Stem Cells
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

First published online October 11, 2007
Stem Cells Vol. 26 No. 1 January 2008, pp. 3 -16
doi:10.1634/stemcells.2007-0194; www.StemCells.com
© 2008 AlphaMed Press

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
2007-0194v1
26/1/3    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Serafimidis, I.
Right arrow Articles by Gavalas, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Serafimidis, I.
Right arrow Articles by Gavalas, A.

EMBRYONIC STEM CELLS

Novel Effectors of Directed and Ngn3-Mediated Differentiation of Mouse Embryonic Stem Cells into Endocrine Pancreas Progenitors

Ioannis Serafimidisa, Irini Rakatzia, Vasso Episkopoub, Mina Goutia, Anthony Gavalasa

aDevelopmental Biology Laboratory, Biomedical Research Foundation of the Academy of Athens, Athens, Greece;
bMammalian Neurogenesis Group, MRC Clinical Sciences Centre, Imperial College School of Medicine, Hammersmith Hospital, London, United Kingdom

Key Words. Embryonic stem cells • Directed differentiation • Endocrine pancreas • Microarrays • Gene regulation • Signaling pathways

Correspondence: Correspondence: Anthony Gavalas, Ph.D., Biomedical Research Foundation of the Academy of Athens, Athens, Greece; Telephone: 0030-210-6597209; Fax: 0030-210-6597209; e-mail: agavalas{at}bioacademy.gr

Received on March 19, 2007; accepted for publication on September 28, 2007.

First published online in STEM CELLS EXPRESS  October 11, 2007.

   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 
The delineation of regulatory networks involved in early endocrine pancreas specification will play a crucial role in directing the differentiation of embryonic stem cells toward the mature phenotype of β cells for cell therapy of type 1 diabetes. The transcription factor Ngn3 is required for the specification of the endocrine lineage, but its direct targets and the scope of biological processes it regulates remain elusive. We show that stepwise differentiation of embryonic stem cells using successive in vivo patterning signals can lead to simultaneous induction of Ptf1a and Pdx1 expression. In this cellular context, Ngn3 induction results in upregulation of its known direct target genes within 12 hours. Microarray gene expression profiling at distinct time points following Ngn3 induction suggested novel and diverse roles of Ngn3 in pancreas endocrine cell specification. Induction of Ngn3 expression results in regulation of the Wnt, integrin, Notch, and transforming growth factor β signaling pathways and changes in biological processes affecting cell motility, adhesion, the cytoskeleton, the extracellular matrix, and gene expression. Furthermore, the combination of in vivo patterning signals and inducible Ngn3 expression enhances ESC differentiation toward the pancreas endocrine lineage. This is shown by strong upregulation of endocrine lineage terminal differentiation markers and strong expression of the hormones glucagon, somatostatin, and insulin. Importantly, all insulin+ cells are also C-peptide+, and glucose-dependent insulin release was 10-fold higher than basal levels. These data suggest that bona fide pancreas endocrine cells have been generated and that timely induction of Ngn3 expression can play a decisive role in directing ESC differentiation toward the endocrine lineage.

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


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 
Recent advances in stem cell biology, nuclear transfer, and the development of human embryonic stem cell (hESC) lines have opened up new prospects in cell replacement therapy for several diseases including diabetes. Clinical studies have suggested that insulin dependence in type 1 diabetes could be reversed if a renewable source of β cells or pancreatic islets were available [1]. Thus, directed differentiation of stem cells to endocrine pancreas fates is particularly important for the development of cell therapies for the treatment of diabetes. ESCs represent an attractive choice because they can differentiate into derivatives of all three germ layers of the embryo and they can respond appropriately to signals used in vivo during embryo development. Elucidation of the genetic networks involved in pancreas specification and their interaction with extracellular signals will be essential for devising rational protocols to efficiently differentiate ESCs or pancreas progenitor cells to fully differentiated endocrine subtypes [2–4]. Identification of pancreas stem and progenitor cell markers will also help in this process.

The developmental mechanisms involved in tissue specification during embryo development provide valuable guidelines for the in vitro differentiation of ESCs toward specific fates [2–5]. The pancreas develops from the definitive endoderm (DE) germ layer that is generated during the gastrulation stage of embryogenesis in the primitive streak [6]. DE initially consists of a flat layer of cells that folds to form the gut tube. The primordia of the different endoderm-derived organs are specified along the length of the gut tube according to their anteroposterior (AP) address. A combination of retinoic acid (RA), bone morphogenetic protein (BMP) signals from neighboring tissues [6, 7], and inhibition of the shh signaling [8] specifies the pancreas domain in the early embryonic endoderm as cells simultaneously expressing Ptf1a and Pdx1 transcription factors [9]. After initial bud formation, further pancreatic growth, branching, and differentiation depend on mesenchymal signals such as Fgf10 [10]. Expression of the basic helix-loop-helix (bHLH) transcription factor Ngn3 in this domain specifies the pancreatic endocrine lineage in a cell-autonomous manner [11, 12]. Ectopic expression of Ngn3 during gut development in vivo results in the expansion of pancreas endocrine cells [13, 14], and its expression in pancreatic ductal cells can activate aspects of the pancreas endocrine developmental program [15]. Ngn3 has been shown to regulate expression of Neurod1, Pax4, Nkx2.2, and IA1 [16–19]. Microarray studies have implicated potentially Ngn3-regulated genes in the developing pancreas [20, 21], in pancreatic ductal cells [15], and in embryoid bodies (EBs) [22]. These studies demonstrated the potential of Ngn3 to mobilize specific networks involved in endocrine specification. However, the analysis was performed either long after Ngn3 expression [20–22], in cellular context different from that of pancreas progenitors [15, 22], or Ngn3 upregulation was relatively low [15, 20]. Therefore, direct Ngn3 targets and the wider regulatory networks regulated by Ngn3 remain elusive.

Several studies have demonstrated that in vitro generation of pancreatic islets or β cells from mouse and human ESCs may be within reach [4, 23–28]. Despite the progress achieved, several limitations still exist, including generation of neuronal insulin-producing cells, uptake of insulin from the culture medium, inefficiency of differentiation and mixed lineages, low insulin content of the insulin-producing cells, lack of directed differentiation, and therefore unknown lineage. A number of criteria confirming true β cell identity, such as Pdx1, insulin I, and II expression and synthesis of the C-peptide, have not been simultaneously fulfilled [29] in these studies. Simulating the known in vivo sequence of events in the specification of the pancreatic endocrine lineage could provide a more efficient alternative route. A recent report following this approach demonstrated generation of pancreatic hormone-expressing endocrine cells from human ESCs, albeit with reduced efficiency of β cell generation [30]. This could be partly attributed to low expression of Ngn3 in intermediate steps.

To explore alternative routes of ESC differentiation and identify genetic networks for endocrine pancreas specification, we used directed ESC differentiation, using in vivo patterning signals and inducible Ngn3 expression in an orderly fashion. The approach combines the sequential induction of (a) endoderm character, (b) simultaneous induction of Ptf1a and Pdx1 expression through the use of appropriate stimuli, and (c) Ngn3-inducible expression. This combination enhanced the generation of endocrine pancreas progenitors and hormone-expressing cells. Furthermore, using this approach, we have identified several novel potential direct Ngn3 target genes through microarray gene expression profiling. These findings set the stage for the generation of bona fide pancreatic islets from ESCs, the identification of gene networks involved in endocrine lineage specification, and the analysis of the interplay between intrinsic determinants and extracellular signals in this developmental process.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 
Generation of Constructs
To generate the targeting vector R26/nls rtTA, we used the pBigT [31] and pROSA26-1 targeting vectors [32]. The pBigT vector was modified by removal of the HindIII/KpnI fragment containing the transcriptional stop, polylinker, and bovine growth hormone (bGH) polyadenylation signal (pA) and insertion of a linker restoring the AscI site to yield plasmid pBigT2. The XbaI/EcoRI sites at the 3' end of the splice acceptor sites of pBigT2 were removed, and an EcoRI/SmaI linker was inserted in their place to yield plasmid pBigT3. The nls rtTA cDNA/SV40 late pA signal cassette [33] was cloned as an EcoRI/PvuII fragment into the EcoRI/SmaI sites of pBigT3 to yield plasmid pBigT/nls rtTA. The PacI/AscI fragment containing the splice acceptor site, PGK-Neo-pA selection marker, and nls rtTA-pA fragment was removed from pBigT/nls rtTA and cloned in the PacI/AscI sites of a modified pROSA26-1 targeting vector plasmid to yield pROSA26-1/rtTA plasmid. The modification of the pROSA26-1 targeting vector consisted of a PacI/AscI polylinker insertion in its XbaI site.

