Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Dev Cell. Author manuscript; available in PMC 2009 November 1.
Published in final edited form as:
PMCID: PMC2610857

Bmp2 signaling regulates the hepatic vs pancreatic fate decision


Explant culture data have suggested that the liver and pancreas originate from common progenitors. We used single cell lineage tracing in zebrafish to investigate this question in vivo as well as to analyze the hepatic vs pancreatic fate decision. At early somite stages, endodermal cells located at least two cells away from the midline can give rise to both liver and pancreas. In contrast, endodermal cells closer to the midline give rise to pancreas and intestine, but not liver. Loss- and gain-of-function analyses show that Bmp2b, expressed in the lateral plate mesoderm, signals through Alk8 to induce endodermal cells to become liver. When Bmp2b was overexpressed, medially located endodermal cells, fated to become pancreas and intestine, contributed to the liver. These data provide in vivo evidence for the existence of bipotential hepatopancreatic progenitors, and indicate that their fate is regulated by the medio-lateral patterning of the endodermal sheet, a process controlled by Bmp2b.


Understanding how distinct cell types arise from common progenitors is a major quest in developmental biology. Such information will have broad implications not only in terms of basic knowledge but also for cellular therapy (Gouon-Evans et al., 2006).

Previous work has suggested that two distinct endodermal organs, the liver and the ventral pancreas, share a common developmental origin (Deutsch et al., 2001; Rossi et al., 2001). Using a mouse embryonic tissue explant system, Deutsch et al. (2001) and Rossi et al. (2001) showed that Fibroblast growth factors (Fgfs) from the cardiac mesoderm and Bone morphogenetic proteins (Bmps) from the septum transversum mesenchyme (STM) initiate the hepatic induction process in the adjacent endoderm. Interestingly, when the hepatic induction process was prevented in this explant system by blocking Bmp and Fgf signaling, the cultured endoderm turned on the expression of Pdx1 (Deutsch et al., 2001; Rossi et al., 2001), a gene normally expressed in the foregut endoderm, including the stomach, duodenum and pancreas, but not in the liver (Offield et al., 1996; Fujitani et al., 2006). Further incubation of these Pdx1 expressing endodermal explants led to the appearance of pancreatic endocrine and exocrine cells (Deutsch et al., 2001). These data led to the proposal that the ventral endoderm has the potential to give rise to multiple tissues, including the liver and pancreas. However, the existence, location and fate of these presumptive bipotential hepatopancreatic progenitors have not been investigated in vivo. We took advantage of the ability to undertake single-cell lineage studies in zebrafish to address these questions as well as the identity and origin of the signals that determine the fate of these progenitors.

In zebrafish, the endoderm forms two converging sheets of cells by the end of gastrulation (10 hours post fertilization (hpf)) (Ober et al., 2003; Chung and Stainier, 2008). Subsequently, these cells condense at the midline to form a rod-like structure by 24 hpf (Ober et al., 2003). The patterning and regionalization of the endodermal sheet are evident from early stages. For example, in the region between somites 1 and 4, high levels of pdx1 expression can be detected in the most medial cells of the bilateral sheets at the 10-somite stage (Biemar et al., 2001), and these cells will form the dorsal pancreatic bud derived endocrine cells (Chung and Stainier, 2008). In addition, low levels of pdx1 expression can be detected at 18 hpf in laterally located endodermal cells which presumably give rise to the intestine and ventral pancreas (Roy et al., 2001; Field et al., 2003). In contrast to the early expression of pdx1, hepatic specification has been thought to occur at approximately 22 hpf when the expression of the transcription factor genes hhex and prox1 can be detected in a condensed endodermal cluster anterior to the pdx1 expressing cells (Ober et al., 2006; Shin et al., 2007).

Previous data have shown that Wnt2bb expressed in the lateral plate mesoderm (LPM) is required for an early step in liver formation in zebrafish (Ober et al., 2006). In addition, blocking Bmp or Fgf signaling, by overexpressing dominant negative receptors, from 18 hpf strongly reduces hhex and prox1 expression in the liver forming region (Shin et al., 2007). However, the identity and relevant expression pattern of the Bmp and Fgf ligands involved in hepatic induction are unknown.

Here, by labeling single cells within the zebrafish endodermal sheet at early post-gastrulation stages, we identified cells that contributed to both the liver and ventral pancreas. Interestingly, these cells were located close to the LPM and at least two cells away from the midline, while most of the medial endodermal cells at the same A-P level gave rise to pancreas and intestine. We also found that bmp2b is expressed in a very discrete region of the LPM adjacent to the liver progenitors starting at the 10-somite stage. Loss- and gain-of-function analyses indicate that Bmp2b is essential for the medio-lateral (M-L) patterning of the endodermal sheet, and consequently liver induction.


