DTD 5
ARTICLE IN PRESS YDBIO-02043; No. of pages: 13; 4C: 3, 4, 6, 7, 8, 9
Developmental Biology xx (2005) xxx – xxx www.elsevier.com/locate/ydbio
Dorsal pancreas agenesis in retinoic acid-deficient Raldh2 mutant mice Merce` Martı´n a, Jabier Gallego-Llamas b,c, Vanessa Ribes b,c, Miche`le Kedinger a,c, Karen Niederreither b,d, Pierre Chambon b,c,e, Pascal Dolle´ b,c,e,*, Ge´rard Gradwohl a,c,* a Inserm, U682, Strasbourg, France Institut de Ge´ne´tique et de Biologie Mole´culaire et Cellulaire, CNRS UMR7104, Inserm, U596, Illkirch, France c Universite´ Louis Pasteur Strasbourg, France d Departments of Medicine and Molecular and Cellular Biology, Center for Cardiovascular Development, Baylor College of Medicine, Houston, TX 77030, USA e Institut Clinique de la Souris, Illkirch, France
b
Received for publication 4 January 2005, revised 10 May 2005, accepted 26 May 2005
Abstract During embryogenesis, the pancreas arises from dorsal and ventral pancreatic protrusions from the primitive gut endoderm upon induction by different stimuli from neighboring mesodermal tissues. Recent studies have shown that Retinoic Acid (RA) signaling is essential for the development of the pancreas in non-mammalian vertebrates. To investigate whether RA regulates mouse pancreas development, we have studied the phenotype of mice with a targeted deletion in the retinaldehyde dehydrogenase 2 (Raldh2) gene, encoding the enzyme required to synthesize RA in the embryo. We show that Raldh2 is expressed in the dorsal pancreatic mesenchyme at the early stage of pancreas specification. RA-responding cells have been detected in pancreatic endodermal and mesenchymal cells. Raldh2-deficient mice do not develop a dorsal pancreatic bud. Mutant embryos lack Pdx1 expression, an essential regulator of early pancreas development, in the dorsal but not the ventral endoderm. In contrast to Pdx1-deficient mice, the early glucagon-expressing cells do not develop in Raldh2 knockout embryos. Shh expression is, as in the wild-type embryo, excluded from the dorsal endodermal region at the site where the dorsal bud is expected to form, indicating that the dorsal bud defect is not related to a mis-expression of Shh. Mesenchymal expression of the LIM homeodomain protein Isl1, required for the formation of the dorsal mesenchyme, is altered in Raldh2 / embryos. The homeobox gene Hlxb9, which is essential for the initiation of the pancreatic program in the dorsal foregut endoderm, is still expressed in Raldh2 / dorsal epithelium but the number of HB9-expressing cells is severely reduced. Maternal supplementation of RA rescues early dorsal pancreas development and restores endodermal Pdx1 and mesenchymal Isl1 expression as well as endocrine cell differentiation. These findings suggest that RA signaling is important for the proper differentiation of the dorsal mesenchyme and development of the dorsal endoderm. We conclude that RA synthesized in the mesenchyme is specifically required for the normal development of the dorsal pancreatic endoderm at a stage preceding Pdx1 function. D 2005 Elsevier Inc. All rights reserved. Keywords: Pancreas; Raldh2; Retinoids; Foregut; Endoderm; Mesenchyme; Shh; Isl1; Hlxb9
Introduction * Corresponding authors. G. Gradwohl is to be contacted at Inserm, U682, Development and Pathophysiology of the Intestine and Pancreas, 3 avenue Molie`re, F-67200, Strasbourg, France. P. Dolle´ , Institut de Ge´ne´tique et de Biologie Mole´culaire et Cellulaire, 1 rue Laurent Fries, BP 10142, F-67404 Illkirch Cedex, France. E-mail addresses:
[email protected] (P. Dolle´),
[email protected] (G. Gradwohl). 0012-1606/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2005.05.035
The specification of the pancreas in mouse embryo takes place between 8.0 and 8.5 embryonic days (E8.0 –E8.5). At E9 – 9.5, dorsal and ventral protrusions (the pancreatic buds) arise from the primitive gut endoderm, which will grow, branch and fuse to form the definitive organ. Just before and during budding, pancreatic endodermal progenitor cells
ARTICLE IN PRESS 2
M. Martı´n et al. / Developmental Biology xx (2005) xxx – xxx
express the homeodomain protein Pdx1/Ipf1, which is also found in duodenal progenitors (Jonsson et al., 1995). Lineage tracing studies demonstrated that endocrine and exocrine cells derive form Pdx1-expressing progenitors (Gu et al., 2002). Pdx1 is required for the pancreatic buds to grow and differentiate but not for the initial commitment of the endoderm to a pancreatic fate (Ahlgren et al., 1996; Jonsson et al., 1994; Offield et al., 1996). Another transcription factor, the bHLH (basic helix –loop –helix) protein Ptf1a, is essential for the acquisition of the pancreatic fate by the foregut endoderm (Kawaguchi et al., 2002). Due to their spatially different positions, the endodermal cells that will form the dorsal and ventral pancreatic buds are exposed sequentially to different inductive stimuli from neighboring tissues of mesodermal origin and will thus execute different differentiation programs (Edlund, 2002; Johansson and Grapin-Botton, 2002; Kumar and Melton, 2003). Dorsally, the pre-pancreatic endoderm is initially in close contact with the notochord, which is required for the early specification of the dorsal pancreas (Kim et al., 1997) likely by locally inhibiting the expression of Shh (Hebrok et al., 1998). Shh exclusion from the pancreatic primordium both dorsally and ventrally is required for pancreas development (Apelqvist et al., 1997; Hebrok et al., 1998). Activin-hB and FGF2 have been identified as the potential notochord secreted factors that repress Shh in the dorsal pro-pancreatic endoderm (Hebrok et al., 1998). At approximately E9.0, the notochord is then separated from the endoderm by midline fusion of the paired dorsal aortas. Signaling from the vascular endothelium induces the expression of Pdx1 and insulin in the pre-specified pancreatic endoderm (Lammert et al., 2001), as well as of the pancreatic transcription factor Ptf1a (Yoshitomi and Zaret, 2004). Then, mesodermal cells are recruited dorsally, surround the dorsal pancreatic bud and promote the development of the underlying pancreatic endoderm. FGF10 has been identified as a mesenchymal factor stimulating the proliferation of pancreatic endodermal progenitor cells and inhibiting their differentiation probably by maintaining Pdx1 expression (Bhushan et al., 2001; Hart et al., 2003; Norgaard et al., 2003). The endoderm that will form the ventral pancreas initially receives instructive signals from the lateral plate mesoderm (Kumar et al., 2003). Later, ventral endoderm sufficiently distant from the cardiac and septum tranversum mesoderm escapes hepatic inducing signals from these structures allowing the initiation of the differentiation program of the ventral pancreas (Deutsch et al., 2001). The analysis of various knockout mouse models has revealed additional differences between the development of the dorsal and ventral pancreatic buds. Several genes such as Hlxb9 (Harrison et al., 1999; Li et al., 1999), NCadherin (Esni et al., 2001) and Islet-1 (Isl1) (Ahlgren et al., 1997) have been shown to be specifically required for dorsal pancreatic development. The Hlx9 homeobox gene