To generate the Hprt-pBI-Ngn3 vector, the hemagglutinin (HA)-Ngn3-SV40pA cassette was cloned from plasmid pcDNA-Ngn3 (a gift from F. Guillemot) as an MluI/EcoRV fragment in the same sites of plasmid pBI2 to generate plasmid pBI-Ngn3. Plasmid pBI2 was derived from plasmid pBI (Clontech, Palo Alto, CA, http://www.clontech.com) by the insertion of a PmeI linker into its SapI site and the insertion of an AscI site into its AstII site. The Hprt targeting vector pSKB1 [34] was modified first by BamHI/PmeI digestion and religation to remove the BamHI, AscI, and PmeI sites at the 5' end of the Hprt homology region and subsequently by the insertion of a PmeI/AscI linker in the MluI/NotI sites of cloning polylinker to yield plasmid Hprt2. The AscI/PmeI fragment of plasmid pBI-Ngn3 was inserted into the same sites of plasmid Hprt2 to generate the Hprt-pBI-Ngn3 targeting vector. In this construct the Tet-Ngn3 transgene and Hprt locus have the same transcriptional orientation.

ESC Culture and Generation of Knock-in Lines
We grew HM-1 ESCs (a gift from D. Melton) in high-glucose Knockout D-MEM (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) supplemented with 10% fetal calf serum (PAN Biotech, Aidenbach, Germany, http://www.pan-biotech.com), 1,000 units of leukemia inhibitory factor (LIF) (Chemicon, Temecula, CA, http://www.chemicon.com) per ml, 100 mM minimal essential medium nonessential amino acids, 0.55 mM 2-mercaptoethanol, penicillin/streptomycin, and L-glutamine (all from Invitrogen) (ESC medium) on mouse embryonic fibroblasts treated with mitomycin (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) at 37°C and 5% CO2. HM-1 cells were electroporated with 30 µg of pROSA26-1/rtTA plasmid linearized with XhoI. After 10 days of selection in ESC medium containing 250 µg/ml neomycin (Invitrogen), clones were picked and analyzed by genomic Southern blot analysis as described for the ROSA26 locus [32]. The resulting ESTet-On cells were electroporated with 30 µg of PvuI-linearized Hprt-pBI-Ngn3 targeting vector. After 8 days of selection in hypoxauthine aminopterin thymidine (Sigma-Aldrich), medium colonies were analyzed by polymerase chain reaction (PCR) [35] and genomic Southern blots [34]. The resulting ESTet-On/Ngn3 clones were analyzed by Western blot analysis and immunofluorescence on ESCs grown on gelatin-coated plates, using the HA antibody in the presence (for 24 hours) and in the absence of 1 µg/ml doxycycline (dox) (Clontech).

ESC Differentiation Conditions
For all differentiation experiments, low-passage ESCs were used and passaged twice after thawing before initiating differentiation. ESCs were collected by trypsinization and separated from feeder cells by plating them onto tissue culture plates for two short (20–30 minutes) successive periods.

A short protocol was used to generate pancreas progenitor-like cells for Ngn3 target gene screening. ESCs were plated at 105 cells per milliliter in 10-cm bacterial dishes and allowed to form embryoid bodies for 4 days in ESC medium without LIF, changing the medium once, at 2 days. EBs were then transferred in Knockout D-MEM containing 10% Serum Replacement (SR) (Invitrogen), 1x penicillin/streptomycin/glutamine (Sigma-Aldrich), 0.5 mM ascorbic acid, and 0.45 mM monothioglycerol (Sigma-Aldrich) in the presence of Activin A (100 ng/ml) (R&D Systems Inc., Minneapolis, http://www.rndsystems.com) for 24 hours. RA was then added at 1 µM for another 24 hours, at which time EBs were dissociated into single cells by trypsinization and plated in Matrigel-coated (BD Biosciences, San Diego, http://www.bdbiosciences.com) tissue culture plates in modified N2 medium. Modified N2 medium contained 1x N2, 0.6% glucose, 1% GlutaMax, 0.4 mg/ml insulin, 0.1% laminin, and 0.5% B27 in Dulbecco's modified Eagle's medium (DMEM)-Ham's F-12 medium (reagents from Invitrogen), 10 ng/ml bFGF (R&D Systems) in the presence or absence of 1 µg/ml dox. Cells densities were 1 x 105, 1.75 x 105, and 2.5 x 105 cells per cm2 for samples collected by trypsinization at 48, 24, or 12 hours, respectively, after dox addition. To evaluate the effect of Ngn3 expression on differentiation into pancreas endocrine progenitors, a long protocol was used. ESCs were initially allowed to form EBs in suspension as described above for 2 days. They were then taken in Knockout D-MEM containing 10% SR, 100 ng/ml Activin A (R&D Systems), and 10 ng/ml Fgf4 (R&D Systems) for 3 days, renewing half the medium and adding fresh signaling molecules every 24 hours. In the next step, EBs were taken in Knockout D-MEM containing 10% SR, 100 ng/ml Activin A, 10 ng/ml Fgf4, 10 ng/ml Fgf10 (R&D Systems), 100 ng/ml BMP4 (R&D Systems), 1µM RA, and 2.5 µM cyclopamine (BIOMOL International LP, Plymouth Meeting, PA, http://www.biomol.com) for 2 days, renewing half the medium and adding fresh signaling molecules every 24 hours. EBs were then transferred in low-glucose (5.55 mM) DMEM containing 10% SR, 100 ng/ml BMP4, 10 ng/ml Fgf10, and 1 µg/ml dox (when applicable) for 2 days, changing the medium at 24 hours. EBs were finally taken in normal-glucose (25 mM) DMEM containing 10% SR medium in the presence of 10 mM nicotinamide for 5 days, changing medium every 2 days. Differentiated EBs were harvested at various stages of the protocol and processed for RNA extraction and reverse transcription (RT)-PCRs or for immunocytochemistry.

Immunofluorescence and Antibodies
Cells and EBs were collected at various stages of the differentiation protocol and fixed for 5 minutes at room temperature in 4% paraformaldehyde in phosphate-buffered saline (PBS). They were then washed in PBST (PBS with 0.1% Triton X-100) and subsequently permeabilized in 80% methanol for 15 minutes at –20°C. For immunofluorescence on paraffin sections, EBs were collected at stage V of the differentiation protocol and fixed overnight in 10% formalin in double-distilled H2O. Following a series of 10-minute sequential dehydration steps in ethanol (50%, 70%, 90%, and finally 100%), samples were incubated in xylene for 10 minutes and then in liquid paraffin for an additional 10 minutes. Sections of 6-µm thickness were cut on TESPA slides using a Leica RM2265 microtome (Heerbrugg, Switzerland, http://www.leica.com). For immunofluorescence, slides were immersed in xylene for 10 minutes for deparaffinization, and samples were rehydrated in a reverse series of ethanol incubations (100%, 95%, and 80%) followed by washes in PBS. Blocking was for 1 hour in PBST containing 10% normal goat serum (NGS). Primary and secondary antibodies were diluted in PBST containing 1% NGS. Cells and EBs were incubated with primary antibodies overnight at 4°C and with secondary antibodies at room temperature for 1 hour and mounted with 4,6-diamidino-2-phenylindole (DAPI)-containing Prolong Antifade (Molecular Probes, Eugene, OR, http://probes.invitrogen.com), and fluorescent images were taken using an inverted Leica SP5 confocal microscope. Technical controls were used where either primary or secondary antibodies were omitted, and these resulted in no fluorescence. Antibodies were as follows: Rabbit anti-Pdx1, 1:5,000 (a gift from C. Wright); rabbit anti-Ptf1a, 1:1,000 (a gift from H. Edlund); mouse anti-Ngn3, 1:1,000 (Developmental Studies Hybridoma Bank, Iowa City, IA, http://www.uiowa.edu/~dshbwww); mouse anti-HA, 1:200 (Cell Signaling Technology, Beverly, MA, http://www.cellsignal.com); rabbit anti-Gapdh, 1:2,000 (Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com); mouse anti-Nkx2-2, 1:100 (Developmental Studies Hybridoma Bank); mouse anti-Isl1, 1:500 (Developmental Studies Hybridoma Bank); mouse anti-insulin, 1:1,000 (Sigma-Aldrich); mouse anti-glucagon, 1:500 (Sigma-Aldrich); mouse anti-synaptophysin, 1:200 (Sigma-Aldrich); rabbit anti-C-peptide, 1:200 (Linco Research, St. Charles, MO, http://www.lincoresearch.com), and rabbit anti-somatostatin, 1:250 (Invitrogen, Carlsbad, CA, http://www.invitrogen.com). Secondary antibodies were anti-rabbit and anti-mouse Alexa 488-conjugated goat antibodies, 1:500 (Molecular Probes).