Lineage Tracing Reveals the Existence of Bipotential Hepatopancreatic Progenitors

In order to test for the existence of bipotential hepatopancreatic progenitors in vivo, we labeled single endodermal cells at early post-gastrulation stages (Table 1). Tg(sox17:GFP)s870 embryos were injected at the one-cell stage with the photoactivatable lineage tracer DMNB-caged fluorescein dextran conjugate (Keegan et al., 2005; Vogeli et al., 2006), and single endodermal cells were uncaged using a 405nm laser at the 6–8 somite stage. We marked endodermal cells at 4 different A-P levels (somites 1, 2, 3 and 4) since our previous results indicated that this region comprises liver and pancreas progenitors (Chung and Stainier, 2008), and at 3 different medio-lateral (M-L) positions: the most medial cells (medial), cells immediately adjacent to the medial cells (lateral 1) and cells one cell away from the medial cells (lateral 2). After uncaging, we fixed control embryos at the 10-somite stage and experimental embryos at 50 hpf. Control embryos showed efficient and specific labeling of single endodermal cells at medial (Figure 1A), lateral 1 (Figure 1B) and lateral 2 (Figure 1C) positions, as well as specific somite levels (data not shown). Consistent with earlier lineage tracing data (Chung and Stainier, 2008), we found that medial endodermal cells at the 6–8 somite stage gave rise mostly to pancreatic endocrine cells at 50 hpf (Figures 1D and 1G; n= 17, and the percentage of embryos showing incorporation into endocrine cells: 100%, and intestine: 29%), suggesting an early fate restriction of these cells. In contrast, the fate of the lateral 1 cells was variable as these cells gave rise to pancreatic exocrine cells and intestinal cells, as well as a small number of pancreatic endocrine cells (Figures 1E and 1H; n= 29, and the percentage of embryos showing incorporation into endocrine cells: 100%, exocrine cells: 86%, intestine: 89%, and liver: 10%). It should be noted that most of the descendants of the lateral 1 cells populate the pdx1 positive organs, i.e. the pancreas and intestine (Field et al., 2003). Most interestingly, we found that single endodermal cells in the lateral 2 position between somites 1 and 3 gave rise not only to pancreatic exocrine cells and intestinal cells, but also to liver cells (Figures 1F and 1I; n=18, and the percentage of embryos showing incorporation into endocrine cells: 11%, exocrine cells: 100%, intestine: 83%, and liver: 89%). These data indicate that endodermal cells at the same AP level can contribute to different organs depending on their M-L position (Figure 1J), and furthermore that lateral 2 cells can give rise to liver as well as pdx1 positive organs including the ventral pancreas and intestine (Figure 1K).

Figure 1
Lineage tracing evidence for hepatopancreatic progenitors
Table 1
Lineage tracing experiments reveal the existence of bipotential hepatopancreatic progenitors.

bmp2b Is Expressed in the Lateral Plate Mesoderm at Early Stages

Since Bmp signaling is essential for hepatic induction in mouse (Deutsch et al., 2001; Rossi et al., 2001), chick (Zhang et al., 2004) and zebrafish (Shin et al., 2007), we examined the expression pattern of several bmp’s at the time of hepatic induction and earlier.

bmp4 is mainly expressed in the cardiac mesoderm at 21 hpf (Figure S1A), and bmp7 is expressed in the head region and otic vesicles at 24 hpf (Figure S1B). In contrast, bmp2a and bmp2b are expressed in the LPM adjacent to the liver forming region at 24 hpf (Figures S1C and S1D), a pattern that resembles that of wnt2bb, which is essential for an early step in liver development (Ober et al., 2006). bmp2a does not appear to be expressed until 24 hpf but bmp2b expression appears earlier and is highly dynamic. At 75% epiboly, bmp2b is highly expressed in the ventral domain of the embryo (Figure S1E) consistent with its role in regulating ventral fate during gastrulation stages (Kishimoto et al., 1997; Nguyen et al., 1998; Tucker et al., 2008). At the 5-somite stage, bmp2b is expressed mainly in the epidermis (Figure S1F) (Kaji and Artinger, 2004). Importantly, we observed bmp2b expression in the LPM starting from the 10-somite stage (Figures 2A and 2B); an initially broad expression pattern becomes more discrete by the 14-somite stage (Figures 2C and 2D). By performing double in-situs with bmp2b and myod1, which marks the forming somites, we found that bmp2b is expressed in the LPM adjacent to somites 1 through 3 (Figure 2E), a region where liver progenitors are located (Figure 1). Notably, lateral 2 endodermal cells are located close to the LPM at the 10-somite stage (Figure 2F).