product, HB9, is initially expressed in pancreatic epithelial progenitor cells and is later restricted to the h-cell lineage. Mice lacking Hlxb9 gene fail to initiate the dorsal pancreatic program resulting in the agenesis of the dorsal pancreatic lobe whereas the ventral pancreas develops but h-cell differentiation is perturbed (Harrison et al., 1999; Li et al., 1999). Mice deficient for the N-Cadherin cell adhesion molecule, initially expressed in the pancreatic mesenchyme and later in the pancreatic endoderm, similarly exhibit a selective dorsal pancreas agenesis (Esni et al., 2001). At early stages (E9.5), the transcription factor Isl1 is expressed in the mesenchyme in a dorso-ventral gradient, as well as in precursors of endocrine cells. Analysis of Isl1deficient embryos revealed that Isl1 has two independent functions during pancreas development (Ahlgren et al., 1997). First, Isl1 is required for dorsal pancreatic mesenchyme formation. The absence of dorsal mesenchyme in Isl1-deficient embryos blocks exocrine differentiation specifically in the dorsal bud. Second, Isl1 is required, cell autonomously, for the differentiation of endocrine cells. Interestingly, in Isl1- and N-Cadherin-deficient mice, although markedly reduced, Pdx1 expression is detected in endodermal progenitors in the initial stage of dorsal bud formation (Ahlgren et al., 1997; Esni et al., 2001). In contrast, Pdx1 expression is specifically lost in the dorsal pancreatic epithelium in embryos lacking Hlxb9 gene suggesting that HB9 is important for early cell fate decisions in the dorsal endoderm (Abraham et al., 2002; Harrison et al., 1999). Recently, a role for retinoic acid (RA) in pancreas development has been proposed in zebrafish, Xenopus and quail embryos (Chen et al., 2004; Kumar et al., 2003; Stafford et al., 2004; Stafford and Prince, 2002). RA, the active metabolite of vitamin A, plays essential roles in morphogenesis and organogenesis (Ross et al., 2000). This signaling molecule binds to two families of nuclear receptors, the RARs (Retinoic Acid Receptors) and RXRs (Retinoid Receptors), which act as heterodimers to control the transcription of genes containing RAREs (RA Responsive Elements) in their promoter region (Chambon, 1996; Ross et al., 2000). Within the embryo, RA is synthesized from circulating retinol (vitamin A) in a twostep reaction involving specific alcohol dehydrogenases (ADH) and aldehyde dehydrogenases (ALDH), respectively. Among the three retinaldehyde dehydrogenases (RALDH) identified in vertebrates, RALDH2 is the earliest and most broadly expressed during embryogenesis (Niederreither et al., 1997). Knockout of the mouse Raldh2 gene is early embryonic lethal due to severe trunk and heart defects (Niederreither et al., 1999, 2000, 2001, 2003). In zebrafish, RA signaling is required for pancreas and liver specification, and treatment with exogenous RA induces an anterior ectopic expression of pancreatic and liver markers in addition to their normal expression domains (Stafford and Prince, 2002). More recently, it has been demonstrated that an inhibition of RA signaling at the gastrula stage in
ARTICLE IN PRESS M. Martı´n et al. / Developmental Biology xx (2005) xxx – xxx
Xenopus results in the loss of dorsal pancreas, correlating with an expansion of Shh expression into the prospective dorsal pancreatic endoderm (Chen et al., 2004; Stafford et al., 2004). Moreover, in avian, RA-deficient embryos also lack dorsal pancreas (Stafford et al., 2004) and RA signaling is sufficient to induce Pdx1 in anterior endoderm (Kumar et al., 2003) suggesting a conserved role of RA in the specification of the pancreas during evolution. To gain insight into the role of RA signaling during mouse pancreas organogenesis, we characterized the expression pattern of Raldh2 at different developmental stages by in situ hybridization, and identified RAresponding cells in the early pancreas anlage of RAREhsp68-LacZ transgenic mice. Our data indicate that Raldh2 is expressed in the developing dorsal pancreatic mesenchyme during the early steps of pancreas specification and that both mesenchymal and endodermal cells respond to RA. Moreover, we report defective dorsal pancreatic development in Raldh2 knockout mice, providing evidence that, in mice, RA signaling controls early steps of pancreas morphogenesis and that Raldh2 is critical for the proper development of the dorsal pancreatic anlage.
3
Results and discussion Raldh2 is expressed in the dorsal pancreatic mesenchyme To study the role of RA signaling during pancreas ontogeny, we initially characterized the expression patterns of Raldh1, 2 and 3 at E9.5 by in situ hybridization. Whereas Raldh2 transcripts were detected in the pancreatic region, we did not detect any expression of Raldh1 and Raldh3 in the pancreatic mesoderm or endoderm at this stage (data not shown), suggesting that RALDH2 is the unique enzyme responsible for RA synthesis during early pancreas development. At E8.75– E9.0, which correspond approximately to the onset of pancreas specification, Raldh2 mRNA was first detected in the dorso-lateral mesenchyme in the vicinity and in contact with the dorsal pancreatic bud expressing Pdx1 in committed pancreatic epithelial cells (Figs. 1A and B). Slightly later, at E9.5, the mesenchymal cells originating from the lateral gut mesoderm are then thought to converge dorsally and form the dorsal pancreatic mesenchyme where Raldh2 is still expressed (Figs. 1E – H and C). At this stage, some Raldh2 expressing cells are found in direct contact with the most dorsal pancreatic endodermal cells whereas
Fig. 1. Raldh2 is expressed in the dorsal mesenchyme at early stages of pancreas development. In situ hybridization with a Raldh2 RNA antisense probe on transverse sections through the developing pancreas at E8.75 (14 somites) (A), E9.0 (B), E9.5 (E – H) and sagittal sections at E9.5 (C) and E10.5 (D). The developing pancreatic endoderm has been identified by immunohistochemistry, using a Pdx1 antibody on the same sections. Raldh2 transcripts are detected in the mesoderm in the vicinity of the dorsal pancreatic bud. Expression is first detected in the dorso-lateral mesenchyme (A, B) and then dorsally between the dorsal aorta and the dorsal pancreatic bud such as some Raldh2-expressing cells contact the dorsal endodermal pancreatic progenitors (F). At E10.5, strong Raldh2 expression is detected in the intestinal mesenchyme (D). E – H are series of transversal sections along the antero-posterior axis of an E9.5 embryo to show the broader expression of Raldh2 in the mesenchyme surrounding the gut endoderm anterior and posterior to the pancreas. dp, dorsal pancreas; vp, ventral pancreas; i, intestine; nc, notochord; s, somite. * marks the aorta. Scale bars = 50 AM.