RT-PCR Analysis
Total RNA was isolated from ESCs, differentiated ESCs and EBs using the RNeasy kit (Qiagen, Hilden, Germany, http://www1.qiagen.com) according to the manufacturer's instructions and then digested by RQ1 DNase (Promega, Madison, WI, http://www.promega.com) to remove genomic DNA. First-strand cDNA synthesis was performed with Superscript II reverse transcriptase (Invitrogen) using random primers. Amplification of cDNAs was performed by PCR using 25–35 cycles with denaturation at 94°C for 30 seconds, annealing at 52°C–62°C for 60 seconds, and extension at 72°C for 60 seconds. With the exception of single-exon genes, all primers were designed to span introns using the MacVector software and conditions were optimized using 10.5-days post coitum mouse embryo total RNA. The primers and temperature used for specific PCRs are shown in supplemental online Table 1. PCR products were analyzed by agarose gel electrophoresis.

Microarray Gene Expression Profiling
RNA from three time points was used to generate biotinylated cRNA that was used to probe the Affymetrix Mouse Genome 430A array (Santa Clara, CA, http://www.affymetrix.com). Gene expression profiling was carried out in each time point for biological triplicates (cells taken independently through the differentiation procedure) for both dox-induced (Ngn3+) and uninduced (Ngn3–) cells. The growth of cells during the induction period was taken into account to harvest and profile cells in comparable cell densities. Total RNA was prepared using the RNeasy kit (Qiagen) according to the manufacturer's instructions, and it was digested by RQ1 DNase (Promega) to remove genomic DNA. The quantity and quality of the total RNA were assessed by UV spectroscopy using the NanoDrop ND-1000 (Wilmington, DE, http://www.nanodrop.com) and HPLC using the Agilent 2100 Bioanalyzer. Twenty micrograms of RNA were converted to double-stranded cDNA using the Roche Microarray cDNA Synthesis Kit and purified using the Roche Microarray Target Purification Kit as recommended (Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com). The cDNA was converted to biotinylated cRNA using the GeneChip IVT Labeling Kit as recommended (Affymetrix). The biotin-cRNA was purified using the RNeasy Mini Kit RNA cleanup protocol and analyzed by UV spectroscopy (NanoDrop ND-1000) and HPLC (Agilent 2100 Bioanalyzer). The purified biotin-cRNA was fragmented and hybridized to the GeneChip Mouse Genome 430A Array as recommended (Affymetrix), and the GeneChip was washed and scanned using the GeneChip Fluidics Station 400 and the GeneArray Scanner (Agilent Technologies) at the MRC Clinical Science Centre Genomics Core Facility. All data have been deposited in the ArrayExpress database (http://www.ebi.ac.uk/arrayexpress) and are accessible through accession ID/Username E-MIMR-461 and password agveo1. The robust multiarray average (RMA) [36] method was used for signal normalization at the probe level and across triplicate chips for each condition. For each time point (12, 24, and 48 hours), the untreated (–dox) samples were defined as the reference against which the treated (+dox) samples were compared to generate differential expression values. Statistical analysis was performed using SAM (Statistical Analysis for Microarrays) [37] and multiple testing correction using Benjamini and Hochberg false discovery rate (FDR) [38]. A list of statistically significant differentially expressed genes (FDR = 0) was generated, and fold-change values were calculated based on the RMA analysis.

Insulin Release Assay
Terminally differentiated EBs were collected at stage V of the protocol and were washed twice in PBS and once in Krebs-Ringer solution (120 mM NaCl, 5 mM KCl, 2.5 mM CaCl2, 1.1 mM MgCl2, 25 mM NaHCO3, and 0.1% bovine serum albumin). EBs were then incubated for 1 hour in Krebs-Ringer solution, which was then replaced by fresh buffer containing either low (2 mM) glucose or high (20 mM) glucose, or low glucose in the presence of 100 µM tolbutamide (Sigma-Aldrich). Incubation was for 1 hour at 37°C, after which the supernatants were collected and assayed for insulin release using an ultrasensitive mouse insulin enzyme-linked immunosorbent assay (ELISA) kit (Mercodia, Uppsala, Sweden, http://www.mercodia.com) according to the manufacturer's instructions. For protein content measurements, EBs were collected and sonicated in 150 µl of Krebs-Ringer solution, and the lysate was assayed using Bradford reagent (Sigma-Aldrich), according to the manufacturer's instructions.


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 
ESC Lines with Inducible Expression of Ngn3
To be able to express Ngn3 in a timely manner, we needed to generate ESC lines allowing its inducible expression. We chose the rtTA system because of its good inducibility characteristics [33] and incorporated its components in the genome, using knock-ins in constitutive loci (HPRT, ROSA26) to ensure accessibility of the transgenes upon differentiation and maximized phenotypic reproducibility due to isogenicity.

We used the HM-1 mouse ESC line, which carries a deletion in the HPRT locus [39]. Using a knock-in strategy, we incorporated the Tet-On cDNA [40] in the ROSA26 locus [32], generating the line ESTet-On (data not shown). Using a similar approach, a Tet-responsive transgene was incorporated in the HPRT locus of the ESTet-On cells to generate the line ESTet-On/Ngn3 (Fig. 1A, 1B). The Ngn3 cDNA was tagged with an HA tag, and dox-mediated inducibility of the Ngn3 Tet-responsive transgene was verified by immunofluorescence and Western blot analysis (Fig. 1C, 1D). Up to 70% of the ESCs (Fig. 1C) or ESC-derived pancreas progenitors (data not shown) express Ngn3 after dox induction. Thus, we generated ESCs affording tight regulation of the Ngn3 transgene.


Figure 1
View larger version (40K):
[in this window]
[in a new window]

  		Figure 1. Generation of mouse ESCs affording tight regulation of an Ngn3 transgene. (A): Schematic diagram of the HPRT knock-in strategy. Numbered boxes represent exons, the area in red represents the missing part of the endogenous HPRT gene, large arrows represent direction of transcription, and numbered arrows represent genotyping primers. (B): PCR genotyping of selected clones using specific primers for the restored wt or the del allele. (C, D): Expression of Ngn3 is tightly regulated by dox as shown by immunocytochemistry (C) and by Western blot analysis using an HA antibody (D). Scale bar = 100 µm. Abbreviations: bp, base pairs; DAPI, 4,6-diamidino-2-phenylindole; del, deleted; dox, doxycycline; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HA, hemagglutinin; Kd, kilodaltons; wt, wild-type.

 
Successive Activin A and RA Signals Generated an ESC-Derived Cell Population Responsive to Inducible Ngn3 Expression
To identify early Ngn3 downstream events in the specification of endocrine pancreas progenitors, the short protocol was used (as described in Materials and Methods). In this protocol, we set up a stepwise differentiation of ESCs to generate cells simulating the cellular context of pancreas progenitors. To achieve this, we used a sequential application of extracellular signals to initially induce and/or potentiate specific endoderm character while repressing alternative cell fates (mesoderm and ectoderm) and subsequently inducing pancreatic progenitor fates.

ESCs were grown as embryoid bodies for 4 days in ESC medium in the absence of LIF (stage I; Fig. 2A), and the expression of markers for ESCs (Rex1) [41], early mesendoderm (Gsc, Mixl1) [42, 43], endoderm (Gata4, Hnf3β, Hex1 and Sox17) [44], definitive endoderm (Cxcr4) [45], mesoderm (Gata1, Bra) [44, 46], and neuroectoderm (Sox1, Sox2, Pax6) [47–49] was assayed by semiquantitative RT-PCR. Collectively, the data suggested that this treatment induced endoderm and mesoderm character but not neuroectoderm character (Fig. 2B). In the next step (stage II; Fig. 2A), we sought to enhance mesendoderm and endoderm fates. Nodal signaling during early embryo development is necessary for definitive endoderm induction [50, 51]. Activin A acts through the same receptors and thus can mimic this induction in EBs [44, 50, 51]. Accordingly, subsequent treatment with Activin A for a single day, in the presence of serum replacement, enhanced the mesendoderm (upregulation of Mixl1 and Gsc expression) and endoderm character of the cells (retention of Hex1 and upregulation Hnf3β expression). A very small Sox1 upregulation was also seen (Fig. 2).


Figure 2
View larger version (32K):
[in this window]
[in a new window]

  		Figure 2. Differentiation of ESCs toward endocrine pancreas progenitors for the identification of novel Ngn3 target genes. (A): Schematic diagram of the differentiation procedure and Ngn3 induction. (B, C): Expression profiling by reverse transcription-polymerase chain reaction of several marker genes at each stage of the differentiation procedure (B) and of several known Ngn3 target genes in response to Ngn3 induction after successive use of ActA and RA (C). Abbreviations: ActA, Activin A; bFGF, basic fibroblast growth factor; d, days; dox, doxycycline; ES, embryonic stem; h, hours; hrs, hours; LIF, leukemia inhibitory factor; RA, retinoic acid; SR, serum replacement.