Figure 2
bmp2b is expressed in the lateral plate mesoderm from an early stage

Since our lineage tracing data showed that lateral 2 endodermal cells give rise to the liver as well as pdx1 positive tissues, we sought to determine the initial time point of pdx1 expression in the intestine and ventral pancreas progenitors. It has been previously shown that pdx1 expression in the dorsal pancreatic progenitors first appears around the 10-somite stage (Biemar et al., 2001), but the time point of initial pdx1 expression in the other progenitors has not been defined. To investigate this issue, we analyzed pdx1 expression in smoothened (smo) mutants since the dorsal pancreatic bud derived endocrine cells fail to form in these embryos (Roy et al., 2001; diIorio et al., 2002; Chung and Stainier, 2008). Low level pdx1 expression, characteristic of intestine and ventral pancreas progenitors, is clearly present in smo mutants from 18 hpf onwards (Roy et al., 2001; Chung and Stainier, 2008). Further analyses showed that low level pdx1 expression is present in smo mutants as early as the 10-somite stage, while high level pdx1expression is completely missing (Figures 2G and 2H). We further tested whether this pdx1 ‘gradient’ correlates with the initial M-L position of cells in the endodermal sheet. After labeling a single endodermal cell at medial, lateral 1 and lateral 2 positions at the 6–8 somite stage, we fixed embryos at the 14-somite stage and analyzed pdx1 expression (Figure 3). The most medial cells, which give rise mainly to pancreatic endocrine cells, express high levels of pdx1 (Figures 3A––3A″), whereas lateral 1 cells, which give rise mainly to intestine and ventral pancreas, express low levels of pdx1 (Figures 3B––3B″). Some lateral 2 cells, which populate the liver, exocrine pancreas and intestine, did not express pdx1 (Figures 3C––3C″) leading to the hypothesis that the subset of their descendants which move closer to the source of Bmp signals keep pdx1 expression off and contribute to the liver, while those which move away from the Bmp signals turn on pdx1 expression and contribute to the exocrine pancreas and intestine. This hypothesis is further supported by lineage tracing analyses in mouse showing that liver cells and their progenitors never express Pdx1 (Gu et al., 2002; Fujitani et al., 2006). Altogether, these data suggest a model whereby all L1 and some L2 endodermal cells start expressing low levels of pdx1 at the 10-somite stage; at the same stage, bmp2b signaling from the LPM downregulates, or blocks, pdx1 expression in a subset of L2 cells, and these cells will give rise to the liver.

Figure 3
The pdx1 ‘gradient’ correlates with the medio-lateral position of endodermal cells

A Decrease in Bmp Signaling Leads to an Increase in pdx1 Expressing Cells

In order to test the role of bmp2b in directing cells towards the liver or ventral pancreas, we first analyzed the differentiation of these cells in laf/alk8 mutants, as several Bmp ligands, including Bmp2b, signal through Alk8 (Bauer et al., 2001; Mintzer et al., 2001). Gut morphology appears to be unaffected in alk8 mutants as assessed by the expression of pan-endodermal markers such as foxa3 (Shin et al., 2007). However, hhex expression in alk8 mutants is downregulated in the liver region (black arrows in Figures 4A and 4B; Shin et al., 2007), while it appears unaffected in the dorsal pancreatic bud (white arrows in Figures 4A and 4B). In contrast, pdx1 expression appears to be expanded into the liver region (Figures 4C and 4D). This expansion of pdx1 expression was more evident when embryos were stained with Prox1 and Pdx1 antibodies (Figures 4E to 4F′). Compared to wild-type, alk8 mutants show a reduction of the Prox1 expression domain (Figures 4E′ and 4F′), and an expansion of the Pdx1 expression domain (compare areas outlined by yellow dashed lines in Figures 4E′ and 4F′). BrdU pulse-chase experiments indicated that this expansion of the Pdx1 domain in alk8 mutants was not due to increased proliferation (data not shown). Next, we examined the initial pdx1 ‘gradient’ in the endodermal sheet in alk8 mutants as well as in the Tg(hsp70l:dnBmpr-GFP)w30 line, which has been shown to block Bmp signaling effectively (Pyati et al., 2005; Shin et al., 2007). Compared to their wild-type siblings, at 18 hpf, alk8 mutants show a dramatic lateral expansion of pdx1 expression (Figures 4G and 4H). Similarly, when Tg(hsp70l:dnBmpr-GFP)w30 embryos were heat-shocked at the 6-somite stage, a time before the initiation of endogenous bmp2b and pdx1 expression, pdx1 expression expanded laterally at 18 hpf (Figures S2A and S2B). Altogether, these data indicate that in embryos with reduced Bmp signaling, more endodermal cells exhibit pdx1 expression than in wild-type.

