Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Jun 2.
Published in final edited form as: Cell Stem Cell. 2016 Apr 28;18(6):755–768. doi: 10.1016/j.stem.2016.03.015

Genome editing of lineage determinants in human pluripotent stem cells reveals mechanisms of pancreatic development and diabetes

Zengrong Zhu 1, Qing V Li 1,2, Kihyun Lee 1,3, Bess P Rosen 4, Federico González 1,5, Chew-Li Soh 1, Danwei Huangfu 1,*
PMCID: PMC4892994  NIHMSID: NIHMS773607  PMID: 27133796

Summary

Directed differentiation of human pluripotent stem cells (hPSCs) into somatic counterparts is a valuable tool for studying disease. However, examination of developmental mechanisms in hPSCs remains challenging given complex multi-factorial actions at different stages. Here, we used TALEN and CRISPR-Cas-mediated gene editing and hPSC directed differentiation for a systematic analysis of the roles of 8 pancreatic transcription factors (PDX1, RFX6, PTF1A, GLIS3, MNX1, NGN3, HES1 and ARX). Our analysis not only verified conserved gene requirements between mice and humans, but also revealed a number of previously unsuspected developmental mechanisms with implications for type 2 diabetes. These include a role of RFX6 in regulating the number of pancreatic progenitors, a haploinsufficient requirement for PDX1 in pancreatic β cell differentiation, and a potentially divergent role of NGN3 in humans and mice. Our findings support use of systematic genome editing in hPSCs as a strategy for understanding mechanisms underlying congenital disorders.

Introduction

The advancements in next-generation sequencing and genome-wide association studies have led to the identification of hundreds of disease-associated sequence variants. Thus, there is an urgent need for a functional evaluation platform to rapidly identify disease-causing mutations. A promising strategy involves the use of human pluripotent stem cells (hPSCs), including both embryonic and induced pluripotent stem cells (hESCs and hiPSCs) for disease modeling. However, the limited access to patient material and the relatively low genome-editing throughput has been a bottleneck for increasing the output of hPSC-based models. Furthermore, most hPSC studies so far have focused on generating disease-relevant cell types for studying disease phenotypes that are manifested at the cellular level, whereas the utility of hPSCs for studying more complex biological processes such as a multistep developmental process remains uncertain.

A unique challenge of modeling developmental defects lies in the need for faithful recreation of the complexity of embryonic development in a petri dish. Despite considerable progress, it remains challenging to perfectly recapitulate the contexts of embryonic development such as complex tissue-tissue interactions; and many in vivo biologists remain skeptical of the relevance of hPSCs for studying developmental disorders. In comparison, to study the cellular phenotype of a disease, some deviation from in vivo development can be tolerated; for instance, one may generate disease-relevant cell types without mimicking development at all through direct lineage reprogramming (Qiang et al., 2014). There are also technical concerns of using hPSCs for developmental studies. Developmental phenotypes are typically manifested as changes in the efficiencies of hPSCs to differentiate into a specific lineage of interest, which could be obscured by variations in differentiation propensity among hPSC lines from different genetic backgrounds (Bock et al., 2011; Osafune et al., 2008).

We have recently established an efficient genome-editing platform in hPSCs named iCRISPR through the use of TALENs (transcription activator like effector nucleases) and CRISPR/Cas (clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated) system (González et al., 2014). Combining the power of genome editing and stem cell biology, we set out to systematically probe transcriptional control of pancreatic development and the developmental defects involved in permanent neonatal diabetes mellitus (PNDM), a rare monogenic form of diabetes that occurs during the first 6 months of life (Aguilar-Bryan and Bryan, 2008). Our analysis not only defines the specific developmental step(s) affected by these mutations, but also revealed a number of insights into disease mechanisms, including a role of RFX6 in regulating the number of pancreatic progenitors, a dosage-sensitive requirement for PDX1 in pancreatic endocrine development, and a potentially divergent role of NGN3 in humans and mice. Taking full advantage of the power of genome editing, we further performed temporal rescue studies to investigate the competence window for NGN3-dependent pancreatic β cell formation as well as gene correction experiments to verify CRISPR targeting specificity. Thus the hESC-based genetic model not only offers the speed and scale necessary to meet the growing demand for validating disease causality, but also enables sophisticated genetic manipulations for mechanistic investigations. The systematic probing of transcription factors that regulate pancreatic development also forms the foundation for understanding the transcriptional circuitry involved in human pancreatic development, and provides the much-needed information for advancing hPSC-based β cell replacement therapies for treatment of diabetes.

Results

Inducible gene expression for gain-of-function studies in hESCs

To harness the power of hESCs for functional human genetics, it is necessary to first establish efficient genetic approaches for knocking out or ectopically expressing a gene of interest. Although viral vectors offer a convenient method for transgene expression, low infection efficiency, random integration and transgene silencing have limited its application in hESCs (Saha and Jaenisch, 2009). Building upon previous findings (DeKelver et al., 2010), we targeted the AAVS1 transgene safe harbour locus with a pair of TALENs for simultaneous integration of two transgenes in trans, one with a constitutive promoter driving the expression of an optimized form of reverse tetracycline-controlled transactivator (M2rtTA) and the other with a tetracycline-response element (TRE) driving the expression of a gene of interest (Figure 1A). This system conveniently generates hESC lines for doxycycline-inducible gene expression, which is critical for studying developmental mechanisms in a stage-specific manner. In a typical targeting experiment, as shown for the creation of inducible GFP expressing lines (named iGFP), about half of the clonal lines had correct biallelic transgene integration without random insertions based on PCR genotyping and Southern blot analysis (Figure S1A–D). Because of this high efficiency, only 6–12 clonal lines need to be analyzed after electroporation, making it feasible to generate inducible expression lines for multiple genes in parallel in just a month (Figure S1B). We also made this system compatible with the Gateway® recombination cloning technology, reducing the time and effort needed for generating the targeting constructs (Figure S1E).

Figure 1. Gain-of-function studies in hESCs.

Figure 1

Open in a new tab

A) Schematics of the doxycycline-inducible gene expression platform in hESCs. Homology directed repair of the DNA DSB induced by a pair of TALENs led to targeted integration of the transgene and M2rtTA into the AAVS1 locus. After the establishment of a clonal line, transgene expression can be induced upon doxycycline treatment. SA: Splice acceptor; 2A: Self-cleavage 2A peptide; Puro: Puromycin resistant gene; TRE: tetracycline response element; Neo: Neomycin resistant gene; CAG, constitutive synthetic promoter; M2rtTA, reverse tetracycline transactivator; DOX: doxycycline (also indicated by the red dot). B, C) Southern blotting (B) and qRT-PCR (C) analysis of the iNotchIC and iNGN3 lines. Correctly targeted lines without random integrations are indicated in red. WT: Wild-type control; 3’ EXT: 3’ external probe; 5’ INT: 5’ external probe; hESC: undifferentiated hESC; PP: pancreatic progenitor. D) Representative immunofluorescence staining of iNotchIC and iNGN3 cells at PH-β cell stage with or without doxycycline treatment at pancreatic progenitor stage. DE: Definitive endoderm; PP: pancreatic progenitors; PH-β: Polyhormonal β cells; CPEP; C-peptide; GCG: glucagon; SST: somatostatin. E, F) Representative FACS plots (E) and quantification of iNGN3 hESC-derived INS+ cells (F) with or without doxycycline treatment. Number in the FACS plots indicates the percentage of target cells. n = 4: two independent experiments were performed on 2 iNGN3 lines. G) qRT-PCR analysis of endocrine markers and endocrine specific transcription factors in iNGN3 hESCs without and with doxycycline treatment. n = 4. Unless otherwise indicated, scale bar = 100 µm in all figures; error bars indicate standard error of the mean (SEM); and P values by unpaired two-tailed student t-test <0.05, 0.01, and 0.0001 are indicated by one, two, and four asterisks, respectively. For qRT-PCR results, P values are not indicated in graphs due to the large number of bars, but are mentioned in text when relevant. (See also Figure S1)