 
The objective of the next step was to restrict the character of the cells to that of early pancreas progenitors. RA is necessary for induction of the pancreatic primordium. Retinaldehyde dehydrogenase 2 (RALDH2) mutants fail to form the dorsal pancreatic bud [52, 53], and exposure of early endoderm to RA results in expansion of the presumptive pancreas territory [54]. Application of RA in the next step (stage III; Fig. 2A) had a dual effect. First, it further restricted cell fate by suppressing the expression of the mesendoderm markers Gsc and Mixl1, suggesting a transition of the endoderm character to a more mature state and to some degree repressing alternative mesoderm (Bra downregulation) and neuroectoderm (Sox1 downregulation) fates (Fig. 2B), respectively. However, expression of Gata1 was maintained and Pax6 was upregulated (Fig. 2B), suggesting the continued presence of mesoderm and ectoderm fates. Second, RA application induced expression of Pdx1 and Ptf1a, simultaneous expression of which marks the pancreas primordium in the early embryo [9]. This suggested that the proper cellular context has been generated to identify Ngn3 direct target genes relevant for endocrine pancreas development.

To directly test this assumption, cells were dissociated and plated on Matrigel in N2 medium containing bFGF. The following day, Ngn3 expression was induced by addition of dox, and expression of several markers was determined at 12, 24, and 48 hours postinduction (stage IV; Fig. 2C). Induction of Ngn3 expression was rapid and reached saturation as early as 12 hours, the earliest time point checked, after dox addition. As a response to Ngn3 transgene activation, expression of the known direct target genes in the pancreas, such as NeuroD, Pax4, Nkx2.2, and IA1 [16–19], as well as Dll1 [55], was strongly upregulated. Upregulation of NeuroD, Nkx2.2, Pax4, and Dll1 expression was observed as early as 12 hours postinduction (Fig. 2C). It was interesting to note that the kinetics of induction were different for the genes examined. Pax4 showed fast upregulation at 12 hours post-dox addition and subsequent downregulation, whereas IA1 was induced relatively late, at 48 hours post-dox induction. On the other hand, Neuro D and Nkx2-2 showed consistent regulation at all three time points (Fig. 2C). Importantly, ESTet-On/Ngn3 cells that were taken through the same differentiation procedure in the absence of Activin A and RA retained the capacity to upregulate the Ngn3 transgene in response to dox but failed to show appropriate regulation of NeuroD, Pax4, Nkx2.2, IA1, and Dll1 (Fig. 2C). Therefore, through the successive application of Activin A and RA, a cellular context mimicking that of pancreas progenitors was generated.

Gene Regulation in Response to Ngn3-Inducible Expression
The upregulation of known Ngn3 target genes 12 or 24 hours post-dox induction of the Ngn3 transgene suggested that this approach can be used to identify potential novel direct Ngn3 target genes. Toward this end, we performed microarray gene expression profiling at 12, 24, and 48 hours after addition of dox in ESC-derived putative pancreas progenitors. The short time frame and the fact that expression of Ngn3 and expression of genes later in the regulatory cascade may be mutually exclusive restricted the list of regulated probe sets.

Analysis of the microarray data using stringent statistical criteria showed that the number of regulated probe sets steadily increased with time (342 probe sets at 12 hours, 1,349 at 24 hours, and 2,299 at 48 hours). Venn diagrams identified 214 probe sets regulated in all three time points, thus adding statistical weight to these particular probe sets. Applying an additional fold regulation filter <0.7 and >1.4, the number of regulated probe sets dropped to 262 at 12 hours, 681 at 24 hours, and 1,847 at 48 hours. Interestingly, at 12 hours, there were twice as many upregulated probe sets as downregulated (172 vs. 90), whereas at 48 hours, probe sets up- and downregulated were evenly split (924 and 923 respectively).

Gene Ontology (GO) analysis performed using the GeneSpring software at 48 hours for the 1,847 probe sets suggested that Ngn3-regulated genes are primarily related to cellular morphogenesis, including cell adhesion, extracellular space, and cytoskeletal function. Specific signaling pathways identified were the Wnt pathway, the integrin-mediated signaling pathway, and the Notch pathway (Table 1). Strikingly, the relative importance of these categories remained virtually unchanged for all three time points, with the notable exception of regulation of cell proliferation and cytoskeleton organization (Table 1). Based on the GO analysis, the list of genes regulated at all three time points, and the strength of gene regulation, we compiled a partial list of novel genes regulated by Ngn3 expression (Table 2Go).


View this table:
[in this window]
[in a new window]

  		Table 1. Ngn3-regulated biological processes, molecular functions, and cellular components identified using systematic GO analysis of 1,847 Ngn3-regulated probe sets after 48 hours of induction of Ngn3 expression

 


View this table:
[in this window]
[in a new window]

  		Table 2. Partial list of novel genes regulated in response to Ngn3 induction

 


View this table:
[in this window]
[in a new window]

  		Table 2. (Continued)

 
To validate our approach, we performed semiquantitative RT-PCR for several genes that were regulated according to the microarray gene expression analysis. These included known Ngn3 targets (NeuroD1, IA1, Nkx2-2); members of the Notch signaling pathway, which is known to be involved in endocrine specification (Notch1, Dll1 and3, HeyL, Mfng, Hes5, Hes6); islet terminal differentiation markers (Isl1, Sst, Sstr2, Scg3); genes involved in G-protein signaling (Rgs4, Chn1), which has not been implicated in pancreas development so far; and extracellular matrix proteins (Spon1, Ctgf). To exclude the possibility that regulation is due to dox-specific effects and not Ngn3-specific effects, we took ESTet-On cells through the same differentiation procedure (including application of Activin A and RA), exposed them to dox, and assayed expression of the genes in question. The RT-PCR results (Fig. 3) confirmed the regulation of several genes, suggesting that this microarray gene expression analysis can be used as a starting point to identify and analyze novel genes in endocrine pancreas development.


Figure 3
View larger version (29K):
[in this window]
[in a new window]

  		Figure 3. Partial list of genes regulated by Ngn3 induction, as indicated by the microarray expression profiling and semiquantitative reverse transcription-polymerase chain reaction (RT-PCR). For the RT-PCR analysis, samples were collected from ESTet-On/Ngn3 cells taken through the differentiation procedure described in Fig. 2A in the presence or absence of dox for 48 h. Control samples were from ESTet-On cells taken through the same procedure. The list of genes shown includes known Ngn3 target genes, regulated components of the Notch pathway, endocrine pancreas terminal differentiation marker genes, and novel genes that could be involved in signaling during endocrine differentiation. Abbreviations: d, doxycycline; h, hours.

 
Elimination of Alternative Fates by Using a Broad Range of Signals in Pancreas Development
The short differentiation protocol described above generated cells with pancreas progenitor character that responded robustly to induction of Ngn3 expression. Thus, it was sufficient to conduct a screen for potential novel Ngn3 target genes, but it did not incorporate the full range of signals implicated in vivo in the generation of pancreatic progenitors. As a result, mesoderm fates were not fully suppressed, as evidenced by the continued expression of Gata1 and the appearance of a small number of beating muscle clusters at the last stage.

To more efficiently eliminate alternative cell fates, we used a long protocol that incorporated successive inductive signals used in vivo during pancreas development. Cells were maintained in suspension to allow for the formation of a three-dimensional structure and thus mimic cell interactions occurring in vivo. In stage I (Fig. 4A), we generated embryoid bodies in the presence of serum for 2 days because RT-PCR data suggested that endoderm fates predominated at that time point (compare expression of Gata1, Gsc, and Hnf3β in Figs. 2B and 4B; data not shown). In stage II (Fig. 4A), we used SR with a combination of Activin A to potentiate endoderm specification [44] and Fgf4 to induce anterior gut fates [56]. This treatment enhanced endoderm (upregulation of Hnf3β and Sox17 expression) and mesendoderm (upregulation of Mixl1 expression) character while partially repressing mesoderm and ectoderm fates (downregulation of Gata1, Pax6, and Sox1 expression) (Fig. 4B).


