• zebrafish;
  • vasculogenesis;
  • angiogenesis;
  • vascular endothelial;
  • microarray;
  • etsrp;
  • etv2;
  • ER71;
  • myeloid


  1. Top of page
  2. Abstract
  6. Acknowledgements
  8. Supporting Information

Signaling pathways controlling vasculogenesis, angiogenesis, and myelopoiesis are still poorly understood, in part because not all genes important for vasculature or myeloid cell formation have been characterized. To identify novel potential regulators of vasculature and myeloid cell formation we performed microarray analysis of zebrafish embryos that overexpress Ets1-related protein (Etsrp/Etv2/ER71), sufficient to induce vasculogenesis and myelopoiesis (Sumanas and Lin [2006] Development 121:3141–3150; Lee [2008] Cell Stem Cell 2:497–507; Sumanas et al. [2008] Blood 111:4500–4510). We performed sequence homology and expression analysis for up-regulated genes that were novel or previously unassociated with the zebrafish vasculature formation. Angiotensin II type 2 receptor (agtr2), src homology 2 domain containing E (she), mannose receptor C1 (mrc1), endothelial cell-specific adhesion molecule (esam), yes-related kinase (yrk/fyn), zinc finger protein, multitype 2b (zfpm2b/fog2b), and stabilin 2 (stab2) were specifically expressed in vascular endothelial cells during early development while keratin18 expression was localized to the myeloid cells. Identification of vasculature and myeloid-specific genes will be important for dissecting molecular mechanisms of vasculogenesis/angiogenesis and myelopoiesis. Developmental Dynamics 238:1836–1850, 2009. © 2009 Wiley-Liss, Inc.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  8. Supporting Information

Vascular endothelial cells are among the first types of cells that undergo differentiation during embryonic development. In zebrafish, which has recently emerged as an advantageous model system to study early embryonic development, angioblasts originate within the lateral plate mesoderm in close association with hematopoietic precursors (Zhong,2005). The main axial vessels form by the process of vasculogenesis whichinvolves differentiation of vascular endothelial progenitor cells, their migration and aggregation into vascular cords. The secondary vessels form by angiogenesis, which involves migration and proliferation of differentiated vascular endothelial cells (Risau and Flamme,1995). Although significant progress has been made toward identifying molecular mechanisms responsible for vasculogenesis, angiogenesis, and hematopoiesis, only a subset of genes involved in vascular and hematopoietic development have been identified so far.

We have previously isolated a novel zebrafish ets1-related protein, etsrp, which encodes a novel ETS domain transcription factor (Sumanas et al.,2005; Sumanas and Lin,2006). Etsrp is one of the earliest markers for the vascular endothelial progenitor cells, and its function is both necessary and sufficient to initiate vasculogenesis (Sumanas and Lin,2006). In addition, Etsrp function is necessary for the initiation of myelopoiesis. These results argue that Etsrp functions within the common endothelial–myeloid precursors, hemangioblasts, in the anterior lateral plate mesoderm (Sumanas et al.,2008). Overexpression of Etsrp results in broad ectopic expression of multiple vasculature as well as early myeloid-specific markers, including the hemangioblast marker scl/tal1 (Robb et al.,1995; Shivdasani et al.,1995; Visvader et al.,1998), the vascular endothelial markers flk1/vegfr2 (Thompson et al.,1998), fli1 (Brown et al.,2000), VE-cadheri cdh5 (Larson et al.,2004; Sumanas et al.,2005), and the early myeloid marker pu.1 (Lieschke et al.,2002). However, erythroid-specific markers such as gata1 have not been affected by Etsrp overexpression. Etsrp function is conserved during vertebrate development. Mouse ER71/Etv2 and human ER71/Etv2 proteins represent functional orthologs of Etsrp (Lee et al.,2008; Sumanas et al.,2008; Ferdous et al.,2009). Overexpression of both mEtv2 and hEtv2 but not of the related hEts1 protein induces expansion of vascular endothelial markers in zebrafish embryos, similar to Etsrp overexpression. A recent study has demonstrated that Etv2 protein associates with Forkhead transcription factors and binds to FOX:ETS enhancer region present in multiple vasculature-specific gene promoters (De Val et al.,2008).

In the current study, we used Etsrp overexpression as a tool to identify novel genes associated with vasculogenesis or myelopoiesis. We performed microarray analysis of Etsrp-overexpressing embryos which resulted in identification of approximately 300 genes that were up-regulated more than twofold upon Etsrp overexpression. We further performed sequence homology and expression analysis of eight genes that were either novel or previously unassociated with the zebrafish vascular or myeloid development. Seven of these genes displayed vascular endothelial-specific expression pattern, while one of them was expressed in the myeloid precursor cells. Identification of novel genes is an important step toward further understanding of mechanisms of vasculogenesis, angiogenesis, and myelopoiesis, which are controlled by Etsrp function.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  8. Supporting Information