To model pancreatic development, we adapted a directed differentiation protocol initially established by D’Amour and colleagues and now widely used in the original or modified forms (D’Amour et al., 2006; Nostro et al., 2011; Rezania et al., 2012). Using HUES8 hESCs, we routinely generate ~50–80% definitive endoderm (DE) cells co-expressing SOX17 and FOXA2 after 3 days of differentiation, and ~50% PDX1-expressing (PDX1+) pancreatic progenitor (PP) cells (also co-expressing additional PP markers such as SOX9 and HNF6) after 6 additional days of differentiation (Figure S1F). Further differentiation generates pancreatic endocrine cells that express endocrine hormones such as insulin (INS) and glucagon (GCG) characteristic of β and α cells respectively. As observed by others (D’Amour et al., 2006), the INS-expressing (INS+) cells often co-express hormones such as GCG and somatostatin (SST), thus resembling immature polyhormonal β cells, which we refer to as PH-β hereafter. Using iGFP hESC lines, we confirmed relatively uniform GFP expression upon doxycycline treatment at each differentiation step (Figure S1G). Thus we went on to perturb NOTCH and NGN3 during hESC differentiation because of their well-demonstrated roles from murine studies in inhibiting and promoting endocrine differentiation, respectively (Shih et al., 2013).

We generated iNotchIC and iNGN3 hESC lines for inducible expression of NotchIC (Notch intracellular domain, the constitutively active form of Notch1) and NGN3 respectively (Figure 1B and C). iNotchIC and iNGN3 hESC cells were differentiated to the PP stage, treated with doxycycline and then examined at the PH-β stage for expression of endocrine markers (Figure 1D). Activation of Notch signaling completely blocked the formation of endocrine cells as indicated by immunofluorescence staining for C-peptide (CPEP), INS, GCG and SST at the PH-β stage (Figure 1D). In contrast, expression of these endocrine markers was greatly promoted by NGN3 overexpression (Figure 1D). Due to concerns of potential uptake of insulin from the culture media (Rajagopal et al., 2003), the expression of C-peptide, the byproduct of insulin biosynthesis, was used for quantification of insulin-producing cells hereafter.

Overexpression of NGN3 led to a ~10 fold increase of INS+ cells based on CPEP intracellular staining and fluorescence-activated cell sorting (FACS) analysis (Figure 1E and F). The pro-endocrine effect of NGN3 was also supported by quantitative RT-PCR (qRT-PCR) analysis which showed an increase in all pancreatic endocrine hormone genes as well as key transcription factor genes such as NKX2.2, NEUROD1, ISL1 and MAFB (Figure 1G). These findings mirror results from murine studies and support the employment of conserved genetic networks in hESC differentiation and in vivo biology.

Streamlined creation of knockout hESC lines

In parallel to the gain-of-function approach, we set out to generate an array of mutant lines affecting 8 key transcription factor genes involved in pancreatic development (Figure 2B). Six of the genes (PDX1, RFX6, PTF1A, GLIS3, MNX1 and NGN3) are associated with PNDM, and biallelic inactivation of these genes is thought to be responsible for the absence of pancreatic endocrine cells including β cells in patients (Flanagan et al., 2014; Rubio-Cabezas et al., 2011; Sellick et al., 2004; Senee et al., 2006; Smith et al., 2010; Stoffers et al., 1997b). Mutations in a subset of these genes (e.g. PDX1 and PTF1A) are associated with severe deficiencies in both endocrine and exocrine pancreas, a condition known as pancreatic agenesis. Two additional genes (ARX and HES1) were chosen because in contrast to the pro-endocrine roles of PNDM-associated genes above, Hes1 inhibits endocrine development whereas Arx is necessary for the formation of α cells but not β cells in mice (Collombat et al., 2003; Jensen et al., 2000).

Figure 2. iCRISPR-mediated creation of knockout hESC lines for loss-of-function studies.

Figure 2

Open in a new tab

A) Schematics of the iCRISPR system. In established iCas9 hESCs, Cas9 expression is induced by doxycycline treatment. After transfection of gRNA, Cas9 is guided to the target locus via Watson-Crick base pairing and induces DNA DSBs. In the absence of repair templates, error-prone non-homologous end joining often results in Indel (insertion/deletion) mutations. In the presence of repair templates, HDR can generate patient-specific mutations or correct a mutation in mutant lines. B) Schematics of the experiment design. 18 CRISPR/gRNAs were used to target 8 genes to generate 62 hESC mutants using iCRISPR. By stepwise differentiation of the hESC mutants into pancreatic lineage, the effects of these mutations were extensively examined by immunofluorescence staining, flow cytometry, qRT-PCR and Western blotting at each developmental step. C) Percentage of hESC clonal lines carrying different types of mutations generated with CRISPR gRNA targeting. +/+: no mutation in either allele; +/Indel: with Indels in one allele; Indel/Indel: with Indels in both alleles or Indel/Y in the case of ARX. *: ARX is on the X chromosome and the parental HUES8 line is from a male donor. D) Western blot analysis for validation of the loss of wild-type proteins in hESC knockout mutants in PDX1, MNX1 and HES1. Clonal names and mutant genotypes are labeled on the top, and the CRISPR/gRNAs used to generate the mutants are labeled in the bottom. The PDX1 antibody detected two protein bands likely associated with post-translational modification of PDX1 (Frogne et al., 2012). (See also in Figure S2)

We previously established the iCRISPR platform based on the creation of iCas9 hESCs that express the RNA-guided DNA endonuclease Cas9 upon doxycycline treatment (Figure 2A) (González et al., 2014). Transfecting doxycycline-treated iCas9 hESCs with synthetic chimeric guide RNAs (gRNAs) directs Cas9 for site-specific DNA cleavage and enables efficient generation of mutant hESC lines. To further increase the throughput, we designed a PCR-based strategy for in vitro gRNA synthesis compatible with multi-well format (Figure S2A), enabling the generation of a large number of gRNAs in a single day. For each gene of interest, we used two gRNAs to offset potential CRISPR/Cas off-target effects, and targeted distinct sequences either shortly after the start codon or within sequences corresponding to critical functional domains (Figure S2B, 2C and Table S1). To verify the generation of knockout alleles, we identified mutant lines with frameshift mutations and performed Western blot analysis for the corresponding protein when reliable antibodies were available (Figure 2D). In protein lysates collected from the PH-β stage, PDX1, MNX1 and HES1 proteins were detected in wild-type but not in the corresponding biallelic mutant cells. A decrease in protein levels was also noticed in PDX1 and MNX1 heterozygous mutant lines. These findings support the generation of null or strong loss-of-function alleles. For simplicity, we refer to biallelic and monoalleic frameshift mutations as the “−/−” and “+/−” genotypes respectively henceforward.

Interrogation of developmental defects underlying neonatal diabetes

The large collection of knockout hESC lines presented us the opportunity to systematically interrogate the pathogenesis of PNDM by characterizing mutant phenotypes at each differentiation stage (i.e. DE, PP and PH-β stages) using a panel of assays including FACS analysis (Figure 2B). To mitigate potential line-to-line variations, we examined 4 −/− mutant lines per gene (2 gRNAs per gene and 2 lines per gRNA) except for HES1, for which we analyzed two −/− lines from one gRNA (Table S3). Two lines from the same targeting experiment without mutations in the target gene were randomly chosen as isogenic wild-type controls. As consistent phenotypes were observed in all analyses, results of mutant lines affecting the same gene were combined for statistical analysis. In the undifferentiated state, all clonal lines displayed typical hESC morphology and expressed key pluripotency proteins such as NANOG and OCT4 (POU5F1) as determined by immunofluorescence staining (Figure S3A and B). None of the mutations affected DE formation as shown by examination of DE markers SOX17 and FOXA2 (co-immunostaining) and CXCR4 (FACS) (Figure 3A, S3C and S3D). At the PP stage, no PDX1+ cells were detected in PDX1−/− mutant lines by immunostaining and FACS analysis as expected (Figure 3A, S3E and S3F). All other mutants formed PDX1+ pancreatic progenitors normally except for RFX6−/− mutants, which showed a ~40% reduction of PDX1+ cells compared with wild-type lines (Figure 3A and S3F).

