Insm1 promotes endocrine cell differentiation by ... - Development

© 2014. Published by The Company of Biologists Ltd | Development (2014) 141, 2939-2949 doi:10.1242/dev.104810

RESEARCH ARTICLE

STEM CELLS AND REGENERATION

Insm1 promotes endocrine cell differentiation by modulating the expression of a network of genes that includes Neurog3 and Ripply3

ABSTRACT Insulinoma associated 1 (Insm1) plays an important role in regulating the development of cells in the central and peripheral nervous systems, olfactory epithelium and endocrine pancreas. To better define the role of Insm1 in pancreatic endocrine cell development we generated mice with an Insm1GFPCre reporter allele and used them to study Insm1-expressing and null populations. Endocrine progenitor cells lacking Insm1 were less differentiated and exhibited broad defects in hormone production, cell proliferation and cell migration. Embryos lacking Insm1 contained greater amounts of a non-coding Neurog3 mRNA splice variant and had fewer Neurog3/Insm1 co-expressing progenitor cells, suggesting that Insm1 positively regulates Neurog3. Moreover, endocrine progenitor cells that express either high or low levels of Pdx1, and thus may be biased towards the formation of specific cell lineages, exhibited cell typespecific differences in the genes regulated by Insm1. Analysis of the function of Ripply3, an Insm1-regulated gene enriched in the Pdx1high cell population, revealed that it negatively regulates the proliferation of early endocrine cells. Taken together, these findings indicate that in developing pancreatic endocrine cells Insm1 promotes the transition from a ductal progenitor to a committed endocrine cell by repressing a progenitor cell program and activating genes essential for RNA splicing, cell migration, controlled cellular proliferation, vasculogenesis, extracellular matrix and hormone secretion. KEY WORDS: Pancreas development, Endocrine progenitor cells, Gene expression, Transcription factors, Mouse

INTRODUCTION

The genetic program responsible for the generation of pancreatic endocrine cells from endocrine progenitors remains incompletely understood (Oliver-Krasinski and Stoffers, 2008; Rieck et al., 2012; Stanger and Hebrok, 2013). Beginning at around embryonic day (E) 9-9.5 in the mouse, and peaking around E15.5, pancreatic epithelial 1

Center for Stem Cell Biology, Vanderbilt University School of Medicine, Nashville, 2 TN 37232, USA. Department of Molecular Physiology and Biophysics, Vanderbilt 3 University School of Medicine, Nashville, TN 37232, USA. Department of Animal 4 Science, Cornell University, Ithaca, NY 14850, USA. Penn Center for Bioinformatics, Department of Genetics, University of Pennsylvania School of 5 Medicine, Philadelphia, PA 19104, USA. Department of Laboratory Animal Science, Kitasato University School of Medicine, Sagamihara, 252-0374, Japan. 6 Okazaki Institute for Integrative Bioscience, National Institutes of Natural Sciences, Okazaki, Aichi, 444-8787, Japan. *Author for correspondence ([email protected]) This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.

Received 10 October 2013; Accepted 26 May 2014

cells express Neurog3, which encodes a master regulatory factor that activates a cascade of secondary transcription factor genes, including Insm1, Neurod1, Pax4, Arx and Rfx6. Together, these and other transcription factors orchestrate the formation of the five endocrine cell types found in adult islets (Huang et al., 2000; Schwitzgebel et al., 2000; Gu et al., 2002; Collombat et al., 2003; Smith et al., 2003; Mellitzer et al., 2006; Soyer et al., 2010). Neurog3 triggers preendocrine cells to delaminate from a pancreatic ductal epithelium by an epithelial-to-mesenchymal transition (EMT), lose basal-apical cell polarity and epithelial adhesions, then migrate and aggregate to form islets of Langerhans (Cole et al., 2009; Gouzi et al., 2011). Moreover, Neurog3 expression coincides both with a markedly lower rate of cell proliferation and an increase in the expression of cyclin-dependent kinase inhibitors such as Cdkn1c ( p57Kip2) and Cdkn1a ( p21Cip1) (Georgia et al., 2006; Miyatsuka et al., 2011). Multiple transcription factors regulate pancreatic endocrine cell development, and they have interacting and sometimes opposing functions. For instance, Arx drives the formation of glucagonproducing α-cells. In its absence, there is a preponderance of insulinproducing β-cells and somatostatin-producing δ-cells. Similarly, Pax4 opposes the effect of Arx and is essential for the formation of βcells, since mice lacking this factor are characterized by an expansion in α-cells (Sosa-Pineda et al., 1997; Collombat et al., 2003). Moreover, nascent β-cells express higher amounts of Pdx1, a transcription factor crucial for the early specification of pancreatic epithelium, compared with other pre-endocrine cells (Ohlsson et al., 1993; Ahlgren et al., 1998; Fujitani et al., 2006; Nishimura et al., 2006; Gannon et al., 2008). Other transcription factors important for β-cell specification and development, such as Nkx2.2, Neurod1, Nkx6.1, Mafb and Mafa, also function in an interrelated manner (Sosa-Pineda et al., 1997; Sussel et al., 1998; Nishimura et al., 2006; Nelson et al., 2007; Schaffer et al., 2013). The expression of Insm1, which encodes a zinc-finger transcription factor, is activated by the binding of Neurog3 to sites within its promoter (Mellitzer et al., 2006). Endocrine cell differentiation is markedly perturbed in mice lacking Insm1 (Gierl et al., 2006). In the absence of this factor, there is a reduction in the number of insulinexpressing cells, with many cells lacking any hormone expression. In addition to being expressed in developing endocrine cells throughout the gut, Insm1 is also expressed in the developing central nervous system, where it contributes to the formation and expansion of intermediate (basal) neural progenitors from early apical progenitor cells (Farkas et al., 2008), in the peripheral neural system and in the olfactory epithelium, where it is involved in regulating the differentiation of neurogenic progenitor cells (Wildner et al., 2008; Rosenbaum et al., 2011). The acquisition of robust quantitative global gene transcription datasets, which are necessary for understanding the gene regulatory 2939

DEVELOPMENT

Anna B. Osipovich1,2, Qiaoming Long3, Elisabetta Manduchi4, Rama Gangula1, Susan B. Hipkens1, Judsen Schneider1,2, Tadashi Okubo5, Christian J. Stoeckert, Jr4, Shinji Takada6 and Mark A. Magnuson1,2, *

RESEARCH ARTICLE

network that dictates the formation and function of endocrine cells, requires the combined use of fluorescent reporter alleles, fluorescence-activated cell sorting (FACS) and next-generation sequencing technology. To this end, we have derived mice containing an Insm1GFPCre reporter allele that enabled us to isolate highly purified populations of Insm1-expressing and -deficient pre-endocrine cells, and to characterize these cell populations by RNA-Seq. In addition, we performed multi-channel FACS to isolate and differentially characterize Insm1-positive cell populations that express either high or low levels of Pdx1. In doing so, we identified many genes that are likely to contribute to both the formation and function of pancreatic endocrine cell types. In addition, the function of Ripply3 and the alternative RNA processing of Neurog3 mRNA were examined. Together, these studies provide multiple new insights into the gene regulatory network controlling pancreatic endocrine cell formation and function. RESULTS Generation of Insm1GFPCre reporter mice

A two-step strategy utilizing both gene targeting and recombinasemediated cassette exchange (RMCE) was used to derive mice that express a green fluorescent protein-Cre fusion protein (GFPCre) under control of the endogenous Insm1 gene locus (Fig. 1A; supplementary material Fig. S1A-F). Insertion of GFPCre sequences into the Insm1 gene locus disrupted Insm1 protein expression, as confirmed by western blot analysis of homozygous null embryos (supplementary material Fig. S1F). Mice heterozygous for this allele (hereafter termed Insm1+/−) appeared normal, whereas animals that were homozygous for Insm1GFPCre (hereafter termed Insm1−/−) died between E15.5 and E18.5 due to defects in catecholamine biosynthesis and secretion (Wildner et al., 2008). Green fluorescence was observed in pancreatic pre-endocrine cells (Fig. 1C) and GFPCre was readily detectable by immunofluorescence in islet cells after birth [at postnatal day (P) 1, P7 and P60; data not shown] and in the developing central nervous system between E9.5 and E18.5 (Fig. 1B). GFPCre expression was also detected in the peripheral nervous system and gut endocrine cells (data not shown).