Global Expression Profiling

To identify novel genes associated with the early vasculogenesis and myelopoiesis, we performed global expression analysis of Etsrp-overexpressing embryos. We have previously reported that Etsrp overexpression resulted in ectopic expression of multiple vasculature and hemangioblast-specific markers that include scl and flk1 (Sumanas and Lin,2006). We investigated if Etsrp overexpression would result in the precocious induction of hemangioblast and vascular endothelial-specific genes. During normal development, scl expression is first observed at the one-somite stage while flk1 is not expressed until the seven-somite stage (Liao et al.,1998; Thompson et al.,1998). After microinjection of either etsrp RNA or DNA expression construct, ectopic expression of scl was observed by in situ hybridization at as early as the 30% epiboly stage, while flk1 expression could be reliably detected after the 50% epiboly stage (data not shown). Thus etsrp overexpression induced precocious expression of hemangioblast and vascular endothelial markers. Ectopic expression of both scl and flk1 was strong at the tail bud stage, while endogenous expression of scl and flk1 was either very weak or undetectable at this stage as determined by reverse transcriptase-polymerase chain reaction (RT-PCR) analysis (data not shown). The tail bud stage was then chosen to perform the global expression profiling in etsrp RNA-overexpressing embryos.

Global expression profiling of control uninjected and etsrp RNA-overexpressing embryos was performed using the zebrafish Nimblegen expression array that contains more than 30,000 genes based on Zv6 genomic sequence assembly (results have been submitted to the GEO database, accession no. GSE12909). Results were averaged from three independent hybridizations using separate RNA samples from different batches of embryos. Approximately 300 genes were up-regulated twofold or more in Etsrp RNA-overexpressing embryos. Because only a fraction of the genes present in the microarray have been annotated, we performed manual annotation of the genes that were up-regulated more than fourfold in Etsrp-overexpressing embryos. Genes that showed P values greater than 0.05 were excluded from further analysis. As we performed manual annotation, we found that in several cases genes or accession numbers that were listed separately in the microarray annotation corresponded to the same gene. In summary, expression of 53 genes was increased greater than fourfold, after duplicate entries were eliminated (Table 1). Nineteen of them have been previously associated with vascular or hemangioblast development in zebrafish. The greatest change was observed in the expression level of etsrp itself, which is expected, as most of the probes would detect injected etsrp RNA still present at the tail bud stage. However, expression of endogenous etsrp was up-regulated as well by 29.4-fold, calculated as an average from three different oligonucleotides that were specific to the etsrp 3′ untranslated region sequence, which was absent in the injected synthetic etsrp mRNA. Other genes up-regulated in etsrp-overexpressing embryos included scl/tal1, crl, aqp8, egfl7, fli1a, fli1b, cdh5, mcam, lmo2, sox7, dusp5, flk1, plexin D1, nrp1b, admr, clic2, hey2/gridlock, all of which have been previously associated with vascular development in the zebrafish embryos (Zhong et al.,2000; Larson et al.,2004; Martyn and Schulte-Merker,2004; Parker et al.,2004; Chan et al.,2005; Dooley et al.,2005; Patterson et al.,2005; Sumanas et al.,2005; Weber et al.,2005; Covassin et al.,2006; Cermenati et al.,2008; Herpers et al.,2008; Pendeville et al.,2008). The remaining genes included genes that were associated with vasculature formation in other organisms but were not previously characterized in zebrafish, genes that were isolated in zebrafish and/or other organisms but were not associated with vascular or myeloid development and novel genes. The majority of known zebrafish vasculature-specific genes showed significant up-regulation in Etsrp-overexpressing embryos. Notable exceptions were ets1a and flk1b/kdrb, expression of which was not significantly up-regulated, suggesting that ets1a and flk1b transcription is regulated in Etsrp-independent manner (Table 1). Genes expressed within hemangioblast cells such as scl, lmo2, and fli1a showed very strong up-regulation, which suggests that some of them are likely to be direct transcriptional targets of Etsrp. Lycat expression, which is associated with the cloche mutation, was not affected by Etsrp overexpression, as lycat functions upstream of etsrp and is expressed ubiquitously (Xiong et al.,2008). Expression of the myeloid marker pu.1/spi1 was up-regulated in the etsrp-overexpressing embryos, while later myeloid markers such as lcp1 and mpx showed no difference or were not expressed. Expression of genes associated with early erythroid development such as gata1 and ZBP-89 was not affected, which supports our earlier report that Etsrp function does not affect erythroid lineage formation (Sumanas and Lin,2006).

Table 1. Genes Affected in Etsrp-Overexpressing Embryos as Analyzed by Global Expression Profilinga
Genes up-regulated greater than 4-fold
SEQ_ID Ratio etsrp/wtP value
  • a

    All genes that showed greater than 4-fold induction, and selected vasculogenesis and hematopoiesis-associated genes that showed lower or no induction, are listed. The same genes listed under different accession numbers were grouped together, and average values were calculated. Regular font, genes that have been previously associated with vasculature or myeloid cell formation in zebrafish. Italics, novel genes or genes that have not been previously associated with zebrafish vascular or myeloid development. Bold, genes chosen for further analysis.