Figure 3. Phenotypic analysis of hESCs mutants to determine the developmental basis for PNDM.

Figure 3

Open in a new tab

A) FACS quantification of CXCR4+ cells, PDX1+ cells and INS+ cells in hESC mutant and wild-type (WT) cells at the DE, PP and PH-β cell stage respectively. FACS quantification results were normalized to the wild-type controls from the same experiment and combined together for statistic analysis. n = 8 for all genes except for HES1 (n = 4). Results from 4 (2 for HES1) clonal lines each with two differentiation experiments were combined. B) Representative immunofluorescence staining of hESC mutants and wild-type cells at the PH-β cell stage. C) Representative FACS plots of INS+ cells in hESC mutants and wild-type cells at the endocrine cell stage. D) qRT-PCR analysis of endocrine markers in hESC mutants and wild-type cells at the endocrine cell stage. n = 8 for all genes except for HES1 (n = 4). (See also in Figure S3)

At the PH-β stage, all −/− mutant lines exhibited a defect with the exception of PTF1A−/−, MNX1−/− and GLIS3−/− mutants. PDX1−/−, RFX6−/− and NGN3−/− mutants all exhibit a severe impairment in forming INS+ or GCG+ cells by immunofluorescence staining and FACS analysis (Figure 3B and C). qRT-PCR analysis also detected a significant reduction in the expression of additional endocrine hormone genes SST and GHRL (Figure 3D). These findings demonstrate critical requirements for PDX1, RFX6 and NGN3 in proper pancreatic endocrine differentiation, and establish causal roles of the corresponding mutations in PNDM. Distinct phenotypes were observed in HES1−/− and ARX−/Y mutants. A significant increase of endocrine cells was observed in HES1−/− mutants based on INS and GCG immunofluorescence staining and qRT-PCR analysis (P < 0.05), and FACS analysis detected a ~2 fold increase of INS+ cells (Figure 3A–D). These findings are in agreement with conclusions from murine studies that HES1 inhibits pancreatic endocrine differentiation (Apelqvist et al., 1999; Jensen et al., 2000). In ARX−/Y mutants, GCG+ cells were absent as determined by immunofluorescence staining (Figure 3B), which is also observed from murine studies (Collombat et al., 2003). We also noticed a ~30% reduction in the number of INS+ cells as determined by FACS analysis (Figure 3A and C). qRT-PCR analysis confirmed corresponding changes in GCG and INS mRNA expression (P < 0.01) (Figure 3D). This finding supports a previously proposed role for Gcg in β cell differentiation (Prasadan et al., 2002), though the reduction of INS+ cells could also reflect the loss of immature polyhormonal β cells in ARX−/Y mutant lines.

RFX6 regulates the formation of early pancreatic progenitor cells

Our study revealed an unexpected reduction of PDX1+ pancreatic progenitor cells in RFX6−/− mutants. RFX6 is known to regulate the differentiation of pancreatic progenitors into endocrine cells during embryonic development (Smith et al., 2010; Soyer et al., 2010), and it is also important for adult β cell function (Chandra et al., 2014; Piccand et al., 2014). However, an earlier requirement for RFX6 in the specification of pancreatic progenitor cells is unknown. The reduction of pancreatic progenitor cells was not caused by reduced proliferation or increased apoptosis. In fact, by immunofluorescence staining and FACS analysis at the PP stage, RFX6−/− mutants showed a slightly higher percentage of PDX1+ cells expressing the mitotic marker Phospho-Histone H3 and a reduced percentage of PDX1+ cells expressing the apoptosis marker cleaved Caspase-3 (Figure 4A–C and S4A). Instead, we observed decreased PDX1 immunofluorescence intensity in RFX6−/− pancreatic progenitors, which was also corroborated by a leftward shift in the FACS histogram compared to wild-type cells (Figure 4A and D). Thus, RFX6 regulates PDX1 expression either direct or indirectly, and without RFX6, the formation of pancreatic progenitor cells is impaired or delayed. A reduced number of pancreatic progenitor cells in embryos would predict a smaller pancreas in adults (Stanger et al., 2007). Indeed, a reduction in the size of the pancreas was noted in Rfx6−/− mice as well as in patients carrying biallelic RFX6 mutations (Concepcion et al., 2014; Smith et al., 2010). We speculate that a similar phenotype is likely present during mouse development, and the sensitivity of the FACS assay contributed to the detection of this early phenotype in the hESC system.

Figure 4. RFX6 and PDX1 in pancreatic differentiation.

Figure 4

Open in a new tab

A) Representative immunofluorescence staining of PDX1, Phospho-Histone H3 (PHH3) and cleaved Caspase 3 (Casp3) expression in wild-type and RFX6−/− mutants at the pancreatic progenitor stage. B, C) FACS quantification of the percentage of PHH3 (B) and Casp3-expressing cells (C) in PDX1+ cells formed from wild-type and RFX6−/− mutants at the PP stage. n = 4 for wild-type lines and n = 8 for RFX6−/− lines. D) FACS histogram for PDX1 expression in wild-type and RFX6−/− mutants at the pancreatic progenitor stage. E) Schematics illustrating the protein structure of the wild-type PDX1 protein and the predicted truncated proteins expressed from PDX1L36fs and PDX1A34fs alleles. Transactivation and DNA-binding homeobox domains are indicated in blue and red respectively. New sequences resulting from frameshift translation are indicated in hatched boxes. F, G) Representative immunofluorescence staining (F) and FACS plots (G) of the wild-type, PDX1−/− (PDX1L36fs/L36fs and PDX1L36fs/A34fs) and PDX1+/− (PDX1L36fs/+ and PDX1A34fs/+) mutant cells at the PH-β cell stage. H) FACS quantification of INS+ cells in wild-type, PDX1−/− and PDX1+/− mutants at the PH-β cell stage. n = 4: two independent experiments were performed on 2 lines for each genotype. I) qRT-PCR analysis of endocrine markers in wild-type control, PDX1−/− and PDX1+/− mutants at the PH-β cell stage. n = 4. J-L) Representative immunofluorescence (J) Western blot analysis (K) and FACS analysis (L) is shown for the wild-type PDX1+/+ control, PDX1−/− (PDX1L36fs/L36fs and PDX1L36fs/A34fs) and PDX1+/− (PDX1L36fs/+ and PDX1A34fs/+) mutant cells The red bar indicates differential PDX1 expression levels in the FACS histogram. (See also in Figure S4)

A haploinsufficient requirement for PDX1 in endocrine differentiation

In analyzing PDX1+/− mutants, a ~65% reduction of INS+ cells was detected at the PH-β stage by FACS analysis (Figure 4E–H). Further immunofluorescence staining confirmed the reduction of both INS+ and GCG+ cells (Figure 4F), and qRT-PCR analysis revealed a similar decrease in all pancreatic endocrine hormone genes, with INS and SST being the most affected (Figure 4I). The haploinsufficiency of PDX1 in endocrine differentiation is surprising. Although PDX1 heterozygous patients exhibit diabetes with the age of onset ranging from 17 to 67 years (Stoffers et al., 1997a), it is thought to be caused by defects in β-cell function and/or the maintenance of β-cell mass in adults (Brissova et al., 2005; Brissova et al., 2002; Johnson et al., 2003). No developmental abnormalities are known to occur in Pdx1+/− mice (Jonsson et al., 1994; Offield et al., 1996).