Development (2014) 141, 2939-2949 doi:10.1242/dev.104810

Co-staining with anti-GFP and anti-Insm1 antibodies at E15.5-18.5 in Insm1+/− pancreata showed that the majority of Insm1-positive cells co-expressed both proteins (supplementary material Fig. S2). Functionality of the Cre portion of the GFPCre fusion was also confirmed (data not shown). Insm1 knockout mice have altered pancreatic hormone cell differentiation, replication, size and migration

To investigate the role of Insm1 in pancreas development we quantified the percentage of different pancreatic hormone-positive cells among GFPCre-expressing endocrine cells in Insm1 heterozygous and knockout animals at E18.5 (supplementary material Fig. S3). Consistent with the results of Gierl et al. (2006), 54% of endocrine cells expressed insulin in heterozygous animals, whereas only 8% of GFPCre-expressing cells were insulin positive (a 6-fold decrease) in Insm1−/− embryos. There were also less pronounced but significant decreases in cells expressing glucagon (from 24% to 11%), somatostatin (from 11% to 7%) and ghrelin (from 8% to 5%) in the null embryos. Also, the number of pancreatic polypeptide-positive cells increased from 7% to 12% in the knockout animals, as is also consistent with the findings of Gierl et al. (2006). Since it has been suggested that Insm1 inhibits the progression of endocrine cells through the cell cycle (Zhang et al., 2009), we quantified the proliferation rates of GFPCre-positive endocrine cells by immunostaining for Ki67, a marker of cell proliferation (Fig. 2A). At E18.5 we observed up to a 7-fold decrease (from 11.6% to 1.5%) in the number of Ki67/GFPCre double-positive cells in heterozygous versus knockout pancreata (Fig. 2B). Concurrently, the size of the Insm1−/− endocrine progenitor cells is increased (Fig. 2C,D) and their shape becomes irregular. No differences in the number of apoptotic cells were observed between heterozygous and homozygous null animals by TUNEL assay (data not shown). To assess whether the proliferation defect is visible at an earlier stage, we also quantified proliferation rates at E15.5 (supplementary material Fig. S4). Although the majority of endocrine cells are postmitotic at E15.5, we were still able to detect a slight decrease in proliferation from 1.4% in Insm1+/− cells to 0.9% in Insm1−/− cells. In addition to a reduced proliferation rate of Insm1−/− cells, we also observed that the GFPCre-positive endocrine cells remained closely associated with pancreatic ductal cells. We quantified the apparent cell migration defect by calculating the percentage of GFPCre-expressing cells that were in contact with ductal cells (DBA-positive cells) and found an increase (from 24% to 55%) in the knockout animals compared with heterozygous animals (Fig. 2E,F). This finding suggests that Insm1, in addition to its other roles, is necessary for the migration of the differentiating pre-endocrine cell away from pancreatic ductal regions.

Fig. 1. The Insm1GFPCre allele. (A) Schematic of the Insm1GFPCre allele. Insm1 coding sequences were replaced with those encoding GFPCre using combined gene targeting/recombinase-mediated cassette exchange (RMCE) as described in supplementary material Fig. S1. The triangles represent heterotypic loxP sites and the circle a remnant FLP recognition target (FRT) site. (B) Green fluorescence in a whole mouse embryo at E11.5 broadly marks the neural system. (C) Green fluorescence in a pancreas at E15.5 marks pre-endocrine cells. Fluorescence images were overlaid with images taken with white light.

2940

To quantitatively assess the gene expression changes brought about by the absence of Insm1, we performed RNA-Seq using FACS-purified progenitor cells isolated at E15.5. On average, we obtained 13,456 (±2886) cells per Insm1+/− embryo and 8505 (±1602) cells per Insm1−/− embryo at E15.5, consistent with a proliferation defect. RNA-Seq resulted in the detection of 13,865 individual mRNAs with an RPKM >0.1. Pairwise comparison of the transcriptional profiles from Insm1+/− and Insm1−/− mice revealed that the expression levels for the majority of genes were highly correlated, but with a subset of the genes being differentially expressed (Fig. 3B; supplementary material Table S1). After counting only those transcripts altered in

DEVELOPMENT

Identification of Insm1-dependent genes in developing endocrine cells

RESEARCH ARTICLE


Fig. 2. Impaired proliferation, cell size and migration of pre-endocrine cells in Insm1 knockout mice. (A) Immunofluorescence labeling of pancreatic tissues from Insm1+/− and Insm1−/− mice at E18.5 with anti-GFP antibodies (green) and antibodies against cell proliferation marker Ki67 (red). Arrows indicate cells co-expressing GFP and Ki67. (B) Percentage of Ki67-positive cells among GFP-positive cells demonstrates a proliferation defect in Insm1−/− pancreas. (C) Immunofluorescence labeling of pancreatic tissues from Insm1+/− and Insm1−/− mice at E18.5 with anti-GFP antibodies (green) and antibodies against cell surface marker E-cadherin (red) enabled the measurement of endocrine cell area. Representative cells for each phenotype are outlined. (D) Quantification of cell area shows increased cell size in Insm1−/− endocrine cells. (E) Immunofluorescence labeling of pancreatic tissues from Insm1+/− and Insm1−/− mice at E18.5 with anti-GFP antibodies (green) and the ductal marker DBA (red). (F) Quantification of the percentage of GFP-positive cells in contact with DBA-positive cells shows a defect in the migration of Insm1−/− endocrine cells away from the ducts. Error bars indicate s.e.m. (n≥6); P-values were determined by Student’s t-test. Scale bars: 50 μm.

expression of transcription factors that play important roles in the maintenance of progenitor cells (Bhlhe22, Rest, Nr5a2, Hes1, Acsl1, Foxp2, Sox6). Markers of EMT (Zeb2, Mmp14, Cldn11, Cdh2, Vim, Cdh11, Fn) and components of Notch signaling pathway (Hes1, Hes5, Hey1, Dll1, Notch2, Notch1) were also increased in the Insm1−/− progenitor cells. Similarly, we observed an upregulation in genes involved in other signaling pathways controlling cell progenitor states, such as TGF/BMP (Smad3, Bmpr1b, Tgfb2, Bmp1), FGF (Fgfr1, Fgfr4) and IGF (Igfbpl1, Igfbp5). Several genes involved in cell proliferation (Cdkn1c, Bcl2, Mertk, Anxa2, Ccnd1) were increased in Insm1−/− cells, whereas others (Cdkn1b, Manf and Akt3) were more highly expressed in the Insm1+/− cells. Insm1+/− cells also showed greater expression of genes involved in cytoskeleton reorganization and cell migration (Gng12, Pak3, Filip1), vascularization-promoting growth factors (Pgf, Vegfa) and a β-cell-specific MPK8 scaffold protein (Mapk8ip2). We performed chromatin immunoprecipitation (ChIP) analysis of the Insm1-regulated genes Rest, Cdkn1c, Cdkn1b and Ripply3, all of which contain putative Insm1 binding sites within their promoters based on the JASPAR database (Mathelier et al., 2014) (supplementary material Table S2). All genes showed enrichment that was similar or greater than that of the Insm1 promoter itself, which was previously shown to be downregulated by Insm1 (Breslin et al., 2002) (supplementary material Fig. S6). Differential expression of alternatively spliced transcripts of Neurog3

Since alternative RNA splicing plays an important role in the differentiation of embryonic stem cells (ESCs) and the specification of many cell lineages (Fu et al., 2009; Grabowski, 2011; Llorian and Smith, 2011), and our results showed a strong Insm1-dependent increase in the expression of many RNAbinding splicing factors, we examined the datasets for splice 2941

DEVELOPMENT

expression by more than 1.5-fold (with confidence values ≥70%), we identified 228 genes that were increased in expression and another 1115 genes that were decreased in expression in Insm1+/− versus Insm1−/− samples. Gene ontology analysis revealed that the majority of the more highly expressed transcripts in the Insm1+/− cells are involved in secretory cell functions, including vesicular traffic, protein processing and degradation. By contrast, the transcripts that were more highly expressed in the knockout cells were grouped into broader categories, including cytoskeleton remodeling, transcriptional regulation and components of developmental, differentiation and morphogenesis pathways (Fig. 3C; supplementary material Fig. S5). Genes most dependent on Insm1 for expression include those encoding pancreatic hormones, especially Ins1 and Ins2, as well as components of the hormone secretory machinery such as Chgb, Scg3, Snap25, Nnat and Vatl1. Consistent with our immunohistochemical findings, Ppy was also more highly expressed in the Insm1−/− cells, as was Cartpt, which encodes a neuropeptide. In the absence of Insm1 there was also a reduction in the expression of several transmembrane channels and transporters (Kctd12b, Slc2a2, Slc30a8, Cacna1a), gap junction proteins (Gjd2), a calcium-sensing protein (Hpca), several extracellular matrix proteins (Cryba2, Muc4, Spock2), and other less well characterized transmembrane proteins (Tmem215, Lrp11, Amigo2). Insm1−/− cells showed an increase in the expression of receptors previously associated with different progenitor cell types (Tspan8, Prom1, Itgam, Cd59a, Lgr4) and neural progenitor cells (Nrp2, Pmp22, Fat4, Gfra1, Lamc1). Interestingly, among the nucleic acid-binding proteins in Insm1+/− cells, there was a marked increase in the expression of genes encoding RNA-processing proteins such as Elavl4, Celf3 and Rod1, all of which have been shown to be involved in alternative RNA splicing (Chapple et al., 2007; Spellman et al., 2007; Ince-Dunn et al., 2012). Genes for DNA-binding proteins, such as Fev, St18, Ripply3, Lmo3 and Jazf1, were upregulated to a lesser extent. Conversely, Insm1−/− cells exhibited a marked increase in the

RESEARCH ARTICLE


junctions for a number of pro-endocrine genes. Interestingly, Neurog3 mRNA was found to be alternatively spliced in endocrine progenitors at E15.5, with two splice variants present – a long coding and a short non-coding, with the shorter variant having about twice as many sequence tags in Insm1−/− than in Insm1+/− cells. To validate the presence of alternatively spliced Neurog3 transcripts, we designed primers to simultaneously amplify both splicing variants or just the long or short variant (Fig. 4A), and performed semi-quantitative RT-PCR. This experiment, as well as RT-qPCR, confirmed the expression of both Neurog3 mRNA isoforms (Fig. 4B,C; supplementary material Fig. S7), with the ratio of long to short variant expression being 2-fold higher in Insm1+/− than in Insm1−/− cells. Immunostaining of E15.5 embryos also showed a decrease in the number of GFP/Neurog3 co-expressing cells (19% in Insm1 +/− compared with 12% in Insm1 −/− tissues; Fig. 4D,E), as might be expected if there was a reduction in Neurog3 protein levels. Validation of RNA-Seq results by RT-qPCR and further characterization of novel genes

To further validate and quantify the differential expression of selected genes, we performed RT-qPCR using 96-well TaqMan 2942

array cards that were designed to detect genes showing higher expression in Insm1+/− versus Insm1−/− cells as well as some known pancreatic endocrine genes as a reference (supplementary material Table S3). Cells were isolated at E15.5 and E18.5, with the later time point serving to better detect changes occurring as the endocrine progenitor cells became more differentiated. This analysis confirmed the upregulation or downregulation of all these genes at E15.5 (supplementary material Fig. S8A) and allowed us to distinguish between those required for later stages of endocrine differentiation and maturation (upregulated in Insm1+/− versus Insm1−/− samples at E18.5, e.g. Mnx1, Mafa, Ripply3 and Fev) and those active at earlier stages of endocrine specification (downregulated in Insm1+/− versus Insm1−/− samples at E18.5, e.g. Neurod1, Runx1t1, Pax6 and Pax4). The expression of some of these genes in pancreatic endocrine cells was confirmed by examining in situ RNA hybridization images in the EUREXPRESS database obtained from E14.5 mouse embryos (Diez-Roux et al., 2011) (supplementary material Fig. S8B). In addition, we performed staining with antibodies against surface proteins, such as cadherin Celsr3, adhesion molecule Amigo2, and low-density lipoprotein receptor-related protein Lrp11 (supplementary material Fig. S9), and found preferential staining in each case for pre-endocrine cells.