Figure 4
Decrease in Bmp signaling leads to a decrease in liver tissue and increase in pdx1 expressing cells

Next, we injected bmp2b MO to specifically knock down this gene. Injection of 210 pg, or more, led to dorsalization of the embryos consistent with published data (Imai and Talbot, 2001). When we injected lower doses (100–150 pg) of bmp2b MO, hhex expression was slightly reduced in the liver region (data not shown). This reduction in hhex expression was more pronounced when we injected the same amount of bmp2b MO into alk8 heterozygous embryos (Figures 4I and 4J). These data, together with the analysis of alk8 mutants, indicate that Bmp2b, signaling through Alk8, regulates hepatic specification, and that reducing Bmp2b signaling leads to an expansion of the pdx1 expression domain.

Bmp2b Overexpression Causes a Pancreatic to Liver Fate Switch

We next examined the liver and pdx1 expression domains in embryos with increased bmp2b expression. To overexpress bmp2b at specific time points, we utilized the Tg(hsp70l:bmp2b)f13 line where bmp2b can be induced by heat-shock treatment (Chocron et al., 2007). Since pdx1 expression in the lateral endodermal cells and bmp2b expression in the LPM appear at the 10-somite stage, Tg(hsp70l:bmp2b)f13 embryos were heat shocked at the 8- and 14-somite stages, i.e. before and after the initiation of endogenous bmp2b expression. Embryos were heat-shocked for 15 minutes and fixed at 44 hpf to examine hhex and pdx1 expression. At this stage in wild-type, hhex is expressed in the liver and dorsal pancreatic bud (Figure 5A), while pdx1 is expressed in a broad region of the foregut endoderm, including the dorsal and ventral pancreas as well as the intestine, but not in the liver (Figure 5D). When bmp2b expression was induced at the 8-somite stage, hhex expression at 44 hpf was greatly expanded and occupied most of the endoderm anterior to the dorsal pancreatic bud (Figure 5B). In these embryos, pdx1 expression was severely reduced (Figure 5E) while its expression in pancreatic endocrine cells appeared unaffected (Figure 5E: area outlined by white dashed line). Consistently, when we examined pdx1 expression at 18 hpf after inducing bmp2b at the 8-somite stage, medial-most cells, as their counterpart in control embryos (Figure S2C), exhibited high levels (Figure S2D), whereas lateral cells exhibited much reduced levels compared to wild-type. However, when bmp2b expression was induced at the 14-somite stage, hhex expression was only slightly expanded (Figure 5C) and pdx1 expression appeared unaffected (Figure 5F), although the morphology of the gut tube was disrupted. To investigate whether the ventral pancreas gets specified in the bmp2b overexpressing embryos, we examined the expression of Tg(ptf1a:GFP)jh1 which marks the developing ventral pancreas (Godinho et al., 2005), along with Prox1 and Islet1 to mark the liver and pancreatic endocrine cells, respectively. In control embryos at 50 hpf, Prox1 is highly expressed in the liver and developing ventral pancreas, and weakly expressed in the hepatopancreatic ductal primordium (Figure 5G). However, the ventral pancreas can be easily distinguished because Tg(ptf1a:GFP)jh1 is only expressed in this tissue (Figure 5G). Also, at this stage, the ventral pancreas has engulfed the dorsal pancreatic bud which can be marked by Islet1 expression (Figure 5G). Compared to control embryos, we found that in the embryos where bmp2b expression was induced at the 8-somite stage, the domain of high level Prox1 expression was dramatically expanded in the endoderm anterior to the pancreatic endocrine cells (Figure 5H). In these embryos, Tg(ptf1a:GFP)jh1 expression was completely absent, indicating that all the Prox1 expressing cells are liver cells (Figure 5H). This complete absence of the ventral pancreas is likely due to the severe reduction of pdx1 expression in the developing gut (Figures 5E and S2D). In sharp contrast, when bmp2b expression was induced at the 14-somite stage, Prox1 expression was only slightly expanded and Tg(ptf1a:GFP)jh1 expressing cells were clearly present (Figure 5I), consistent with the fact that pdx1 expression was maintained in these embryos (Figure 5F). Results from inducing bmp2b expression at additional stages (6-, 10- and 18-somite; data not shown) indicate that the ability of bmp2b to induce hepatic specification and suppress pdx1 expression is most efficient around the 10-somite stage. These results thus suggest that after the initiation of bmp2b expression in the LPM and the patterning of the endodermal sheet in the M-L axis, increasing Bmp signaling cannot affect this patterning, while it can disrupt gut morphogenesis.