Figure 4
View larger version (39K):
[in this window]
[in a new window]

  		Figure 4. Differentiation of ESCs toward endocrine pancreas-like progenitors. (A): Schematic diagram of the differentiation procedure and Ngn3 induction. (B): Expression profiling by reverse transcription-polymerase chain reaction (RT-PCR) of several marker genes at each stage of the differentiation procedure. (C): Expression of Pdx1 and Ptf1a at the same stage was attributed to the use of ActA, RA, and cyc and not to the other signals used as shown by immunofluorescence. (D, E): Immunofluorescence using antibodies against Ngn3, Isl1, and Nkx2.2 in dox-untreated (–dox) and dox-treated (+dox) cells (D) and RT-PCR for Ngn3 and known Ngn3-target genes (E) at the end of stage IV of the differentiation procedure. Scale bar = 100 µm. Abbreviations: ActA, Activin A; BMP, bone morphogenetic protein; cyc, cyclopamine; dox, doxycycline; ES, embryonic stem; Fgf, fibroblast growth factor; LIF, leukemia inhibitory factor; RA, retinoic acid; SR, serum replacement.

 
In stage III (Fig. 4A), we maintained the presence of Activin A and Fgf4 while adding Fgf10 and BMP4 to expand putative pancreatic progenitors [10, 57]. Pancreas specification depends on RA [53, 58] and suppression of shh signals [59, 60]. Accordingly, at this step, we also included RA (1 µM) and cyclopamine (cyc), a shh inhibitor. As a result, we observed repression of mesendoderm and mesoderm markers Gsc, Mixl1, Gata1, and Bra (Fig. 4B). The combined use of RA/cyc is necessary for the nearly complete repression of the mesodermal markers Gata1 and Bra, as well as the mesendoderm markers Gsc and Mixl1. Robust expression of Bra, Mixl1, and Gsc in stage III EBs not treated with RA/cyc suggested that under these culture conditions, early mesendoderm/mesoderm fates are maintained, disallowing maturation of these early populations. Furthermore, this protocol succeeds in stronger repression of the mesoderm (Gata1, Bra) and neuroectoderm markers (Pax6, Sox1, Sox2) tested as compared with the earlier protocol (compare Fig. 4B with Fig. 2B). Repression of mesoderm fates was also manifested by the complete absence of fibroblast-like cells and beating muscle clusters. Cells that have grown in the presence of SR alone fail to upregulate many endoderm markers and retain expression of mesendoderm and mesoderm markers.

Importantly, the combination of these treatments resulted in strong simultaneous upregulation of Pdx1 and Ptf1a expression (Fig. 3B). Their combined expression was stronger than in the earlier protocol (data not shown), and this was probably due to longer exposure to Activin A, combined use of RA/cyc, or both. The other signaling molecules did not appear to play a significant role in upregulation of these genes (Fig. 4C). This is the first time that simultaneous expression of Pdx1 and Ptf1a has been achieved in ESC-derived pancreas progenitor cells and suggests that this protocol can be used as a stepping stone toward terminal differentiation to pancreas endocrine cells.

Expression of Ngn3 Enhances Differentiation of ESC-Derived Pancreas Progenitors Toward the Endocrine Lineage
In the next step, we investigated whether inducible Ngn3 expression could enhance the differentiation of these cells toward the endocrine lineage. Expression of Ngn3 was induced by the addition of dox in low glucose medium and in the presence of BMP4 and Fgf10, which may help expand the endocrine pancreas progenitor population [10, 57]. We used low-glucose medium at this stage because glucose restriction may initiate generation of insulin-producing cells [61]. Induction of Ngn3 is strong and correlates with strong induction of NeuroD1, Nkx2.2, Pax4, Dll1, IA1, and Isl1, as shown by immunofluorescence (Fig. 4D) and RT-PCR (Fig. 4E). Terminal differentiation was induced at stage V (5 days) in the presence of normal glucose levels, withdrawal of BMP4 and Fgf10, and inclusion of nicotinamide [62]. To determine whether induction of Ngn3 enhances the potential of the cells for endocrine differentiation, we compared, at the end of stage V, cells exposed and not exposed to dox. Immunofluorescence for glucagon, somatostatin, insulin, and C-peptide suggested that Ngn3+ cells were capable of endocrine pancreas terminal differentiation in contrast to Ngn3– cells (Fig. 5A, 5B). Confocal images from end of stage V EBs whole mount-stained for glucagon and somatostatin suggested that cells positive for the two hormones are distinct (Fig. 5A). Furthermore, colocalization of insulin and C-peptide (Fig. 5B) suggested that insulin immunoreactivity was not due to insulin absorption from the medium. Semiquantitative RT-PCR showed that all the endocrine terminal differentiation markers examined were strongly upregulated in the cells exposed to dox (Fig. 5C). We then examined whether these cell clusters were capable of glucose-stimulated insulin release. Insulin released in the medium was measured using an ELISA for insulin, and released insulin was expressed as ng/mg of cell protein. Cells exposed to low, nonstimulatory levels of glucose (low, 2 mM) responded very poorly, if at all, irrespective of whether they were derived from cells exposed to dox or not (Fig. 5D). High levels of glucose (high, 20 mM) elicited a strong insulin release in the cells exposed to dox but not in the cells that were not exposed to dox (Fig. 5D). When the insulin release agonist tolbutamide, an ATP-dependent K+ channel inhibitor, was added in the presence of low levels of glucose, strong insulin release was observed. Again, the secretion was observed only in cells previously exposed to dox (Fig. 5D). Taken together, these data suggested that bona fide endocrine pancreatic cells have been generated.


Figure 5
View larger version (38K):
[in this window]
[in a new window]

  		Figure 5. Ngn3 expression enhances differentiation of ESC-derived pancreas progenitors toward the endocrine lineage. All analyses were performed at the end of stage V on embryoid bodies that had (+dox) or had not (–dox) been exposed to Ngn3 expression. (A): Expression of glucagon and somatostatin detected by whole mount immunofluorescence. The dotted line follows the outline of the EBs. (B): Expression and colocalization of insulin and C-peptide detected by immunofluorescence on paraffin sections at the end of stage V. (C): Semiquantitative reverse transcription-polymerase chain reaction of islet-specific markers. (D): Gluc-dependent insulin release assays. Scale bars = 100 µm (A) and 45 µm (B). Abbreviations: DAPI, 4,6-diamidino-2-phenylindole; dox, doxycycline; Gluc, glucose.

 
Thus, we generated an ESC-derived pancreas progenitor population simultaneously expressing Pdx1 and Ptf1a and demonstrated that inducible expression of Ngn3 greatly enhanced its potential to differentiate into bona fide endocrine pancreas cells. This system will allow the dissection of the direct effects of Ngn3 interaction with extracellular signals involved in endocrine pancreas specification and the development of efficient protocols toward endocrine differentiation.


   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosure of Potential...
 Acknowledgments
 References
 
The aim of this study was to set up an ESC-based model system that would more closely simulate in vivo events to: (a) elucidate the interactions of Ngn3+ pancreatic progenitor cells with extracellular signals, (b) begin to unravel gene regulatory networks involved in pancreas endocrine specification, and (c) drive ESC-derived pancreas progenitors to terminal endocrine differentiation. Re-creation of the appropriate cellular context is crucial in the undertaking of these studies. We have generated ESC-derived pancreas progenitor-like cells that simultaneously express the key transcription factors in pancreas specification, Pdx1 and Ptf1a [9]. Induction of Ngn3 expression at this stage results in rapid upregulation of known Ngn3 target genes. Gene expression profiling suggested that Ngn3 is implicated in regulating diverse cellular processes, and further differentiation experiments demonstrated that the combination of extracellular signals and timely Ngn3 expression enhances differentiation of ESCs toward the endocrine lineage.

It is becoming increasingly evident that the molecular mechanisms used in vivo for lineage specification and tissue development are a valuable guide in directing the in vitro differentiation of tissue-specific and embryonic stem cells. Of particular importance are the mechanisms underlying lineage specification decisions of stem and progenitor cells and the role that key transcription factors play in this process. However, crucial progenitor and stem cell populations are found in vivo only transiently and/or in small numbers, making their isolation and study problematic. The analysis of tissue lacking or overexpressing key transcription factors for these populations masks their immediate effects because of the accumulation of secondary events and the often small numbers of cells normally expressing them. The same reasons limit the application of chromatin immunoprecipitation experiments, particularly for genome-wide studies.

An ESC-derived progenitor-like population would overcome these limitations. The study of such progenitor cells could generate new hypotheses testable in vivo and provide further clues regarding the signaling pathways involved in cell specification that would help in directing cells to specific lineages. It would also allow study of the interactions of transcription factors with extracellular signals in determining cell fate decisions. Furthermore, the timely induction of a key specification transcription factor may enhance the directed differentiation of ESCs or even promote otherwise inaccessible alternative differentiation routes. An ESC-derived progenitor cell population should express molecular markers that characterize its in vivo counterpart, and induction of the transcription factor in question should result in rapid activation of known target genes.