DEVELOPMENT

Fig. 3. RNA-Seq analysis of Insm1+/− and Insm1−/− endocrine progenitor cell populations at E15.5. (A) Representative FACS profiles of GFPCre-positive Insm1+/− and Insm1−/− endocrine cell populations isolated at E15.5 and used for RNA isolation and RNA-Seq. (B) Analysis of RNA-Seq results demonstrates differential gene expression between Insm1+/− and Insm1−/− cells. Correlation plot of RPKM (reads per kilobase per million mapped reads, log10) values for identified genes in Insm1+/− and Insm1−/− cell populations shows that although the majority of reads fall along the line representing equal expression, a number of transcripts are upregulated in Insm1+/− or Insm1−/− cells. (C) Functional groupings of genes upregulated in Insm1+/− and Insm1−/− cell populations. Enrichment scores for functional gene ontology groups represent the geometric mean (in −log scale) of P-values in the corresponding annotation cluster.

RESEARCH ARTICLE


Temporal expression profiling of endocrine populations expressing high and low levels of Pdx1

Beginning at ∼E14.5 during pancreas development, Pdx1 is expressed at higher levels in β-cells than in non-β-cells (Ohlsson et al., 1993). Thus, to identify Insm1-regulated genes that might be specifically important for β-cell development and function we used a Pdx1CFP allele (Potter et al., 2011) to isolate populations of Insm1-positive endocrine cells that express either high (HI) or low (LO) levels of Pdx1 at time points from E15.5 to E18.5. Confocal imaging of direct fluorescence confirmed that Insm1GFPCre/+;Pdx1CFP/+ mice expressed both GFP and CFP in endocrine cells, with different levels of CFP expression marking endocrine subpopulations (Fig. 5A), which we were able to separate by FACS (Fig. 5B). This allowed us to assess changes in the relative proportions of different cell types during pancreas development over a 4 day interval (Fig. 5C). Insm1/Pdx1 double-positive endocrine cells comprised ∼20-30% of total sorted cells, with the greatest increase in cell numbers occurring between E16.5 and E17.5. As expected, the proportion of Pdx1-HI cells in the double-positive cell populations increased between E15.5 and E18.5, mirroring the increase in β-cells during pancreas development. To explore the differences between the Insm1/Pdx1-HI and Insm1/Pdx1-LO pre-endocrine cell populations as a function of time, we performed both RT-qPCR and RNA-Seq analysis. Pairwise comparison of RNA-Seq results between these cell populations at E15.5 and E18.5 revealed an increase in the expression of a number of genes important for β-cell specification

(supplementary material Tables S4 and S5). Gene ontology analysis showed that most of the Insm1/Pdx1-HI enriched genes are involved in hormone secretion as well as ion transport and glucose metabolism (supplementary material Fig. S10A,B). Ninety-three genes were commonly upregulated in Insm1/Pdx1-HI versus Insm1/ Pdx1-LO cells at E15.5 and E18.5. Correlation analysis showed that many of these genes have similar differential expression levels and many are known to be important for β-cell function (supplementary material Fig. S10C). RT-qPCR analysis of pre-endocrine cell populations from E15.5 to E18.5 allowed hierarchical clustering based on relative levels of expression of the genes in the Insm1/Pdx1-HI versus Insm1/Pdx1LO cells (supplementary material Fig. S11). Genes expressed at higher levels in the Insm1/Pdx1-LO cells included the α-cell transcription factor Arx, the hormones Glc and Ghrl, as well as many transcription factors known to be involved in early endocrine specification, such as Neurod1, Neurog3, Isl1 and Pax4. These results confirmed our prediction that the Insm1/Pdx1-LO population consists mostly of non-β-cells and non-committed pre-endocrine cells. Interestingly, somatostatin and Ppy were expressed at a higher level in the Insm1/Pdx1-HI population, especially at E15.5. Other genes that were more highly expressed in the Insm1/Pdx1-HI cells included known β-cell genes and transcription factors such as Ins2, Glut2, Mafa, Mnx1 and Nkx6.1. This population also contained the transcriptional regulators Mafg and Ripply3 and the signaling molecule Igfbp7. 2943

DEVELOPMENT

Fig. 4. Neurog3 is alternatively spliced in endocrine progenitor cells. (A) Neurog3 gene and mRNA splicing isoforms. Primer pair N1+N3 is designed to amplify both isoforms, N1+N2 the long coding isoform, and N1+N4 the short non-coding isoform. (B) Semi-quantitative PCR analysis for Neurog3 mRNA isoforms. The proportion of the short non-coding isoform is higher in Insm1−/− than in Insm1+/− cells. (C) Quantification of relative expression of the Neurog3 long versus short isoforms by real-time PCR. (D) Immunofluorescence labeling of pancreatic tissues from Insm1+/− and Insm1−/− mice at E15.5 with anti-GFP antibodies (green) and anti-Neurog3 antibodies (red). (E) The percentage of Neurog3-positive cells among GFP-positive cells is decreased in Insm1−/− pancreas, implying changes in Neurog3 expression on a protein level. Error bars indicate s.e.m. (n≥6); P-values were determined by Student’s t-test. Scale bar: 50 μm.

RESEARCH ARTICLE


β-cell and α-cell areas were increased (Fig. 6C,D; supplementary material Fig. S12C). In addition, there was a notable (although not statistically significant) increase in somatostatin-positive and Ppypositive cell area (data not shown). Co-staining of insulinexpressing and glucagon-expressing cells for Ki67 indicated that the proliferation of β-cells was increased in the Ripply3 knockout pancreata (Fig. 6E,F). The increase in proliferation was not accompanied by the loss of differentiated β-cell phenotype since knockout tissues expressed levels of the maturation markers Glut2, Mafa, Mafa and Nkx6.1 similar to those of the heterozygote (supplementary material Fig. S12D). TUNEL assay showed no increase in apoptosis of β-cells in Ripply3 null animals at E18.5 (data not shown).

Fig. 5. Isolation of Insm1-positive endocrine cell populations expressing high (Insm1/Pdx1-HI) and low (Insm1/Pdx1-LO) levels of Pdx1. (A) Confocal image of direct fluorescence in a pancreas from animals that were doubly heterozygous for both the Insm1GFPCre and Pdx1CFP alleles. GFP and CFP expression marks pre-endocrine cells, with different levels of CFP marking heterogeneous endocrine populations expressing Pdx1 at high and low levels. Based on the known expression patterns of Insm1 and Pdx1 (Guz et al., 1995; Gu et al., 2002; Mellitzer et al., 2006), the perdurance of CFP for 46 h, and immunohistochemical staining of pancreatic tissues with anti-Pdx1 and anti-hormone antibodies (data not shown), these populations are thought to represent the following cells: Pdx1-LO, predominantly acinar cells with a small proportion of ductal cells; Insm1/Pdx1-HI, endocrine cells including mostly β-cells but also with some δ-cells; Insm1/Pdx1-LO, endocrine cells including uncommitted endocrine cells and other non-β-cells; and Insm1positive neural cells of non-pancreatic origin. (B) FACS profile of pancreatic populations sorted from Insm1GFPCre/+;Pdx1CFP/+ animals. Four distinct populations are sorted: GFP positive, CFP positive and two GFP/CFP doublepositive populations expressing high (Pdx1-HI) and low (Pdx1-LO) levels of CFP. (C) Temporal changes in the relative proportions of the different sorted pancreatic cell populations. Scale bar: 50 μm.