ARTICLE IN PRESS 4
M. Martı´n et al. / Developmental Biology xx (2005) xxx – xxx
others are more distant (Figs. 1F and G). In contrast to the dorsal pancreatic domain, Raldh2 was not expressed in mesenchymal cells in the vicinity of the ventral foregut endoderm at the level of the ventral pancreatic bud (Figs. 1A, B and G). In regions anterior (Fig. 1E) and posterior (Fig. 1H) to the pancreatic domain, the gut endoderm is surrounded by Raldh2+ mesodermal cells. Raldh2 expression becomes very strong in intestinal (midgut) mesenchymal cells by E10.5 (Fig. 1D) where some Raldh2+ cells are in contact with the ventral pancreas endoderm. At E12.5, when the pancreatic epithelium starts to branch, the dorsal pancreatic mesenchymal expression of Raldh2 is almost lost (not shown). Raldh2 was not found expressed in the foregut endoderm of the early pancreatic buds (Fig. 1), neither was it expressed later in the developing pancreatic epithelium (not shown). Both mesenchymal and epithelial cells of the developing dorsal pancreas respond to RA signaling During early pancreas development, RA could act directly on endodermal epithelial cells by diffusing from its mesodermal site of synthesis, or indirectly via the mesodermal cells. To identify the cells that respond to the RA signal, we used an RA-sensitive transgenic line in which the LacZ reporter gene is under the transcriptional control of a minimal promoter linked to 3 copies of the RA response element (RARE) from the Rarb gene (the RARE-hsp68LacZ line; Rossant et al., 1991). This line is commonly used as a sensitive reporter for the detection of RA-responsive cells in the embryo (e.g. Niederreither et al., 2002; Yashiro et al., 2004; and refs therein). Analysis of h-galactosidase (h-Gal) activity on transversal sections of E9.5 RARE-
hsp68-LacZ embryos at the level of the developing pancreas revealed a general dorso-ventral gradient in the intensity of the responsiveness of cells to RA throughout tissues. (Fig. 2A). RA-responsive cells are clearly found both in the Pdx1+ epithelium (white arrows) of the dorsal pancreatic bud and surrounding mesenchyme (dark arrow) (Fig. 2B). Some weaker RA signaling activity was also observed in the ventral endoderm and adjacent mesenchyme. It is likely that RA activity in the ventral endoderm results from RA diffusing from posterior, Raldh2+, intestinal mesenchymal cells (Figs. 1C and H). Double immunofluorescence using antibodies against h-galactosidase and insulin + glucagon revealed that some of the h-Gal-expressing cells at E9.5 and E10.5 co-stain for the endocrine hormones providing evidence that islet cells derive from RA-responding epithelial cells (data not shown). At E15.5, when islet cells differentiate massively, we did not detect any RA-responding cell in the developing pancreas (data not shown) in agreement with the loss of Raldh2 expression. Taken together, these results suggest that RA, synthesized in the dorso-lateral mesodermal cells where Raldh2 is expressed, could act during normal dorsal pancreas development either by signaling directly to the adjacent dorsal endodermal cells or by controlling important features of mesenchymal cells, or both. This latter hypothesis is supported by previous studies which have identified RA receptors, by RT-PCR, both in the epithelium and the mesenchyme of mouse embryonic pancreas (Tulachan et al., 2003). Raldh2 knockout embryos lack the dorsal pancreatic bud The early expression of Raldh2 in the dorso-lateral and then in the most dorsal pancreatic mesenchymal cells, as
Fig. 2. RA-responsive cells are detected in both the mesenchyme and endoderm of the pancreatic anlage. (A – D) RA signaling was visualized in RARE-hsp68LacZ transgenic E9.5 embryos on transversal sections by the detection of h-galactosidase activity using X-Gal-staining. The pancreatic epithelium was revealed by immunohistochemistry using Pdx1 antibody. (A – C) In Raldh2+/+ wild-type background, a decreasing dorso-ventral gradient of the RARE-hsp68LacZ transgene expression is observed throughout the embryo with a stronger RA activity in the dorsal compared to ventral pancreatic mesenchyme (black arrows in panels B and C, respectively, compare intensity of X-Gal stainings). RA-responsive cells were also detected in the dorsal and ventral pancreatic epithelium (white arrows in panels B and C, respectively show X-Gal and Pdx1 double stained cells). Panels B and C are high-power magnifications of the boxed zones in panel A. (D) In RA-treated E9.5 Raldh2 / ;RARE-hsp68-LacZ embryos, Pdx1 expression is restored in the dorsal endoderm and dorsal pancreas development is rescued; we observed a strong RA activity in pancreatic endodermal cells (white arrows) and weaker in surrounding mesenchyme (black arrows). Ep: epithelium, Me: mesenchyme, vp: ventral pancreas, dp: dorsal pancreas. Scale bars = 50 AM.
ARTICLE IN PRESS M. Martı´n et al. / Developmental Biology xx (2005) xxx – xxx
well as the presence of RA-responding cells in the dorsal pancreatic anlagen suggests that RA signaling could be involved in the specification and/or growth of the dorsal bud. Since neither Raldh1 nor Raldh3 expression has been detected in the developing pancreatic region, we further examined the role of RA signaling on pancreas development by analyzing Raldh2-null mutant mice. Raldh2 / embryos die at E10.5, due to defects in cardiovascular development (Lai et al., 2003; Niederreither et al., 1999, 2001). Therefore, we initially examined E9.5 (20 – 25 somites) mutant embryos (n > 20) and used Pdx1 homeoprotein as a marker of the developing pancreatic buds. Pdx1 is expressed when the foregut endoderm commits into a pancreatic fate and is required for pancreas development after the formation of both the ventral and dorsal pancreatic buds in mouse (Jonsson et al., 1994, 1995; Offield et al., 1996). E9.5 Raldh2 / embryos were found to completely lack Pdx1+ cells in the prospective dorsal pancreatic region (Figs. 3B, F and N). Pdx1 was also not detected at 14 somites (E8.75, n = 3) in the dorsal endoderm, the earliest stage at which we detect Pdx1 by immuno-histochemistry (Fig. 3, compare C and D). In contrast, the ventral pancreatic bud formed properly beneath the developing heart with a clear expansion of Pdx1+ cells in the ventral pancreatic endoderm (Figs. 3B, D, F and N). Similarly, the specification of the liver is not affected in mutant embryos (n = 2), as suggested by the maintained expression of the liver and ventral pancreatic bud marker, the homeobox gene Hex (Bort et al., 2004) in endodermal progenitors at the level of the formation of the liver bud (Figs. 3O and P). We also examined the presence of the transcription factor Ptf1a, which has been shown using Cre recombination, to be specifically present in ventral and dorsal pancreatic progenitors at E10.5 (Kawaguchi et al., 2002). Ptf1a transcripts are detected in the dorsal pancreatic endoderm of wild-type embryos at E9.5 but are absent or strongly reduced in Raldh2 knockout mice (n = 3, Figs. 3I– L). We could not detect Ptf1a in the ventral bud of wildtype embryos at E9.5 (Fig. 3I) although Pdx1 protein is already present (Fig. 3M) probably resulting from the limited sensitivity of the in situ hybridization technique since Ptf1a mRNA as been detected previously by RT-PCR in the ventral bud (Yoshitomi and Zaret, 2004). In agreement with the absence of formation of the dorsal pancreatic bud in Raldh2-deficient mice, endocrine cell differentiation is impaired (n = 3) judged by the total lack of glucagonpositive cells which are normally found in the dorsal bud at this developmental stage (Fig. 3, compare panels G and H). This contrasts with the phenotype of Pdx1-deficient mice which lack a pancreas at birth but pancreas specification occurs and the developing buds initially contain islet cells but subsequent growth and morphogenesis are blocked (Ahlgren et al., 1996; Offield et al., 1996). Similarly, some insulin- and glucagon-expressing cells form dorsally in Ptf1a knockout mice (Kawaguchi et al., 2002). Thus, Raldh2 is required during the initiation of pancreas
5