BC124743, NM_001037375etsrp171.42 × 10−11
NM_213237stem cell leukemia; scl; tal155.04.8 × 10−6
XM_697015; NM_199786complement receptor C1q-related; crl39.67.0 × 10−7
XM_679988; NM_001004661; XM_702651aquaporin 8; aqp819.41.6 × 10−9
NM_001076746; NM_001005231; NM_001040385EGF-like-domain, multiple 7; egfl719.43.3 × 10−16
NM_001013350angiotensin II type 2 receptor; agtr218.11.9 × 104
XM_682920Kelch-like 4, klhl416.75.4 × 107
XM_684895Src homology 2 domain containing E; she15.51.2 × 104
XM_686481similar to mannose receptor C1; mrc114.15.8 × 104
NM_131348friend leukemia virus integration 1b; fli1a15.23.2 × 10−4
XM_684029endothelial cell-specific adhesion molecule; ESAM12.30.00654
XM_684323cdc42 guanine nucleotide exchange factor (GEF) 9; arhgef911.10.0011
XM_696527; NM_001008780friend leukemia virus integration 1b; fli1b10.41.2 × 10−6
NM_001080660novel protein; contains neurogranin-similar region10.15.9 × 104
XM_683132lysozyme g; lyg9.40.0058
NM_001003983cadherin 5; VE-cadherin; cdh59.27.1 × 10−5
BC081616; NM_212781yes-related kinase; yrk8.72.2 × 107
NM_001012367melanoma cell adhesion molecule; mcam; cd1468.70.0011
NM_131705major histocompatibility complex class I UEA; mhc1uea8.40.024
NM_001039524zinc finger protein, multitype 2b; zfpm2b; fog2b7.93.3 × 104
NM_001002561chloride intracellular channel 2; clic27.10.0033
BC093136; NM_131111LIM domain only; lmo26.91.2 × 10−6
NM_001080750; XM_704797SRY-box containing gene 7; sox76.97.7 × 10−7
XM_685498synaptotagmin VI; sytVI6.90.0042
NM_212565dual specificity phosphatase 5; dusp56.80.0081
NM_200568; BC066541keratin 186.61.8 × 105
XM_688479stabilin 2; CD44-like6.60.0087
NM_131472; XM_687969; BC129158kinase insert domain receptor; kdr; Vegfr2; flk16.49.6 × 10−10
XM_682490PR domain containing 16; prdm166.30.0061
XM_686639ADAM metallopeptidase domain 125.60.013
NM_205697plexin D15.60.0053
XM_686228pleckstrin homology domain protein; plekhg55.40.0012
NM_205674neuropilin 1b; nrp1b5.30.012
XM_691752ral-guanine nucleotide dissociation stimulator-like 2; rgl25.30.0048
NM_001035261supervillin; svil5.00.0040
XM_686503myocyte-specific enhancer factor 2A; MEF2A5.00.0053
BC124471G-protein coupled receptor 182; adrenomedullin receptor; admr4.90.009
XM_682373myoferlin isoform b; FER1L34.90.014
NM_001024448; XM_679456smooth muscle myosin, heavy polypeptide 11; myh114.94.4 × 105
BC127569; NM_131372; BC124092; BC129254; BC110099invariant chain-like protein 2; iclp24.72.9 × 1016
XM_688540Contactin associated protein-like 2 isoform a; Caspr24.77.6 × 104
XM_690985A-kinase anchor protein 2; PRKA2; AKAP-2; AKAP-KL4.60.0032
XM_684579beta-1,3-galactosyltransferase; B3GNT34.57.0 × 104
NM_213158tropomyosin 4; tpm44.40.0018
XM_679193microrchidia 3; morc34.40.0016
XM_679975fibroblast growth factor 5; fgf54.20.0021
XM_687580; NM_001002624FYVE, RhoGEF and PH domain-containing protein 5; FGD54.27.0 × 105
XM_687671; XM_687593similar to lymphatic vessel endothelial hyaluronan receptor 1;LYVE-14.21.6 × 108
NM_205670transmembrane protein 88; tmm884.20.041
XM_687270Anthrax toxin receptor 1 precursor; Tumor endothelial marker 8; Antr14.10.0017
NM_131622; BC114263hairy/enhancer-of-split related 2; hey2; gridlock4.11.5 × 10−7
NM_131207LIM homeobox 1b; lhx1b4.10.0035
BC071290engrailed 2b/eng34.00.0056
Other selected genes associated with vasculogenesis and hematopoiesis
BC067622spleen focus forming virus (SFFV) proviral  integration oncogene; spi1/pu.12.60.05
NM_130945fms-related tyrosine kinase 4; flt42.20.02
NM_130934hematopoietically expressed homeobox; hhex2.00.07
NM_131461endothelium-specific receptor tyrosine kinase 2; tie21.70.05
NM_001024653kinase insert domain receptor; flk1b/kdrb1.70.31
NM_131233GATA-binding protein; gata21.50.07
NM_001017558v-ets erythroblastosis virus E26 oncogene homolog 1a; ets1a1.30.22
NM_213270lysocardiolipin acyltransferase; lycat1.10.29
NM_131320lymphocyte cytosolic plastin 1; lcp10.930.49
BC124483GATA-binding protein 1; gata10.910.61
NM_001080180zinc finger protein 148; znf148/ZBP-890.840.31
NM_212779myeloid-specific peroxidase; mpxno expression 