We therefore performed additional experiments to confirm that the unexpected phenotype observed in PDX1+/− hESCs was due to haploinsufficiency rather than the frameshift alleles having dominant negative effects. The two PDX1 heterozygous mutant lines carry different frameshift mutations, PDX1L36fs and PDX1A34fs (Figure 4E). To verify that no functional protein was made from the mutant allele, we examined two biallelic mutant lines carrying the same mutations, PDX1L36fs/L36fs and PDX1L36fs/A34fs at the PP stage so that the consequence of these mutations could be determined without interference of any endocrine differentiation phenotype. No PDX1 expression was detected in PDX1L36fs/L36fs or PDX1L36fs/A34fs mutants by Western blot and immunofluorescence staining (Figure 4J and K). Notably, a reduction in PDX1 expression was observed by Western blot analysis in heterozygous mutants (PDX1L36fs/+ and PDX1A34fs/+) (Figure 4K), which corresponded to a leftward shift in the FACS histogram comparing heterozygous vs wild-type cells (Figure 4L). These results suggest that PDX1L36fs and PDX1A34fs are both null alleles, and losing one functional PDX1 allele results in decreased PDX1 protein expression, which in turn compromises endocrine differentiation. This dosage-sensitive requirement for PDX1 in the formation of endocrine cells adds to the well-established roles of PDX1 in regulating the specification of pancreatic progenitors and the function of adult β cells. Considering the association of PDX1 heterozygous mutations and genetic variants to type 2 diabetes (Hani et al., 1999; Macfarlane et al., 1999; Stoffers et al., 1997a), our results suggest that compromised β cell development could predispose an individual to diabetes.

Previous studies in mice and other model organisms have identified a gene regulatory cascade of transcription factors governing pancreatic development (Shih et al., 2013). The largely consistent mutant phenotypes observed in hESC mutants suggest a conserved gene regulatory network in humans (Figure S4C, black lines), a conclusion also supported by gene expression analysis (Figure S4B). In addition,observations such as the phenotypes observed with RFX6−/− and PDX1+/− mutant lines, led us to propose a number of potentially human-specific regulations that could involve either direct or indirect transcriptional regulation (Figure S4C, red lines).

An important but not absolute requirement for NGN3 in human endocrine differentiation

When analyzing endocrine differentiation phenotypes, we noticed a few differences between PDX1−/−, RFX6−/− and NGN3−/− mutants. For example, no GCG+ cells were detected in RFX6−/− mutants, whereas a small number of GCG+ cells were observed in PDX1−/− and NGN3−/− mutants (Figure S4D). Additionally, although the number of INS+ cells was greatly reduced in NGN3−/− mutants, a small number of INS+ cells (~0.5%) were consistently detected by immunofluorescence staining and FACS analysis; whereas no INS+ cells were detected in PDX1−/− or RFX6−/− mutants (Figure 3C and S4D). In comparison, murine studies also identified Gcg+ cells in Pdx1−/− and Ngn3−/− mutants, but no Ins+ cells were detected in Ngn3−/− mutants at all developmental stages examined (Gradwohl et al., 2000; Offield et al., 1996; Wang et al., 2008). Thus NGN3−/− hESCs exhibited a distinct phenotype from the mouse knockouts.

This finding is intriguing considering phenotypes of patients carrying biallelic NGN3 mutations reported to date (Pauerstein et al., 2015; Pinney et al., 2011; Rubio-Cabezas et al., 2014; Rubio-Cabezas et al., 2011; Sayar et al., 2013; Unlusoy Aksu et al., 2015; Wang et al., 2006). Although all 11 patients suffer from congenital malabsorptive diarrhea due to impaired development of enteroendocrine cells, only 3 patients were diagnosed with permanent or transient neonatal diabetes, and all 3 patients had low but detectable blood C-peptide levels indicative of the presence of residual β cells. This could be explained if a patient carried hypomorphic instead of null mutations as suggested by some studies (Jensen et al., 2007; Pauerstein et al., 2015), and indeed ~10% of normal NGN3 activity appears sufficient to support endocrine differentiation (McGrath et al., 2015). On the other hand, our findings with NGN3−/− hESCs suggest that a small number of β cells could form even in the absence of any NGN3 activity.

We generated additional NGN3 mutant lines to further validate this conclusion. First, to make sure that true null phenotypes are analyzed, we examined a NGN3 mutant line, designated as NGN3Δ/Δ, which carries a 526-bp homozygous deletion that removes most of the NGN3 protein-coding sequences including the start codon and the sequences corresponding to the DNA-binding bHLH domain (Figure 5A and S2B). In addition, we established disease-mimicking hESC lines through recreating the NGN3R107S/R107S genotype of a patient born with no overt diabetic symptoms (Wang et al., 2006) (Figure 5A and B, S5A and B, and Table S2). We observed a great reduction in the number of INS+ cells at the PH-β stage in the NGN3Δ/Δ hESC line, 2 NGN3R107S/R107S hESC lines, and 4 NGN3−/− mutants with frameshift mutations (generated using NGN3-cr3 and cr4) (Figure S2B and Table S3). Nevertheless, ~0.5% INS+ cells were readily detected in all NGN3 mutant lines by immunostaining and FACS analysis (Figure 5C, 5D and 3C). Although the NGN3R107S mutation showed some activity in over-expression assays (Pauerstein et al., 2015), it behaved similarly to a null allele when expressed from the endogenous locus. Importantly, the consistent detection of INS+ cells in all 11 NGN3 mutant lines examined could explain why C-peptide was detected even in NGN3 patients with the most severe symptoms.

Figure 5. An important but not essential requirement for NGN3 in human endocrine differentiation.

Figure 5

Open in a new tab

A) Schematics showing the genomic sequences of the wild-type NGN3 allele (in NGN3+/+), the NGN3 mutant allele with a 526 bp deletion (in NGN3Δ/Δ), the NGN3 Indel mutant alleles generated using NGN3-cr3, cr4, cr5 and cr6 (in NGN−/−) and the NGN3 allele carrying a patient-specific mutation (in NGN3R107S/R107S). Coding sequences are indicated in grey, with the sequence corresponding to the bHLH DNA-binding domain highlighted in red. C>A substitution (red, also indicated by an asterisk) was introduced using the ssDNA HDR template, resulting in an R107S amino acid substitution. B) Sanger sequencing results of NGN3 wild-type cells and NGN3 mutants homozygous for NGN3R107S. C>A substitution is labeled in red box. C) Representative FACS plots of INS+ cells showing the appearance of INS+ cells only in NGN3Δ/Δ but not PDX1−/− mutants. D) Representative immunofluorescence staining showing the residue GCG+ and/or INS+ cells in NGN3−/− (generated using NGN3-cr3 and cr4), NGN3Δ/Δ and NGN3R107S/R107S cells. Arrows point to the same cells in corresponding images in the upper and lower panel. Intact nuclei DAPI staining (without fragmentation) argues against non-specific staining from dead cells. (See also in Figure S4 and S5)

NGN3 is important for the generation of glucose-responsive β-like cells

Our findings thus far suggest that patients with no NGN3 activity could still form fetal β cells, but it is unclear whether functionally mature β cells could also be formed. To address this question, we adapted two recently developed protocols (Pagliuca et al., 2014; Rezania et al., 2014) to differentiate hESCs into glucose-responsive insulin secreting cells, which we refer to as β-like cells hereafter. To also rule out potential CRISPR/Cas off-target effects, we further corrected the mutation in an NGN3−/− line through homology-directed repair (HDR) using an ssDNA donor, creating 2 NGN3Cr/Cr lines (Figure 6A, S5F, S5G). The differentiation protocol enabled us to differentiate PDX1+NKX6.1- early pancreatic progenitor (designated as PP1) into PDX1+NKX6.1+ progenitor cells (designated as PP2), a key intermediate step necessary for the generation of functional β cells (Kelly et al., 2011; Rezania et al., 2013). No significant differences were observed among wild-type, NGN3Cr/Cr and NGN3−/− lines up to the PP2 stage, when ~95% PDX1+ and ~40% PDX1+NKX6.1+ cells were detected in all conditions (Figure 6B, S6B, S6C).