formation for Pdx1 and Ptf1a expression in the dorsal pancreatic endoderm, as well as for early endocrine differentiation. These results suggest that Raldh2-mediated RA signaling controls the specification of the dorsal pancreatic fate within the mouse foregut endoderm. The absence of Raldh2 function results in altered gene expression in the mesoderm and endoderm of the prospective dorsal pancreas of Raldh2 / embryos To gain further insight into the role of Raldh2 in the development of the dorsal pancreatic bud, we have analyzed the expression of genes known to be involved in this process (Fig. 4). One of these genes, the transcription factor Hlxb9, encodes the HB9 homeodomain protein. Homozygous disruption of Hlxb9 leads to a phenotype similar to that described herein, in which the dorsal pancreatic bud fails to develop and Pdx1 expression is lost dorsally (Harrison et al., 1999; Li et al., 1999). Between E8 and E9.5, HB9 is expressed in the dorsal foregut endoderm -including the dorsal pancreas anlage- and its expression precedes that of Pdx1 (Harrison et al., 1999; Li et al., 1999). One day later, HB9 is also expressed in the ventral bud. In E9.5 Raldh2 / embryos, some HB9-expressing cells were detected along the dorsal foregut endoderm organized in a single cell layer in contrast to wild-type budding epithelium (Fig. 4B compared to Fig. 4A). However, their number was severely reduced (from 50 T 24 HB9+ cells/section in wild type to 15 T 7 in mutant embryos; n = 4) and the HB9-positive cells failed to expand and bud out. This result suggests that the dorsal endoderm is specified properly and that Raldh2 function and by extension RA signaling might be required for the growth of the dorsal pancreatic epithelium. The knockout of another homedomain transcription factor, Islet1 (Isl1), also results in a complete dorsal pancreas agenesis (Ahlgren et al., 1997). Isl1 is initially expressed (E9) in the dorso-lateral mesenchymal cells of the developing pancreas and later (E9.5 – E10) also in the mesodermal cells surrounding the dorsal pancreatic epithelium (Figs. 4C and E). In Isl1 / mice, the dorso-lateral cells fail to condensate dorsally and the dorsal mesenchyme does not develop (Ahlgren et al., 1997). Raldh2 deficient embryos (n = 4) exhibited a very similar phenotype with fewer mesenchymal cells (40% reduction in average/ section) present laterally and dorsally to the dorsal endoderm compared to wild-type littermates (Fig. 4, compare panels C, E with D, F). The remaining mesenchymal cells express weakly or not at all Isl1 (Figs. 4D and F, green and blue arrowheads, respectively). Indeed, 72% T 15 of all the mesenchymal cells located dorsally to the ventral bud are positive for Isl1 in wild types compared to 26% T 8 in Raldh2 / embryos. In addition to its mesenchymal expression, Isl1 is also expressed in the nuclei of cells of the endocrine lineage, and is required for their differentiation (Ahlgren et al., 1997). Interestingly, although no glucagon-expressing cells (the first hormone-
ARTICLE IN PRESS 6
M. Martı´n et al. / Developmental Biology xx (2005) xxx – xxx
Fig. 3. Dorsal pancreatic bud agenesis in Raldh2 knockout embryos. Pdx1 was used as an early marker of the pancreatic epithelium by immunohistochemistry on whole mounts (A, B) or transverse sections (C – F) of E9.5 (22 – 25S) (A, B, E, F) and E8.75 (12 – 14S) (C, D) embryos. Unlike in wild-type embryos (A, C, E), in Raldh2 / mutants (B, D, F), only the ventral Pdx1 expression domain, delineating the ventral pancreatic endoderm, is observed (white arrows in panels A – B and vp). Dashed arrows in panels B, D, F and H indicate the expected site of dorsal pancreas development. Glucagon immunohistochemistry (G, H) on adjacent transversal sections of the E9.5 embryos (E, F) show the lack of Glucagon-expressing cells in the dorsal endoderm at the prospective pancreatic dorsal domain in Raldh2 / embryos (H) compared to wild-type (G). (I – L) Expression of pancreatic epithelial-specific transcription factor Ptf1a is lost in dorsal endoderm of Raldh2 / embryos at E9.5. Panels K and L are high magnifications of the boxed zones in panels I and J; the blue arrow points to a Ptf1aexpressing cell. Panels M and N are Pdx1 immunochemistry on adjacent sections of panels I and J. (O, P) Hex1 is expressed in the developing hepatic bud in wild-type and Raldh2 / embryos at E9.5. Dashed lines delineate the foregut endoderm. de: dorsal endoderm, dp: dorsal pancreas, vp: ventral pancreas, li: liver. Scale bars = 50 AM.
producing cells to appear during normal pancreas ontogeny) were detected in the dorsal presumptive pancreatic region of Raldh2 / embryos (Fig. 3H), Isl1-expressing cells could clearly be detected in quite a large number in the mutant dorsal foregut endoderm (Figs. 4D and F, green arrows). Some of these Isl1+ cells co-express HB9 in Raldh2-
deficient embryos. To determine whether RA signaling is important for the proliferation of the pancreatic bud, BrdU staining (16 h pulse) was performed. Surprisingly, BrdU labeled nuclei are clearly identified in wild-type and mutant endodermal epithelium in absence of Raldh2 (Figs. 4G and H). One possible explanation would be that these cells could
ARTICLE IN PRESS M. Martı´n et al. / Developmental Biology xx (2005) xxx – xxx
7
Fig. 4. Impaired dorsal pancreatic endoderm expansion and mesenchyme differentiation in Raldh2 knockout embryos. (A – F) Double immunofluorescence using antibodies against HB9 and Isl1 on transverse sections of E9.5 embryos. Some HB9-expressing cells are detected in the dorsal foregut endoderm of the mutant embryos (yellow arrow in panel B), but their number is severely decreased with respect to control embryos (compare red nuclei in panels A and B). Mesenchymal development and Isl1 expression in mesenchymal cells surrounding the dorsal HB9-positive endoderm are severely impaired in mutant embryos (D, F), in contrast to wild-type littermates (C, E). Green and blue arrowheads point to Isl1+ and Isl1 mesenchymal cells, respectively. Immature Isl1+ endocrine cells are found in both the wild-type and mutant dorsal endoderm (green arrows in panels C and D). Yellow arrows point to HB9/Isl1 double positive endodermal cells. (G, H) BrdU immuno-histochemistry at E9.5 revealed proliferating cells in dorsal pancreatic endoderm of wild-type (G) and Raldh2 / dorsal endoderm where the pancreas was expected to develop (H). Dashed lines delineate the foregut endoderm. de: endoderm, nc: notochord, dp: dorsal pancreas, sc: spinal chord.
divide during the labeling period, and prior analysis, but their further expansion is blocked by the lack of appropriate proliferative signals from the mesenchyme. Indeed mesenchymal growth factors are known to control the proliferation of pancreatic epithelial progenitor cells (Bhushan et al., 2001). The expression of Isl1, a marker of post-mitotic
endocrine cells, in the mutant epithelium is in agreement with this hypothesis. Taken together, these results suggest that the dorsal mesenchyme does not differentiate properly and that the dorsal bud agenesis in Raldh2 / embryos might be secondary to deficient signals from the mesoderm. Moreover, they suggest that a Pdx1-independent pancreatic
ARTICLE IN PRESS 8
M. Martı´n et al. / Developmental Biology xx (2005) xxx – xxx
exocrine and endocrine epithelial pancreas differentiation with a pancreatic to duodenal mesenchymal transformation (Apelqvist et al., 1997). It has been shown that dorsally, notochord-derived signals are important for active Shh repression in the underlying pancreatic endoderm (Hebrok et al., 1998). However, the mechanisms that maintain Shh repression after notochord/endoderm separation by the dorsal aorta remain unknown, and RA retinoic acid could potentially regulate this process. Supporting this hypothesis, Chen et al. (2004) have recently demonstrated that, in Xenopus, inhibition of RA signaling using the chemical inhibitor BMS453 results in an expansion of Shh expression in the territory of the future pancreas. Altered expression of Shh in the pancreatic region could thus explain the dorsal pancreas agenesis observed in Raldh2 embryos. When examined at E9.5, Shh expression was absent in endodermal cells of the presumptive pancreatic regions in wild-type embryos as expected (Fig. 5A). Shh expression is strongly reduced in the duodenal region of Raldh2-deficient embryos and no expression was observed in the endoderm where the dorsal bud was expected to develop (Fig. 5B). Like Shh, Ihh is also excluded from the pancreatic endoderm (Fig. 5C and Li et al., 1999), and this expression pattern was maintained in RA-deficient embryos (Fig. 5D). This result suggests that the dorsal pancreatic bud agenesis is not due to an ectopic expression of Shh or Ihh in the dorsal pancreatic endoderm.