Etsrp overexpression resulted in induction of a subset of genes normally associated with muscle development that include mef2a, myoferlin, myh11, and tropomyosin4 (Table 1). However, myoferlin displays strong expression in mouse endothelial cells where it regulates flk1 stability and function (Bernatchez et al.,2007); mef2a is detected in the endothelium of coronary arteries (Edmondson et al.,1994; Wang et al.,2003); mutations in myh11 have been associated with aortic defects in humans (Zhu et al.,2006); and an isoform of zebrafish tropomyosin4, tpm4-tv2, is specifically expressed in the embryonic vasculature (Zhao et al.,2008). These results argue that Etsrp is involved in regulating vasculature-specific expression of these genes.

We did not find any genes that were down-regulated greater than threefold with a statistical significance (P < 0.05) in Etsrp-overexpressing embryos. Although a few genes showed greater than twofold down-regulation, these results have not been yet validated with an independent method. Induction of multiple genes caused by Etsrp overexpression is consistent with Etsrp functioning as a transcriptional activator.

We further validated microarray results for 14 genes that showed the largest increase in etsrp-overexpressing embryos. These genes were either novel or previously unassociated with the vasculature formation or myelopoiesis in zebrafish. We obtained cDNA and performed whole-mount in situ hybridization analysis to characterize expression of these genes. Expression results are summarized in Table 2. Of 14 cDNAs analyzed, 8 displayed expression pattern associated with the vascular or myeloid development, while 6 were expressed ubiquitously. We further verified induction of the novel genes using RT-PCR. Expression of each gene analyzed was strongly increased in Etsrp-overexpressing embryos, similar to the microarray results (Table 2). It is possible that Etsrp RNA overexpression caused nonspecific induction of some of the genes that normally are not related to the vasculature formation or myelopoiesis. To test if overexpression of a related human ets1 gene caused similar up-regulation of downstream targets, hets1 DNA overexpression construct was microinjected into embryos at the one-cell stage. Expression levels of the same target genes were analyzed by the real-time RT-PCR at the tail bud stage. hEts1 and Etsrp share highly similar ETS DNA binding domain (Sumanas and Lin,2006) and would be expected to result in similar nonspecific effects if they were caused by the off-target DNA binding. Expression of hets1 itself was up-regulated by 67,600-fold in hets1-injected embryos confirming that the gene was expressed. Hets1 overexpression resulted in the increased expression of only selected number of genes while others were induced only by etsrp but not hets1 (Table 2). This is consistent with the previous reports showing that Ets1 can up-regulate certain genes associated with vasculature formation (Dittmer,2003). Thus, etsrp overexpression resulted in the up-regulation of a specific set of genes that most likely represent its endogenous transcriptional targets. The likely explanation for induction of certain ubiquitously expressed genes is that Etsrp participates in the regulation of their expression during normal development. Ubiquitous expression pattern includes expression in vascular endothelial cells. As Etsrp overexpression causes precocious formation of vascular endothelial cell precursors that express multiple vascular markers, it also induces precocious expression of ubiquitous genes in those cells which is observed by the microarray analysis.

Table 2. Confirmation of Expression Profiling Data Using In Situ Hybridization and/or RT-PCR
GeneEmbryonic expressionRatio etsrp / wtRatio hets1 / wt
  1. aEmbryonic expression summary is listed for 14 novel genes that displayed biggest induction according to microarray data. Eight vascular or myeloid-specific genes are marked in bold. RT-PCR was performed using tail bud stage RNA purified from etsrp DNA or human ets1 DNA overexpressing embryos, and normalized to EF1α expression. Note that etsrp and human ets1 overexpression results in up-regulation of distinct sets of genes, arguing for the specificity of observed induction.

  2. ND, not determined; RT-PCR, reverse transcriptase-polymerase chain reaction; wt, wild-type.

agtr2vasculature5.9 ± 2.134 ± 3.7
klhl4ubiquitous41.7 ± 30.01.8 ± 0.2
shevasculature, lateral line2.9 ± 1.12.5 ± 0.8
esamvasculature, otic vesicle, neural2.9 ± 0.12.8 ± 0.2
arhgef9/collybistinubiquitous10.3 ± 3.80.8 ± 0.2
novel; contains neurogranin-similar regionubiquitous4.7 ± 0.70.7 ± 0.1
lygubiquitous18.8 ± 7.70.6 ± 0.1
yrk/fynvasculature13 ± 5.58.7 ± 4
MHC class I UEAubiquitous7.0 ± 3.60.6 ± 0.1
zfpm2b/fog2bvasculature, neural18 ± 13.523.8 ± 16.4
sytVIubiquitous6.3 ± 0.20.9 ± 0.1
ker18neural, lateral line, olfactory, myeloid15.3 ± 0.81.9 ± 0.1

We further performed sequence analysis and analyzed expression pattern for eight novel genes expressed in specific tissues during early development. In all cases, their expression domains included vasculature or myeloid cells.

