Hematopoietic and vascular endothelial cells constitute the circulatory system and are both generated from the ventral mesoderm. However, the molecules and signaling pathways involved in ventral mesoderm formation and specification remain unclear. We found that zebrafish etv5a was expressed in the ventral mesoderm during gastrulation. Knockdown of Etv5a using morpholinos increased the proliferation of ventral mesoderm cells and caused defects in hematopoietic derivatives and in vascular formation. By contrast, the formation of other mesodermal derivatives, such as pronephros, somites and the gut wall, was not affected. Knockdown specificity was further confirmed by overexpression of an etv5a construct lacking its acidic domain. In conclusion, our data reveal that etv5a is essential for the inhibition of ventral mesoderm cell proliferation and for the formation of the hemato-vascular lineage.

The circulatory system is one of the first functioning organ systems to develop in vertebrate embryos. The coordinated development of the heart, blood vessels and blood cells allows for the establishment of blood circulation, which is crucial for early survival. During gastrulation, the mesoderm is induced and patterned across the dorsal–ventral axis; later, the ventral mesoderm cells proliferate and differentiate to form a transient progenitor cell termed the hemangioblast, which gives rise to both endothelial and hematopoietic cell lineages (Amatruda and Zon, 1999; Lugus et al., 2005; Orkin and Zon, 2008; Patan, 2000; Xiong, 2008). Subsequently, during both vasculogenesis and angiogenesis, the endothelial cells line the vasculature, whereas hematopoietic progenitors differentiate into distinct blood lineages through hematopoiesis. Studies in different vertebrate models showed that these developmental processes and the genetic program that controls them are generally conserved throughout vertebrate evolution (Orkin and Zon, 2008; Paik and Zon, 2010).

Several molecules play a vital role in the formation of the circulatory system. The GATA family of transcription factors is essential for both primitive and definitive hematopoiesis. Gata1 is required for the proper maturation of erythroid, mast and megakaryocytic precursors, and for the specification of eosinophils, whereas Gata2 is necessary for mast cell development and for the maintenance and expansion of hematopoietic stem cells (Fujiwara et al., 2004; Ohneda and Yamamoto, 2002). Gata3 is required for the development of T lymphocytes (Hosoya et al., 2009). Other critical factors in hematopoiesis are Rag1, which is essential for the differentiation of T and B lymphocytes (Petrie-Hanson et al., 2009), and Runx1, which regulates the formation of hematopoietic progenitors, myeloid cells and the vasculature (Jin et al., 2012; Kalev-Zylinska et al., 2002). The development of endothelial cells and vascular patterning is regulated by the vascular endothelial growth factor (VEGF) receptors Flt1 and Flk1 (Hirashima, 2009; Orkin and Zon, 1997; Siekmann et al., 2008), whereas the basic helix–loop–helix transcription factor Scl is crucial for both hematopoiesis and vasculogenesis (Gering et al., 1998). Although the molecules that regulate the formation of blood cells and blood vessels are well characterized, those involved in the specification of the hemangioblastic fate in ventral mesoderm are less well characterized.

Ets transcription factors contain a conserved winged helix–turn–helix ETS domain that is essential for DNA binding. Members of this family play multiple roles in cell proliferation, differentiation and migration, and many of them are essential for hematopoiesis and blood vessel development (Liu and Patient, 2008). For example, Ets1, Ets2 and Fli1 regulate both vaculogenesis and angiogenesis (Randi et al., 2009). Etsrp is required to drive haemangioblasts towards a vascular fate and is essential for the formation of myeloid cells (Ellett et al., 2009; Sumanas and Lin, 2006). Erg is essential for normal adult hematopoietic stem cell homeostasis as well as endothelial differentiation and vascular development (Kruse et al., 2009; Randi et al., 2009). Similarly, defects in many human ETS genes have been associated with cancer, such as leukemia (Sharrocks, 2001).

Zebrafish (Danio rerio) has emerged as an excellent vertebrate model for the study of many aspects of the developmental process of the circulatory system, not only because of the transparency of the embryos, which facilitates in vivo imaging, but also because the embryos can survive for many days without circulating erythrocytes. In addition, most of the molecules involved in the development of the circulatory system are evolutionary conserved in this species (Belele et al., 2009). Here, we show that one Ets gene, etv5a, is expressed in the ventral mesoderm in zebrafish embryos. Knockdown of endogenous etv5a by antisense morpholinos resulted in increased proliferation of ventral mesoderm cells and defective hematopoietic cell and vascular endothelium cell formation, without affecting other mesodermal derivatives. The effect of the morpholino knockdown was phenocopied by overexpression of a dominant-negative etv5a construct. Our data provide further insights into the roles of Etv5a in the formation of the ventral mesoderm and hemato-vascular cell lineages.

etv5a is orthologous to mammalian ETV5

Zebrafish etv5a (previously named etv5) and etv5b (erm) are both homologous to mammalian ETV5 and arose from a gene duplication in ray-finned fish (Roussigné and Blader, 2006). Sequence comparison showed that zebrafish Etv5a displays more similarities with other vertebrate ETV5 proteins than Etv5b (supplementary material Fig. S1A) (also see Roussigné and Blader, 2006; Liu and Patient, 2008). A syntenic comparison of etv5a with mammalian homologs showed that gene loci flanking the zebrafish etv5a gene (chromosome 9) are highly conserved in human chromosome 3, mouse chromosome 16 and rat chromosome 11 (supplementary material Fig. S1B). Conversely, zebrafish etv5b is located in chromosome 6 and lacks significant synteny (supplementary material Fig. S1C). These results substantiate the hypothesis that zebrafish etv5a is orthologous to mammalian ETV5.