To generate pancreas progenitor-like cells, we have simulated the in vivo sequence of events leading to this cell population through successive cell fate restrictions. Initially, we combined EB formation and Activin A to potentiate definitive endoderm specification [44] of ESCs. Activin A was combined with Fgf4 to induce anterior gut fates [56]. Subsequent pancreatic endoderm specification in vivo is mediated through RA signaling [53, 58] and repression of shh signaling [59, 60]. Accordingly, in the next step, ESC-derived definitive endoderm was treated with RA and the shh inhibitor cyclopamine. To enhance the expansion of the forming pancreatic progenitors, Fgf10 (10 ng/ml) and BMP4 (100 ng/ml) [10, 57] were included. The sequential use of these factors succeeded for the first time in inducing simultaneous expression of both Pdx1 and Ptf1a, key transcription factors that mark the pancreas primordium in the developing embryo [9]. In addition, successive application of these signals was necessary to generate progenitor cells that responded properly to Ngn3 induction by activation of downstream Ngn3 target genes. This demonstrated the importance of appropriate cellular context and developmental history of ESC-derived progenitor populations. In addition, cells exposed to Ngn3 expression were able to give rise to somatostatin+, glucagon+, and insulin+ cells upon terminal differentiation. Terminal differentiation was accompanied by strong induction of all terminal endocrine pancreas markers examined by RT-PCR. Furthermore, cells responded robustly to glucose by insulin secretion, and virtually all insulin+ cells also stained with C-peptide, suggesting bona fide endocrine pancreas differentiation. Thus, sequential combination of appropriate extracellular signals and inducible Ngn3 expression can greatly enhance endocrine pancreas differentiation. Improvements in this method could result in high rates of ESC conversion to endocrine pancreas hormone-producing cells.

Pancreas progenitors go through competence phases in their ability to generate endocrine cells in a mechanism intrinsic to the pancreatic epithelium [63]. It will be interesting to investigate whether the ESC-generated pancreas progenitor-like cells can recapitulate this competence. We showed that induction of Ngn3 expression enhances the potential of these cells to differentiate toward the endocrine lineage, but the optimal length of Ngn3 induction for subsequent differentiation remains to be determined. The finding that known direct target genes of Ngn3 show different kinetics of activation after Ngn3 induction suggests that the length of Ngn3 expression may affect the final outcome of the differentiation.

In our study, ESC-derived pancreas-like progenitors simultaneously express pdx1 and ptf1a, similarly to their in vivo counterparts [9], and the activation of known Ngn3 target genes occurs within 12 hours of its induction. The gene expression profiling at three different time points (12, 24, and 48 hours) after Ngn3 induction suggested several novel direct targets, such as components of the Notch signaling pathway. It will also be instrumental to further delineate immediate and later events in Ngn3-mediated transcriptional regulation. Gene regulation at later time points may represent secondary effects, or it may imply the requirement of Ngn3-induced factor(s) in addition to Ngn3 itself. Many of the regulated genes have already been implicated in pancreas development. On the other hand, the data suggest extensive involvement of Ngn3 in diverse processes and networks, some of which have not previously been implicated in pancreas endocrine development. A direct comparison of our microarray data with those of similar studies is not feasible due to significant differences in the experimental approach. Studies using Ngn3 mutant animals [20, 21] look at global effects resulting from the loss of the endocrine lineage several days after initiation of Ngn3 expression. The study by Treff et al. [22] uses inducible Ngn3 expression in ESC-derived cells as well, but the earliest analysis was performed 3 days after Ngn3 induction and in a heterogeneous cell context (EBs). The study by Gasa et al. [15] overexpressed Ngn3 in yet another cell context, pancreatic duct cells. It is important to note that most known Ngn3 direct target genes and members of the Notch signaling pathway known to be involved in endocrine pancreas development were identified in the studies using expression of Ngn3 in a cellular context [15, 22], but not in the studies using tissue from transgenic and Ngn3 knock-out mice [20, 21]. This suggests that a cell-based approach for identification of bona fide Ngn3 direct targets and enhancer characterization on a genome-wide scale would be preferable. The crucial issue for the success of this approach will be homogeneity of the cell population and close similarity to the cellular context of Pdx1+/Ptf1a+ pancreas progenitors.

As endocrine progenitors differentiate, they migrate out of the epithelium of the branching pancreas in the mesenchyme to form interstitial clusters. One of the surprising aspects of our GO analysis is that Ngn3 emerges as a key regulator of this process because changes in cell adhesion, cell motility, cell [64] membrane, and cytoskeleton correlate strongly with Ngn3 induction. Changes in cell adhesion may also mark the segregation of committed endocrine cells from progenitors and cells of the exocrine lineage. Furthermore, the GO analysis suggested that the integrin signaling pathway strongly correlates as an Ngn3 regulated process. Integrin signaling enables cells to detach from neighboring cells, reorientate their polarity during migration, and survive and proliferate in a changing microenvironment [65]. The polarity complex controls a variety of cellular processes, such as asymmetric cell division and establishment of epithelial apico-basal polarity, and polarizes cell migration [66]. It is worth noting that all the main components of the polarity complex [66] are also regulated in our system (Table 2Go; data not shown) in response to Ngn3 expression. Furthermore, CD133/prominin, a cell surface polarity marker associated with tissue-specific stem cells, is expressed in Ngn3+ progenitors [67] and was upregulated in our system by Ngn3 (Table 2Go).

Several signaling pathways are regulated in response to Ngn3 induction. Wnts acting through the canonical (β-catenin) and noncanonical pathways can regulate cell adhesion, polarity, morphology, proliferation, migration, and tissue structural remodeling. Both pathways have been implicated in several aspects of pancreas development, such as exocrine development [68, 69], endocrine development [70], and insulin cell migration [71], and both pathways are regulated in our system in response to Ngn3 expression in a statistically significant manner (Table 2Go). It is well-established that the Notch signaling pathway in pancreas development is essential for progenitor maintenance [72, 73] and lineage specification [74–76]. Our data suggest for the first time that several components of the Notch pathway are directly regulated by Ngn3, given their rapid induction. Hes-1 has been implicated in progenitor maintenance [73] and negative regulation of endodermal endocrine differentiation [75] and does not appear in our screen, suggesting that it lies strictly upstream of Ngn3 expression. All three branches of the transforming growth factor β pathway have been implicated in pancreas development. Activins are required for the specialization of pancreatic precursors from the gut endoderm during midgestation [77]. BMP4 can mediate expansion of pancreas progenitors via Id2 [57], whereas GDF11 modulates the number of Ngn3+ progenitors and promotes B-cell differentiation [78]. We found that mainly extracellular and membrane components of the pathway, as well as Smad1 and Smad6, are regulated in ESC-derived pancreas progenitors in response to Ngn3 expression.

The GO analysis confirmed that another important Ngn3 function is the control of the expression of other transcription factors. Several transcription factors were regulated in all three time points examined in response to Ngn3 expression. Many of these are also bHLH proteins. Intriguingly, some of them (bhlhb5, Id1-3) are expected to act as inhibitors of transcriptional activity, suggesting that the execution of endocrine differentiation may be transiently inhibited in Ngn3+ progenitors. Subsequent Ngn3+ autorepression [79] may release this transient inhibition. Sry-box-containing genes have been implicated in pancreas development [73, 80, 81], and it appears that Ngn3 may play at least an indirect role in regulating sox4 and sox18.

In summary, using an ESC based model system, we have identified novel genes and pathways that may constitute targets of Ngn3 function in endocrine pancreas specification and demonstrated that combination of extracellular signals and timely Ngn3 expression may enhance differentiation of ESCs toward the pancreas endocrine lineage.


   DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
Acknowledgments
 References
 
The authors indicate no potential conflicts of interest.


   ACKNOWLEDGMENTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
References
 
We thank Kian-Leong Lee and Anthony Brooks for help with the microarray experiments and Alexandros Polyzos and Thanassis Spathis for advice on the microarray data analysis. We also thank George Chroussos for advice and stimulating discussions. This work was supported by grants from the General Secretariat of Research and Technology (EPAn YB/18) and an Innovative Juvenile Diabetes Research Foundation grant (5-2006-238) to A.G. I.S. and I.R. contributed equally to this work.