Ripply3 is expressed in pancreatic β-cells and contributes to their proliferation

Since the function of Ripply3, a gene we found to be both upregulated by Insm1 and enriched in Insm1/Pdx1-HI cell populations, had not been previously explored in pancreas, we analyzed mouse pancreatic tissues from Ripply3β-Gal mice, which express β-galactosidase instead of Ripply3 (Okubo et al., 2011). Co-staining for β-galactosidase, chromogranin A and islet hormones confirmed that Ripply3 is expressed in nascent endocrine cells, with higher levels detected in β-cells at E18.5 (Fig. 6A; supplementary material Fig. S12A). RT-qPCR on embryonic and adult tissues showed that Ripply3 is strongly expressed in heart and lung, and to a lesser degree in liver and kidney. In the adult pancreas the expression of Ripply3 was highest in islets (Fig. 6B). Based on RNA-Seq profiling of distinct pancreatic developmental populations, Ripply3 is expressed in early endoderm at E8.0, with lower levels detected in pancreatic endoderm and MPCs, and increased expression in Neurog3-positive endocrine progenitor cells at E15.5 and further in nascent β-cells at E16.5 (supplementary material Fig. S12B). Although pancreatic endocrine cell types were present at E18.5 in Rippy3β-Gal heterozygous and knockout mice, both 2944

In agreement with previously published data (Gierl et al., 2006), mice without Insm1 have a reduced number of endocrine cells and hormone expression in these cells is either lacking or imbalanced in terms of specific endocrine cell types. Given that Insm1 has long been known to contain a SNAG repressor domain (Chiang and Ayyanathan, 2013), our finding that approximately twice as many genes in Insm1−/− than in Insm1+/− mice were increased in expression is consistent with the major role of Insm1 being as a repressor. In the absence of Insm1, pancreatic pre-endocrine cells display many features of earlier stage progenitor cells, including an increase in the expression of transcription factors and cell surface receptors associated with the progenitor state, EMT markers, and components of morphogenic signaling pathways (e.g. Notch, Tgf/ BMP and Wnt). Although EMT markers increased in Insm1 null cells, they failed to migrate away from the ducts. This might be due to incomplete EMT, since some the genes upregulated in Insm1+/− cells are involved in actin remodeling and cell migration. Moreover, signaling molecules and extracellular matrix proteins that are expressed upon Insm1-dependent endocrine differentiation might be necessary for the migration and aggregation of newly differentiated endocrine cells into islets. Insm1 null cells also have decreased rates of proliferation that starts at E15.5 and is well pronounced at E18.5. This decrease was not accompanied by an increase in apoptosis, and the knockout cells appeared larger in size, a characteristic that often accompanies cell cycle arrest at G1/G2 phase (Bjorklund et al., 2006). Consistent with cell cycle arrest without apoptosis, Insm1 null cells express more of the cell cycle inhibitor Cdkn1c and the pro-survival gene Bcl2. Interestingly, Insm1-expressing cells make more of the cell cycle inhibitor Cdkn1b ( p27Kip1), which is involved in controlling the proliferation of newly differentiated β-cells (Georgia and Bhushan, 2006). This suggests that Insm1 promotes a switch in the mode of cell cycle regulation from being Cdkn1c based to being Cdkn1b regulated, with both genes being directly regulated by Insm1. A similar developmental switch in cell cycle regulators has been observed during pituitary development, in which noncycling precursors express Cdkn1c and more differentiated cells express Cdkn1b (Bilodeau et al., 2009). Although the differentiation of endocrine cells in the presence of Insm1 is associated with an increase in proliferation, cell proliferation itself is under the control of multiple genes, including cell cycle inhibitors such as Cdkn1b and Ripply3. The complex regulation of endocrine progenitor cell proliferation is not surprising given that overexpression of Insm1 in the developing neural cortex has been shown to decrease the proliferation of less differentiated apical progenitor cells but to increase the proliferation of more differentiated intermediate basal progenitor cells (Farkas et al., 2008).

DEVELOPMENT

DISCUSSION Role of Insm1 in pancreatic pre-endocrine cell differentiation

RESEARCH ARTICLE


Fig. 6. Ripply3 is involved in the regulation of endocrine cell replication. (A) Ripply3 is expressed in pancreatic endocrine cells at E18.5. Immunofluorescence staining of pancreata from Ripply3β-Gal heterozygous mice with anti-β-galactosidase and antibodies against the pancreatic hormones insulin (Ins), glucagon (Gcg), pancreatic polypeptide (Ppy) and somatostatin (Sst). (B) Ripply3 is expressed in adult pancreatic islets. RT-qPCR analysis of Ripply3 relative expression in different tissues and during embryonic development (normalized to expression in ESCs). (C) Immunofluorescence staining for insulin and glucagon for representative sections of Ripply3β-Gal heterozygous and homozygous mice. The hormone-positive endocrine area appeared to be expanded in Ripply3 knockout animals. (D) Quantification of total insulin-positive and glucagon-positive areas in Ripply3β-Gal heterozygous and knockout animals. Both areas are significantly increased in knockout animals. (E) Immunofluorescence staining for insulin, glucagon and Ki67 for representative sections of Ripply3β-Gal heterozygous and homozygous mice. (F) The percentage of Ki67-positive cells among total insulin-positive or glucagon-positive cells shows an increase in the proliferation of endocrine cells in Ripply3 knockout animals. Error bars indicate s.e.m. (n≥6); P-values were determined by Student’s t-test. Scale bars: 50 μm.

Identification of novel endocrine surface markers

This study revealed a number of surface proteins that mark early endocrine cells. For example, mRNAs for Amigo2, Celsr3 and Lrp11 were strongly upregulated in Insm1-positive cells, and immunostaining for the encoded proteins confirmed their localization along the membrane of endocrine cells. Amigo2 showed the strongest endocrinespecific staining pattern and may function in cell adhesion and migration (Rabenau et al., 2004). Celsr3, the cadherin associated with planar polarity, displayed localized membrane staining and was recently shown to promote endocrine cell differentiation (Tissir and Goffinet, 2006; Cortijo et al., 2012). Another early endocrine cell marker validated by our study is Lrp11, as was previously suggested by Hald et al. (2012). Interestingly, although Lrp11 is present in all endocrine cells, it seems to be most highly expressed in δ-cells. Neurog3 is alternatively spliced in endocrine progenitor cells

RNA-Seq revealed that ∼30% of the differentially expressed genes had two or more differentially expressed RNA variants (data not shown). This suggests that the Insm1-dependent differentiation program might also function at a post-transcriptional level by promoting endocrine lineage-specific alternative splicing. Although the mechanisms involved in this regulation are unclear, this could occur through the activation of genes encoding RNA-binding 2945

DEVELOPMENT

This study also brings to light several genes that might play important but previously unrecognized roles in the formation of islet cells. Indeed, although a key function of Insm1 might be to repress genes (e.g. Rest) that maintain pancreatic ductal epithelial cells in a progenitor-like state, which would be expected to promote endocrine cell differentiation, many genes were also found to be upregulated in Insm1-expressing cells. These include genes necessary for hormone secretion, membrane transporters and channels, extracellular matrix proteins, RNA-processing proteins and cell migration proteins. For example, Insm1-expressing progenitor cells are characterized by higher levels of transcription factors that are found in differentiating β-cells (Mafa, Mafb, Mnx1, Vdr, Hmgn3, Fev, Lmx1b), as well as novel genes that have not been studied in the pancreas (Lmo3, Ripply3, Jazf1, Nhlh1). Among the Myt1 family, St18, Myt1 and Myt1l1 were all upregulated, suggesting that all three of these genes are expressed in differentiating endocrine cells, contrary to a previous report (Gu et al., 2004). Considering that many of the Insm1-regulated transcription factors have also been shown to play key roles during neural development, and that neural and islet cells share many phenotypic properties, these genes probably also contribute to endocrine development. Interestingly, one such gene, the transcription factor Jazf1, is associated with the development of type 2 diabetes (Zeggini et al., 2008).

RESEARCH ARTICLE


proteins, such as Elavl4, Celf3 and Rod1. All of these transcripts were strongly upregulated in Insm1-positve cells and so could be important in regulating splicing changes during endocrine differentiation. In particular, we found that Neurog3 mRNA is alternatively spliced in endocrine progenitor cells, with a long coding and short non-coding isoform present. The latter is formed through splicing to the alternative splice acceptor site located in the second exon of Neurog3 (supplementary material Fig. S7). Interestingly, the ratio of the Neurog3 coding to non-coding isoform was higher in Insm1-positive cells, suggesting that alternative splicing of the Neurog3 pre-mRNA in an Insm1-dependent manner provides a positive-feedback mechanism for maintaining Neurog3 expression during development. Transcriptional profiling of subpopulations of pre-endocrine cells

Ripply3 as a downstream modulator of endocrine cell proliferation

In this study we discovered that Ripply3 is also expressed in endocrine progenitor cell populations and that Insm1 binds to the Ripply3 promoter. Ripply3 belongs to a family of related proteins, the other two members of which, Ripply1 and Ripply2, are necessary for the proper formation of somite boundaries and function by antagonizing T-box transcription factor activity during development (Kawamura et al., 2005; Hitachi et al., 2009; Takahashi et al., 2010). Ripply3 has also been shown to regulate pharyngeal apparatus development through suppression of Tbx1 activity (Okubo et al., 2011). The importance of other Ripply gene family members prompted us to explore the role of Ripply3 in pancreas development. Although our analysis did not reveal any defects in endocrine cell differentiation in Ripply3 null pancreatic tissues, we observed that both β-cell and α-cell areas and proliferation were increased. The increase in proliferation in Ripply3 null tissues suggests that, similar to Cdkn1b, Ripply3 might control the proliferation of newly committed β-cells in order to allow them to properly differentiate. 2946

Fig. 7. Model for the role of Insm1 in endocrine progenitor cells. (A) The Insm1 transcriptional program regulates a progression of pre-endocrine cells from Neurog3-positive endocrine progenitors residing in a duct to islet-bound committed endocrine cells. (B) Endocrine progenitor cells are characterized by Notch signaling and other stemness-controlling pathways. As endocrine cell differentiation occurs, the proliferation of these cells slows as they exit the cell cycle and delaminate from the duct through the activation of the EMT program. Insm1 promotes these changes by repressing progenitor cell functions while simultaneously activating genes involved in hormone secretion, formation of extracellular matrix, cell migration, and promotion of alternative splicing and vascularization. The Insm1-dependent endocrine differentiation program is accompanied by an increase in the coding isoform of Neurog3 and expression of Ripply3, a gene we have shown to regulate β-cell proliferation.