Expression of etv5a in the ventral mesoderm and its derivatives

etv5a expression was analyzed by whole-mount in situ hybridization and quantitative real-time PCR of zebrafish embryos at developmental stages ranging from one-cell stage to 48 hours post fertilization (hpf). A strong and ubiquitous signal was detected from one-cell to the oblong stage, which indicated that etv5a was expressed as a maternal mRNA (Fig. 1A–C). Ubiquitous expression was observed throughout the whole embryo before gastrulation (Fig. 1D,E). During early gastrulation (75% epiboly), expression localized to the ventral mesoderm (arrowhead in Fig. 1F,G). At later stages, etv5a expression became confined to the lateral mesoderm as demonstrated by the presence of two longitudinal stripes at the six-somite stage (Fig. 1H–J). A previous study showed that cells present in this region are multipotent progenitors that give rise to hematopoietic, endothelial and pronephric derivatives (Gering et al., 1998). etv5a expression was maintained in the posterior lateral mesoderm (Fig. 1K–N) during segmentation, but after 24 hpf it became restricted to the developing pronephric progenitors (Fig. 1O–T). In addition to mesodermal tissues, etv5a expression was detected in the developing nervous system, specifically in the brain and spinal cord, at the six-somite stage, and was maintained until the latest stage analyzed (48 hpf, Fig. H–U). The dynamic expression of etv5a suggests its importance in the developing central nervous system and mesodermal derivatives. By contrast, etv5b (erm) is expressed in the developing nervous system but not in the ventral mesoderm, lateral mesoderm or pronephric ducts (Münchberg et al., 1999; Raible and Brand, 2001). Mouse Etv5 is also strongly expressed in the developing nervous system and in the intermediate and lateral plate mesoderm, but it is expressed late in the mesonephros (Chotteau-Lelièvre et al., 1997; Chotteau-Lelièvre et al., 2001). The results presented here reveal that the expression of Etv5 and etv5a orthologs is evolutionary conserved.

Fig. 1.

etv5a expression in the developing zebrafish embryo.etv5a expression was analyzed by in situ hybridization. Embryo stages are shown in the bottom right corner of each panel. Lateral view, dorsal to the right (E–F, H, K, M, O, Q); dorsal view (I, J, L, N, P. R, T). G is a coronal section of the ventral etv5a-stained region, showing that etv5a is expressed in the mesoderm (arrowhead in G). (J) A double in situ hybridization of etv5a (purple) and scl (red) showing that etv5a is expressed in the posterior lateral mesoderm. etv5a expression first appears in the ventral mesoderm at 75% epiboly (arrowhead in F) and later becomes restricted to the lateral mesoderm (arrows in I, J, L and N). The transcripts were detected in the pronephric ducts at 24 hpf (O and P) and were retained until the latest stage analyzed (48 hpf) (O–T). pd, pronephric ducts. (U) etv5a expression analyzed by quantitative real-time PCR. Bottom bands show tubulin as a loading control.

Fig. 1.

etv5a expression in the developing zebrafish embryo.etv5a expression was analyzed by in situ hybridization. Embryo stages are shown in the bottom right corner of each panel. Lateral view, dorsal to the right (E–F, H, K, M, O, Q); dorsal view (I, J, L, N, P. R, T). G is a coronal section of the ventral etv5a-stained region, showing that etv5a is expressed in the mesoderm (arrowhead in G). (J) A double in situ hybridization of etv5a (purple) and scl (red) showing that etv5a is expressed in the posterior lateral mesoderm. etv5a expression first appears in the ventral mesoderm at 75% epiboly (arrowhead in F) and later becomes restricted to the lateral mesoderm (arrows in I, J, L and N). The transcripts were detected in the pronephric ducts at 24 hpf (O and P) and were retained until the latest stage analyzed (48 hpf) (O–T). pd, pronephric ducts. (U) etv5a expression analyzed by quantitative real-time PCR. Bottom bands show tubulin as a loading control.

Defects in the circulatory system after knockdown of Etv5a

To delineate the role of Etv5a during embryonic development, the morpholino (MO) knockdown approach was used to interfere with Etv5a expression. To block protein production, two 25 bp antisense morpholinos (MO1 and MO2) were synthesized to target different regions located upstream of the translation start site of etv5a mRNA. The specificity of the MOs was confirmed by rescue experiments in which the MOs were coinjected with cRNA for etv5a, as described for each experiment below. To confirm the efficacy of the MO knockdown approach, each of the two etv5a MOs was co-injected with the cRNA of a reporter construct that contained the etv5a MO1 and MO2 binding sequences upstream of an enhanced green fluorescent protein reporter (5′etv5a-EGFP). Effective knockdown, as revealed by the loss of EGFP protein, was observed upon co-injection with either of the two etv5a MOs, whereas no reduction in EGFP protein expression was observed upon co-injection of a control MO (supplementary material Fig. S2A).

Embryos injected with MO1 or MO2 were analyzed at 24 hpf, 2 days post-fertilization (dpf), 3 dpf and 4 dpf for morphological defects. The injection of 2 ng of MO1 or 10 ng of MO2 resulted in identical phenotypes (supplementary material Fig. S3) and therefore only embryos injected with 2 ng of MO1 are shown. Although knockdown of etv5a did not result in significant morphological abnormalities (supplementary material Fig. S2B), it caused a reduced heart beat rate (control, 75.28±5.82 beats/minutes; etv5a morphants, 32.43±4.53 beats/minutes at 24 hpf) and a lower blood cell count at all stages analyzed. Concomitant injection of MOs with etv5a cRNA rescued these phenotypes, indicating that the MO-induced defect was due to loss of Etv5a function.

Defective hemato-vascular derivatives in Etv5a-knockdown embryos

The reduced number of blood cells observed in etv5a morphants suggested a hematopoietic defect; thus, we examined the formation of erythroid cells by assessing the expression of gata1, a marker of erythroid precursors (Detrich et al., 1995), by in situ hybridization and quantitative real-time PCR. The results showed that gata1 expression was downregulated in etv5a-knockdown embryos (Fig. 2). We also found that injection of etv5a MOs downregulated markers for granulocytes (mpo) (Bennett et al., 2001) and lymphoid cells (rag1) (Willett et al., 1997) (Fig. 2). These results suggested that the common progenitor for these hematopoietic derivatives was affected by loss of Etv5a. Further analysis of gata2, which marks hematopoietic stem cells at the ten-somite stage (Detrich et al., 1995; Yamauchi et al., 2006), demonstrated that its expression was decreased in etv5a-knockdown embryos (Fig. 2), indicating that Etv5a is required for the formation of hematopoietic stem cells.

Fig. 2.

Decreased levels of hemato-vascular progenitors and derivatives in Etv5a-knockdown embryos. (A) In situ hybridization showing downregulation of markers for hemangioblasts and hematopoietic lineages in etv5a MO-injected embryos. The phenotypes observed were rescued by co-injection with etv5a cRNA but not a tp53 MO. The insert panels are enlargements of the stained areas. The fraction of embryos exhibiting these phenotypes are shown in the top right corner and embryo stages are shown in the bottom right corner of each panel. scl, hemangioblast; gata2, hematopoietic stem cell; gata1, erythroid progenitor; mpo, granulocytes; rag1, lymphocyte. (B) The results shown in A were quantitatively confirmed by qPCR analysis. Error bars represent s.e.m. ***P<0.001.

Fig. 2.