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 Acknowledgments
 References
 

  1. Soria B. In-vitro differentiation of pancreatic β-cells. Differentiation 2001;68:205–219.[CrossRef][Medline]

  2. Edlund H. Pancreatic organogenesis—Developmental mechanisms and implications for therapy. Nat Rev Genet 2002;3:524–532.[CrossRef][Medline]

  3. Stoffel M, Vallier L, Pedersen RA. Navigating the pathway from embryonic stem cells to beta cells. Semin Cell Dev Biol 2004;15:327–336.[CrossRef][Medline]

  4. Bonner-Weir S, Weir GC. New sources of pancreatic beta-cells. Nat Biotechnol 2005;23:857–861.[CrossRef][Medline]

  5. Keller G. Embryonic stem cell differentiation: Emergence of a new era in biology and medicine. Genes Dev 2005;19:1129–1155.[Abstract/Free Full Text]

  6. Wells JM, Melton DA. Vertebrate endoderm development. Annu Rev Cell Dev Biol 1999;15:393–410.[CrossRef][Medline]

  7. Kumar M, Melton D. Pancreas specification: A budding question. Curr Opin Genet Dev 2003;13:401–407.[CrossRef][Medline]

  8. Hebrok M, Kim SK, Melton DA. Notochord repression of endodermal Sonic hedgehog permits pancreas development. Genes Dev 1998;12:1705–1713.[Abstract/Free Full Text]

  9. Kawaguchi Y, Cooper B, Gannon M et al. The role of the transcriptional regulator Ptf1a in converting intestinal to pancreatic progenitors. Nat Genet 2002;32:128–134.[CrossRef][Medline]

  10. Hart A, Papadopoulou S, Edlund H. Fgf10 maintains notch activation, stimulates proliferation, and blocks differentiation of pancreatic epithelial cells. Dev Dyn 2003;228:185–193.[CrossRef][Medline]

  11. Gradwohl G, Dierich A, LeMeur M et al. neurogenin3 is required for the development of the four endocrine cell lineages of the pancreas. Proc Natl Acad Sci U S A 2000;97:1607–1611.[Abstract/Free Full Text]

  12. Gu G, Dubauskaite J, Melton DA. Direct evidence for the pancreatic lineage: Ngn3+ cells are islet progenitors and are distinct from duct progenitors. Development 2002;129:2447–2457.[Medline]

  13. Schwitzgebel VM, Scheel DW, Conners JR et al. Expression of neurogenin3 reveals an islet cell precursor population in the pancreas. Development 2000;127:3533–3542.[Abstract]

  14. Grapin-Botton A, Majithia AR, Melton DA. Key events of pancreas formation are triggered in gut endoderm by ectopic expression of pancreatic regulatory genes. Genes Dev 2001;15:444–454.[Abstract/Free Full Text]

  15. Gasa R, Mrejen C, Leachman N et al. Proendocrine genes coordinate the pancreatic islet differentiation program in vitro. Proc Natl Acad Sci U S A 2004;101:13245–13250.[Abstract/Free Full Text]

  16. Watada H, Scheel DW, Leung J et al. Distinct gene expression programs function in progenitor and mature islet cells. J Biol Chem 2003;278:17130–17140.[Abstract/Free Full Text]

  17. Huang HP, Liu M, El-Hodiri HM et al. Regulation of the pancreatic islet-specific gene BETA2 (neuroD) by neurogenin 3. Mol Cell Biol 2000;20:3292–3307.[Abstract/Free Full Text]

  18. Smith SB, Gasa R, Watada H et al. Neurogenin3 and hepatic nuclear factor 1 cooperate in activating pancreatic expression of Pax4. J Biol Chem 2003;278:38254–38259.[Abstract/Free Full Text]

  19. Mellitzer G, Bonne S, Luco RF et al. IA1 is Ngn3-dependent and essential for differentiation of the endocrine pancreas. EMBO J 2006;25:1344–1352.[CrossRef][Medline]

  20. Gu G, Wells JM, Dombkowski D et al. Global expression analysis of gene regulatory pathways during endocrine pancreatic development. Development 2004;131:165–179.[Abstract/Free Full Text]

  21. Petri A, Ahnfelt-Ronne J, Frederiksen KS et al. The effect of neurogenin3 deficiency on pancreatic gene expression in embryonic mice. J Mol Endocrinol 2006;37:301–316.[Abstract/Free Full Text]

  22. Treff NR, Vincent RK, Budde ML et al. Differentiation of embryonic stem cells conditionally expressing neurogenin 3. STEM CELLS 2006;24:2529–2537.[Abstract/Free Full Text]

  23. Segev H, Fishman B, Ziskind A et al. Differentiation of human embryonic stem cells into insulin-producing clusters. STEM CELLS 2004;22:265–274.[Abstract/Free Full Text]

  24. Assady S, Maor G, Amit M et al. Insulin production by human embryonic stem cells. Diabetes 2001;50:1691–1697.[Abstract/Free Full Text]

  25. Blyszczuk P, Czyz J, Kania G et al. Expression of Pax4 in embryonic stem cells promotes differentiation of nestin-positive progenitor and insulin-producing cells. Proc Natl Acad Sci U S A 2003;100:998–1003.[Abstract/Free Full Text]

  26. Lumelsky N, Blondel O, Laeng P et al. Differentiation of embryonic stem cells to insulin-secreting structures similar to pancreatic islets. Science 2001;292:1389–1394.[Abstract/Free Full Text]

  27. Hori Y, Rulifson IC, Tsai BC et al. Growth inhibitors promote differentiation of insulin-producing tissue from embryonic stem cells. Proc Natl Acad Sci U S A 2002;99:16105–16110.[Abstract/Free Full Text]

  28. Kania G, Blyszczuk P, Wobus AM. The generation of insulin-producing cells from embryonic stem cells—A discussion of controversial findings. Int J Dev Biol 2004;48:1061–1064.[CrossRef][Medline]

  29. Rajagopal J, Anderson WJ, Kume S et al. Insulin staining of ES cell progeny from insulin uptake. Science 2003;299:363.[Free Full Text]

  30. D'Amour KA, Bang AG, Eliazer S et al. Production of pancreatic hormone-expressing endocrine cells from human embryonic stem cells. Nat Biotechnol 2006;24:1392–1401.[CrossRef][Medline]

  31. Srinivas S, Watanabe T, Lin CS et al. Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev Biol 2001;1:4.[CrossRef][Medline]

  32. Soriano P. Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat Genet 1999;21:70–71.[CrossRef][Medline]

  33. Gossen M, Bonin AL, Bujard H. Control of gene activity in higher eukaryotic cells by prokaryotic regulatory elements. Trends Biochem Sci 1993;18:471–475.[CrossRef][Medline]

  34. Bronson SK, Plaehn EG, Kluckman KD et al. Single-copy transgenic mice with chosen-site integration. Proc Natl Acad Sci U S A 1996;93:9067–9072.[Abstract/Free Full Text]

  35. McEwan C, Melton DW. A simple genotyping assay for the Hprt null allele in mice produced from the HM-1 and E14TG2a mouse embryonic stem cell lines. Transgenic Res 2003;12:519–520.[CrossRef][Medline]

  36. Irizarry RA, Bolstad BM, Collin F et al. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res 2003;31:e15.[Abstract/Free Full Text]

  37. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A 2001;98:5116–5121.[Abstract/Free Full Text]

  38. Storey G. A direct approach to false discovery rates. J R Stat Soc Ser B 2002;64:479–498.[CrossRef]

  39. Doetschman T, Gregg RG, Maeda N et al. Targetted correction of a mutant HPRT gene in mouse embryonic stem cells. Nature 1987;330:576–578.[CrossRef][Medline]

  40. Gossen M, Bujard H. Tight control of gene expression in mammalian cells by tetracycline-responsive promoters. Proc Natl Acad Sci U S A 1992;89:5547–5551.[Abstract/Free Full Text]

  41. Rogers MB, Hosler BA, Gudas LJ. Specific expression of a retinoic acid-regulated, zinc-finger gene, Rex-1, in preimplantation embryos, trophoblast and spermatocytes. Development 1991;113:815–824.[Abstract]

  42. Blum M, Gaunt SJ, Cho KW et al. Gastrulation in the mouse: The role of the homeobox gene goosecoid. Cell 1992;69:1097–1106.[CrossRef][Medline]

  43. Hart AH, Hartley L, Sourris K et al. Mixl1 is required for axial mesendoderm morphogenesis and patterning in the murine embryo. Development 2002;129:3597–3608.[Abstract/Free Full Text]

  44. Kubo A, Shinozaki K, Shannon JM et al. Development of definitive endoderm from embryonic stem cells in culture. Development 2004;131:1651–1662.[Abstract/Free Full Text]

  45. Yasunaga M, Tada S, Torikai-Nishikawa S et al. Induction and monitoring of definitive and visceral endoderm differentiation of mouse ES cells. Nat Biotechnol 2005;23:1542–1550.[CrossRef][Medline]