The mechanism by which Ripply3 functions in pre-endocrine cells is unclear. As mentioned above, members of this class of proteins have been shown to bind and suppress the activity of T-box transcription factors. Previous reports (Begum and Papaioannou, 2011; Oropez and Horb, 2012) have shown that Tbx2 is expressed in pre-endocrine cells, and our current findings indicate that it is upregulated in the absence of Insm1. Thus, given that Tbx2 was previously shown to promote cell proliferation (Vance et al., 2005), Ripply3 might function by antagonizing Tbx2 activity in differentiating endocrine cells. In any case, the functional interactions between Ripply3 and Tbx2, and how each gene might contribute to the growing network of genes that promote and/or regulate the proliferation of pancreatic endocrine progenitor cells, merit additional study. Conclusions

These studies provide valuable new insights into the role that Insm1 plays in developing pancreatic endocrine cells (as summarized in the model shown in Fig. 7). We expect that the RNA-Seq datasets, besides identifying many genes that are affected in expression by Insm1, will serve as a platform for further discoveries, especially as more comprehensive models of the gene regulatory network in endocrine cells are built. MATERIALS AND METHODS Antibodies and primers

Antibodies employed in immunohistochemistry and ChIP and primers used for genotyping, qPCR and ChIP analysis are described in the Methods in the supplementary material.

DEVELOPMENT

Combinatorial FACS using two or more fluorescent reporter alleles is a powerful means of isolating specific progenitor cell subpopulations. For this reason, we also made use of a recently derived Pdx1CFP fluorescent reporter allele to explore the differences in gene expression between developing endocrine progenitor cells that express either high or low levels of Pdx1. It is well known that Pdx1 marks all endocrine cells (Herrera, 2000; Gu et al., 2002) and that developing β-cells, as well as a small fraction of δ-cells, express high levels of Pdx1 (Leonard et al., 1993; Ohlsson et al., 1993; Guz et al., 1995). As expected, Insm1/Pdx1-HI cells express more β-cell-specific genes, such as Ins2, Slc2a2 and Mafa. However, we were surprised to find that this cell population (at E15.5) also contains many somatostatin and Ppy transcripts. This suggests that the Insm1/Pdx1-HI cell population might contain a common β/δ-progenitor, as well as newly formed δ-cells, the existence of which has been predicted by Pax4 and Arx knockout studies (Collombat et al., 2003, 2005). Our data are also consistent with previous lineage-tracing studies that have shown that a Ppy-Cre transgene marks both β-cells and Ppy-producing cells (PP-cells) (Herrera et al., 1994; Herrera, 2000), a finding that suggests that PPcells and a common β/δ-cell might share common a progenitor, or that Ppy is expressed in the common β/δ-cell progenitor. Analysis of differential expression between Insm1/Pdx1-HI and Insm1/Pdx1LO endocrine subpopulations allowed us to define Insm1-dependent genes that are enriched in Pdx1-HI populations and therefore potentially important for specific functions of β-cells and/or δ-cells.

RESEARCH ARTICLE

The Insm1GFPCre allele was made using a sequential gene targeting and RMCE strategy using methods and vectors that have been described previously (Chen et al., 2011). Insm1GFPCre and Pdx1CFP mice (Potter et al., 2011) were maintained on a CD1 outbred background (90% or higher) to extend the survival of the embryos, as previously described (Gierl et al., 2006). Approximately 10% of the Insm1 homozygous null pups survived to E18.5, consistent with the results of Gierl et al. (Gierl et al., 2006), with all other embryos dying between E15.5 and E18.5. Ripply3β-Gal mice were of mixed C57BL6/129SV background. All animal procedures and husbandry protocols were approved by the Vanderbilt University Institutional Animal Care and Use Committee. Embryo isolation and imaging, tissue dissociation and FACS

Timed matings were carried out by designating noon on the day vaginal plugs were identified as E0.5. Embryos at different stages were obtained by dissection and imaged by direct fluorescence visualization using a Leica MZ 16 FA stereoscope equipped with a QImaging RETIGA 4000R camera. For FACS, embryos were isolated at E15.5-18.5 and initially genotyped by direct fluorescence. Pancreata were dissected in cold PBS and dissociated with Accumax solution (Sigma-Aldrich) containing 5 μg/ml DNase I (SigmaAldrich) at 37°C for 1 h with periodic resuspension. Quenching solution [L15 medium, 10 mM Hepes (pH 7.0), 1 mg/ml bovine serum albumin (Life Technologies), 5 μg/ml DNase I] was added to stop the reaction. The cell suspension was filtered through a 35 μm filter into FACS tubes, washed with quenching solution and resuspended in 0.5 ml quenching solution containing 7AAD viability dye (1:1000; Life Technologies). Fluorescent cells were sorted directly into Trizol LS reagent (Life Technologies) containing 40 μg/ml mussel glycogen (Sigma-Aldrich) using a FACSAria I cell sorter (BD Biosciences). RNA isolation, library construction and RNA-Seq

Three independent RNA isolates from each genotype were used for cDNA amplification and sequencing. Cells from multiple embryos were pooled to obtain a sufficient amount of RNA for analysis. The methods used to isolate RNA, validate integrity and purity, amplify cDNA and make libraries have been described previously (Choi et al., 2012). Single-end sequencing (110 bp) was performed on an Illumina HiSeq2000 genome analyzer. Read alignment to the mouse genome (mm9) was performed using RNA-Seq Unified Mapper (RUM) (Grant et al., 2011). Expression level was quantified as reads per kilobase of gene model per million mapped reads (RPKM). Genome alignment of sequencing data yielded total mapped reads in a range from 30 to 47 million per sample, from which 66-70% of reads were uniquely mapped. RPKM values of the different samples were highly correlated between replicates (R2=0.99). Differential expression analysis was performed using Patterns of Gene Expression (PaGE) software (Grant et al., 2005). Summaries of the differential expression analyses are provided in supplementary material Tables S1, S3 and S4. Gene ontology analysis was performed using DAVID Bioinformatics Resources v.6.7 (Huang et al., 2009). RNA-Seq data are available at http://genomics.betacell.org (study numbers 4294 and 4674) and are also deposited at ArrayExpress (accession number E-MTAB-1929). RT-qPCR analysis

Ninety-six gene custom TaqMan Low-Density Array (TLDA; Life Technologies) cards were used for validation of gene expression. The TaqMan probe and primer sets were selected from predesigned TaqMan gene expression assays (Life Technologies). Three independent RNA isolates for each genotype or developmental time point were assayed. cDNA was prepared from 30 μg total RNA using a high-capacity cDNA Archive Kit (Life Technologies) and divided between two ports of the TLDA cards. PCR was performed with an ABI 7900HT real-time PCR system (Life Technologies) and amplification data were analyzed using Sequence Detection System version 2.1 (Life Technologies) and Excel software (Microsoft). 18S ribosomal RNA and Gapdh were used as endogenous housekeeping controls for normalization, and the comparative Ct method was used to calculate relative fold expression by 2−ΔΔCt (supplementary material Table S2). Hierarchical clustering analysis of the data was performed using GenePattern software (Reich et al., 2006). The clustering

was performed on relative expression levels (log10) using Euclidean distance and pairwise average linkage. The Hierarchical Clustering Viewer from GenePattern was used for graphical representation, with coloring relative for each gene row. For analysis of Neurog3 alternative splicing variants, 1 ng amplified cDNA (used for RNA-Seq) was amplified using either real-time PCR with Power SYBR Green PCR Master Mix (Life Technologies) or semi-quantitative PCR using Phusion Taq (New England Biolabs). Immunohistochemistry

Immunofluorescence staining of frozen sections was essentially as previously described (Burlison et al., 2008). Ductal cells were stained with biotinylated Dolichos biflorus agglutinin (DBA; Vector Laboratories) followed by streptavidin-Cy5 conjugate (Vector Laboratories). Coverslips were mounted using ProLong Gold antifade reagent with DAPI (Life Technologies). Fluorescence images were acquired using an Axioplan2 microscope (Zeiss) with 20× and 40× objectives and QImaging RETIGA EXi camera. Confocal microscopy images were acquired using an LSM 510 Meta microscope and imaging software (Zeiss). Images taken with different color channels were merged using Photoshop (Adobe Systems). Morphometric analysis

The fractions of GFP-positive or β-galactosidase-positive endocrine cells co-expressing other proteins were calculated by manual cell counting on step sections (50 μm apart) of pancreatic tissues from at least six different animals of the same age and genotype. At least 500 cells were counted for each phenotype in each experiment. The percentage of GFP-positive cells adjacent to ducts was calculated based on the number of these cells in contact with DBA-positive cells. The cell area of GFP-positive endocrine cells was measured using ImageJ (NIH) software to outline E-cadherinpositive cell borders using a total of 1000 cells from three different animals of each phenotype. The insulin-positive and glucagon-positive areas of Ripply3 heterozygous and mutant animals were measured using every fifth section from three different animals of the same phenotype. ChIP

Pancreata were dissected from E15.5 wild-type embryos, cross-linked with 1% formaldehyde in PBS for 8 min, quenched with 125 mM glycine and washed with PBS. Tissues were lysed in 100 μl lysis buffer [50 mM TrisHCl ( pH 8.0), 10 mM EDTA, 1% SDS, 1× protease inhibitor mix (Sigma), 1 mM PMSF], sonicated (Bioraptor, Diagenode) and lysates diluted to 1 ml with RIPA buffer [10 mM Tris-HCl ( pH 7.5), 140 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 1% Triton X-100, 0.1% (w/v) SDS, 1× protease inhibitor mix, PMSF]. 20 μg sonicated chromatin was incubated with 8 μg antibody coupled to IgG magnetic beads (Cell Signaling Technology) overnight at 4°C. The beads were washed five times with RIPA buffer and once with TE buffer [10 mM Tris ( pH 8.0), 1 mM EDTA]. After washing, bound DNA was eluted at 65°C in elution buffer [20 mM Tris-HCl ( pH 7.5), 5 mM EDTA, 50 mM NaCl, 1% SDS, 50 μg/ml proteinase K] for 2 h. After crosslinking reversal, the immunoprecipitated DNA was purified by phenolchloroform extraction and ethanol precipitation and resuspended in 30 μl TE buffer. Enrichment at target promoters was determined by real-time PCR with Power SYBR Green PCR Master Mix. Relative fold enrichment at different sites was calculated by the 2−ΔΔCT method from five different ChIP experiments. Acknowledgements We thank Guoqiang Gu, Jennifer Stancill and Hannah Worchel for critically reviewing the manuscript; David K. Flaherty and Brittany Matlock for providing FACS expertise and technical assistance; Jennifer Skelton for help with deriving the mice; and Travis Clark for assistance with the next-generation sequencing.