Fig. 5. Expression of Shh and Ihh in E9.5 Raldh2 / deficient embryos. Double-labeling by in situ hybridization and immunohistochemistry using Shh RNA antisense probe and Pdx1 antibodies, or single in situ hybridization using Ihh RNA antisense probe was performed on transverse sections of wild-type (A, C) and Raldh2 / (B, D) E9.5 embryos. In wildtype embryos, both Shh and Ihh are specifically expressed in the lateral duodenal endoderm (brackets in panels A and C) and not in the Pdx1positive pancreatic endoderm (dp and vp). In the Raldh2 / endoderm, although strongly reduced (brackets in B and D), both Shh and Ihh are typically excluded from the pancreatic anlagen ventrally but also dorsally in the region where the dorsal bud is expected to develop. de: dorsal endoderm, dp: dorsal pancreas, vp: ventral pancreas, nc: notochord.
program has been started since some endodermal cells adopt an endocrine fate but are blocked and cannot complete differentiation into mature hormone-expressing cells. Sonic hedgehog is normally excluded from the presumptive dorsal pancreatic foregut epithelium in Raldh2 / embryos At early developmental stages, the Sonic hedgehog (Shh) gene is expressed throughout the embryonic gut endoderm but selectively excluded from the prospective pancreatic epithelium, suggesting that the Shh signal is not required for the initiation of pancreas development, and could inhibit pancreas organogenesis (Hebrok et al., 1998). Accordingly, forced ectopic expression of Shh in the pancreatic buds under the control of the Pdx1 promoter results in a lack of
Restoration of dorsal endodermal and mesodermal pancreas development in RA-rescued Raldh2 / embryos The early lethality of Raldh2 / embryos can be rescued through short-term supplementation of the maternal food with RA at subteratogenic doses (Niederreither et al., 1999, 2002). A minimal RA supplementation between E7.5 and E8.5 extends the viability of most of the Raldh2 / mutants until variable fetal stages and rescues some, but not all, of their phenotypic defects (Lai et al., 2003; Niederreither et al., 2001, 2002). Considering that the specification of the pancreas begins at about E8.5, this window of RA administration seemed appropriate to analyze whether dorsal pancreatic development may be rescued by exogenous RA in Raldh2 / embryos. As shown in Fig. 6, RArescued Raldh2 / embryos, analyzed at E9.5, develop two pancreatic buds based on Pdx1 expression which is restored in the dorsal endoderm (Figs. 6A –D). Interestingly, maternal RA administration also restores Isl1 expression and condensation of dorsal mesenchyme (Figs. 6C –D). In agreement with the rescue of endodermal and mesodermal pancreas development, hormone-expressing cells were detected in the dorsal pancreas of minimal RA-rescued E10.5 Raldh2 / embryos (Fig. 6F). However, when compared to wild-type embryos (Fig. 6E), the number of endocrine cells was decreased by 60% in average/section and their location was altered (hormone-expressing cells were found at the center, rather than as clusters at the periphery of the pancreas). When the RA-rescue treatment
ARTICLE IN PRESS M. Martı´n et al. / Developmental Biology xx (2005) xxx – xxx
9
Fig. 6. Rescue of dorsal pancreas development and islet cell differentiation by maternal RA supplementation of Raldh2-deficient embryos. Different doses (see Materials and methods) and windows (indicated at the bottom of the figure) of maternal RA supplementation have been administrated to rescue the Raldh2 / pancreatic phenotype. Anti-Pdx1 immunofluorescence was used on whole-mounts (A, B) or transversal cryosections (C – H) to visualize pancreatic bud formation at E9.5 (A – D) and E10.5 (E – H). Glucagon and insulin immunostaining was used to evaluate endocrine differentiation. Dorsal pancreatic bud formation and Pdx1 expression (B, D, F, H) as well as endocrine differentiation (F, H) are restored by RA treatment, but the proper number of hormoneexpressing cells in the dorsal pancreas is dependent on higher doses and extended period of RA administration (compare green cells in panels F and H). (C, D) Pdx1/Isl1 double immunofluorescence on E9.5 transversal cryosections. RA treatment rescues dorsal pancreatic mesenchymal cell condensation and mesenchymal Isl1 expression in Raldh2-deficient embryos (compare panels C and D). Histological analysis of frontal sections of minimally rescued (see Materials and methods) E18.5 embryos (I, J) shows that mutant pancreas are organized normally, in exocrine acini and endocrine islets of Langerhans. Amylase (K, L) and double insulin and Pdx1 immunostaining (M, N) were used as specific markers to study exocrine and endocrine differentiation, respectively, in RArescued (E7.5 – E10.5) E18.5 embryos. Differentiation of both types of tissues is clearly observed in the rescued E18.5 Raldh2 / mice. ac: acini, is: islet of Langerhans, vp: ventral pancreas, dp: dorsal pancreas. Scale bars = 50 AM.