Figure 5
Bmp2b overexpression causes a pancreatic to liver fate switch

Our analyses in wild-type, mutant and transgenic embryos show that liver genes and pdx1 exhibit complementary expression patterns, leading us to hypothesize that bmp2b regulates the liver versus pdx1 expressing fate decision by affecting their specification rather than by affecting cell proliferation and/or survival. To directly test this hypothesis, we performed lineage tracing experiments in the Tg(hsp70l:bmp2b)f13 line to examine possible cell fate changes in vivo.

Tg(hsp70l:bmp2b)f13;Tg(sox17:GFP)s870 embryos were injected with the photoactivatable DMNB-caged fluorescein dextran conjugate at the one-cell stage. At the 6–8 somite stage, a single lateral 1 cell at the level of somite 2 was uncaged (Figure 5J). Following uncaging, the embryos were heat-shocked at the 8-somite stage. As expected from earlier data (Figures 1E and 1H), in control embryos, most of the lateral 1 cells at the level of somite 2 gave rise to intestine and pancreas, but not liver (Figure 5K; 90.9%, n = 10/11). In embryos where bmp2b expression was induced at the 8-somite stage, lateral 1 cells at the level of somite 2 contributed to the liver (Figure 5L; 87.5%, n = 7/8). Altogether, these data show that increased Bmp2b signaling caused ventral pancreas and intestine progenitors to become liver cells, further indicating the importance of Bmp2b signaling in the M-L patterning of the endoderm sheet as well as in hepatic fate decision.


Our data provide in vivo evidence that some endodermal cells can give rise to both the liver and ventral pancreas. By single cell lineage tracing, we show that these common progenitors are located at least two cells away from the midline between somites 1 and 3 when the endoderm consists of a monolayer. Furthermore, our data indicate that bmp2b, which is expressed in the LPM from the 10-somite stage onwards, patterns the endodermal sheet in the M-L axis, such that the descendants of the lateral 2 cells, which are closest to the Bmp signal, differentiate into liver, while the descendants of the more medial cells (medial and lateral 1 cells), which are further away from the Bmp signal, contribute to the pdx1 expression domain which gives rise to the pancreas and intestine.

Previous endodermal fate maps in various organisms have indicated that the liver arises from paired lateral regions (Rosenquist, 1971; Warga and Nüsslein-Volhard, 1999; Chalmers and Slack, 2000; Tremblay and Zaret, 2005; Ward et al., 2007) as well as a small anterior population at the ventral midline in mouse (Tremblay and Zaret, 2005). In zebrafish, we find that at the 6–8 somite stage liver progenitors are located in discrete bilateral domains long before the expression of the known liver markers, hhex and prox1. These lateral cells also contribute to non-liver tissues including ventral pancreas and intestine, suggesting that their fate has not yet been determined, and that the signals responsible for fate specification function after this stage. In contrast, within the region between somites 1 and 4, many (12<17) of the medial endodermal cells labeled at the 6–8 somite stage give rise exclusively to pancreatic endocrine cells suggesting that this lineage is specified earlier.

Previous data have suggested that Bmp signaling regulates hepatic specification as well as liver bud formation (Deutsch et al., 2001; Rossi et al., 2001; Zhang et al., 2004). Rossi et al. (2001) showed that Bmp4 knock-out mouse embryos exhibit a severe delay in liver bud formation, providing in vivo evidence for a role for Bmp signaling in this process. However, although several protocols have used Bmp ligands to drive the differentiation of endodermal cells into hepatocytes in culture (Lavon and Benvenisty, 2005; Gouon-Evans et al., 2006), genetic evidence supporting a role for Bmp signaling in hepatic specification has been lacking. In a genetic screen for mutations affecting endodermal organ formation (Ober et al., 2006), we identified the Bmp receptor gene alk8 as playing a positive role in hepatic induction. Because Alk8 function is maternally provided, we subsequently used a dominant negative Bmp receptor to further analyze the role of this signaling pathway in hepatic induction (Shin et al., 2007). Data from these analyses indicated that Bmp signaling was essential for hepatic specification but the identity and relevant expression pattern of the inducer(s) was not determined (Shin et al., 2007).

In an effort to identify the Bmp ligand(s) that bind(s) to Alk8 to regulate hepatic specification, we found that bmp2b is specifically expressed in the LPM between somites 1 and 3 from the 10-somite stage onwards. Knocking down bmp2b function in a sensitized background led to a reduction in liver gene expression and a concomitant increase in pancreatic gene (i.e. pdx1) expression. Conversely, bmp2b overexpression at early somite stages caused a dramatic expansion of liver gene expression and a severe reduction in pdx1 expression. In addition, single cell lineage tracing in these experimental embryos showed that bmp2b overexpression could redirect pancreatic and intestinal progenitors to a liver fate, further indicating the important role of Bmp signaling in determining the fate of endodermal cells towards liver or pdx1 expressing tissues.