  46. Orkin SH. GATA-binding transcription factors in hematopoietic cells. Blood 1992;80:575–581.[Free Full Text]

  47. Wood HB, Episkopou V. Comparative expression of the mouse Sox1, Sox2 and Sox3 genes from pre-gastrulation to early somite stages. Mech Dev 1999;86:197–201.[CrossRef][Medline]

  48. Hill RE, Hanson IM. Molecular genetics of the Pax gene family. Curr Opin Cell Biol 1992;4:967–972.[CrossRef][Medline]

  49. Graham V, Khudyakov J, Ellis P et al. SOX2 functions to maintain neural progenitor identity. Neuron 2003;39:749–765.[CrossRef][Medline]

  50. Matzuk MM, Kumar TR, Vassalli A et al. Functional analysis of activins during mammalian development. Nature 1995;374:354–356.[CrossRef][Medline]

  51. Vassalli A, Matzuk MM, Gardner HA et al. Activin/inhibin beta B subunit gene disruption leads to defects in eyelid development and female reproduction. Genes Dev 1994;8:414–427.[Abstract/Free Full Text]

  52. Molotkov A, Molotkova N, Duester G. Retinoic acid generated by Raldh2 in mesoderm is required for mouse dorsal endodermal pancreas development. Dev Dyn 2005;232:950–957.[CrossRef][Medline]

  53. Martin M, Gallego-Llamas J, Ribes V et al. Dorsal pancreas agenesis in retinoic acid-deficient Raldh2 mutant mice. Dev Biol 2005;284:399–411.[Medline]

  54. Stafford D, Prince VE. Retinoic acid signaling is required for a critical early step in zebrafish pancreatic development. Curr Biol 2002;12:1215–1220.[CrossRef][Medline]

  55. Heremans Y, Van De Casteele M, in't Veld P et al. Recapitulation of embryonic neuroendocrine differentiation in adult human pancreatic duct cells expressing neurogenin 3. J Cell Biol 2002;159:303–312.[Abstract/Free Full Text]

  56. Dessimoz J, Opoka R, Kordich JJ et al. FGF signaling is necessary for establishing gut tube domains along the anterior-posterior axis in vivo. Mech Dev 2006;123:42–55.[CrossRef][Medline]

  57. Hua H, Zhang YQ, Dabernat S et al. BMP4 regulates pancreatic progenitor cell expansion through Id2. J Biol Chem 2006;281:13574–13580.[Abstract/Free Full Text]

  58. Lipscomb K, Schmitt C, Sablyak A et al. Role for retinoid signaling in left-right asymmetric digestive organ morphogenesis. Dev Dyn 2006;235:2266–2275.[CrossRef][Medline]

  59. Kim SK, Melton DA. Pancreas development is promoted by cyclopamine, a hedgehog signaling inhibitor. Proc Natl Acad Sci U S A 1998;95:13036–13041.[Abstract/Free Full Text]

  60. Kawahira H, Ma NH, Tzanakakis ES et al. Combined activities of hedgehog signaling inhibitors regulate pancreas development. Development 2003;130:4871–4879.[Abstract/Free Full Text]

  61. Hori Y, Gu X, Xie X et al. Differentiation of insulin producing cells from human neural progenitor cells. PLoS Med 2005;2:e103.[CrossRef][Medline]

  62. Vaca P, Berna G, Martin F et al. Nicotinamide induces both proliferation and differentiation of embryonic stem cells into insulin-producing cells. Transplant Proc 2003;35:2021–2023.[CrossRef][Medline]

  63. Johansson KA, Dursun U, Jordan N et al. Temporal control of neurogenin3 activity in pancreas progenitors reveals competence windows for the generation of different endocrine cell types. Dev Cell 2007;12:457–465.[CrossRef][Medline]

  64. Herrera PL, Huarte J, Sanvito F et al. Embryogenesis of the murine endocrine pancreas; early expression of pancreatic polypeptide gene. Development 1991;113:1257–1265.[Abstract]

  65. Guo W, Giancotti FG. Integrin signalling during tumour progression. Nat Rev Mol Cell Biol 2004;5:816–826.[CrossRef][Medline]

  66. Bose R, Wrana JL. Regulation of Par6 by extracellular signals. Curr Opin Cell Biol 2006;18:206–212.[CrossRef][Medline]

  67. Sugiyama T, Rodriguez RT, McLean GW et al. Conserved markers of fetal pancreatic epithelium permit prospective isolation of islet progenitor cells by FACS. Proc Natl Acad Sci U S A 2007;104:175–180.[Abstract/Free Full Text]

  68. Wells JM, Esni F, Boivin GP et al. Wnt/beta-catenin signaling is required for development of the exocrine pancreas. BMC Dev Biol 2007;7:4.[CrossRef][Medline]

  69. Murtaugh LC, Law AC, Dor Y et al. Beta-catenin is essential for pancreatic acinar but not islet development. Development 2005;132:4663–4674.[Abstract/Free Full Text]

  70. Dessimoz J, Bonnard C, Huelsken J et al. Pancreas-specific deletion of beta-catenin reveals Wnt-dependent and Wnt-independent functions during development. Curr Biol 2005;15:1677–1683.[CrossRef][Medline]

  71. Kim HJ, Schleiffarth JR, Jessurun J et al. Wnt5 signaling in vertebrate pancreas development. BMC Biol 2005;3:23.[CrossRef][Medline]

  72. Murtaugh LC, Stanger BZ, Kwan KM et al. Notch signaling controls multiple steps of pancreatic differentiation. Proc Natl Acad Sci U S A 2003;100:14920–14925.[Abstract/Free Full Text]

  73. Seymour PA, Freude KK, Tran MN et al. SOX9 is required for maintenance of the pancreatic progenitor cell pool. Proc Natl Acad Sci U S A 2007;104:1865–1870.[Abstract/Free Full Text]

  74. Apelqvist A, Li H, Sommer L et al. Notch signalling controls pancreatic cell differentiation. Nature 1999;400:877–881.[CrossRef][Medline]

  75. Jensen J, Pedersen EE, Galante P et al. Control of endodermal endocrine development by Hes-1. Nat Genet 2000;24:36–44.[CrossRef][Medline]

  76. Zecchin E, Filippi A, Biemar F et al. Distinct delta and jagged genes control sequential segregation of pancreatic cell types from precursor pools in zebrafish. Dev Biol 2007;301:192–204.[CrossRef][Medline]

  77. Zhang YQ, Cleary MM, Si Y et al. Inhibition of activin signaling induces pancreatic epithelial cell expansion and diminishes terminal differentiation of pancreatic beta-cells. Diabetes 2004;53:2024–2033.[Abstract/Free Full Text]

  78. Harmon EB, Apelqvist AA, Smart NG et al. GDF11 modulates Ngn3+ islet progenitor cell number and promotes beta-cell differentiation in pancreas development. Development 2004;131:6163–6174.[Abstract/Free Full Text]

  79. Wilson ME, Scheel D, German MS. Gene expression cascades in pancreatic development. Mech Dev 2003;120:65–80.[CrossRef][Medline]

  80. Mavropoulos A, Devos N, Biemar F et al. sox4b is a key player of pancreatic alpha cell differentiation in zebrafish. Dev Biol 2005;285:211–223.[CrossRef][Medline]

  81. Wilson ME, Yang KY, Kalousova A et al. The HMG box transcription factor Sox4 contributes to the development of the endocrine pancreas. Diabetes 2005;54:3402–3409.[Abstract/Free Full Text]




This article has been cited by other articles:


Home page
 Proc. Natl. Acad. Sci. USAHome page
M. D. Kinkel, E. M. Sefton, Y. Kikuchi, T. Mizoguchi, A. B. Ward, and V. E. Prince
Cyp26 enzymes function in endoderm to regulate pancreatic field size
PNAS, May 12, 2009; 106(19): 7864 - 7869.
[Abstract] [Full Text] [PDF]

Home page
 GutHome page
M Mimeault and S K Batra
Recent progress on normal and malignant pancreatic stem/progenitor cell research: therapeutic implications for the treatment of type 1 or 2 diabetes mellitus and aggressive pancreatic cancer
Gut, October 1, 2008; 57(10): 1456 - 1468.
[Abstract] [Full Text] [PDF]

Home page
 Stem CellsHome page
M. Gouti and A. Gavalas
Hoxb1 Controls Cell Fate Specification and Proliferative Capacity of Neural Stem and Progenitor Cells
Stem Cells, August 1, 2008; 26(8): 1985 - 1997.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
2007-0194v1
26/1/3    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Serafimidis, I.
Right arrow Articles by Gavalas, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Serafimidis, I.
Right arrow Articles by Gavalas, A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
STEM CELLS THE ONCOLOGIST CME ALPHAMED PRESS JOURNALS