was extended for an additional day until E9.5 (see Materials and methods), the number and distribution of hormoneexpressing cells in E10.5 Raldh2 / embryos were restored (Fig. 6H). Most of the endocrine cells in the fetal mouse pancreas arise during the secondary transition, a wave of differentiation starting at E13.5 and exocrine cells are first detected at E14.5. We thus also examined pancreas development at E18.5, the latest stage at which rescued Raldh2 / mutants can be recovered. Only 4% of rescued
embryos by a minimal RA treatment (E7.5 – E8.5) reach E18.5. Histological analysis of the pancreatic tissue of these mutants (n = 2), which is mainly located between the liver lobules, revealed well-formed acini as well as islets of Langerhans (Figs. 6I and J). When the RA supplementation was extended until E10.5, about 15% of the embryos reach E18.5. Acinar cells differentiated properly in rescued Raldh2 / mutants (n = 4) and expressed the exocrine enzyme amylase as expected (Figs 6K and L). Insulin-
ARTICLE IN PRESS 10
M. Martı´n et al. / Developmental Biology xx (2005) xxx – xxx
producing cells were detected as well co-expressing Pdx1, which regulates the maintenance of beta cells identity in mature cells (Ahlgren et al., 1998). To identify the RAresponding cells in rescued mice, we analyzed the expression of RARE-hsp68-LacZ transgene in RA-rescued Raldh2 / E9.5 embryos (n = 2). LacZ expression was clearly detected in the pancreatic endodermal cells and to a lower level in surrounding mesoderm in proximity of Pdx1+ epithelial cells (Fig. 2D). These findings suggest that both mesodermal and endodermal cells receive RA during maternal RA treatment and may act in concert to rescue dorsal pancreas development. In the present study, we show that Raldh2 is required for dorsal, but not ventral pancreas bud formation. Our data also revealed that Raldh2 is the only Raldh gene expressed during pancreas specification and that this expression is restricted to the dorsal pancreatic mesenchyme prior and during early bud formation. By extension, it is likely that the pancreatic mesenchyme is the source of RA needed for proper pancreas development. While this manuscript was in revision, Molotkov et al. (2005) reported that Raldh2 was required for mouse dorsal endodermal development in agreement with our findings. Our data provide further evidence that the dorsal bud defect observed in Raldh2 knockout embryos is not related to an expansion of Shh in the pancreatic endoderm. Moreover, the altered pattern of expression of Isl1 in the dorsal mesenchyme supports a function for RA signaling in the differentiation of the pancreatic mesenchyme. These mesenchymal cells and related signals would then control the expansion of the dorsal pancreatic endoderm and/or induction of Pdx1 expression in dorsal pancreatic progenitors. However, our results do not exclude that the RA synthesized in close proximity of the prepancreatic endodermal cells signals directly to these epithelial cells resulting in Pdx1 induction. Alternatively, both mechanisms might operate simultaneously or sequentially, a hypothesis supported by the fact that both endodermal and mesenchymal cells respond to RA. Recent studies using quail/chick explants support the hypothesis that RA is acting on the endoderm indirectly via the mesoderm. Indeed, in elegant experiments, Kumar et al. (2003) demonstrated that RA can induce, in a mesodermal-dependent fashion, Pdx1 and other pancreas markers in anterior endodermal explants, which otherwise would not adopt a pancreatic fate. However, as suggested by the authors, these data do not completely exclude a possible direct effect of RA on endodermal cells, concomitantly with mesodermal derived signals. Interestingly, it has recently been reported that RA can induce the commitment of mouse ES cells to Pdx1+ endoderm (Micallef et al., 2004). Previous studies have reported an important role of RA signaling in pancreas development in zebrafish, Xenopus and quail (Chen et al., 2004; Kumar et al., 2003; Stafford and Prince, 2002; Stafford et al., 2004), demonstrating that RA-mediated specification of the pancreas is a conserved mechanism from fish to mammals. Despite this conserved
requirement of RA signaling in pancreas formation, some differences in RA function do exist between these vertebrate species. In contrast to the mouse, where only dorsal bud development is affected, Raldh2 inactivation in zebrafish results in a complete lack of the pancreas (Stafford and Prince, 2002). As previously mentioned by others who also observed a differential requirement of RA in dorsal and ventral pancreas, the development of the pancreatic primordia in the zebrafish is not strictly equivalent to the pancreatic bud ontogeny in other vertebrates which could explain this discrepancy (Chen et al., 2004). In zebrafish, Raldh2mediated RA signaling has been suggested to control endoderm patterning and pancreas specification during gastrulation before the onset of Pdx1 expression (Stafford and Prince, 2002). Our studies in mouse rather suggest a later role for RA, but also prior Pdx1 expression, in the induction of the dorsal pancreatic program based on the expression pattern of Raldh2 in the pancreatic mesenchyme at the early bud stage (E8.75 – E9.5) and loss of Pdx1 expression in mutant embryos. However, we cannot completely rule out an earlier function of RA since the dorsal bud agenesis can be rescued by maternal RA treatment between E7.5 and E8.5 corresponding to the end of gastrulation in mouse. Another difference between the fish and mouse is that like in Xenopus, RA is not required for liver specification. RA signaling has also been reported to control endocrine versus exocrine differentiation in Xenopus, with increased RA resulting in expansion of endocrine cell population at the expense of exocrine cells (Chen et al., 2004). Such a conclusion could not be reached in our RA loss of function study in mice where Raldh2deficient embryos die at E10.5 prior to the major wave of endocrine differentiation which starts at E13.5 and prior exocrine differentiation which starts at E14.5 – E15.5 and. However, we observed proper acini formation and exocrine marker expression as well as hormones producing islet cells in E18.5 RA-rescued (E7.5 – E10.5) embryos. Since it is likely that exogenous RA is eliminated rapidly after treatment, this result suggests that, in mouse, Raldh2 is not essential for endocrine and exocrine development when massive differentiation occurs. In conclusion, our findings identified RA as an important pathway controlling dorsal pancreas development in mouse. However, the downstream target genes executing the RAmediated differentiation program are still unclear. Furthermore, the potential crosstalk between RA and the other signaling pathways known to control pancreas development remains to be elucidated.
Materials and methods Animals To collect embryos, mice were mated overnight and midday of the day of vaginal plug detection was considered
ARTICLE IN PRESS M. Martı´n et al. / Developmental Biology xx (2005) xxx – xxx
as E0.5. Embryos were collected in cold phosphatebuffered saline (PBS) and fixed in 4% paraformaldehyde in PBS for 2 h 30 min at 4-C, washed three times for 10 min in cold PBS, equilibrated in 20% sucrose for 1 h 30 min at 4-C, 30% sucrose for 1 h 30 min at 4-C and embedded in OCT. The RARE-hsp68-lacZ reporter transgenic line has been described (Rossant et al., 1991). The generation of Raldh2 / null mutant mice and the RA supplementation protocol have been described previously (Niederreither et al., 1999, 2002). RARE-hsp68-lacZ; Raldh2+/ mice were generated by crossing RAREhsp68-lacZ transgenic males with Raldh2+/ females. Mating of these animals with Raldh2+/ mice resulted in the generation of Raldh2 / embryos heterozygote for the RARE-hsp68-lacZ transgene. A supplementation of 100 Ag RA per gram of food between E7.5 and E8.5 is referred as a minimal rescue. Extended rescue has also been performed, using 250 Ag RA/g food from E8.5 onwards to E9.5 or E10.5. Quantitative analysis Mesenchymal cells quantification and percentage of Isl1 positive cells have been calculated after counting Dapi- and Isl1-positive nuclei in the dorso-lateral mesenchyme of E9.5 embryos (dorsal to the expression of Pdx1 in the ventral pancreas). A total number of 4102 (8 sections) and 3548 (12 sections) dapi- and 2852 (8 sections) and 917 (12 sections) Isl1-positive cells have been counted for Raldh2+/+ and Raldh2 / embryos, respectively; 4 embryos analyzed for each genotype in 3 independent experiments. Total number of HB9-positive cells per section has been counted in the dorsal endoderm of E9.5 embryos. A total number of 404 (8 sections) and 242 (16 sections) cells have been counted for Raldh2+/+ and Raldh2 / ; 4 embryos per genotype in 4 independent experiments. Total number of insulin and glucagon labeled cells per section has been counted on E10.5 rescued embryos. A total number of 458 (14 sections, 3 embryos), 229 (17 sections, 4 embryos) and 182 (6 sections, 2 embryos) cells have been counted for Raldh2+/+ and Raldh2 / minimally and largely rescued, respectively, corresponding to 3 independent experiments. b-galactosidase assays Embryos harvested in cold PBS were fixed in 0.2% glutaraldehyde in 5 mM EGTA, 2 mM MgCl2, 0.1 M sodium phosphate pH 7.3 for 20 min at RT, equilibrated and permeabilized in LacZ washing buffer (2 mM MgCl2 in 0.1 M sodium phosphate pH 7.3, 0.02% NP-40) for 1 h at RT. They were then processed as described previously (Jenny et al., 2002). After X-Gal staining, embryos were dehydrated, embedded in paraffin and cut in 6 Am sections. The pancreatic buds were identified by Pdx1 immunohistochemistry as described below.