Decreased levels of hemato-vascular progenitors and derivatives in Etv5a-knockdown embryos. (A) In situ hybridization showing downregulation of markers for hemangioblasts and hematopoietic lineages in etv5a MO-injected embryos. The phenotypes observed were rescued by co-injection with etv5a cRNA but not a tp53 MO. The insert panels are enlargements of the stained areas. The fraction of embryos exhibiting these phenotypes are shown in the top right corner and embryo stages are shown in the bottom right corner of each panel. scl, hemangioblast; gata2, hematopoietic stem cell; gata1, erythroid progenitor; mpo, granulocytes; rag1, lymphocyte. (B) The results shown in A were quantitatively confirmed by qPCR analysis. Error bars represent s.e.m. ***P<0.001.

We next analyzed the formation of blood vessels by using Tg(fli1:eGFP) zebrafish embryos in which vascular endothelial cells are labeled with green fluorescent protein (Bennett et al., 2001). In etv5a-knockdown embryos, the main artery and vein initially formed at 24 hpf; however, the GFP fluorescence was significantly weaker than in the controls, as shown by direct observation and quantification of GFP by western blot analysis (Fig. 3). The defect in blood vessel formation became more severe as angiogenesis progressed, as evidenced by the weaker GFP signals in the intersomitic and sub-intestinal venous vessels, the formation of which is used as an indicator of proper angiogenesis (Isogai et al., 2001) (Fig. 3A).

Fig. 3.

Knockdown of Etv5a results in defective blood vessel formation. (A) Tg (fli1:eGFP) embryos showing defective GFP-labeled blood vessels as a result of etv5a MO injection. The sub-intestinal venous vessels (middle panels) and intersomitic vessels (bottom panels) are indicated by arrows and arrowheads, respectively. (B) Western blot analysis confirming that the levels of fli1-driven eGFP expression in etv5a morphants were downregulated compared with levels in the control. Error bars represent s.e.m. *P<0.05.

Fig. 3.

Knockdown of Etv5a results in defective blood vessel formation. (A) Tg (fli1:eGFP) embryos showing defective GFP-labeled blood vessels as a result of etv5a MO injection. The sub-intestinal venous vessels (middle panels) and intersomitic vessels (bottom panels) are indicated by arrows and arrowheads, respectively. (B) Western blot analysis confirming that the levels of fli1-driven eGFP expression in etv5a morphants were downregulated compared with levels in the control. Error bars represent s.e.m. *P<0.05.

During development, blood and endothelial cells develop from a common progenitor, the hemangioblast (Gering et al., 1998). Because our results showed defects in both blood cells and vessels, we examined the expression of scl, which marks hemangioblasts in lateral plate mesoderm (Gering et al., 1998). During normal development, transcripts for scl were seen in two pairs of lateral stripes flanking the mesoderm. By contrast, scl expression was downregulated in Etv5a-knockdown embryos (Fig. 2A,B). Taken together, these results show that knockdown of Etv5a is sufficient to inhibit the formation of hemato-vascular progenitors, which results in defective hematopoiesis and vasculogenesis.

Previous studies have shown that MOs can cause off-target apoptosis mediated by p53 activation (Robu et al., 2007). To rule out this possibility, all of the etv5a MOs were co-injected with a tp53 MO. The results showed no significant differences between the phenotypes of embryos co-injected with etv5a and tp53 MOs or those injected with etv5a MO alone (Fig. 2). Because the phenotypes caused by injection of etv5a MO were also rescued by concomitant injection of etv5a cRNA as noted above, the phenotypes of the etv5a morphants were not the result of p53 activation but the result of specific inhibition of Etv5a function.

Increased proliferation of ventral mesoderm cells after knockdown of Etv5a

In zebrafish embryos, hemangioblasts originate from the lateral plate mesoderm, which is itself derived from the ventral mesoderm (Bockamp et al., 2009; Davidson and Zon, 2004). We next examined the expression of eve1, which is specifically restricted to the ventral mesoderm during gastrulation (Seebald and Szeto, 2011). Intriguingly, the expression of eve1 was significantly upregulated in etv5a morphants (Fig. 4A,B). The increase in the expression of eve1 was not a result of embryo ventralization because head and notochord structures, which are the first structures affected in ventralized embryos, were morphologically normal (Neave et al., 1997) (supplementary material Fig. S2). The observed increase in eve1 could, however, result from increased cell proliferation, reduced cell death, or inhibition of differentiation of ventral mesoderm cells. We first examined cell proliferation by phospho-histone-H3 antibody and counterstained with eve1, which revealed increased proliferation of ventral mesoderm cells in etv5a morphants (Fig. 4C–F). We next assessed apoptosis by detecting the presence of activated caspase-3 using immunohistochemistry. No significant differences were observed between etv5a morphants and controls at 75% epiboly (Fig. 5). Intriguingly, significantly increased numbers of apoptotic cells were detected 2 hours later at the tail-bud stage and this abnormally increased apoptosis ceased 1 hour later (six-somite stage), suggesting that the apoptosis was triggered by deregulation of proliferation. These results demonstrate that knockdown of Etv5a increases the proliferation of ventral mesoderm cells.

Fig. 4.

Increased proliferation of ventral mesoderm cells after knockdown of Etv5a. (A) In situ hybridization showing upregulation of eve1 in etv5a MO-injected embryos. (B) This phenotype was rescued by co-injection with etv5a cRNA but not a tp53 MO, as confirmed by qPCR analysis. (C) etv5a morphants showed an increase in cell proliferation. Proliferating cells were detected using an anti-phospho-histone-H3 antibody (brown) and were double-labeled with eve1 (purple). (D) The proportions of cells positive for phospho-histone-H3 and eve1 among the total eve1-positive cells were quantified. (E) Ventral view of flat-mounted embryos. Right panels are enlargements of boxed areas on the left. etv5a-expressing cells are pseudocolored in red and counterstained with phospho-histone-H3 (green) and DAPI (blue) to locate the cell nuclei. Cells double labeled with etv5a and phospho-histone-H3 are indicated by white arrowheads. (F) Quantification of proliferating etv5a-positive cells in E. Error bars represent s.e.m. **P<0.01; ***P<0.001.

Fig. 4.