While this manuscript was in preparation, a related study was published that describes identification of novel vascular and hematopoietic genes by using microarray analysis of etsrp overexpressing embryos (Gomez et al.,2009). As expected, up-regulation of some previously uncharacterized genes including arhgef9, yrk, and fgd5 was observed in both studies. However, most of the genes presented in the current study have not been described in any of the previous studies.

Angiotensin Type II Receptor

We obtained a cDNA clone (IMAGE clone ID 5604078) that corresponds to NM_001013350 sequence which was up-regulated more than 18-fold in Etsrp-overexpressing embryos. Its sequence completely matched to the NM_001013350 sequence and encodes a protein of 360 amino acids with the highest homology to mammalian Angiotensin Type II Receptor (AT2, Agtr2; Supp. Fig. S1, which is available online). Zebrafish and human Agtr2 protein sequences display 41.4% identity. Agtr2 encodes a G-protein coupled receptor that contains 7 transmembrane domains (Supp. Fig. S1A). AT2 is thought to function as a modulator of complex biological programs involved in embryonic development, cell differentiation, tissue repair, and programmed cell death (Unger,1999; Blume et al.,2001). AT2 is widely expressed in the fetal tissues in the rat while its expression is restricted to a few organs after birth that include brain, adrenal, heart, kidney, myometrium, and ovary (Grady et al.,1991). Within the heart tissue, AT2 protein is present in coronary arteries, cardiomyocytes, and the ventricular myocardium (Wang et al.,1998; Busche et al.,2000). Zebrafish AT2 homologs have not been previously characterized. A distantly related gene zebrafish angiotensin II receptor-like (agtrl1a) is expressed in multiple organs, including venous vasculature during early embryonic development (Tucker et al.,2007).

We analyzed expression pattern of agtr2 by whole-mount in situ hybridization at the 10-somite to 36 hours postfertilization (hpf) stages. At the 10- and 15-somite stages agtr2 expression was observed in the presumptive angioblasts positioned bilaterally in the trunk region within posterior lateral plate mesoderm (PLPM), which are known to migrate toward midline during mid-somitogenesis stages (Fig. 1A, and data not shown). Some of agtr2-expressing angioblasts were observed at the midline at the 15-somite stage, and most the agtr2-expressing cells were located at the midline within the trunk region by the 18-somite stage, marking the location of dorsal aorta (Fig. 1B). Agtr2 expression was localized to the dorsal aorta at the 22-somite stage (Fig. 1C,D). By 24 hpf, agtr2 was expressed within the dorsal aorta and head vessels (Fig. 1E,F). Expression was also observed in the cardinal vein tail plexus region and intersegmental vessels. Much weaker expression was observed also in the cardinal vein. Similar pattern but weaker expression was apparent at 36 hpf (Fig. 1G). Thus agtr2 expression is localized exclusively to the vascular endothelial cells during early zebrafish embryonic development.

thumbnail image

Figure 1. Zebrafish angiotensin II type 2 receptor (agtr2) is expressed in vascular endothelial cells during early embryonic development as analyzed by in situ hybridization. A–C: Dorsal view, anterior to the left. D–G: Lateral view, anterior to the left. AG: A flat-mounted 15-somite embryo (A), 18-somite (B), 22-somite (C,D), 24 hours postfertilization (hpf; E,F); 36 hpf stage whole-mount embryos (G). F,G: Enlarged views of the tail region. Note agtr2 expression in bilateral endothelial cell precursors (arrowheads, A) and in the axial vessels (arrows, A–C). D: Agtr2 expression is limited mostly to the dorsal aorta (da). E,F: Agtr2 is expressed in the dorsal aorta, head (hv), and intersegmental vessels (isv) and the cardinal vein (cv). G: Agtr2 expression is mostly confined to the cardinal vein plexus.

Download figure to PowerPoint

Mannose Receptor, C Type 1

We obtained and sequenced cDNA corresponding to the gene XM_686481, expression of which was increased more than 14-fold by Etsrp overexpression (GenBank accession number of the complete cDNA sequence is FJ205703). The assembled sequence encodes a protein of 1,272 amino acids with the greatest similarity to mammalian mannose receptor, C type 1 (Mrc1), also known as the mannose receptor (MR, CD206) and the macrophage receptor (54.2% and 54.5% similarity and 41.9% and 41.5% identity to mouse and human MR proteins, respectively; Supp. Fig. S2). Mrc1 is comprised of an N-terminal, cysteine-rich domain, a fibronectin type II domain, eight C-type lectin-like domains, a transmembrane domain, and a cytoplasmic domain (Supp. Fig. S2A). A second Mrc1 homolog (Mrc1b) is predicted within the zebrafish genome (Supp. Fig. S2). Mouse MR is expressed predominantly in macrophages, and additional sites of expression are in astrocytes, perivascular microglia, glomerular mesangial cells, retinal pigment epithelium, lymphatic endothelium, and endothelial cells within the spleen and liver (Linehan et al.,1999; Regnier-Vigouroux,2003; Jansen and Pavlath,2006). Of interest, MR is first expressed in the vascular endothelium in the yolk sac on E9 (Takahashi et al.,1998). MR plays a role in the innate immune system by participating in the phagocytosis of microorganisms, and it may play a role in the antigen recognition and processing and in regulating serum glycoprotein homeostasis (East and Isacke,2002; Lee et al.,2002). Zebrafish homologs of MR have not been previously characterized.