Competing interests The authors declare no competing financial interests.

Author contributions A.B.O., Q.L., J.S., C.J.S. and M.A.M. were involved in the conception and design of the experiments and co-wrote the manuscript. A.B.O. performed most of the experiments and analyzed the data; Q.L. designed the gene targeting vector; R.G. performed embryo dissections and stereoscopy; S.B.H. performed the gene

2947

DEVELOPMENT

Mouse lines


targeting and RMCE procedures; E.M. and C.J.S. performed initial RNA-Seq data analysis; T.O. and S.T. provided tissues from Ripply3 knockout mice. All authors were provided with the opportunity to comment on the manuscript.

Funding These studies were supported by National Institutes of Health (NIH) grants [DK72473 and DK89523] to M.A.M. Use of shared resources at Vanderbilt University School of Medicine was supported by NIH grants [CA68485, DK58404 and DK20593]. Deposited in PMC for immediate release.

Supplementary material Supplementary material available online at http://dev.biologists.org/lookup/suppl/doi:10.1242/dev.104810/-/DC1

References Ahlgren, U., Jonsson, J., Jonsson, L., Simu, K. and Edlund, H. (1998). beta-cellspecific inactivation of the mouse Ipf1/Pdx1 gene results in loss of the beta-cell phenotype and maturity onset diabetes. Genes Dev. 12, 1763-1768. Begum, S. and Papaioannou, V. E. (2011). Dynamic expression of Tbx2 and Tbx3 in developing mouse pancreas. Gene Expr. Patterns 11, 476-483. Bilodeau, S., Roussel-Gervais, A. and Drouin, J. (2009). Distinct developmental roles of cell cycle inhibitors p57Kip2 and p27Kip1 distinguish pituitary progenitor cell cycle exit from cell cycle reentry of differentiated cells. Mol. Cell. Biol. 29, 1895-1908. Bjorklund, M., Taipale, M., Varjosalo, M., Saharinen, J., Lahdenpera, J. and Taipale, J. (2006). Identification of pathways regulating cell size and cell-cycle progression by RNAi. Nature 439, 1009-1013. Breslin, M. B., Zhu, M., Notkins, A. L. and Lan, M. S. (2002). Neuroendocrine differentiation factor, IA-1, is a transcriptional repressor and contains a specific DNA-binding domain: identification of consensus IA-1 binding sequence. Nucleic Acids Res. 30, 1038-1045. Burlison, J. S., Long, Q., Fujitani, Y., Wright, C. V. E. and Magnuson, M. A. (2008). Pdx-1 and Ptf1a concurrently determine fate specification of pancreatic multipotent progenitor cells. Dev. Biol. 316, 74-86. Chapple, J. P., Anthony, K., Martin, T. R., Dev, A., Cooper, T. A. and Gallo, J.-M. (2007). Expression, localization and tau exon 10 splicing activity of the brain RNAbinding protein TNRC4. Hum. Mol. Genet. 16, 2760-2769. Chen, S. X., Osipovich, A. B., Ustione, A., Potter, L. A., Hipkens, S., Gangula, R., Yuan, W., Piston, D. W. and Magnuson, M. A. (2011). Quantification of factors influencing fluorescent protein expression using RMCE to generate an allelic series in the ROSA26 locus in mice. Dis. Model. Mech. 4, 537-547. Chiang, C. and Ayyanathan, K. (2013). Snail/Gfi-1 (SNAG) family zinc finger proteins in transcription regulation, chromatin dynamics, cell signaling, development, and disease. Cytokine Growth Factor Rev. 24, 123-131. Choi, E., Kraus, M. R.-C., Lemaire, L. A., Yoshimoto, M., Vemula, S., Potter, L. A., Manduchi, E., Stoeckert, C. J., Jr, Grapin-Botton, A. and Magnuson, M. A. (2012). Dual lineage-specific expression of Sox17 during mouse embryogenesis. Stem Cells 30, 2297-2308. Cole, L., Anderson, M., Antin, P. B. and Limesand, S. W. (2009). One process for pancreatic β-cell coalescence into islets involves an epithelial-mesenchymal transition. J. Endocrinol. 203, 19-31. Collombat, P., Mansouri, A., Hecksher-Sørensen, J., Serup, P., Krull, J., Gradwohl, G. and Gruss, P. (2003). Opposing actions of Arx and Pax4 in endocrine pancreas development. Genes Dev. 17, 2591-2603. Collombat, P., Hecksher-Sorensen, J., Broccoli, V., Krull, J., Ponte, I., Mundiger, T., Smith, J., Gruss, P., Serup, P. and Mansouri, A. (2005). The simultaneous loss of Arx and Pax4 genes promotes a somatostatin-producing cell fate specification at the expense of the alpha- and beta-cell lineages in the mouse endocrine pancreas. Development 132, 2969-2980. Cortijo, C., Gouzi, M., Tissir, F. and Grapin-Botton, A. (2012). Planar cell polarity controls pancreatic beta cell differentiation and glucose homeostasis. Cell Rep. 2, 1593-1606. Diez-Roux, G., Banfi, S., Sultan, M., Geffers, L., Anand, S., Rozado, D., Magen, A., Canidio, E., Pagani, M., Peluso, I. et al. (2011). A high-resolution anatomical atlas of the transcriptome in the mouse embryo. PLoS Biol. 9, e1000582. Farkas, L. M., Haffner, C., Giger, T., Khaitovich, P., Nowick, K., Birchmeier, C., Pä ä bo, S. and Huttner, W. B. (2008). Insulinoma-associated 1 has a panneurogenic role and promotes the generation and expansion of basal progenitors in the developing mouse neocortex. Neuron 60, 40-55. Fu, R.-H., Liu, S.-P., Ou, C.-W., Yu, H.-H., Li, K.-W., Tsai, C.-H., Shyu, W.-C. and Lin, S.-Z. (2009). Alternative splicing modulates stem cell differentiation. Cell Transplant. 18, 1029-1038. Fujitani, Y., Fujitani, S., Boyer, D. F., Gannon, M., Kawaguchi, Y., Ray, M., Shiota, M., Stein, R. W., Magnuson, M. A. and Wright, C. V. E. (2006). Targeted deletion of a cis-regulatory region reveals differential gene dosage requirements for Pdx1 in foregut organ differentiation and pancreas formation. Genes Dev. 20, 253-266.