Increased proliferation of ventral mesoderm cells after knockdown of Etv5a. (A) In situ hybridization showing upregulation of eve1 in etv5a MO-injected embryos. (B) This phenotype was rescued by co-injection with etv5a cRNA but not a tp53 MO, as confirmed by qPCR analysis. (C) etv5a morphants showed an increase in cell proliferation. Proliferating cells were detected using an anti-phospho-histone-H3 antibody (brown) and were double-labeled with eve1 (purple). (D) The proportions of cells positive for phospho-histone-H3 and eve1 among the total eve1-positive cells were quantified. (E) Ventral view of flat-mounted embryos. Right panels are enlargements of boxed areas on the left. etv5a-expressing cells are pseudocolored in red and counterstained with phospho-histone-H3 (green) and DAPI (blue) to locate the cell nuclei. Cells double labeled with etv5a and phospho-histone-H3 are indicated by white arrowheads. (F) Quantification of proliferating etv5a-positive cells in E. Error bars represent s.e.m. **P<0.01; ***P<0.001.

Fig. 5.

Knockdown of Etv5a causes a temporary increased apoptosis of ventral mesoderm cells. (A) Control and Etv5a-knockdown embryos were labeled by in situ hybridization of eve1 (left panels) and anti-activated caspase-3 antibody (red in the right panels). In the right panels, eve1-expressing cells are pseudocolored in green to give a better presentation of the counterstaining. Note that significantly increased cell apoptosis can only be detected at the tail-bud stage. (B) Quantification of apoptotic eve1-positive cells in A. Error bars represent s.e.m. **P<0.01.

Fig. 5.

Knockdown of Etv5a causes a temporary increased apoptosis of ventral mesoderm cells. (A) Control and Etv5a-knockdown embryos were labeled by in situ hybridization of eve1 (left panels) and anti-activated caspase-3 antibody (red in the right panels). In the right panels, eve1-expressing cells are pseudocolored in green to give a better presentation of the counterstaining. Note that significantly increased cell apoptosis can only be detected at the tail-bud stage. (B) Quantification of apoptotic eve1-positive cells in A. Error bars represent s.e.m. **P<0.01.

In addition to hemato-vascular progenitors, the ventral mesoderm also gives rise to somites, pronepheric cells and the gut wall. To examine whether knockdown of etv5a affects other ventral mesoderm derivatives, the expression of myoD (a marker for early somitic mesoderm) (Weinberg et al., 1996), pax2a (a marker of pronephric ducts) (Majumdar et al., 2000) and foxf1 (a marker of the gut wall) (Madison et al., 2009) were analyzed. The results show that the expression of these markers was not affected by Etv5a knockdown (supplementary material Fig. S4), suggesting that Etv5a specifically regulates the formation of the hemato-vascular lineage. Taken together, these findings demonstrate that the proliferation of ventral mesoderm cells and the differentiation of hemato-vascular progenitors are dependent on the proper expression of Etv5a.

Abrogation of the acidic domain in etv5a causes a dominant-negative effect

Widespread expression of Etv5a by injection of etv5a cRNA did not show any significant phenotype compared with the controls (Fig. 6B,C). To gain insight into the structural requirements for Etv5a function, we created two deletion variants: one lacking the region containing the ETS DNA-binding domain up to the C-terminus of Etv5a (etv5aΔETS) and the other lacking the N-terminus, including the acidic domain (etv5aΔacidic) (Fig. 6A). Injection of etv5aΔETS cRNA did not result in any significant change in the phenotype compared with controls (Fig. 6B,C). However, injection of etv5aΔacidic resulted in reduced levels of all hematopoietic derivatives including hematopoietic stem cells and hemangioblasts (Fig. 6B,C) and increased the expression of the ventral mesoderm marker eve1 (Fig. 6B,C). The phenotypes caused by injection of etv5aΔacidic were identical to those observed in Etv5a-knockdown embryos. These results indicate that although deletion of the N-terminus including the acidic domain (etv5aΔacidic), abolishes Etv5a transactivation, the remaining ETS motif might compete with endogenous Etv5a for the DNA binding site, thereby causing a dominant-negative effect. By contrast, deletion of the ETS domain (etv5aΔETS) would produce a non-functional protein that is incapable of competing with endogenous Etv5a. These results confirm the specificity of the phenotypes conferred by injection of etv5a MO.

Fig. 6.

Resemblance between the phenotypes of etv5aΔacidic-injected embryos and etv5a morphants. (A) Schematic illustration of etv5aΔETS and etv5aΔacidic deletion constructs. (B) Injection of etv5aΔacidic was sufficient to downregulate markers for hemangioblasts and hematopoietic derivatives, and to increase the expression of eve1. By contrast, overexpression of the fully coding sequence of etv5a or the etv5aΔETS deletion mutant did not cause a significant phenotype. These results were confirmed by qPCR (C). Error bars represent s.e.m. ***P<0.001.

Fig. 6.

Resemblance between the phenotypes of etv5aΔacidic-injected embryos and etv5a morphants. (A) Schematic illustration of etv5aΔETS and etv5aΔacidic deletion constructs. (B) Injection of etv5aΔacidic was sufficient to downregulate markers for hemangioblasts and hematopoietic derivatives, and to increase the expression of eve1. By contrast, overexpression of the fully coding sequence of etv5a or the etv5aΔETS deletion mutant did not cause a significant phenotype. These results were confirmed by qPCR (C). Error bars represent s.e.m. ***P<0.001.

In this study, we used etv5a MOs and a dominant-negative etv5a variant to show that interference with etv5a expression in zebrafish embryos results in an increase in the proliferation of ventral mesoderm cells. The formation of hemato-vascular derivatives was also defective, whereas other derivatives developed normally. These abnormally proliferated hemangioblastic ventral mesoderm cells later underwent apoptosis during a very short time interval and consequently caused defective hematopoietic cells and vessels; however, cell differentiation seemed unaffected because those cells that escaped apoptosis differentiated normally according to the residual expression of hematopoietic and vascular markers. Taken together, our results suggest that Etv5a specifically regulates the proliferation of ventral mesoderm cells with hemato-vascular potential.

Molecules and regulatory mechanisms involved in the development of mesoderm cells and their progeny have attracted increasing attention because of their potential applications in regenerative medicine and stem-cell-based therapies for human diseases. The mechanisms involved in hematopoiesis are well characterized; however, only a small number of factors involved in ventral mesoderm and hemato-vascular specification have been identified. Only a few mutant lines showing defects in mesoderm and mesodermal derivatives are available in zebrafish. Moreover, the defects in these cell lines are not restricted to the hemato-vascular lineage but also affect many other mesodermal derivatives (Davidson et al., 2003; Gering et al., 1998; Griffin et al., 1998; Thompson et al., 1998). Our results showing that Etv5a specifically regulates the formation of the hemato-vascular lineage without affecting other mesodermal derivatives, suggests that etv5a is one of the earliest players in the generation of the hemato-vascular progenitors from the ventral mesoderm. By contrast, the cloche (clo) mutant displays severe deficiencies in the development of both endothelial and all hematopoietic cells, without affecting pronephros or somite formation, indicating that cloche regulates mesoderm cell specification and that the cloche mutation specifically affects hemangioblast formation (Stainier et al., 1995). The phenotype caused by Etv5a deficiency is similar to that of the cloche mutant, suggesting a possible interaction between Etv5a and Cloche. However, whether Etv5a and Cloche interact and work synergetically to regulate proliferation of ventral mesoderm requires further confirmation.