As analyzed by in situ hybridization, zebrafish mrc1 expression was localized to vascular endothelial cell precursors starting from the 18-somite stage (Fig. 2A). Expression was limited to the cardinal vein region in the posterior part of an embryo at this stage. At 22–26 hpf, mrc1 expression was observed within the cardinal vein, the tail plexus region, and venous head vessels (Fig. 2B,C). Expression was mostly absent from the arterial vessels. During 36–72 hpf stages, endothelial mrc1 expression gradually disappeared (data not shown). We did not observe mrc1 expression in macrophages during these developmental stages.

thumbnail image

Figure 2. Vascular-endothelial expression of the zebrafish mannose receptor C1 (mrc1) and src homology 2 domain containing E (she) as analyzed by in situ hybridization. A–C,E,F: Lateral view, anterior to the left. D: Anterior view. AC: mrc1 expression. DF: she expression. A–F: 18-somite (A), 22 hours postfertilization (hpf; B), 26 hpf (C), 20-somite (D,E), 24 hpf (F) stage embryos. A–C: Note mrc1 expression in the posterior cardinal vein and venous head vessels. D–F: She expression is observed in the dorsal aorta, intersegmental and head vessels, the posterior cardinal vein plexus region, and weakly in the cardinal vein. In addition, she is expressed in the lateral line primordia (arrowhead, F) and a cluster of neural cells within the dorsal hindbrain (arrow, F). CaDI, caudal division of the internal carotid artery; hb, neural expression within the hindbrain; hv, head vessels; da, dorsal aorta; isv, intersegmental vessels; lda, lateral dorsal aorta; llp, lateral line primordia; MceV, middle cerebral vein; pcv, posterior cardinal vein; phbc, primordial hindbrain channel; pmbc, primordial midbrain channel.

Download figure to PowerPoint

Src Homology 2 Domain Containing E

Expression of gene XM_684895 was up-regulated more than 15-fold in etsrp-overexpressing embryos. The gene encodes a protein of 458 amino acids with significant homology to mammalian SH2 domain-containing protein E (She). She and a related protein Shd have been identified in a yeast two-hybrid screen as proteins binding to Abl tyrosine kinase (Oda et al.,1997). She is expressed in mouse heart, lung, brain, and skeletal muscle. Embryonic she expression or function has not been previously examined.

We obtained a zebrafish cDNA corresponding to the predicted she gene (GenBank accession no. AL928447). The predicted protein sequence contains a highly conserved C-terminal SH2 domain (Supp. Fig. S3). Zebrafish She is also homologous to more distant SH2 domain family members Shd and Shb (Supp. Fig. S3B). As analyzed by in situ hybridization, she expression was observed in vascular endothelial cell precursors starting from the 16-somite stage (data not shown). It was expressed in anterior lateral plate mesoderm (ALPM) and PLPM at the 20-somite stage in the regions that correspond to head and axial vessels (Fig. 2D,E). She expression was significantly stronger in the dorsal aorta than in the cardinal vein. At 24 hpf, she expression was observed in the dorsal aorta, head vessels, the tail plexus region, and weakly in the cardinal vein (Fig. 2F). In addition, she was expressed in the lateral line primordia and a group of cells in the dorsal hindbrain (Fig. 2F). She expression in vascular endothelial cells was also observed at 36 and 48hpf (data not shown).

Endothelial Cell-Specific Adhesion Molecule

Expression of the predicted gene XM_684029 was up-regulated more than 12-fold in etsrp-overexpressing embryos. We obtained and fully sequenced a cDNA clone (ID 3818382) corresponding to the XM_684029 sequence (GenBank accession number of the complete cDNA sequence is FJ205704). The predicted protein of 428 amino acids displays the highest homology to endothelial cell-specific adhesion molecule (ESAM) subfamily (Supp. Fig. S4). Two separate ESAM homologs are predicted within the zebrafish genome that we refer to as ESAMa and ESAMb. Zebrafish ESAMa displays 32.7% and 30.9% identity and 47.3% and 47.1% similarity to human and mouse ESAM proteins, respectively. Esam is a glycoprotein containing two immunoglobulin domains, a transmembrane and cytoplasmic domains (Hirata et al.,2001). Mouse Esam expression is restricted to endothelial cells in the embryonic and adult vasculature (Hirata et al.,2001; Nasdala et al.,2002). It is thought to mediate homophilic and calcium-independent adhesion of expressing cells. Zebrafish Esam homologs have not been previously characterized.