Figure 6. NGN3 is important for the generation of glucose-responsive β-like cells.

Figure 6

Open in a new tab

A) Schematics and Sanger sequencing results demonstrating the generation of knockout NGN3−/− hESC lines, and the subsequent creation of the corrected NGN3Cr/Cr hESC lines, in which both NGN3 mutant alleles were reversed to wild-type. B) FACS quantification of CXCR4+, PDX1+NKX6.1+ and CPEP+ (total and monohormonal) cells formed from wild-type, NGN3Cr/Cr and NGN3−/− lines at the DE, PP2 and β-like stage, respectively. n = 4. Results from 2 clonal lines each with 2 differentiation experiments were combined. C, D) Representative immunofluorescence staining (C) and FACS plots (D) for pancreatic endocrine hormone expression in wild-type, NGN3Cr/Cr lines and NGN3−/− mutants at the β-like stage. E, F) Representative immunofluorescence staining (E) and FACS plots (F) for C-peptide, PDX1 and NKX6.1 expression in wild-type, NGN3Cr/Cr lines and NGN3−/− mutants at the β-like stage. G) Glucose-stimulated insulin secretion assay for wild-type, NGN3Cr/Cr and NGN3−/− hESCs at the β-like stage. The fold change of insulin secretion with high glucose (16.7 mM) relative to low glucose (2.8 mM) treatment is shown. H, I) FACS quantification of the percentage of monohormonal and polyhormonal cells (H) and NKX6.1+ cells (I) in CPEP+ cells in wild-type, NGN3Cr/Cr and NGN3−/− hESCs. n = 4. (See also in Figure S5 and S6)

The PP2 cells were then cultured at air-liquid interface for further differentiation (Figure S6A, S6D). Wild-type and NGN3Cr/Cr hESCs formed ~15% CPEP+ β-like cells, most of which (~80%) were monohormonal (CPEP+GCG-SST-) by immunofluorescence staining and FACS analysis (Figure 6B, C and D). The majority of CPEP+ cells expressed key β cell transcription factors, PDX1, NKX6.1, NKX2.2 and NEUROD1 (Figure 6E, 6F, S6E). Comprehensive qRT-PCR analysis revealed that β-like cells expressed key β cell markers at levels comparable to human islets (Figure S6G). Importantly, these β-like cells also exhibited glucose-stimulated insulin secretion (GSIS): the ratio of insulin secreted in high glucose (16.7 mM) to low glucose (2.8 mM) was around 2 fold (Figure 6G and S6F).

Unlike wild-type and NGN3Cr/Cr hESCs, the formation of β-like cells from NGN3−/− hESCs was severely affected, yet ~0.5% CPEP+ cells were still detected (Figure 6B–F). The majority of CPEP+ cells were monohormonal (Figure 6H), and the ratio of monohormonal vs polyhormonal cells was similar to those observed in wild-type and NGN3Cr/Cr lines. Notably, however, only ~10% CPEP+ cells co-expressed NKX6.1 (Figure 6I). Insulin secretion was detected from NGN3−/− cells, but there was no significant GSIS (Figure 6G and S6F). Therefore, NGN3 is not absolutely required for the formation of monohormonal CPEP+ cells, but the majority of β cells present in NGN3-deficient patients are likely impaired in function. The small number (~0.05%) of CPEP+NKX6.1+ cells formed from NGN3−/− hESCs could be functional, but to evaluate their function would require more sensitive assays or GSIS on purified CPEP+NKX6.1+ cells.

Temporal control of NGN3 activity reveals an endocrine differentiation competence window

To further explore the competence window responsive to NGN3 expression, we combined gain- and loss-of-function approaches. Modeling after an elegant murine study by Grapin-Botton and colleagues (Johansson et al., 2007), we inducibly expressed NGN3 in the NGN3 knockout background (Figure 7A). We used a pair of gRNAs targeting each end of the Puro-iCas9 cassette in the AAVS1 locus, and replace it with a Hygro-iNGN3 transgene through HDR (Figure S7A). Despite the relatively large size of the Cas9 cassette (~6kb), this strategy turned out to work as efficiently as the original targeting of the AAVS1 locus. All 16 colonies picked after antibiotic selection showed correct Hygro-iNGN3 integration by PCR genotyping (Figure S7B), and 15 clonal lines expressed NGN3 upon doxycycline treatment (Figure S7C). Thus similar strategies could be applied in other genome-editing contexts.

Figure 7. Temporal control of NGN3 activity in NGN3−/− mutants.

Figure 7

Open in a new tab

A) Schematics illustrating gene correction and temporal control of NGN3 expression in NGN3−/− hESCs. B) Schematics of temporal regulation of NGN3 expression at different stages of pancreatic differentiation in the NGN3−/− mutant background. C, D) Representative immunofluorescence staining (C) and FACS plots (D) for the induction of pancreatic endocrine cells corresponding to NGN3 expression at different differentiation stages. E) Quantification of the FACS results in panel D. n = 6: results from 3 clonal lines each with two differentiation experiments were combined. Statistic analysis was indicated for both monohormonal and total CPEP+ cells. (See also in Figure S7)

Hygro-iNGN3; NGN3−/− cells were differentiated to the PP1 stage and then subject to doxycycline treatment at 4 differentiation stages, each of a 3-day interval (Figure 7B). Generation of major endocrine cell types, INS+, GCG+ and SST+ cells, were evaluated by immunofluorescence staining and FACS analysis at the end of the differentiation (Figure 7B, C and D). Expression of the NGN3 transgene in the NGN3 knockout background induces endocrine differentiation in all stages, with the strongest effects corresponding to the S4 and S5 stages (Figure 7C, D and E). These findings support a competence window that coincides with the formation of PDX1+NKX6.1+ pancreatic progenitor cells.

Discussion

Combining stem cell biology and powerful genome-editing technology, our work supports the feasibility of using hPSCs to examine a large number of candidate disease genes for functional validation and mechanistic investigation. Compared with studies focused on analyzing cellular phenotypes of hPSC-derived cells, we are able to further infer the exact developmental step(s) perturbed in a congenital disorder. For instance, PNDM patients with RFX6 mutations are likely affected at two developmental stages: the formation of pancreatic progenitors and their further differentiation into functional endocrine cells. The hESC-based platform also provides knowledge about human biology not readily available from murine studies. Mutations in the same gene could cause divergent phenotypes in humans and mice (Zhu and Huangfu, 2013). Notable examples include monogenic diabetes-associated genes NGN3, GATA6, HNF1A, HNF4A, and HNF1B (Gradwohl et al., 2000; Rodriguez-Segui et al., 2012; Wang et al., 2006). However, the causes for these discrepancies are often debated as in the case of NGN3. While some patients may indeed harbor partial loss-of-function mutations (Jensen et al., 2007; Pauerstein et al., 2015), our extensive analysis of NGN3 mutant hESC lines supports the presence of as-yet-unidentified, NGN3-independent pro-endocrine pathway(s) that are likely not conserved in mice. Our study also uncovered a haploinsufficient requirement for PDX1 in pancreatic endocrine differentiation. Although type 2 diabetes is generally considered a physiological problem, our findings suggest that compromised β cell development may predispose an individual to diabetes later in life. Supporting this idea, there is considerable variation in β cell mass between individuals (Meier et al., 2008). Thus one may use hiPSCs for analysis of β cell differentiation and physiology, and then identify individuals with a genetic susceptibility to type 2 diabetes for early intervention.