Several molecules belonging to the BMP, FGF and Wnt signaling pathways have been implicated as early inducers and patterning factors of the ventral mesoderm; however, how they direct the ventral mesoderm into different cell fates is still unclear. Here, we showed that Etv5a is a novel regulator of ventral mesodermal proliferation and specification. To our knowledge, Etv5a is the first transcription factor shown to be essential for the inhibition of proliferation of ventral mesoderm cells and for hemato-vascular specification. However, the upstream regulator for Etv5a in ventral mesoderm proliferation still remains to be identified. A previous study in mouse embryos demonstrated that BMP signaling positively regulates the expression of another Ets transcription factor, Er71/etsrp, in mesoderm cells, which contribute to hematopoietic and endothelial, skeletal, and smooth muscle cell lineages (Lee et al., 2008). A study in zebrafish also showed that BMP signaling was required for the expression of fli1 (also a member of Ets family) in hemangioblasts of the lateral mesoderm (Liu et al., 2008). We also found that inhibition of BMP signaling upregulated etv5a expression (data not shown). Previous studies in zebrafish showed that etv5a transcription was downregulated by inhibition of the FGF signaling pathway at 24 hpf (Mao et al., 2009; Roussigné and Blader, 2006). By contrast, we found that inhibition of FGF or MEK signaling did not affect the expression of etv5a at 75% epiboly (data not shown). Therefore, how etv5a responds to BMP and FGF signaling and whether this regulation is essential for ventral mesodermal proliferation remains to be examined.

ETV5/ERM regulates cell proliferation in several tumor cell lines and during spermatogenesis, and has been suggested to act as a proto-oncogene (Chen et al., 2005; Oh et al., 2012). Our results demonstrate that Etv5a is essential for the proliferation of mesoderm cells and suggest that Etv5a is a positive regulator of development of the hemato-vascular lineage. However, embryos overexpressing the full etv5a cRNA had a normal phenotype. A possible explanation for this finding is that ETV5/ERM protein requires phosphorylation for its activity (Janknecht et al., 1996). Therefore, it will be interesting to determine whether activated Etv5a is sufficient to induce abnormal hematopoiesis and vasculogenesis. In line with this idea, ETV5 is overexpressed in lymphoid leukemia and lymphoma (Charfi et al., 2011; Korz et al., 2002), which also reinforces the positive correlation between ETV5 expression and hematopoiesis. Whether the embryonic role of Etv5 found in this study is a feature of leukemias and other tumors warrants further investigation.

Ethics statement

All experiments were performed in strict accordance to standard guidelines for zebrafish work and approved by the Institutional Animal Care and Use Committee of Chang Gung University (IACUC approval number: CGU08-86 and CGU11-118).

Fish maintenance and mutants

(wild-type) zebrafish embryos were purchased from the Zebrafish International Resource Center (ZIRC; Eugene, OR) and were raised, maintained and paired under standard conditions. Tg(fli1:eGFP) zebrafish was obtained from Taiwan Zebrafish Core Facility at ZeTH with permission from Zebrafish International Resource Center (ZIRC). The embryos were staged according to the number of somites, hours post fertilization and days post fertilization (Kimmel et al., 1995).

Sequence comparisons and phylogeny

Amino acid sequences were aligned and displayed using Vector NTI (Invitrogen). Phylogenetic trees were constructed with ClustalX (Thompson et al., 1997). The GenBank accession numbers of the compared proteins are as follows: human ETV5/ERM (NM_004454.2); rat ETV5/ERM (NM_001107082.1); mouse ETV5/ERM (NM_023794.2); chicken ETV5 (XM_422651.2); zebrafish Etv5a (NM_001126461.1), Etv5b (NM_131205.1).

Construct generation

The open reading frame of zebrafish etv5a was PCR amplified with high-fidelity Pfu polymerase (Fermentas) and primers (5′-GAATTCGCCACCATGGACGGATTTTATGACC-3′and 5′- GGAATTCCTCAGTACACGTAACCATCAGGG-3′) which were designed according to the GenBank sequence (accession number: NM_001126461.1). etv5aΔETS was created with primers 5′-GGAATTCATGGACGGATTTTATGACCAGCAAG-3′ and 5′-GCTCTAGAGAGATCCGCGCCGCTGATATG-3′, whereas etv5aΔacidic was made with primers 5′-GGAATTCATGTCGGAGAGCTTGATGTTTCATG-3′ and 5′-GCTCTAGAGTCAGTACACGTAACCATCAGG-3′. etv5a MO1 and MO2 binding sequences were inserted upstream of an enhanced green fluorescent protein reporter in the pCS2 vector to create the 5′rgs4-EGFP construct to evaluate the specificity and efficiency of MOs.

Injection of RNA and morpholinos

Capped RNAs encoding the full coding sequence of Etv5a, etv5aΔETS and etv5aΔacidic were prepared as described previously (Chung et al., 2011). Antisense MO oligonucleotides were purchased from Gene Tools, LLC (Philomath, OR). Two morpholinos against etv5a were used: MO1 (TCACCTGGGTCTTCAAAGAGGCTCC), which overlaps the ATG start codon (−26 to −2), and MO2 (GATCTTCGCTTAAAAGCGATAGCTG), which corresponds to −59 to −35 at the translation start site. Blast analysis revealed homology of less than 20 bp identity for MO1 or MO2 to other genomic sequences; none of which corresponded to 5′ UTR or exon–intron splicing site of predicted or characterized genes, suggesting that MO1 and MO2 will act specifically on etv5a. A control MO designed against a random sequence of nucleotides not found in the zebrafish genome (5′-CCTCTTACCTCAGTTACAATTTATA-3′; Gene Tools) and an MO with a five-base mismatch to MO1 (5′-TCAgCTGGgTCTTgAAAGAcGCTgC-3′; mismatched bases are indicated by lowercase letters) was injected in an equal amount of MO1 as a control experiment. All injections were performed at the one- to two-cell stage and cRNAs or morpholinos were introduced into blastomeres.