As analyzed by in situ hybridization, esam (esama) is expressed as early as the six-somite stage in bilateral stripes within ALPM and PLPM regions that correspond to the presumptive vascular endothelial cell precursors (data not shown). At the 15-somite stage esam expression is apparent in the anterior and posterior angioblasts, some of which have coalesced at the midline within the trunk region (Fig. 3A). Esam is expressed in the dorsal aorta, cardinal vein, and head vessels at the 22-somite stage (Fig. 3B). Staining is also apparent in the ventral part of the otic vesicle. At 24 hpf, esam expression is apparent in all vascular endothelial cells that include the dorsal aorta, the cardinal vein, head, and intersegmental vessels as well as endocardium (Fig. 3C,D). In addition, esam is expressed in the otic vesicle and a subset of neurons in the brain region including the epiphysis/pineal gland (Fig. 3C,E,F). Esam expression is apparent in vascular endothelial cells, endocardium, and a subset of neurons including epiphysis at 48 hpf (Fig. 3E–G).

thumbnail image

Figure 3. Expression of endothelial cell-specific adhesion molecule (ESAM) as analyzed by in situ hybridization. A,D: Dorsal view, anterior is to the left. B,C,E: Lateral view, anterior is to the left. AC: A flat-mounted 15-somite embryo (A), whole-mounted 22-somite (B), and 24 hours postfertilization (hpf; C) embryos; higher magnification view of the tail region (inset in C). DG: Anterior region of a flat-mounted 24 hpf embryo (D), 48 hpf stage embryos (E–G), views of esam expression in epiphysis (F) and heart (G). A,B: Note esam expression in the progenitors of head vessels (hv), endocardium (ec), dorsal aorta (da), and posterior cardinal vein (pcv), as well as nonvascular expression in the otic vesicle (ov). C–E,G: Esam is expressed in the axial, intersegmental (isv), and head vessels (C); and in the dorsal longitudinal vessel, aortic arches, and endocardial cells of the atrium (a) and the ventricle (v; D,E,G). C,E,F: In addition, esam is expressed in the otic vesicle (arrows, C,E) and a subset of neurons in the brain region including epiphysis (ep; C,E,F).

Download figure to PowerPoint

Yes-related Kinase

Sequences BC081616 and NM_212781 corresponded to a gene yes-related kinase (yrk), previously identified in the microarray analysis of cloche mutant embryos (Qian et al.,2005). Its vasculature-specific expression has been recently demonstrated (Gomez et al.,2009). However, its detailed expression and phylogenetic analysis have not been performed. Both BC081616 and NM_212781 appear to correspond to the different isoforms of the same yrk gene. We have obtained and completely sequenced a cDNA clone corresponding to yrk gene (submitted to Genbank with accession no. FJ205701). Zebrafish Yrk displays the highest homology to the mammalian Fyn proteins (Supp. Fig. S5) and is more distantly related to Yes1 and Src proteins. Two related Fyn homologs, zebrafish Fyn and Yrk, are present in the zebrafish genome. Fyn/Yrk is a member of the Src family of tyrosine kinases. It contains Src homology domains, which include SH2 phosphotyrosine binding domain and SH1 tyrosine kinase domain. Mammalian Fyn proteins are expressed as two isoforms as a result of alternative splicing (Resh,1998). Fyn phosphorylates numerous intracellular signaling molecules and is involved in diverse biological functions including T-cell receptor signaling (Lowell and Soriano,1996) and mediating VEGF-induced endothelial cell migration (Lamalice et al.,2006).

As analyzed by whole-mount in situ hybridization, yrk is expressed in the presumptive endothelial cell precursors within lateral plate mesoderm (LPM) as early as the 10-somite stage (Fig. 4A). Expression is observed in both anterior and posterior LPM throughout the somitogenesis (Fig. 4A,B). At 24–30 hpf, yrk expression is observed within all vascular endothelial cells and includes the dorsal aorta, cardinal vein, the tail plexus region, intersegmental vessels, head vessels, and the endocardium (Fig. 4C–G).

thumbnail image

Figure 4. Expression of yes-related kinase (yrk) in vascular endothelial cells. A,F,G: Dorsal view of flat-mounted embryos, anterior is to the right. B–E: Lateral view, anterior is to the left. AE: The 10-somite (A), 18-somite (B), anterior (C), trunk (D), and posterior (E) regions of 24 hours postfertilization (hpf) embryos. F,G: Anterior regions of flat-mounted embryos at 24 hpf (F) and 30 hpf (G). A,B: Yrk expression is observed in the endothelial cell precursors in the anterior (ALPM) and posterior lateral plate mesoderm (PLPM). A: Note yrk expression in migrating angioblasts (arrow). C–G: Yrk is expressed in axial, intersegmental, head vessels, the cardinal vein tail plexus region and the endocardium. aa, aortic arch; cv, cardinal vein; ec, endocardium; hv, head vessels; da, dorsal aorta; isv, intersegmental vessels; lda, lateral dorsal aorta; MceV, middle cerebral vein; pcv, posterior cardinal vein; phbc, primordial hindbrain channel; pmbc, primordial midbrain channel.