2948


Gannon, M., Ables, E. T., Crawford, L., Lowe, D., Offield, M. F., Magnuson, M. A. and Wright, C. V. E. (2008). pdx-1 function is specifically required in embryonic β cells to generate appropriate numbers of endocrine cell types and maintain glucose homeostasis. Dev. Biol. 314, 406-417. Georgia, S. and Bhushan, A. (2006). p27 Regulates the transition of β-cells from quiescence to proliferation. Diabetes 55, 2950-2956. Georgia, S., Soliz, R., Li, M., Zhang, P. and Bhushan, A. (2006). p57 and Hes1 coordinate cell cycle exit with self-renewal of pancreatic progenitors. Dev. Biol. 298, 22-31. Gierl, M. S., Karoulias, N., Wende, H., Strehle, M. and Birchmeier, C. (2006). The zinc-finger factor Insm1 (IA-1) is essential for the development of pancreatic beta cells and intestinal endocrine cells. Genes Dev. 20, 2465-2478. Gouzi, M., Kim, Y. H., Katsumoto, K., Johansson, K. and Grapin-Botton, A. (2011). Neurogenin3 initiates stepwise delamination of differentiating endocrine cells during pancreas development. Dev. Dyn. 240, 589-604. Grabowski, P. (2011). Alternative splicing takes shape during neuronal development. Curr. Opin. Genet. Dev. 21, 388-394. Grant, G. R., Liu, J. and Stoeckert, C. J., Jr. (2005). A practical false discovery rate approach to identifying patterns of differential expression in microarray data. Bioinformatics 21, 2684-2690. Grant, G. R., Farkas, M. H., Pizarro, A. D., Lahens, N. F., Schug, J., Brunk, B. P., Stoeckert, C. J., Hogenesch, J. B. and Pierce, E. A. (2011). Comparative analysis of RNA-Seq alignment algorithms and the RNA-Seq unified mapper (RUM). Bioinformatics 27, 2518-2528. Gu, G., Dubauskaite, J. and Melton, D. A. (2002). Direct evidence for the pancreatic lineage: NGN3+ cells are islet progenitors and are distinct from duct progenitors. Development 129, 2447-2457. Gu, G., Wells, J. M., Dombkowski, D., Preffer, F., Aronow, B. and Melton, D. A. (2004). Global expression analysis of gene regulatory pathways during endocrine pancreatic development. Development 131, 165-179. Guz, Y., Montminy, M. R., Stein, R., Leonard, J., Gamer, L. W., Wright, C. V. and Teitelman, G. (1995). Expression of murine STF-1, a putative insulin gene transcription factor, in beta cells of pancreas, duodenal epithelium and pancreatic exocrine and endocrine progenitors during ontogeny. Development 121, 11-18. Hald, J., Galbo, T., Rescan, C., Radzikowski, L., Sprinkel, A. E., Heimberg, H., Ahnfelt-Rønne, J., Jensen, J., Scharfmann, R., Gradwohl, G. et al. (2012). Pancreatic islet and progenitor cell surface markers with cell sorting potential. Diabetologia 55, 154-165. Herrera, P. L. (2000). Adult insulin- and glucagon-producing cells differentiate from two independent cell lineages. Development 127, 2317-2322. Herrera, P. L., Huarte, J., Zufferey, R., Nichols, A., Mermillod, B., Philippe, J., Muniesa, P., Sanvito, F., Orci, L. and Vassalli, J. D. (1994). Ablation of islet endocrine cells by targeted expression of hormone-promoter-driven toxigenes. Proc. Natl. Acad. Sci. USA 91, 12999-13003. Hitachi, K., Danno, H., Tazumi, S., Aihara, Y., Uchiyama, H., Okabayashi, K., Kondow, A. and Asashima, M. (2009). The Xenopus Bowline/Ripply family proteins negatively regulate the transcriptional activity of T-box transcription factors. Int. J. Dev. Biol. 53, 631-639. Huang, H.-P., Liu, M., El-Hodiri, H. M., Chu, K., Jamrich, M. and Tsai, M.-J. (2000). Regulation of the pancreatic islet-specific gene BETA2 (neuroD) by neurogenin 3. Mol. Cell. Biol. 20, 3292-3307. Huang, D. W., Sherman, B. T. and Lempicki, R. A. (2009). Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protoc. 4, 44-57. Ince-Dunn, G., Okano, H. J., Jensen, K. B., Park, W.-Y., Zhong, R., Ule, J., Mele, A., Fak, J. J., Yang, C., Zhang, C. et al. (2012). Neuronal Elav-like (Hu) proteins regulate RNA splicing and abundance to control glutamate levels and neuronal excitability. Neuron 75, 1067-1080. Kawamura, A., Koshida, S., Hijikata, H., Ohbayashi, A., Kondoh, H. and Takada, S. (2005). Groucho-associated transcriptional repressor ripply1 is required for proper transition from the presomitic mesoderm to somites. Dev. Cell 9, 735-744. Leonard, J., Peers, B., Johnson, T., Ferreri, K., Lee, S. and Montminy, M. R. (1993). Characterization of somatostatin transactivating factor-1, a novel homeobox factor that stimulates somatostatin expression in pancreatic islet cells. Mol. Endocrinol. 7, 1275-1283. Llorian, M. and Smith, C. W. J. (2011). Decoding muscle alternative splicing. Curr. Opin. Genet. Dev. 21, 380-387. Mathelier, A., Zhao, X., Zhang, A. W., Parcy, F., Worsley-Hunt, R., Arenillas, D. J., Buchman, S., Chen, C.-y., Chou, A., Ienasescu, H. et al. (2014). JASPAR 2014: an extensively expanded and updated open-access database of transcription factor binding profiles. Nucleic Acids Res. 42, D142-D147. Mellitzer, G., Bonné , S., Luco, R. F., Van De Casteele, M., Lenne-Samuel, N., Collombat, P., Mansouri, A., Lee, J., Lan, M., Pipeleers, D. et al. (2006). IA1 is NGN3-dependent and essential for differentiation of the endocrine pancreas. EMBO J. 25, 1344-1352. Miyatsuka, T., Kosaka, Y., Kim, H. and German, M. S. (2011). Neurogenin3 inhibits proliferation in endocrine progenitors by inducing Cdkn1a. Proc. Natl. Acad. Sci. USA 108, 185-190.

DEVELOPMENT

RESEARCH ARTICLE

Nelson, S. B., Schaffer, A. E. and Sander, M. (2007). The transcription factors Nkx6.1 and Nkx6.2 possess equivalent activities in promoting beta-cell fate specification in Pdx1+ pancreatic progenitor cells. Development 134, 2491-2500. Nishimura, W., Kondo, T., Salameh, T., El Khattabi, I., Dodge, R., Bonner-Weir, S. and Sharma, A. (2006). A switch from MafB to MafA expression accompanies differentiation to pancreatic β-cells. Dev. Biol. 293, 526-539. Ohlsson, H., Karlsson, K. and Edlund, T. (1993). IPF1, a homeodomaincontaining transactivator of the insulin gene. EMBO J. 12, 4251-4259. Okubo, T., Kawamura, A., Takahashi, J., Yagi, H., Morishima, M., Matsuoka, R. and Takada, S. (2011). Ripply3, a Tbx1 repressor, is required for development of the pharyngeal apparatus and its derivatives in mice. Development 138, 339-348. Oliver-Krasinski, J. M. and Stoffers, D. A. (2008). On the origin of the β cell. Genes Dev. 22, 1998-2021. Oropez, D. and Horb, M. (2012). Transient expression of Ngn3 in Xenopus endoderm promotes early and ectopic development of pancreatic beta and delta cells. Genesis 50, 271-285. Potter, L. A., Choi, E., Hipkens, S. B., Wright, C. V. and Magnuson, M. A. (2011). A recombinase-mediated cassette exchange-derived cyan fluorescent protein reporter allele for Pdx1. Genesis 50, 384-392. Rabenau, K. E., O’Toole, J. M., Bassi, R., Kotanides, H., Witte, L., Ludwig, D. L. and Pereira, D. S. (2004). DEGA/AMIGO-2, a leucine-rich repeat family member, differentially expressed in human gastric adenocarcinoma: effects on ploidy, chromosomal stability, cell adhesion/migration and tumorigenicity. Oncogene 23, 5056-5067. Reich, M., Liefeld, T., Gould, J., Lerner, J., Tamayo, P. and Mesirov, J. P. (2006). GenePattern 2.0. Nat. Genet. 38, 500-501. Rieck, S., Bankaitis, E. D. and Wright, C. V. E. (2012). Lineage determinants in early endocrine development. Semin. Cell Dev. Biol. 23, 673-684. Rosenbaum, J. N., Duggan, A. and Garcı́a-Añoveros, J. (2011). Insm1 promotes the transition of olfactory progenitors from apical and proliferative to basal, terminally dividing and neuronogenic. Neural Dev. 6, 6. Schaffer, A. E., Taylor, B. L., Benthuysen, J. R., Liu, J., Thorel, F., Yuan, W., Jiao, Y., Kaestner, K. H., Herrera, P. L., Magnuson, M. A. et al. (2013). Nkx6.1 controls a gene regulatory network required for establishing and maintaining pancreatic Beta cell identity. PLoS Genet. 9, e1003274. Schwitzgebel, V. M., Scheel, D. W., Conners, J. R., Kalamaras, J., Lee, J. E., Anderson, D. J., Sussel, L., Johnson, J. D. and German, M. S. (2000). Expression of neurogenin3 reveals an islet cell precursor population in the pancreas. Development 127, 3533-3542.


Smith, S. B., Gasa, R., Watada, H., Wang, J., Griffen, S. C. and German, M. S. (2003). Neurogenin3 and hepatic nuclear factor 1 cooperate in activating pancreatic expression of Pax4. J. Biol. Chem. 278, 38254-38259. Sosa-Pineda, B., Chowdhury, K., Torres, M., Oliver, G. and Gruss, P. (1997). The Pax4 gene is essential for differentiation of insulin-producing β cells in the mammalian pancreas. Nature 386, 399-402. Soyer, J., Flasse, L., Raffelsberger, W., Beucher, A., Orvain, C., Peers, B., Ravassard, P., Vermot, J., Voz, M. L., Mellitzer, G. et al. (2010). Rfx6 is an Ngn3-dependent winged helix transcription factor required for pancreatic islet cell development. Development 137, 203-212. Spellman, R., Llorian, M. and Smith, C. W. J. (2007). Crossregulation and functional redundancy between the splicing regulator PTB and its paralogs nPTB and ROD1. Mol. Cell 27, 420-434. Stanger, B. Z. and Hebrok, M. (2013). Control of cell identity in pancreas development and regeneration. Gastroenterology 144, 1170-1179. Sussel, L., Kalamaras, J., Hartigan-O’Connor, D. J., Meneses, J. J., Pedersen, R. A., Rubenstein, J. L. and German, M. S. (1998). Mice lacking the homeodomain transcription factor Nkx2.2 have diabetes due to arrested differentiation of pancreatic beta cells. Development 125, 2213-2221. Takahashi, J., Ohbayashi, A., Oginuma, M., Saito, D., Mochizuki, A., Saga, Y. and Takada, S. (2010). Analysis of Ripply1/2-deficient mouse embryos reveals a mechanism underlying the rostro-caudal patterning within a somite. Dev. Biol. 342, 134-145. Tissir, F. and Goffinet, A. M. (2006). Expression of planar cell polarity genes during development of the mouse CNS. Eur. J. Neurosci. 23, 597-607. Vance, K. W., Carreira, S., Brosch, G. and Goding, C. R. (2005). Tbx2 is overexpressed and plays an important role in maintaining proliferation and suppression of senescence in melanomas. Cancer Res. 65, 2260-2268. Wildner, H., Gierl, M. S., Strehle, M., Pla, P. and Birchmeier, C. (2008). Insm1 (IA-1) is a crucial component of the transcriptional network that controls differentiation of the sympatho-adrenal lineage. Development 135, 473-481. Zeggini, E., Scott, L. J., Saxena, R., Voight, B. F., Marchini, J. L., Hu, T., de Bakker, P. I. W., Abecasis, G. R., Almgren, P., Andersen, G. et al. (2008). Meta-analysis of genome-wide association data and large-scale replication identifies additional susceptibility loci for type 2 diabetes. Nat. Genet. 40, 638-645. Zhang, T., Liu, W.-D., Saunee, N. A., Breslin, M. B. and Lan, M. S. (2009). Zinc finger transcription factor INSM1 interrupts cyclin D1 and CDK4 binding and induces cell cycle arrest. J. Biol. Chem. 284, 5574-5581.

