulate stem cell fate in conjunction with other signaling pathways, through its downstream signaling molecule Smads, which might crosstalk with FGF signaling at the level of mitogen-activated protein kinase [45].

It is intriguing that FGF2 turned out to be inhibitory for hESC differentiation into intestinal epithelium. Addition of SU5402, an FGF inhibitor to khES-3 cells grown on MEF cells, enhanced initial differentiation into CDX2-expressing cells (supporting information Fig. S4). To further explore the mechanism, we tested the effects of FGF2 on Caco2, which is a colon cancer cell line widely used to reproduce in vitro the differentiation into absorptive enterocytes (supporting information Fig. S4). In Caco2 cells, FGF2 showed inhibitory effects for differentiation into Cdx2-positive enterocytes. Taken together, the discrepancy in the effects of FGF2 might be due to the stages or target cell type differences between the ESC differentiation systems. In fact, we observed time-dependent differences in sensitivity to FGF in khES-3 cells, in contrast to that in early-stage a high-dose of FGF was required to inhibit differentiation, a lower dose of FGF2 was enough to inhibit differentiation in Caco2 cells at late stage (Supporting Information Fig. S5).

After the intestinal epithelium is formed, BIO and DAPT then further potentiated the generation of differentiated cell types of the intestine. This step requires additional factors derived from M15 or MEF to increase the differentiation efficiency, since under feeder-free condition, the efficiency of differentiation into enterocytes, goblet cells, and enteroendocrine cells was not high compared to that induced under feeder condition.

The Wnt and Notch signalings are well known to maintain the proliferative state of the ISCs and keep them from differentiation [46–49]. It is tempting to hypothesize that BIO and DAPT might function to regulate the differentiation of the ISCs, which appear in the ESC-derived intestinal cells. The expression of Lgr5, an ISC marker [36], was detected in the present hESC differentiation system. Although we tried to manipulate the differentiation of ISC by activation of the Wnt or Notch signaling sequentially in a combinatory manner, but found it difficult. This might be due to the low population of the ISCs generated in the present differentiation procedure. Another possibility is that the ISCs that reside in the adult intestine do not give rise in embryonic stages but emerge later on. The Lgr5-expressing cells in our ESC system might not represent adult ISC. Taken together, our results suggest that ESCs system represents embryonic differentiation into mature cell types of the intestine, which differ from that of the adult stem cell system.

CONCLUSIONS

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

In conclusion, we demonstrated the efficient differentiation of human and mouse ESCs into all four mature intestinal cell types through inhibition of the Notch pathway with DAPT and the activation of the canonical Wnt signaling pathways with BIO. In mouse ESCs, FGF and BMP signaling cooperatively regulate intestinal differentiation, whereas in hESCs, FGF signaling seems to act as an inhibitory signal. These findings would be useful for future therapeutic application and investigation in developmental studies of the intestine.

Acknowledgements

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

We thank Dr. Daisuke Sakano and all the Kume Laboratory members for critical comments. We thank members of the CARD and Gene Technology Center at Kumamoto University for technical assistance. This work was supported by a Grant-in-Aid (#21390280 to S.K. and #21790671 to N.S.), and in part by a Global COE grant (Cell Fate Regulation Research and Education Unit, to S.K.) by “Funding Program for Next Generation World-Leading Researchers” (to S.K.), the Japan Society for the Promotion of Science, and the Realization of Regenerative Medicine from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) Japan.

REFERENCES

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

Supporting Information

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

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
sc-12-0751_sm_SupplFigure1.pdf239KSupplementary Figure 1 Cdx2+ cells are proliferative cells EdU incorporation tests assayed for 3 hours (d12) revealed that 45.6±1.51% (n=3) of the ES cell-derived Cdx2+ (green) cells were EdU positive (red). DAPI (blue) indicate nuclei. Scale bars: 100 μm.
sc-12-0751_sm_SupplFigure2.tif2027KSupplementary Figure 2 MEF cells did not express Cdx2. (A) Immunocytochemical analysis of MEF treated with BIO and DAPT. RT-PCR analyses of intestinal markers on day 20. ES cells differentiated on MEF without BIO and DAPT. Scale bars: 100 μm. (B) Immunocytochemical analysis using specific isotype negative control IgGs. Scale bars: 100 μm.
sc-12-0751_sm_SupplFigure3.tif1054KSupplementary Figure 3 There were no undifferentiated ES on day 15 under feeder free condition. Immunocytochemical analysis of undifferentiated ES cells and ES cells-derived cells on day 15 are shown. Nanog+ cells were not observed on day 15. Scale bar: 100 μm.
sc-12-0751_sm_SupplFigure4.tif136KSupplementary Figure 4 FGF inhibitor SU5402 treatment increased initial induction of CDX2. (A) A schematic presentation of the experiment. khES3 cells were cultured on feeders, added with SU5402 from day 10 to 30, and were assayed for CDX2 transcript by real time PCR. (B) Initial induction of CDX2 was potentiated by SU5402, but CDX2 expression in control caught up later. **p < 0.01 versus control (white bars) by Student's t-test (N=3-5).
sc-12-0751_sm_SupplFigure5.tif2100KSupplementary Figure 5 FGF2 inhibited intestinal differentiation in Caco2 cells. (A) A schematic presentation of the experiment. Caco2 cells were cultured to confluency, then differentiated at the presence or absence of FGF2. (B) Bright field views of Caco2 cells under different FGF2 conditions. FGF2 inhibited dome-like morphology of mature intestinal endoderm. (C, D) Flow cytometeric analysis of Caco2 cells treated with graded concentrations of FGF2. Differentiation into CDX2high population was inhibited by 250 ng/ml FGF. (E) RT-PCR analysis of Caco2 cells treated with FGF2. FGF2 inhibited CDX2 expression in Caco2 cells. **p < 0.01 versus control (white bar) by Student's t-test (N=3). Scale bars: 100 μm.
sc-12-0751_sm_SupplTable1.pdf76KSupplementary Table 1 The primer sequences and number of cycles. The primer sequences, cycle numbers for semi-quantitative RT-PCR and real time PCR are shown.
sc-12-0751_sm_SupplTable2.pdf87KSupplementary Table 2 The growth factors tested for potentiating the induction into Cdx2-positive intestinal cells. The growth factors and suppliers used in the experiments are shown.

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.