It has been previously suggested that in zebrafish gastrulae, the Bmp2b signaling gradient patterns the endodermal cells in the dorso-ventral axis, thereby regulating their anteroposterior identity at later stages (Tiso et al., 2002). Together with our data, these results suggest that Bmp signaling affects the patterning of endodermal cells at multiple times.

Several possibilities can account for the ability of Bmp2b at the 10-somite stage (14 hpf) to regulate liver induction. First, Bmp2b may be involved in keeping cells competent to become liver by suppressing pdx1 expression. In this scenario, endodermal cells near the source of Bmp2b maintain their potential, and are competent to respond to other hepatic induction signals, such as Wnt2bb, which is subsequently expressed in the LPM (Ober et al., 2006). Alternatively, Bmp2b may actively specify liver progenitors as early as the 10-somite stage by regulating unknown target genes which could act upstream of hhex and prox1.

Experimental Procedures

Zebrafish Strains

Embryos and adult fish were raised and maintained under standard laboratory conditions. We used the following mutant and transgenic lines: smohi1640 (Chen et al., 2001), laf/alk8tm110 (Bauer et al., 2001), Tg(sox17:GFP)s870 (Chung and Stainier, 2008), Tg(ptf1a:GFP)jh1 (Godinho et al., 2005), Tg(hsp70l:bmp2b)f13 (Chocron et al., 2007), Tg(hsp70l:dnBmpr-GFP)w30 (Pyati et al., 2005) and Tg(gutGFP)s854 (Field et al., 2003).

Lineage tracing

Tg(sox17:GFP)s870 embryos were injected with 3 nl of 2% photoactivatable DMNB-caged fluorescein dextran conjugate as a lineage tracer (10 kDa; Molecular Probes) in 0.2 M KCl and allowed to develop until the 6-somite stage (12 hpf). After dechorionation, embryos were mounted dorsally in the transplantation mold filled with egg water. Using a Nikon C1si spectral confocal microscope, we visualized the endodermal sheet in live embryos at the 6–8 somite stage and the A-P position of endodermal cells was determined by counting somites. Caged-fluorescein was activated in a single endodermal cell in each embryo with a 405 nm laser focused through a 40X objective lens. The embryos were fixed at 14 or 50 hpf with 4% PFA and stained with antibodies against GFP, the uncaged-fluorescein and Insulin or Prox1.

Injection of bmp2b MO and genotyping alk8 mutants

100–150 pg of bmp2b MO (Imai and Talbot, 2001) was injected into one-cell stage embryos. Genotyping of the alk8tm110 was carried out as previously described (Bauer et al., 2001).

Heat-shock conditions

Tg(hsp70l:bmp2b)f13 and Tg(hsp70l:dnBmpr-GFP)w30 embryos were heat shocked at various stages by transferring them into egg water pre-warmed at 37 and 40º C, respectively. After a 15~30 min heat shock in the 37 or 40º C incubator, the plate containing the embryos was transferred into a 28 º C incubator, and the embryos were harvested at various stages.

In situ hybridization and histochemical methods

Whole-mount in situ hybridization was performed as previously described (Alexander et al., 1998), using the following probes: pdx1 (Biemar et al., 2001), bmp2b (Nikaido et al., 1997) and hhex (Ho et al., 1999). Two-color in situ hybridization was performed as previously described (Hauptmann and Gerster, 2000), using FITC-labeled myod1 (Coutelle et al., 2001) and DIG-labeled bmp2b. To detect pdx1 expression after uncaging endodermal cells, we fluorescently labeled pdx1 mRNA expression using Fast Red (Roche; Novak and Ribera, 2003).

We used the following antibodies: polyclonal chicken anti-GFP (1:1000; Aves Labs, Inc.), monoclonal mouse anti-GFP (1:1000; Molecular Probes), monoclonal mouse anti-Islet1 (1:15; Developmental Studies Hybridoma Bank, clone 39.4D5), polyclonal rabbit anti-Fibronectin (1:200; Sigma), monoclonal mouse IgG anti-Zona Occludin-1 (1:200; BD Transduction Laboratory), polyclonal guinea pig anti-Pdx1 (1:500; gift from C. Wright), polyclonal guinea pig anti-Insulin (1:500; Biomeda), polyclonal rabbit anti-Prox1 (1:1000; Chemicon), polyclonal rabbit anti-fluorescein (1:100; Molecular Probes), polyclonal goat anti-fluorescein (1:100; Molecular Probes) and fluorescently conjugated Alexa antibodies from Molecular Probes. Nuclei were visualized with TOPRO (1:10000; Molecular Probes). Embryos were fixed with 3% formaldehyde in 0.1M Pipes, 1.0mM MgSO4, 2mM EGTA overnight at 4°C. The yolk was manually removed and embryos were blocked for one hour in PBS with 4% BSA and 0.3% Triton X-100. Primary and secondary antibodies were incubated overnight. For transverse sections, whole mount embryos stained with antibodies were embedded in 4% low melting agarose and sectioned on a Leica VT1000S vibratome into 100 μm thick slices. The whole mount embryos and sections were mounted in Vectashield (Vector Laboratories) and imaged with a Zeiss Lumar fluorescent stereomicroscope and LSM5 Pascal confocal microscope.