11
Immunohistochemistry Immunohistochemistry and immunofluorescence analyses were performed on 10 Am cryosections or 6 Am paraffin sections for the X-Gal stained embryos. Slides were hydrated, treated in blocking buffer (5% goat serum in PBS, 0.1% Triton X-100) for 30 min at RT and incubated with primary antibodies in blocking buffer overnight at 4-C. The following primary antibodies were used: rabbit anti-Pdx1 at 1:2000 (kindly provided by Dr Chris Wright, Vanderbilt University, Nashville, TN), mouse anti-insulin at 1:1000 (Sigma), mouse anti-glucagon at 1:2000 (Sigma), guinea pig anti-glucagon at 1:1000 (Linco), rabbit anti-HB9 at 1:8000 (gift from Dr Kathleen A. Harrison, NIH, Bethesda, MD), mouse anti-Isl1 at 1:100 (mix 40.2D6:39.4D5 (1:1)) (Developmental Studies Hybridoma Bank), mouse anti-BrdU (Boehringer Mannheim) at 1:100 and rabbit anti-a-amylase (Sigma) at 1:1000. After washing for 5 min three times in PBS, 0.1% Triton X-100 (PBST), the secondary antibodies were added for 1 h at RT in blocking buffer. Secondary antibodies used were: Alexa 488 anti-mouse at 1:1000 (Molecular Probes), Alexa 568 anti-rabbit at 1:1000 (Molecular probes), Cy3 anti-rabbit and anti-guinea pig at 1:1000 (Jackson Immunoresearch) and biotin-coupled anti-rabbit at 1:200 (Vector Laboratories). In the immunofluorescence assays, after washing for 5 min three times in PBST, nuclei were stained for 5 min with Dapi at 1:10000 in PBS, washed and mounted in Aqua-poly/mount (Polysciences). For immunohistochemistry analysis, endogenous peroxidase activity was blocked by incubation in 0.5% H2O2 diluted in Methanol. After washing in PBST, the signal was revealed using the Vectastain Elite ABC Kit (Vector Laboratories) and the DAB chromogen (DakoCytomation) supplemented with 0.02% H2O2. Slides were dehydrated and mounted in Eukitt (Euromedex). For BrdU detection assays, BrdU was injected to pregnant females at 50 mg/kg body weight, 16 h prior recovering the embryos. Before incubation with primary antibody, antigen retrieval was performed followed by HCl (2N) treatment in PBS for 1 h at RT. For immunostainings on whole mount embryos, the same protocol was used with the following modifications: Triton X-100 was increased to 0.3% in the blocking buffer and the PBST, and antibody incubations were increased up to 60 h for primary and 16 h for secondary antibodies and AB complex (Vector Laboratories). In situ hybridization RNA ISH experiments on embryo cryosections were performed as described previously (Cau et al., 1997; Gradwohl et al., 2000), using template plasmids cloned in our institute or kindly provided by Dr P. Wellauer (Swiss Institute for Cancer Research, Epalinges, Switzerland; Ptf1a) and Dr A. Mc Mahon (Harvard University, Cambridge, MA; Ihh). The whole-mount in situ hybridization procedure is described at
ARTICLE IN PRESS 12
M. Martı´n et al. / Developmental Biology xx (2005) xxx – xxx
http://www.eumorphia.org/EMPReSS/servlet/EMPReSS. Frameset (gene expression section). In some cases, Pdx1 immunostaining was performed after ISH, as described above.
Acknowledgments We are most grateful to A. Grapin-Botton, M. Hebrok, R. Scharfmann and H. Semb for discussion. We thank M. Messmer and V. Hauer for excellent technical assistance, V. Fraulob for help with embryo collection and analysis and O. Lefe`bvre for advise with imaging. This work was supported by founds from the Institut National de la Sante´ et de la Recherche Me´dicale (INSERM) (Avenir Grant), the JDRF Center for h cell therapy in Europe, the NIH Beta Cell Biology Consortium to GG; funds from CNRS, INSERM, the Ministe`re de la Recherche and Institut Universitaire de France to PD. MM was supported by a postdoctoral fellowship from INSERM.
References Abraham, E.J., Leech, C.A., Lin, J.C., Zulewski, H., Habener, J.F., 2002. Insulinotropic hormone glucagon-like peptide-1 differentiation of human pancreatic islet-derived progenitor cells into insulin-producing cells. Endocrinology 143, 3152 – 3161. Ahlgren, U., Jonsson, J., Edlund, H., 1996. The morphogenesis of the pancreatic mesenchyme is uncoupled from that of the pancreatic epithelium in IPF1/PDX1-deficient mice. Development 122, 1409 – 1416. Ahlgren, U., Pfaff, S.L., Jessell, T.M., Edlund, T., Edlund, H., 1997. Independent requirement for ISL1 in formation of pancreatic mesenchyme and islet cells. Nature 385, 257 – 260. Ahlgren, U., Jonsson, J., Jonsson, L., Simu, K., Edlund, H., 1998. Betacell-specific inactivation of the mouse Ipf1/Pdx1 gene results in loss of the beta-cell phenotype and maturity onset diabetes. Genes Dev. 12, 1763 – 1768. Apelqvist, A., Ahlgren, U., Edlund, H., 1997. Sonic hedgehog directs specialised mesoderm differentiation in the intestine and pancreas. Curr. Biol. 7, 801 – 804. Bhushan, A., Itoh, N., Kato, S., Thiery, J.P., Czernichow, P., Bellusci, S., Scharfmann, R., 2001. Fgf10 is essential for maintaining the proliferative capacity of epithelial progenitor cells during early pancreatic organogenesis. Development 128, 5109 – 5117. Bort, R., Martinez-Barbera, J.P., Beddington, R.S., Zaret, K.S., 2004. Hex homeobox gene-dependent tissue positioning is required for organogenesis of the ventral pancreas. Development 131, 797 – 806. Cau, E., Gradwohl, G., Fode, C., Guillemot, F., 1997. Mash1 activates a cascade of bHLH regulators in olfactory neuron progenitors. Development 124, 1611 – 1621. Chambon, P., 1996. A decade of molecular biology of retinoic acid receptors. FASEB J. 10, 940 – 954. Chen, Y., Pan, F.C., Brandes, N., Afelik, S., Solter, M., Pieler, T., 2004. Retinoic acid signaling is essential for pancreas development and promotes endocrine at the expense of exocrine cell differentiation in Xenopus. Dev. Biol. 271, 144 – 160. Deutsch, G., Jung, J., Zheng, M., Lora, J., Zaret, K.S., 2001. A bipotential precursor population for pancreas and liver within the embryonic endoderm. Development 128, 871 – 881.