With the continuous advance of the genome-editing technology, it is now possible to perform sophisticated genetic manipulations for mechanistic studies. The challenge now is to develop better ways to recreate, as closely as possible, what occurs in a human embryo. Advances in the field are bringing us closer to this goal. For instance, when we initiated the study, we used the best protocol available at the time, which enabled us to mimic early stages of pancreatic development relatively well, but it generates polyhormonal β cells of fetal phenotypes (D’Amour et al., 2006). Recent advancements (Nostro et al., 2015; Pagliuca et al., 2014; Rezania et al., 2014; Russ et al., 2015) now enable us to investigate later events such as the formation of monohormonal, glucose-responsive β cells. This has allowed us to investigate the exact roles of NGN3 in the formation of functional β cells, and explore the competence window of NGN3-dependent β cell formation. This protocol may allow us to identify later differentiation defects in PTF1A−/−, MNX1−/− and GLIS3−/− mutant lines, which did not exhibit an overt phenotype at early differentiation stages. When human vs mouse divergence is observed, it would be valuable to also compare findings in hESCs with analogous studies in mouse ESCs. Future studies will likely benefit from incorporating advances in other organ systems, such as the self-organizing organoid structures for a better recapitulation of the complex in vivo tissue milieu. As such, hPSCs can offer a useful system complementing in vivo studies, and there is great potential in utilizing this unique cell culture system for studying disease mechanisms and for discovering disease-causing genes through unbiased screening approaches.

Experimental Procedures

All experiments were performed on HUES8, a well-characterized hESC line (NIH approval number: NIHhESC-09-0021). The generation of HUES8 iCas9 hESCs was described previously (González et al., 2014; Zhu et al., 2014). Clonal hESC mutant lines were established through transfection of HUES8 iCas9 hESCs with specific CRISPR gRNAs, and subsequent sing-cell cloning and verification by Sanger-sequencing. Please refer to the Supplemental Experimental Procedures for more detailed description of these procedures.

HUES8 hESCs were differentiated into PH-β cells using a protocol adapted from published studies (D’Amour et al., 2006; Nostro et al., 2011). In brief, undifferentiated hESCs were cultured on irradiated mouse embryonic fibroblasts until 50–70% confluency to start differentiation (designated as “d0”). The differentiation condition and media recipes are presented in Figure S1F and Table S5.

HUES8 hESCs were differentiated into glucose-responsive β-like cells using a protocol adapted from two recently publications (Pagliuca et al., 2014; Rezania et al., 2014). The main modification was the addition of Wnt antagonist IWP-2 at the S2 and S4 stages (Figure S6A) to improve the consistency of differentiation outcomes (Zhu et al., data not shown). In brief, undifferentiated hESCs were cultured to ~70–80% confluency in E8 condition, and dissociated into single cells using TrypLE Select. Collected cell pellet were re-suspended in E8 medium supplemented with 10 µM Y-27632 and seeded at ~1.4X105 cells/cm2 on VTN-coated plates. E8 medium were changed after 24 hours and differentiation was initiated 48 hours after plating when the culture was ~80–90% in confluency (designated as “d0”). The differentiation condition and media recipes are presented in Figure S6A and Table S6.

For evaluation of differentiation phenotypes, mutant lines and isogenic wild-type controls of similar passage numbers were used. All differentiation experiments were repeated at least twice with at least 2 independent clonal lines. Within each experiment, at least two technical repeats were performed. As a quality control for successful differentiation, experiments were continued only when >40% CXCR4+ cells were detected by FACS analysis at the DE stage. In completion of each differentiation experiment, genomic DNA was collected for sequence analysis to confirm the identity of each mutant line. Microsoft Excel and GraphPad Prism 6 were used for unpaired two-tailed student t-test unless otherwise indicated.

Supplementary Material

Acknowledgments

We thank Rudolf Jaenisch and Douglas Melton for providing vectors through Addgene; Dirk Hockemeyer for the AAVS1 3’ external probe vector; Alireza Rezania and Sebastian Rieck for advice on differentiating hPSCs into glucose-responsive β-like cells; Daniela Georgieva for technical assistance; Hui Zeng for sharing qPCR primer information; and Shuibing Chen and members of the Huangfu laboratory for insightful discussions and critical reading of the manuscript. The MNX1 81.5C10 monoclonal antibody developed by Thomas Jessell and Susan Brenner-Morton was obtained from the Developmental Studies Hybridoma Bank, created by the NICHD of the NIH and maintained at The University of Iowa, Department of Biology, Iowa City, IA 52242. This study was funded in part by NIH/NIDDK (R01DK096239), New York State Stem Cell Science (NYSTEM C029156), Basil O’Connor Starter Scholar Award from March of Dimes Birth Defects Foundation, and Tri-Institutional Stem Cell Initiative. Z.Z. and F.G. were supported by the NYSTEM postdoctoral fellowship from the Center for Stem Cell Biology of the Sloan Kettering Institute.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Author contributions

D.H. and Z.Z. conceived the project, designed experiments, analyzed and interpreted results; Z.Z. performed most experiments; Q.L. and C.S. assisted with the generation of knockout lines; K.L. assisted with the intracellular staining flow cytometric analysis; B.R. performed cryosectioning and immunofluorescence staining at the β-like stage; F.G. constructed the Hygro-iDEST plasmid used in the NGN3 temporal rescue experiment; D.H. and Z.Z. wrote the manuscript; and all other authors provided editorial advice.