Histological analysis

Digoxigenin-UTP-labeled riboprobes were synthesized according to the manufacturer's instructions (Roche) and in situ hybridizations were performed as described previously (Cheng et al., 2012). The color reaction was carried out using NBT/BCIP substrate (Roche). For immunohistochemistry, the embryos were blocked in 5% goat serum and incubated with rabbit phospho-histone H3 antibody or rabbit monoclonal anti-active caspase-3 (1∶200, Abcam). Goat anti-mouse IgG HRP or goat anti-rabbit IgG HRP (Roche) was used to detect the primary antibodies and DAB was used as a substrate for secondary-antibody-conjugated HRP (Amresco). Embryos were mounted with Vectashield mounting medium (Vector Laboratories).

Quantitative analysis

For quantitative real-time PCR (qPCR), embryos were homogenized in TRIzol reagent (Invitrogen) and total RNA was extracted using a standard method. cDNA was synthesized from total RNA with random hexamer priming using RevertAid First Strand cDNA Synthesis Kit (Fermentas). qPCR was performed on an ABI StepOne Real-Time PCR System (Applied Biosystems) with SYBR green fluorescent label (Fermentas). Primers used were eve1: F, 5′-CCCTGGTTAGGTGGTCTTCCA-3′ and R, 5′-GGGTTGTAGGCCTGTCCTAGCT-3′; scl: F, 5′-CGCAGACCTGCACCTTATGA-3′ and R, 5′-AGGGTGTGTTGGGATGAGCTT-3′; gata2: F, 5′-AAGCACGGCTCCAGTTTCCT-3′ and R, 5′-TCCTTTTCGTCCATTCTTGCA-3′; gata1: F, 5′-ACACAGTCCAGTTCGCCAAGT-3′ and R, 5′-TGGAGAGGTGTTTTTGGGAAA-3′; mpo: F, 5′-TCTTTTTGCCTGCCTGATTTC-3′ and R, 5′-ATTCCGGTGTTGTCGCAGAT-3′; rag1: F, 5′-CACTAAGCTCATCCCCACTGAAG-3′ and R, 5′-CCCAAAGCATGGGTGTACCT-3′. Gene expression levels were normalized to gapdh and assessed using the comparative CT (40 cycles) according to the manufacturer's instructions (Applied Biosystems).

For western blot analysis, embryos were homogenized in SDS lysis buffer. 60 µg were loaded on 12% SDS polyacrylamide gel and transferred to a PVDF membrane and detected with anti-GFP monoclonal antibody (1∶1000, Invitrogen). After washing, membranes were incubated with goat anti-Mouse HRP-conjugated secondary antibody (Chemicon) and developed with ECL (Millipore). Band intensities were quantified using Multi Gaugre analysis software.

Statistical analysis was performed by Student's t-test using Microsoft Excel 2007. The significance level was set at P<0.05. All reactions were performed in triplicate for each sample.

We are grateful to Taiwan Zebrafish Core facility at ZeTH and Zebrafish Core in Academia Sinica for providing fish.

Author contributions

Y.-C.C. and C.-D.H. conceived and designed the experiments; S.-Y.C., H.-Y.S., and S.-J.L. performed the experiments; S.-Y.C., H.-Y.S., S.-J.L., C.-D.H., and Z.-C.L. analyzed the data; Y.-C.C. wrote the paper.

Funding

This work was supported by grants from Chang Gung Memorial Hospital [grant numbers CMRPD1B0251, CMRPD3B0041 to Y.-C.C.] and the National Science Council of Taiwan [grant numbers NMRPD180593, NMRPD140413 to Y.-C.C.].