Edlund, H., 2002. Pancreatic organogenesis—Developmental mechanisms and implications for therapy. Nat. Rev., Genet. 3, 524 – 532. Esni, F., Johansson, B.R., Radice, G.L., Semb, H., 2001. Dorsal pancreas agenesis in N-cadherin-deficient mice. Dev. Biol. 238, 202 – 212. Gradwohl, G., Dierich, A., LeMeur, M., Guillemot, F., 2000. Neurogenin3 is required for the development of the four endocrine cell lineages of the pancreas. Proc. Natl. Acad. Sci. U. S. A. 97, 1607 – 1611. Gu, G., Dubauskaite, J., 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. Harrison, K.A., Thaler, J., Pfaff, S.L., Gu, H., Kehrl, J.H., 1999. Pancreas dorsal lobe agenesis and abnormal islets of Langerhans in Hlxb9deficient mice. Nat. Genet. 23, 71 – 75. Hart, A., Papadopoulou, S., Edlund, H., 2003. Fgf10 maintains notch activation, stimulates proliferation, and blocks differentiation of pancreatic epithelial cells. Dev. Dyn. 228, 185 – 193. Hebrok, M., Kim, S.K., Melton, D.A., 1998. Notochord repression of endodermal Sonic hedgehog permits pancreas development. Genes Dev. 12, 1705 – 1713. Jenny, M., Uhl, C., Roche, C., Duluc, I., Guillermin, V., Guillemot, F., Jensen, J., Kedinger, M., Gradwohl, G., 2002. Neurogenin3 is differentially required for endocrine cell fate specification in the intestinal and gastric epithelium. EMBO J. 21, 6338 – 6347. Johansson, K.A., Grapin-Botton, A., 2002. Development and diseases of the pancreas. Clin. Genet. 62, 14 – 23. Jonsson, J., Carlsson, L., Edlund, T., Edlund, H., 1994. Insulin-promoterfactor 1 is required for pancreas development in mice. Nature 371, 606 – 609. Jonsson, J., Ahlgren, U., Edlund, T., Edlund, H., 1995. IPF1, a homeodomain protein with a dual function in pancreas development. Int. J. Dev. Biol. 39, 789 – 798. Kawaguchi, Y., Cooper, B., Gannon, M., Ray, M., MacDonald, R.J., Wright, C.V., 2002. The role of the transcriptional regulator Ptf1a in converting intestinal to pancreatic progenitors. Nat. Genet. 32, 128 – 134. Kim, S.K., Hebrok, M., Melton, D.A., 1997. Notochord to endoderm signaling is required for pancreas development. Development 124, 4243 – 4252. Kumar, M., Melton, D., 2003. Pancreas specification: a budding question. Curr. Opin. Genet. Dev. 13, 401 – 407. Kumar, M., Jordan, N., Melton, D., Grapin-Botton, A., 2003. Signals from lateral plate mesoderm instruct endoderm toward a pancreatic fate. Dev. Biol. 259, 109 – 122. Lai, L., Bohnsack, B.L., Niederreither, K., Hirschi, K.K., 2003. Retinoic acid regulates endothelial cell proliferation during vasculogenesis. Development 130, 6465 – 6474. Lammert, E., Cleaver, O., Melton, D., 2001. Induction of pancreatic differentiation by signals from blood vessels. Science 294, 564 – 567. Li, H., Arber, S., Jessell, T.M., Edlund, H., 1999. Selective agenesis of the dorsal pancreas in mice lacking homeobox gene Hlxb9. Nat. Genet. 23, 67 – 70. Micallef, S.J., Janes, M.E., Knezevic, K., Davis, R.P., Elefanty, A.G., Stanley, E.G., 2004. Retinoic acid induces Pdx1-positive endoderm in differentiating mouse embryonic stem cells. Diabetes. Molotkov, A., Molotkova, N., Duester, G., 2005. Retinoic acid generated by Raldh2 in mesoderm is required for mouse dorsal endodermal pancreas development. Dev. Dyn. 232, 950 – 957. Niederreither, K., McCaffery, P., Drager, U.C., Chambon, P., Dolle, P., 1997. Restricted expression and retinoic acid-induced downregulation of the retinaldehyde dehydrogenase type 2 (RALDH-2) gene during mouse development. Mech. Dev. 62, 67 – 78. Niederreither, K., Subbarayan, V., Dolle, P., Chambon, P., 1999. Embryonic retinoic acid synthesis is essential for early mouse post-implantation development. Nat. Genet. 21, 444 – 448. Niederreither, K., Vermot, J., Schuhbaur, B., Chambon, P., Dolle, P., 2000. Retinoic acid synthesis and hindbrain patterning in the mouse embryo. Development 127, 75 – 85.
ARTICLE IN PRESS M. Martı´n et al. / Developmental Biology xx (2005) xxx – xxx Niederreither, K., Vermot, J., Messaddeq, N., Schuhbaur, B., Chambon, P., Dolle, P., 2001. Embryonic retinoic acid synthesis is essential for heart morphogenesis in the mouse. Development 128, 1019 – 1031. Niederreither, K., Vermot, J., Fraulob, V., Chambon, P., Dolle, P., 2002. Retinaldehyde dehydrogenase 2 (RALDH2)-independent patterns of retinoic acid synthesis in the mouse embryo. Proc. Natl. Acad. Sci. U. S. A. 99, 16111 – 16116. Niederreither, K., Vermot, J., Le, R.I., Schuhbaur, B., Chambon, P., Dolle, P., 2003. The regional pattern of retinoic acid synthesis by RALDH2 is essential for the development of posterior pharyngeal arches and the enteric nervous system. Development 130, 2525 – 2534. Norgaard, G.A., Jensen, J.N., Jensen, J., 2003. FGF10 signaling maintains the pancreatic progenitor cell state revealing a novel role of Notch in organ development. Dev. Biol. 264, 323 – 338. Offield, M.F., Jetton, T.L., Labosky, P.A., Ray, M., Stein, R.W., Magnuson, M.A., Hogan, B.L., Wright, C.V., 1996. PDX-1 is required for pancreatic outgrowth and differentiation of the rostral duodenum. Development 122, 983 – 995. Ross, S.A., McCaffery, P.J., Drager, U.C., De Luca, L.M., 2000. Retinoids in embryonal development. Physiol. Rev. 80, 1021 – 1054. Rossant, J., Zirngibl, R., Cado, D., Shago, M., Giguere, V., 1991.
13
Expression of a retinoic acid response element-hsplacZ transgene defines specific domains of transcriptional activity during mouse embryogenesis. Genes Dev. 5, 1333 – 1344. Stafford, D., Prince, V.E., 2002. Retinoic acid signaling is required for a critical early step in zebrafish pancreatic development. Curr. Biol. 12, 1215 – 1220. Stafford, D., Hornbruch, A., Mueller, P.R., Prince, V.E., 2004. A conserved role for retinoid signaling in vertebrate pancreas development. Dev. Genes Evol. 214, 432 – 441. Tulachan, S.S., Doi, R., Kawaguchi, Y., Tsuji, S., Nakajima, S., Masui, T., Koizumi, M., Toyoda, E., Mori, T., Ito, D., Kami, K., Fujimoto, K., Imamura, M., 2003. All-trans retinoic acid induces differentiation of ducts and endocrine cells by mesenchymal/epithelial interactions in embryonic pancreas. Diabetes 52, 76 – 84. Yashiro, K., Zhao, X., Uehara, M., Yamashita, K., Nishijima, M., Nishino, J., Saijoh, Y., Sakai, Y., Hamada, H., 2004. Regulation of retinoic acid distribution is required for proximodistal patterning and outgrowth of the developing mouse limb. Dev. Cell 6, 411 – 422. Yoshitomi, H., Zaret, K.S., 2004. Endothelial cell interactions initiate dorsal pancreas development by selectively inducing the transcription factor Ptf1a. Development 131, 807 – 817.