References

  1. Aguilar-Bryan L, Bryan J. Neonatal diabetes mellitus. Endocr Rev. 2008;29:265–291. doi: 10.1210/er.2007-0029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Apelqvist A, Li H, Sommer L, Beatus P, Anderson DJ, Honjo T, Hrabe de Angelis M, Lendahl U, Edlund H. Notch signalling controls pancreatic cell differentiation. Nature. 1999;400:877–881. doi: 10.1038/23716. [DOI] [PubMed] [Google Scholar]
  3. Bock C, Kiskinis E, Verstappen G, Gu H, Boulting G, Smith ZD, Ziller M, Croft GF, Amoroso MW, Oakley DH, et al. Reference Maps of human ES and iPS cell variation enable high-throughput characterization of pluripotent cell lines. Cell. 2011;144:439–452. doi: 10.1016/j.cell.2010.12.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brissova M, Blaha M, Spear C, Nicholson W, Radhika A, Shiota M, Charron MJ, Wright CV, Powers AC. Reduced PDX-1 expression impairs islet response to insulin resistance and worsens glucose homeostasis. Am J Physiol Endocrinol Metab. 2005;288:E707–E714. doi: 10.1152/ajpendo.00252.2004. [DOI] [PubMed] [Google Scholar]
  5. Brissova M, Shiota M, Nicholson WE, Gannon M, Knobel SM, Piston DW, Wright CV, Powers AC. Reduction in pancreatic transcription factor PDX-1 impairs glucose-stimulated insulin secretion. J Biol Chem. 2002;277:11225–11232. doi: 10.1074/jbc.M111272200. [DOI] [PubMed] [Google Scholar]
  6. Chandra V, Albagli-Curiel O, Hastoy B, Piccand J, Randriamampita C, Vaillant E, Cave H, Busiah K, Froguel P, Vaxillaire M, et al. RFX6 Regulates Insulin Secretion by Modulating Ca(2+) Homeostasis in Human beta Cells. Cell Rep. 2014;9:2206–2218. doi: 10.1016/j.celrep.2014.11.010. [DOI] [PubMed] [Google Scholar]
  7. Collombat P, Mansouri A, Hecksher-Sorensen J, Serup P, Krull J, Gradwohl G, Gruss P. Opposing actions of Arx and Pax4 in endocrine pancreas development. Genes Dev. 2003;17:2591–2603. doi: 10.1101/gad.269003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Concepcion JP, Reh CS, Daniels M, Liu X, Paz VP, Ye H, Highland HM, Hanis CL, Greeley SA. Neonatal diabetes, gallbladder agenesis, duodenal atresia, and intestinal malrotation caused by a novel homozygous mutation in RFX6. Pediatr Diabetes. 2014;15:67–72. doi: 10.1111/pedi.12063. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. D’Amour KA, Bang AG, Eliazer S, Kelly OG, Agulnick AD, Smart NG, Moorman MA, Kroon E, Carpenter MK, Baetge EE. Production of pancreatic hormone-expressing endocrine cells from human embryonic stem cells. Nat Biotechnol. 2006;24:1392–1401. doi: 10.1038/nbt1259. [DOI] [PubMed] [Google Scholar]
  10. DeKelver RC, Choi VM, Moehle EA, Paschon DE, Hockemeyer D, Meijsing SH, Sancak Y, Cui X, Steine EJ, Miller JC, et al. Functional genomics, proteomics, and regulatory DNA analysis in isogenic settings using zinc finger nuclease-driven transgenesis into a safe harbor locus in the human genome. Genome Res. 2010;20:1133–1142. doi: 10.1101/gr.106773.110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Flanagan SE, De Franco E, Lango Allen H, Zerah M, Abdul-Rasoul MM, Edge JA, Stewart H, Alamiri E, Hussain K, Wallis S, et al. Analysis of transcription factors key for mouse pancreatic development establishes NKX2-2 and MNX1 mutations as causes of neonatal diabetes in man. Cell Metab. 2014;19:146–154. doi: 10.1016/j.cmet.2013.11.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Frogne T, Sylvestersen KB, Kubicek S, Nielsen ML, Hecksher-Sorensen J. Pdx1 is post-translationally modified in vivo and serine 61 is the principal site of phosphorylation. PLoS One. 2012;7:e35233. doi: 10.1371/journal.pone.0035233. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. González F, Zhu Z, Shi ZD, Lelli K, Verma N, Li QV, Huangfu D. An iCRISPR Platform for Rapid, Multiplexable, and Inducible Genome Editing in Human Pluripotent Stem Cells. Cell Stem Cell. 2014;15:215–226. doi: 10.1016/j.stem.2014.05.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gradwohl G, Dierich A, LeMeur M, Guillemot F. 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. doi: 10.1073/pnas.97.4.1607. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hani EH, Stoffers DA, Chevre JC, Durand E, Stanojevic V, Dina C, Habener JF, Froguel P. Defective mutations in the insulin promoter factor-1 (IPF-1) gene in late-onset type 2 diabetes mellitus. J Clin Invest. 1999;104:R41–R48. doi: 10.1172/JCI7469. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Jensen J, Pedersen EE, Galante P, Hald J, Heller RS, Ishibashi M, Kageyama R, Guillemot F, Serup P, Madsen OD. Control of endodermal endocrine development by Hes-1. Nat Genet. 2000;24:36–44. doi: 10.1038/71657. [DOI] [PubMed] [Google Scholar]
  17. Jensen JN, Rosenberg LC, Hecksher-Sorensen J, Serup P. Mutant neurogenin-3 in congenital malabsorptive diarrhea. N Engl J Med. 2007;356:1781–1782. doi: 10.1056/NEJMc063247. author reply 1782. [DOI] [PubMed] [Google Scholar]
  18. Johansson KA, Dursun U, Jordan N, Gu G, Beermann F, Gradwohl G, Grapin-Botton A. 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. doi: 10.1016/j.devcel.2007.02.010. [DOI] [PubMed] [Google Scholar]
  19. Johnson JD, Ahmed NT, Luciani DS, Han Z, Tran H, Fujita J, Misler S, Edlund H, Polonsky KS. Increased islet apoptosis in Pdx1+/− mice. J Clin Invest. 2003;111:1147–1160. doi: 10.1172/JCI16537. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Jonsson J, Carlsson L, Edlund T, Edlund H. Insulin-promoter-factor 1 is required for pancreas development in mice. Nature. 1994;371:606–609. doi: 10.1038/371606a0. [DOI] [PubMed] [Google Scholar]
  21. Kelly OG, Chan MY, Martinson LA, Kadoya K, Ostertag TM, Ross KG, Richardson M, Carpenter MK, D’Amour KA, Kroon E, et al. Cell-surface markers for the isolation of pancreatic cell types derived from human embryonic stem cells. Nat Biotechnol. 2011;29:750–756. doi: 10.1038/nbt.1931. [DOI] [PubMed] [Google Scholar]
  22. Macfarlane WM, Frayling TM, Ellard S, Evans JC, Allen LI, Bulman MP, Ayres S, Shepherd M, Clark P, Millward A, et al. Missense mutations in the insulin promoter factor-1 gene predispose to type 2 diabetes. J Clin Invest. 1999;104:R33–R39. doi: 10.1172/JCI7449. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. McGrath PS, Watson CL, Ingram C, Helmrath MA, Wells JM. The Basic Helix-Loop-Helix Transcription Factor NEUROG3 Is Required for Development of the Human Endocrine Pancreas. Diabetes. 2015;64:2497–2505. doi: 10.2337/db14-1412. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Meier JJ, Butler AE, Saisho Y, Monchamp T, Galasso R, Bhushan A, Rizza RA, Butler PC. Beta-cell replication is the primary mechanism subserving the postnatal expansion of beta-cell mass in humans. Diabetes. 2008;57:1584–1594. doi: 10.2337/db07-1369. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Nostro MC, Sarangi F, Ogawa S, Holtzinger A, Corneo B, Li X, Micallef SJ, Park IH, Basford C, Wheeler MB, et al. Stage-specific signaling through TGFbeta family members and WNT regulates patterning and pancreatic specification of human pluripotent stem cells. Development. 2011;138:861–871. doi: 10.1242/dev.055236. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Nostro MC, Sarangi F, Yang C, Holland A, Elefanty AG, Stanley EG, Greiner DL, Keller G. Efficient generation of NKX6-1+ pancreatic progenitors from multiple human pluripotent stem cell lines. Stem Cell Reports. 2015;4:591–604. doi: 10.1016/j.stemcr.2015.02.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Offield MF, Jetton TL, Labosky PA, Ray M, Stein RW, Magnuson MA, Hogan BL, Wright CV. PDX-1 is required for pancreatic outgrowth and differentiation of the rostral duodenum. Development. 1996;122:983–995. doi: 10.1242/dev.122.3.983. [DOI] [PubMed] [Google Scholar]
  28. Osafune K, Caron L, Borowiak M, Martinez RJ, Fitz-Gerald CS, Sato Y, Cowan CA, Chien KR, Melton DA. Marked differences in differentiation propensity among human embryonic stem cell lines. Nat Biotechnol. 2008;26:313–315. doi: 10.1038/nbt1383. [DOI] [PubMed] [Google Scholar]