Amatruda
J. F.
,
Zon
L. I.
(
1999
).
Dissecting hematopoiesis and disease using the zebrafish.
Dev. Biol.
216
,
1
15
.
Belele
C. L.
,
English
M. A.
,
Chahal
J.
,
Burnetti
A.
,
Finckbeiner
S. M.
,
Gibney
G.
,
Kirby
M.
,
Sood
R.
,
Liu
P. P.
(
2009
).
Differential requirement for Gata1 DNA binding and transactivation between primitive and definitive stages of hematopoiesis in zebrafish.
Blood
114
,
5162
5172
.
Bennett
C. M.
,
Kanki
J. P.
,
Rhodes
J.
,
Liu
T. X.
,
Paw
B. H.
,
Kieran
M. W.
,
Langenau
D. M.
,
Delahaye-Brown
A.
,
Zon
L. I.
,
Fleming
M. D.
 et al. (
2001
).
Myelopoiesis in the zebrafish, Danio rerio.
Blood
98
,
643
651
.
Bockamp
E.
,
Antunes
C.
,
Liebner
S.
,
Schmitt
S.
,
Cabezas-Wallscheid
N.
,
Heck
R.
,
Ohnngemach
S.
,
Oesch-Bartlomowicz
B.
,
Rickert
C.
,
Sanchez
M. J.
 et al. (
2009
).
In vivo fate mapping with SCL regulatory elements identifies progenitors for primitive and definitive hematopoiesis in mice.
Mech. Dev.
126
,
863
872
.
Charfi
C.
,
Voisin
V.
,
Levros
L. C.
 Jr
,
Edouard
E.
,
Rassart
E.
(
2011
).
Gene profiling of Graffi murine leukemia virus-induced lymphoid leukemias: identification of leukemia markers and Fmn2 as a potential oncogene.
Blood
117
,
1899
1910
.
Chen
C.
,
Ouyang
W.
,
Grigura
V.
,
Zhou
Q.
,
Carnes
K.
,
Lim
H.
,
Zhao
G. Q.
,
Arber
S.
,
Kurpios
N.
,
Murphy
T. L.
 et al. (
2005
).
ERM is required for transcriptional control of the spermatogonial stem cell niche.
Nature
436
,
1030
1034
.
Cheng
Y. C.
,
Scotting
P. J.
,
Hsu
L. S.
,
Lin
S. J.
,
Shih
H. Y.
,
Hsieh
F. Y.
,
Wu
H. L.
,
Tsao
C. L.
,
Shen
C. J.
(
2012
).
Zebrafish rgs4 is essential for motility and axonogenesis mediated by Akt signaling.
Cell. Mol. Life Sci.
70
,
935
950
.
Chotteau-Lelièvre
A.
,
Desbiens
X.
,
Pelczar
H.
,
Defossez
P. A.
,
de Launoit
Y.
(
1997
).
Differential expression patterns of the PEA3 group transcription factors through murine embryonic development.
Oncogene
15
,
937
952
.
Chotteau-Lelievre
A.
,
Dolle
P.
,
Peronne
V.
,
Coutte
L.
,
de Launoit
Y.
,
Desbiens
X.
(
2001
).
Expression patterns of the Ets transcription factors from the PEA3 group during early stages of mouse development.
Mech. Dev.
108
,
191
195
.
Chung
P. C.
,
Lin
W. S.
,
Scotting
P. J.
,
Hsieh
F. Y.
,
Wu
H. L.
,
Cheng
Y. C.
(
2011
).
Zebrafish Her8a is activated by Su(H)-dependent Notch signaling and is essential for the inhibition of neurogenesis.
PLoS ONE
6
,
e19394
.
Davidson
A. J.
,
Zon
L. I.
(
2004
).
The ‘definitive’ (and ‘primitive’) guide to zebrafish hematopoiesis.
Oncogene
23
,
7233
7246
.
Davidson
A. J.
,
Ernst
P.
,
Wang
Y.
,
Dekens
M. P.
,
Kingsley
P. D.
,
Palis
J.
,
Korsmeyer
S. J.
,
Daley
G. Q.
,
Zon
L. I.
(
2003
).
cdx4 mutants fail to specify blood progenitors and can be rescued by multiple hox genes.
Nature
425
,
300
306
.
Detrich
H. W.
 III
,
Kieran
M. W.
,
Chan
F. Y.
,
Barone
L. M.
,
Yee
K.
,
Rundstadler
J. A.
,
Pratt
S.
,
Ransom
D.
,
Zon
L. I.
(
1995
).
Intraembryonic hematopoietic cell migration during vertebrate development.
Proc. Natl. Acad. Sci. USA
92
,
10713
10717
.
Ellett
F.
,
Kile
B. T.
,
Lieschke
G. J.
(
2009
).
The role of the ETS factor erg in zebrafish vasculogenesis.
Mech. Dev.
126
,
220
229
.
Fujiwara
Y.
,
Chang
A. N.
,
Williams
A. M.
,
Orkin
S. H.
(
2004
).
Functional overlap of GATA-1 and GATA-2 in primitive hematopoietic development.
Blood
103
,
583
585
.
Gering
M.
,
Rodaway
A. R.
,
Göttgens
B.
,
Patient
R. K.
,
Green
A. R.
(
1998
).
The SCL gene specifies haemangioblast development from early mesoderm.
EMBO J.
17
,
4029
4045
.
Griffin
K. J.
,
Amacher
S. L.
,
Kimmel
C. B.
,
Kimelman
D.
(
1998
).
Molecular identification of spadetail: regulation of zebrafish trunk and tail mesoderm formation by T-box genes.
Development
125
,
3379
3388
.
Hirashima
M.
(
2009
).
Regulation of endothelial cell differentiation and arterial specification by VEGF and Notch signaling.
Anat. Sci. Int.
84
,
95
101
.
Hosoya
T.
,
Kuroha
T.
,
Moriguchi
T.
,
Cummings
D.
,
Maillard
I.
,
Lim
K. C.
,
Engel
J. D.
(
2009
).
GATA-3 is required for early T lineage progenitor development.
J. Exp. Med.
206
,
2987
3000
.
Isogai
S.
,
Horiguchi
M.
,
Weinstein
B. M.
(
2001
).
The vascular anatomy of the developing zebrafish: an atlas of embryonic and early larval development.
Dev. Biol.
230
,
278
301
.
Janknecht
R.
,
Monté
D.
,
Baert
J. L.
,
de Launoit
Y.
(
1996
).
The ETS-related transcription factor ERM is a nuclear target of signaling cascades involving MAPK and PKA.
Oncogene
13
,
1745
1754
.
Jin
H.
,
Li
L.
,
Xu
J.
,
Zhen
F.
,
Zhu
L.
,
Liu
P. P.
,
Zhang
M.
,
Zhang
W.
,
Wen
Z.
(
2012
).
Runx1 regulates embryonic myeloid fate choice in zebrafish through a negative feedback loop inhibiting Pu.1 expression.
Blood
119
,
5239
5249
.
Kalev-Zylinska
M. L.
,
Horsfield
J. A.
,
Flores
M. V.
,
Postlethwait
J. H.
,
Vitas
M. R.
,
Baas
A. M.
,
Crosier
P. S.
,
Crosier
K. E.
(
2002
).
Runx1 is required for zebrafish blood and vessel development and expression of a human RUNX1-CBF2T1 transgene advances a model for studies of leukemogenesis.
Development
129
,
2015
2030
.
Kimmel
C. B.
,
Ballard
W. W.
,
Kimmel
S. R.
,
Ullmann
B.
,
Schilling
T. F.
(
1995
).
Stages of embryonic development of the zebrafish.
Dev. Dyn.
203
,
253
310
.
Korz
C.
,
Pscherer
A.
,
Benner
A.
,
Mertens
D.
,
Schaffner
C.
,
Leupolt
E.
,
Döhner
H.
,
Stilgenbauer
S.
,
Lichter
P.
(
2002
).
Evidence for distinct pathomechanisms in B-cell chronic lymphocytic leukemia and mantle cell lymphoma by quantitative expression analysis of cell cycle and apoptosis-associated genes.
Blood
99
,
4554
4561
.
Kruse
E. A.
,
Loughran
S. J.
,
Baldwin
T. M.
,
Josefsson
E. C.
,
Ellis
S.
,
Watson
D. K.
,
Nurden
P.
,