Supplementary Material



Competing Interests statement The authors declare no competing financial interests.

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


  • Alexander J, Stainier DY, Yelon D. Screening mosaic F1 females for mutations affecting zebrafish heart induction and patterning. Dev Genet. 1998;22:288–299. [PubMed]
  • Bauer H, Lele Z, Rauch GJ, Geisler R, Hammerschmidt M. The type I serine/threonine kinase receptor Alk8/Lost-a-fin is required for Bmp2b/7 signal transduction during dorsoventral patterning of the zebrafish embryo. Development. 2001;128:849–858. [PubMed]
  • Biemar F, Argenton F, Schmidtke R, Epperlein S, Peers B, Driever W. Pancreas development in zebrafish: early dispersed appearance of endocrine hormone expressing cells and their convergence to form the definitive islet. Dev Biol. 2001;230:189–203. [PubMed]
  • Chalmers AD, Slack JM. The Xenopus tadpole gut: fate maps and morphogenetic movements. Development. 2000;127:381–392. [PubMed]
  • Chocron S, Verhoeven MC, Rentzsch F, Hammerschmidt M, Bakkers J. Zebrafish Bmp4 regulates left-right asymmetry at two distinct developmental time points. Dev Biol. 2007;305:577–588. [PubMed]
  • Chung W, Stainier DY. Intra-endodermal interactions are required for pancreatic β-cell induction. Dev Cell. 2008;14:582–593. [PMC free article] [PubMed]
  • Coutelle O, Blagden CS, Hampson R, Halai C, Rigby PW, Hughes SM. Hedgehog signalling is required for maintenance of myf5 and myoD expression and timely terminal differentiation in zebrafish adaxial myogenesis. Dev Biol. 2001;236:136–150. [PubMed]
  • Deutsch G, Jung J, Zheng M, Lora J, Zaret KS. A bipotential precursor population for pancreas and liver within the embryonic endoderm. Development. 2001;128:871–881. [PubMed]
  • diIorio PJ, Moss JB, Sbrogna JL, Karlstrom RO, Moss LG. Sonic hedgehog is required early in pancreatic islet development. Dev Biol. 2002;244:75–84. [PubMed]
  • Field HA, Dong PD, Beis D, Stainier DY. Formation of the digestive system in zebrafish. II. Pancreas morphogenesis. Dev Biol. 2003;261:197–208. [PubMed]
  • Fujitani Y, Fujitani S, Boyer DF, Gannon M, Kawaguchi Y, Ray M, Shiota M, Stein RW, Magnuson MA, Wright CV. Targeted deletion of a cis-regulatory region reveals differential gene dosage requirements for Pdx1 in foregut organ differentiation and pancreas formation. Genes Dev. 2006;20:253–266. [PubMed]
  • Godinho L, Mumm JS, Williams PR, Schroeter EH, Koerber A, Park SW, Leach SD, Wong RO. Targeting of amacrine cell neurites to appropriate synaptic laminae in the developing zebrafish retina. Development. 2005;132:5069–5079. [PubMed]
  • Gouon-Evans V, Boussemart L, Gadue P, Nierhoff D, Koehler CI, Kubo A, Shafritz DA, Keller G. BMP-4 is required for hepatic specification of mouse embryonic stem cell-derived definitive endoderm. Nat Biotechnol. 2006;24:1402–1411. [PubMed]
  • Gu G, Dubauskaite J, Melton DA. Direct evidence for the pancreatic lineage: Ngn3+ cells are islet progenitors and are distinct from duct progenitors. Development. 2002;129:2447–2457. [PubMed]
  • Hauptmann G, Gerster T. Method in Molecular Biology. Vol. 137. Humana Press; 2000. Multicolor Whole-Mount In Situ Hybridization; pp. 139–148. [PubMed]
  • Imai Y, Talbot WS. Morpholino phenocopies of the bmp2b/swirl and bmp7/snailhouse mutations. Genesis. 2001;30:160–163. [PubMed]
  • Kaji T, Artinger KB. dlx3b and dlx4b function in the development of Rohon-Beard sensory neurons and trigeminal placode in the zebrafish neurula. Dev Biol. 2004;276:523–540. [PMC free article] [PubMed]
  • Keegan BR, Feldman JL, Begemann G, Ingham PW, Yelon D. Retinoic acid signaling restricts the cardiac progenitor pool. Science. 2005;307:247–249. [PubMed]