  29. Pagliuca FW, Millman JR, Gurtler M, Segel M, Van Dervort A, Ryu JH, Peterson QP, Greiner D, Melton DA. Generation of functional human pancreatic beta cells in vitro. Cell. 2014;159:428–439. doi: 10.1016/j.cell.2014.09.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Pauerstein PT, Sugiyama T, Stanley SE, McLean GW, Wang J, Martin MG, Kim SK. Dissecting Human Gene Functions Regulating Islet Development With Targeted Gene Transduction. Diabetes. 2015;64:3037–3049. doi: 10.2337/db15-0042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Piccand J, Strasser P, Hodson DJ, Meunier A, Ye T, Keime C, Birling MC, Rutter GA, Gradwohl G. Rfx6 Maintains the Functional Identity of Adult Pancreatic beta Cells. Cell Rep. 2014;9:2219–2232. doi: 10.1016/j.celrep.2014.11.033. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Pinney SE, Oliver-Krasinski J, Ernst L, Hughes N, Patel P, Stoffers DA, Russo P, De Leon DD. Neonatal diabetes and congenital malabsorptive diarrhea attributable to a novel mutation in the human neurogenin-3 gene coding sequence. J Clin Endocrinol Metab. 2011;96:1960–1965. doi: 10.1210/jc.2011-0029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Prasadan K, Daume E, Preuett B, Spilde T, Bhatia A, Kobayashi H, Hembree M, Manna P, Gittes GK. Glucagon is required for early insulin-positive differentiation in the developing mouse pancreas. Diabetes. 2002;51:3229–3236. doi: 10.2337/diabetes.51.11.3229. [DOI] [PubMed] [Google Scholar]
  34. Qiang L, Inoue K, Abeliovich A. Instant neurons: directed somatic cell reprogramming models of central nervous system disorders. Biol Psychiatry. 2014;75:945–951. doi: 10.1016/j.biopsych.2013.10.027. [DOI] [PubMed] [Google Scholar]
  35. Rajagopal J, Anderson WJ, Kume S, Martinez OI, Melton DA. Insulin staining of ES cell progeny from insulin uptake. Science. 2003;299:363. doi: 10.1126/science.1077838. [DOI] [PubMed] [Google Scholar]
  36. Rezania A, Bruin JE, Arora P, Rubin A, Batushansky I, Asadi A, O’Dwyer S, Quiskamp N, Mojibian M, Albrecht T, et al. Reversal of diabetes with insulin-producing cells derived in vitro from human pluripotent stem cells. Nat Biotechnol. 2014;32:1121–1133. doi: 10.1038/nbt.3033. [DOI] [PubMed] [Google Scholar]
  37. Rezania A, Bruin JE, Riedel MJ, Mojibian M, Asadi A, Xu J, Gauvin R, Narayan K, Karanu F, O’Neil JJ, et al. Maturation of human embryonic stem cell-derived pancreatic progenitors into functional islets capable of treating pre-existing diabetes in mice. Diabetes. 2012;61:2016–2029. doi: 10.2337/db11-1711. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Rezania A, Bruin JE, Xu J, Narayan K, Fox JK, O’Neil JJ, Kieffer TJ. Enrichment of human embryonic stem cell-derived NKX6.1-expressing pancreatic progenitor cells accelerates the maturation of insulin-secreting cells in vivo. Stem Cells. 2013;31:2432–2442. doi: 10.1002/stem.1489. [DOI] [PubMed] [Google Scholar]
  39. Rodriguez-Segui S, Akerman I, Ferrer J. GATA believe it: new essential regulators of pancreas development. J Clin Invest. 2012;122:3469–3471. doi: 10.1172/JCI65751. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Rubio-Cabezas O, Codner E, Flanagan SE, Gomez JL, Ellard S, Hattersley AT. Neurogenin 3 is important but not essential for pancreatic islet development in humans. Diabetologia. 2014;57:2421–2424. doi: 10.1007/s00125-014-3349-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Rubio-Cabezas O, Jensen JN, Hodgson MI, Codner E, Ellard S, Serup P, Hattersley AT. Permanent Neonatal Diabetes and Enteric Anendocrinosis Associated With Biallelic Mutations in NEUROG3. Diabetes. 2011;60:1349–1353. doi: 10.2337/db10-1008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Russ HA, Parent AV, Ringler JJ, Hennings TG, Nair GG, Shveygert M, Guo T, Puri S, Haataja L, Cirulli V, et al. Controlled induction of human pancreatic progenitors produces functional beta-like cells in vitro. EMBO J. 2015;34:1759–1772. doi: 10.15252/embj.201591058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Saha K, Jaenisch R. Technical challenges in using human induced pluripotent stem cells to model disease. Cell Stem Cell. 2009;5:584–595. doi: 10.1016/j.stem.2009.11.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Sayar E, Islek A, Yilmaz A, Akcam M, Flanagan SE, Artan R. Extremely rare cause of congenital diarrhea: enteric anendocrinosis. Pediatr Int. 2013;55:661–663. doi: 10.1111/ped.12169. [DOI] [PubMed] [Google Scholar]
  45. Sellick GS, Barker KT, Stolte-Dijkstra I, Fleischmann C, Coleman RJ, Garrett C, Gloyn AL, Edghill EL, Hattersley AT, Wellauer PK, et al. Mutations in PTF1A cause pancreatic and cerebellar agenesis. Nat Genet. 2004;36:1301–1305. doi: 10.1038/ng1475. [DOI] [PubMed] [Google Scholar]
  46. Senee V, Chelala C, Duchatelet S, Feng D, Blanc H, Cossec JC, Charon C, Nicolino M, Boileau P, Cavener DR, et al. Mutations in GLIS3 are responsible for a rare syndrome with neonatal diabetes mellitus and congenital hypothyroidism. Nat Genet. 2006;38:682–687. doi: 10.1038/ng1802. [DOI] [PubMed] [Google Scholar]
  47. Shih HP, Wang A, Sander M. Pancreas organogenesis: from lineage determination to morphogenesis. Annu Rev Cell Dev Biol. 2013;29:81–105. doi: 10.1146/annurev-cellbio-101512-122405. [DOI] [PubMed] [Google Scholar]
  48. Smith SB, Qu HQ, Taleb N, Kishimoto NY, Scheel DW, Lu Y, Patch AM, Grabs R, Wang J, Lynn FC, et al. Rfx6 directs islet formation and insulin production in mice and humans. Nature. 2010;463:775–780. doi: 10.1038/nature08748. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Soyer J, Flasse L, Raffelsberger W, Beucher A, Orvain C, Peers B, Ravassard P, Vermot J, Voz ML, Mellitzer G, et al. Rfx6 is an Ngn3-dependent winged helix transcription factor required for pancreatic islet cell development. Development. 2010;137:203–212. doi: 10.1242/dev.041673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Stanger BZ, Tanaka AJ, Melton DA. Organ size is limited by the number of embryonic progenitor cells in the pancreas but not the liver. Nature. 2007;445:886–891. doi: 10.1038/nature05537. [DOI] [PubMed] [Google Scholar]
  51. Stoffers DA, Ferrer J, Clarke WL, Habener JF. Early-onset type-II diabetes mellitus (MODY4) linked to IPF1. Nat Genet. 1997a;17:138–139. doi: 10.1038/ng1097-138. [DOI] [PubMed] [Google Scholar]
  52. Stoffers DA, Zinkin NT, Stanojevic V, Clarke WL, Habener JF. Pancreatic agenesis attributable to a single nucleotide deletion in the human IPF1 gene coding sequence. Nat Genet. 1997b;15:106–110. doi: 10.1038/ng0197-106. [DOI] [PubMed] [Google Scholar]
  53. Unlusoy Aksu A, Egritas Gurkan O, Sari S, Demirtas Z, Turkyilmaz C, Poyraz A, Dalgic B. Mutant neurogenin-3 in a Turkish boy with congenital malabsorptive diarrhea. Pediatr Int. 2015 doi: 10.1111/ped.12783. In Press. [DOI] [PubMed] [Google Scholar]
  54. Wang J, Cortina G, Wu SV, Tran R, Cho JH, Tsai MJ, Bailey TJ, Jamrich M, Ament ME, Treem WR, et al. Mutant neurogenin-3 in congenital malabsorptive diarrhea. N Engl J Med. 2006;355:270–280. doi: 10.1056/NEJMoa054288. [DOI] [PubMed] [Google Scholar]
  55. Wang S, Hecksher-Sorensen J, Xu Y, Zhao A, Dor Y, Rosenberg L, Serup P, Gu G. Myt1 and Ngn3 form a feed-forward expression loop to promote endocrine islet cell differentiation. Dev Biol. 2008;317:531–540. doi: 10.1016/j.ydbio.2008.02.052. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zhu Z, González F, Huangfu D. The iCRISPR Platform for Rapid Genome Editing in Human Pluripotent Stem Cells. Methods Enzymol. 2014;546:215–250. doi: 10.1016/B978-0-12-801185-0.00011-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Zhu Z, Huangfu D. Human pluripotent stem cells: an emerging model in developmental biology. Development. 2013;140:705–717. doi: 10.1242/dev.086165. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

NIHMS773607-supplement.pdf (7.8MB, pdf)

RESOURCES