Metcalf
D.
,
Hilton
D. J.
,
Alexander
W. S.
 et al. (
2009
).
Dual requirement for the ETS transcription factors Fli-1 and Erg in hematopoietic stem cells and the megakaryocyte lineage.
Proc. Natl. Acad. Sci. USA
106
,
13814
13819
.
Lee
D.
,
Park
C.
,
Lee
H.
,
Lugus
J. J.
,
Kim
S. H.
,
Arentson
E.
,
Chung
Y. S.
,
Gomez
G.
,
Kyba
M.
,
Lin
S.
 et al. (
2008
).
ER71 acts downstream of BMP, Notch, and Wnt signaling in blood and vessel progenitor specification.
Cell Stem Cell
2
,
497
507
.
Liu
F.
,
Patient
R.
(
2008
).
Genome-wide analysis of the zebrafish ETS family identifies three genes required for hemangioblast differentiation or angiogenesis.
Circ. Res.
103
,
1147
1154
.
Liu
F.
,
Walmsley
M.
,
Rodaway
A.
,
Patient
R.
(
2008
).
Fli1 acts at the top of the transcriptional network driving blood and endothelial development.
Curr. Biol.
18
,
1234
1240
.
Lugus
J. J.
,
Park
C.
,
Choi
K.
(
2005
).
Developmental relationship between hematopoietic and endothelial cells.
Immunol. Res.
32
,
57
74
.
Madison
B. B.
,
McKenna
L. B.
,
Dolson
D.
,
Epstein
D. J.
,
Kaestner
K. H.
(
2009
).
FoxF1 and FoxL1 link hedgehog signaling and the control of epithelial proliferation in the developing stomach and intestine.
J. Biol. Chem.
284
,
5936
5944
.
Majumdar
A.
,
Lun
K.
,
Brand
M.
,
Drummond
I. A.
(
2000
).
Zebrafish no isthmus reveals a role for pax2.1 in tubule differentiation and patterning events in the pronephric primordia.
Development
127
,
2089
2098
.
Mao
J.
,
McGlinn
E.
,
Huang
P.
,
Tabin
C. J.
,
McMahon
A. P.
(
2009
).
Fgf-dependent Etv4/5 activity is required for posterior restriction of Sonic Hedgehog and promoting outgrowth of the vertebrate limb.
Dev. Cell
16
,
600
606
.
Münchberg
S. R.
,
Ober
E. A.
,
Steinbeisser
H.
(
1999
).
Expression of the Ets transcription factors erm and pea3 in early zebrafish development.
Mech. Dev.
88
,
233
236
.
Neave
B.
,
Holder
N.
,
Patient
R.
(
1997
).
A graded response to BMP-4 spatially coordinates patterning of the mesoderm and ectoderm in the zebrafish.
Mech. Dev.
62
,
183
195
.
Oh
S.
,
Shin
S.
,
Janknecht
R.
(
2012
).
ETV1, 4 and 5: an oncogenic subfamily of ETS transcription factors.
Biochim. Biophys. Acta
1826
,
1
12
.
Ohneda
K.
,
Yamamoto
M.
(
2002
).
Roles of hematopoietic transcription factors GATA-1 and GATA-2 in the development of red blood cell lineage.
Acta Haematol.
108
,
237
245
.
Orkin
S. H.
,
Zon
L. I.
(
1997
).
Genetics of erythropoiesis: induced mutations in mice and zebrafish.
Annu. Rev. Genet.
31
,
33
60
.
Orkin
S. H.
,
Zon
L. I.
(
2008
).
Hematopoiesis: an evolving paradigm for stem cell biology.
Cell
132
,
631
644
.
Paik
E. J.
,
Zon
L. I.
(
2010
).
Hematopoietic development in the zebrafish.
Int. J. Dev. Biol.
54
,
1127
1137
.
Patan
S.
(
2000
).
Vasculogenesis and angiogenesis as mechanisms of vascular network formation, growth and remodeling.
J. Neurooncol.
50
,
1
15
.
Petrie-Hanson
L.
,
Hohn
C.
,
Hanson
L.
(
2009
).
Characterization of rag1 mutant zebrafish leukocytes.
BMC Immunol.
10
,
8
.
Raible
F.
,
Brand
M.
(
2001
).
Tight transcriptional control of the ETS domain factors Erm and Pea3 by Fgf signaling during early zebrafish development.
Mech. Dev.
107
,
105
117
.
Randi
A. M.
,
Sperone
A.
,
Dryden
N. H.
,
Birdsey
G. M.
(
2009
).
Regulation of angiogenesis by ETS transcription factors.
Biochem. Soc. Trans.
37
,
1248
1253
.
Robu
M. E.
,
Larson
J. D.
,
Nasevicius
A.
,
Beiraghi
S.
,
Brenner
C.
,
Farber
S. A.
,
Ekker
S. C.
(
2007
).
p53 activation by knockdown technologies.
PLoS Genet.
3
,
e78
.
Roussigné
M.
,
Blader
P.
(
2006
).
Divergence in regulation of the PEA3 family of ETS transcription factors.
Gene Expr. Patterns
6
,
777
782
.
Seebald
J. L.
,
Szeto
D. P.
(
2011
).
Zebrafish eve1 regulates the lateral and ventral fates of mesodermal progenitor cells at the onset of gastrulation.
Dev. Biol.
349
,
78
89
.
Sharrocks
A. D.
(
2001
).
The ETS-domain transcription factor family.
Nat. Rev. Mol. Cell Biol.
2
,
827
837
.
Siekmann
A. F.
,
Covassin
L.
,
Lawson
N. D.
(
2008
).
Modulation of VEGF signalling output by the Notch pathway.
Bioessays
30
,
303
313
.
Stainier
D. Y.
,
Weinstein
B. M.
,
Detrich
H. W.
 III
,
Zon
L. I.
,
Fishman
M. C.
(
1995
).
Cloche, an early acting zebrafish gene, is required by both the endothelial and hematopoietic lineages.
Development
121
,
3141
3150
.
Sumanas
S.
,
Lin
S.
(
2006
).
Ets1-related protein is a key regulator of vasculogenesis in zebrafish.
PLoS Biol.
4
,
e10
.
Thompson
J. D.
,
Gibson
T. J.
,
Plewniak
F.
,
Jeanmougin
F.
,
Higgins
D. G.
(
1997
).
The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools.
Nucleic Acids Res.
25
,
4876
4882
.
Thompson
M. A.
,
Ransom
D. G.
,
Pratt
S. J.
,
MacLennan
H.
,
Kieran
M. W.
,
Detrich
H. W.
 III
,
Vail
B.
,
Huber
T. L.
,
Paw
B.
,
Brownlie
A. J.
 et al. (
1998
).
The cloche and spadetail genes differentially affect hematopoiesis and vasculogenesis.
Dev. Biol.
197
,
248
269
.
Weinberg
E. S.
,
Allende
M. L.
,
Kelly
C. S.
,
Abdelhamid
A.
,
Murakami
T.
,
Andermann
P.
,
Doerre
O. G.
,
Grunwald
D. J.
,
Riggleman
B.
(
1996
).
Developmental regulation of zebrafish MyoD in wild-type, no tail and spadetail embryos.
Development
122
,
271
280
.
Willett
C. E.
,
Cherry
J. J.
,
Steiner
L. A.
(
1997
).
Characterization and expression of the recombination activating genes (rag1 and rag2) of zebrafish.
Immunogenetics
45
,
394
404
.
Xiong
J. W.
(
2008
).
Molecular and developmental biology of the hemangioblast.
Dev. Dyn.
237
,
1218
1231
.
Yamauchi
H.
,
Hotta
Y.
,
Konishi
M.
,
Miyake
A.
,
Kawahara
A.
,
Itoh
N.
(
2006
).
Fgf21 is essential for haematopoiesis in zebrafish.
EMBO Rep.
7
,
649
654
.

Supplementary information