  • Kishimoto Y, Lee KH, Zon L, Hammerschmidt M, Schulte-Merker S. The molecular nature of zebrafish swirl: BMP2 function is essential during early dorsoventral patterning. Development. 1997;124:4457–4466. [PubMed]
  • Lavon N, Benvenisty N. Study of hepatocyte differentiation using embryonic stem cells. J Cell Biochem. 2005;96:1193–1202. [PubMed]
  • Mintzer KA, Lee MA, Runke G, Trout J, Whitman M, Mullins MC. lost-a-fin encodes a type I BMP receptor, Alk8, acting maternally and zygotically in dorsoventral pattern formation. Development. 2001;128:859–869. [PubMed]
  • Nguyen VH, Schmid B, Trout J, Connors SA, Ekker M, Mullins MC. Ventral and lateral regions of the zebrafish gastrula, including the neural crest progenitors, are established by bmp2b/swirl pathway of genes. Dev Biol. 1998;199:93–110. [PubMed]
  • Nikaido M, Tada M, Saji T, Ueno N. Conservation of BMP signaling in zebrafish mesoderm patterning. Mech Dev. 1997;61:75–88. [PubMed]
  • Novak AE, Ribera AB. Immunocytochemistry as a tool for zebrafish developmental neurobiology. Methods Cell Sci. 2003;25:79–83. [PubMed]
  • Ober EA, Field HA, Stainier DY. From endoderm formation to liver and pancreas development in zebrafish. Mech Dev. 2003;120:5–18. [PubMed]
  • Ober EA, Verkade H, Field HA, Stainier DY. Mesodermal Wnt2b signalling positively regulates liver specification. Nature. 2006;442:688–691. [PubMed]
  • Offield MF, Jetton TL, Labosky PA, Ray M, Stein RW, Magnuson MA, Hogan BL, Wright CV. PDX-1 is required for pancreatic outgrowth and differentiation of the rostral duodenum. Development. 1996;122:983–995. [PubMed]
  • Pyati UJ, Webb AE, Kimelman D. Transgenic zebrafish reveal stage-specific roles for Bmp signaling in ventral and posterior mesoderm development. Development. 2005;132:2333–2343. [PubMed]
  • Rosenquist GC. The location of the pregut endoderm in the chick embryo at the primitive streak stage as determined by radioautographic mapping. Dev Biol. 1971;26:323–335. [PubMed]
  • Rossi JM, Dunn NR, Hogan BL, Zaret KS. Distinct mesodermal signals, including BMPs from the septum transversum mesenchyme, are required in combination for hepatogenesis from the endoderm. Genes Dev. 2001;15:1998–2009. [PubMed]
  • Roy S, Qiao T, Wolff C, Ingham PW. Hedgehog signaling pathway is essential for pancreas specification in the zebrafish embryo. Curr Biol. 2001;11:1358–1363. [PubMed]
  • Shin D, Shin CH, Tucker J, Ober EA, Rentzsch F, Poss KD, Hammerschmidt M, Mullins MC, Stainier DY. Bmp and Fgf signaling are essential for liver specification in zebrafish. Development. 2007;134:2041–2050. [PubMed]
  • Tiso N, Filippi A, Pauls S, Bortolussi M, Argenton F. BMP signalling regulates anteroposterior endoderm patterning in zebrafish. Mech Dev. 2002;118:29–37. [PubMed]
  • Tremblay KD, Zaret KS. Distinct populations of endoderm cells converge to generate the embryonic liver bud and ventral foregut tissues. Dev Biol. 2005;280:87–99. [PubMed]
  • Tucker JA, Mintzer KA, Mullins MC. The BMP signaling gradient patterns dorsoventral tissues in a temporally progressive manner along the anteroposterior axis. Dev Cell. 2008;14:108–119. [PMC free article] [PubMed]
  • Vogeli KM, Jin SW, Martin GR, Stainier DY. A common progenitor for haematopoietic and endothelial lineages in the zebrafish gastrula. Nature. 2006;443:337–339. [PubMed]
  • Ward AB, Warga RM, Prince VE. Origin of the zebrafish endocrine and exocrine pancreas. Dev Dyn. 2007;236:1558–1569. [PubMed]
  • Warga RM, Nüsslein-Volhard C. Origin and development of the zebrafish endoderm. Development. 1999;126:827–838. [PubMed]
  • Zhang W, Yatskievych TA, Baker RK, Antin PB. Regulation of Hex gene expression and initial stages of avian hepatogenesis by Bmp and Fgf signaling. Dev Biol. 2004;268:312–326. [PubMed]