DEVELOPMENT

RESEARCH ARTICLE

2949

Supplementary Methods

Primers. Genotyping of the Insm1GFPCre allele was performed using primers IA3: 5′GCCCTTGTACAACCGACAGCTCT and IA4: 5′-GCCCTTGTACAACCGACAGCTCT, which amplify a 423 bp fragment from the wild-type allele and a 362 bp fragment from the Insm1GFPCre allele. Genotyping of the Pdx1CFP allele was performed using primers IA3: 5′-GCCCTTGTACAACCGACAGCTCT and IA4: 5′GCCCTTGTACAACCGACAGCTCT, which amplify a 423 bp fragment from the wild-type allele and a 362 bp fragment from the Insm1GFPCre allele. Genotyping of the Ripply3β-Gal allele was performed using primers Rpy3geF1: 5′-AACCTGAGATCGACTACTGC-3′ and Rpy3-geR1: ATCCCTTAAGGTCTGTCTGC, which amplify a 350 bp fragment from the wild-type allele; and lacZ-F1: 5′-TGTTTTGACCGCTGGGATCTGC-3′ and lacZ-R1 5′-CCAGACCAACTGGTAATGGTAGC-3′, which amplify a 550 bp fragment from the Ripply3β-Gal allele. For qPCR analysis of Neurog3 alternative splicing variants the following primers were used: (N1) 5′CAGCTATCCACTGCTGCTTG; (N2) 5′-GAATTGGAACTGAGCACTTC; (N3) 5′GTTTGCTGAGTGCCAACTC; and (N4) 5′-CGGAGCAGTCCCTAGTTCTC. RT-qPCR for Ripply3 gene expression was performed on 50 ng cDNA using primers 5′-TGAGTCTTGGGGAGACCAAC and 5′AGAAATGAATGGTGGCTTGC. Gene expression was normalized using the Hprt gene, which was amplified with primers 5′-TACGAGGAGTCCTGTTGATGTTGC and 5′-GGGACGCAGCAACTGACATTTCTA. For ChIP analysis the following primers were used. Insm1 promoter: 5′-GAGGAGCTGCGGACGCTCTGATTG-3′, 5′-GACGCGCGTGCGGAGCGCAGAG-3′; Rest promoter forward 5′-GATTAGGTAAGTTTCCCCCGAG-3′, reverse 5′-CGCACATTCCAGCACAGGAC-3′; Cdkn1c promoter forward 5′-CTGTGCCCAGCTCCATGGTC3′, reverse 5′-AGAGCTCTGCGGGCGGTAAC-3′; Cdkn1b promoter forward 5′-

1

Development | Supplementary Material

GAGCGGTCAGTCCTGGCTTC-3′, reverse 5′-GGAGGGTATGCGAGAGATTCAG-3′; Ripply3 promoter forward 5′-GCACTGGAGAACAGGTTATTGG-3′, reverse 5′-GTGACAGCTGGGTTTACGAAG-3′; Pck1 promoter forward 5′-CAGGCAGGGTCAAAGTTTAGTC-3′, reverse 5′- GGCAGGCCTTTGGATCATAGCC3′.

Antibodies. For immunohistochemical staining the primary antibodies and dilutions used were as follows: chicken anti-GFP (A10262, Life Technologies), 1:500; rabbit or guinea pig anti-PDX-1 (a gift from Christopher Wright, Vanderbilt University, Nashville, TN, USA), 1:1000; guinea pig anti-INS (A0564, DACO), 1:1000; rabbit anti-GCG (AB932, EMD Millipore), 1:1000; sheep anti-SST (13-2366, American Research Products), 1:1000; rabbit anti-GHRL (19-031-30, Phoenix Pharmaceuticals), 1:1000, rabbit anti-PPY (AB15668, EMD Millipore), 1:1000; mouse anti-MKI67 (550609, BD Biosciences), 1:1000; goat anti-NEUROG3 (AB2774, BCBC Antibody Core), 1:500; mouse anti-INSM1 (AB2154, BCBC Antibody Core), 1:200; rabbit anti-LRP11 (C16508, Assay Biotech), 1:500; rabbit anti-CELSR3 (sc-99200, Santa Cruz Biotechnology), 1:100; rabbit antiAMIGO2 (252046, Ab Biotech), 1:1000; rabbit anti-GLUT2 (07-1402, EMD Millipore), 1:1000; rabbit antiMAFA (A300-611A, Bethyl Laboratories), 1:1000; rabbit anti-MAFB (IHC-0351, Bethyl Laboratories), 1:1000; rabbit anti-NKX6.1 (AB1069, BCBC Antibody Core), 1:1000; chicken anti-β-galactosidase (AB9361, Abcam), 1:500; rabbit anti-Chromogranin A (AB15160, Abcam), 1:1000; rabbit anti-E-cadherin (4065, Cell Signaling), 1:1000. Secondary antibodies (Life Technologies, 1:1000) were as follows: donkey anti-mouse Alexa 488 conjugated (A-21202); goat anti-chicken IgG Alexa 488 (A-11039) or 555(A-21437) conjugated; goat antiguinea pig IgG Alexa 555 (A-21435) or 647 (A-21450) conjugated; donkey anti-rabbit IgG Alexa 555 (A-31572) or 680 (A-10043) conjugated; donkey anti-sheep IgG Alexa 555 (A-21099) conjugated. Staining with anti-

2


INSM1 antibodies required the use of the TSA amplification kit (PerkinElmer). ChIPs were performed using rabbit polyclonal anti-INSM1 antibodies (sc-68355X) and normal rabbit antibodies (sc-2027, Santa Cruz Biotechnology).

3


Supplementary Figure S1. Generation of Insm1GFPCre allele. (A) Gene targeting and RMCE strategy. An InsmLCA allele was made by gene targeting in mESCs. A 5232 bp fragment containing the Insm1 promoter region, single exon and 3’ UTR were replaced by positive-negative selection marker puroDtk surrounded by heterotypic Lox66 and Lox2272 sites. Puromycin-resistant clones were screened by Southern blotting. To screen for homologous recombination on the 5’ end DNA was digested using SpeI restriction enzyme and hybridized with the 5’ probe; for the 3’ end screen DNA was digested with NdeI and hybridized with the 3’ probe. A correctly targeted clone (1G3) containing the Insm1LCA allele was used for RMCE to create Insm1GFPCre allele introducing Insm1 promoter and GFP-Cre fusion gene. Clones resistant to hygromycin were screened for cassette exchange by PCR with primers 1+2 and 3+4 to assess site specific recombination at the 5’- and 3’-ends, respectively. Clone 1G3/3B3 was used to make mice which were then crossed to animals expressing FlpE to delete the hygromycin-resistance cassette. (B) Southern blot analysis. The presence of 11.4 kb band with 5’ probe and 4.5 kb band with 3’ probe indicates positively targeted clones. (C) PCR analysis. Exchanged clones are characterized by double bands (423 and 325 bp) at 5’ end, and a single 567 bp band at the 3’ end. (D) Insm1GFPCre mouse genotyping is done using the same primers as for 5’ RMCE screen. (E) Insm1GFPCre homozygous mice do not express INSM1 protein as confirmed by Western blot with anti-INSM1 antibodies. The PC12 cell line was used as positive control for INSM1 expression.


Supplementary Figure S2. GFP is co-expressed with Insm1 in Insm1GFPCre/+ mice. Immunofluorescence labeling with anti-GFP (green) and anti-INSM1 (red) antibodies shows that GFP is co-expressed with Insm1 in pancreatic endocrine cells from E15.5 to E18.5. Scale bar = 50 mm.


Supplementary Figure S3. Impaired differentiation of endocrine cells in Insm1 knockout mice. (A) Immunofluorescence labeling of pancreatic tissues from Insm1+/- and Insm1-/- mice at E18.5 with anti-GFP antibodies (green) and antibodies against pancreatic hormones (red): insulin (Ins), glucagon (Gcg), somatostatin (Sst), ghrelin (Ghrl) and pancreatic polypeptide (Ppy). (B) Percentage of GFP-positive cells co-expressing different hormones at E18.5 in Insm1+/- and Insm1-/- mice (p≤0.01 for all). Scale bar = 50 mm.


Supplementary Figure S4. Analysis of proliferation of pre-endocrine cells in Insm1 knockout mice at E15.5. (A) Immunofluorescence labeling of pancreatic tissues from Insm1+/- and Insm1-/- mice at E15.5 with anti-GFP antibodies (green) and antibodies against cell proliferation marker Ki-67(red). (B) Percentage of Ki-67 positive cells per GFP positive cells demonstrates slight proliferation defect in Insm1-/- pancreas. (p