Download figure to PowerPoint

Zinc Finger Protein, Multitype 2b

NM_001039524 sequence expression of which was up-regulated almost eightfold in etsrp-overexpressing embryos, corresponds to the zebrafish friend of GATA (fog2b) gene. FOG proteins are multi-type zinc-finger proteins that function as co-factors of transcription factor GATA (Cantor and Orkin,2005). Mice with a targeted disruption in the fog2 gene die in mid-gestation of congenital heart malformations and the failure to form coronary arteries (Svensson et al.,2000; Tevosian et al.,2000). Two zebrafish homologs, fog2a and fog2b, have been previously described (Walton et al.,2006). Embryonic expression of both zebrafish genes in the brain area has been observed. However, expression has not been reported in other tissues, including vasculature.

As analyzed by in situ hybridization, fog2b expression is observed in the presumptive endothelial cell precursors within the lateral plate mesoderm at the 14- to 20-somite stages (Fig. 5A–C). Expression is the strongest at the 16-somite stage, when fog2b expression is apparent in endothelial cell precursors of the ALPM and at the midline within the trunk region. By the 18-somite stage, fog2b is expressed bilaterally at the posterior edge of the PLPM, while expression in other regions of the vasculature system is mostly undetectable. Fog2b expression at the 22-somite stage is limited to a subset of neurons in the central nervous system mid-brain region as previously described (Walton et al.,2006) (Fig. 5D). Endothelial-specific fog2b expression is strongly down-regulated in etsrp morphants as analyzed by in situ hybridization (Fig. 5E,F) and RT-PCR (Supp. Fig. S6).

thumbnail image

Figure 5. Expression of zinc finger protein, multitype 2b (zfpm2b/fog2b). A,B: Dorsal view of anterior (A) and trunk (B) regions at the 16-somite stage, anterior is to the left. C: Posterior view at the 18-somite stage, dorsal is up. DF: Lateral view at the 22-somite stage, anterior is to the left. E,F: Lateral view of the tail region of a control wild-type (wt) embryo (E) and etsrp morphant at the 18-somite stage (F). A–C: Fog2b expression is observed in endothelial cell precursors at the 16–18 somite stages in the anterior (ALPM; A) and posterior lateral plate mesoderm (arrows; B,C). D: Fog2b is expressed in a subset of neurons within central nervous system at the 22-somite stage (arrow). E,F: Endothelial-specific fog2b expression in the tail region is significantly reduced in etsrp morphants (arrows).

Download figure to PowerPoint

Keratin 18

Both NM_200568 and BC066541 sequences correspond to the keratin18 (ker18) gene. Keratin18 is expressed in epithelial cells of multiple tissues including human vascular endothelium (Miettinen and Fetsch,2000). Zebrafish keratin18 has been previously described (Schaffeld et al.,2003), and a transgenic keratin18:RFP line has been established (Wang et al.,2006). Zebrafish keratin18 expression has been observed in the head region, epithelial cells, and the pronephric duct. In addition, expression of the transgene was observed in the dorsal aorta at 2 days postfertilization (dpf; Wang et al.,2006).

We obtained a cDNA (clone ID: 8782598) corresponding to the zebrafish keratin18 which was used for the probe synthesis to analyze ker18 expression by in situ hybridization. At the 15-somite stage, ker18 was expressed in the epidermal cells, similar to the previous report (Wang et al.,2006; data not shown). However, at 24 hpf ker18 was expressed in a tissue-restricted pattern that was not analyzed previously. Ker18 expression was observed in multiple epithelial and neural cells that included neurons in the brain region, lateral line primordial, and the olfactory placodes (Fig. 6A). While ker18 was expressed in the epithelial cells adjacent to the lateral and ventral mesoderm, no apparent expression in the blood vessels was observed at this stage. However, ker18 expression was present in the presumptive myeloid cells scattered over the yolk (Fig. 6D). These ker18-expressing myeloid cells were absent in etsrp morphants and clo mutant embryos, while epithelial and neural expression of ker18 was not affected (Fig. 6). Etsrp morphants and clo mutants display absence of myeloid cells (Stainier et al.,1995; Sumanas et al.,2008), arguing that ker18-expressing cells are indeed myeloid progenitors.

thumbnail image

Figure 6. A–F: Expression of keratin18 (ker18) in control wild-type embryos (A,D), etsrp morphants (B,E), and homozygous cloche mutant embryos (C,F) at 24 hours postfertilization (hpf). Lateral view, anterior is to the left. AC: Ker18 expression in epidermal cells, neural cells including the lateral line primordia and olfactory placodes is not affected in etsrp morphants and clo mutants. Note that ker18 expression is apparent within the superficial epidermal tissue and no vascular-specific staining is observed at these stages. DF: However, multiple ker18-expressing cells that are located on top of the yolk and apparently correspond to the myeloid cells (arrows, D), are absent in etsrp morphants and clo mutants (D–F).

Download figure to PowerPoint

Stabilin 2

Expression of the sequence XM_688479 that encodes a zebrafish Stabilin 2 homolog, was up-regulated more than sixfold in etsrp RNA-injected embryos. A complete coding sequence of stab2 was isolated by PCR using cDNA from 24 hpf embryos and sequenced (GenBank accession no. FJ945321). Zebrafish Stab2 protein contains 2508 amino acids and displays high similarity to the mouse and human Stabilin 2 (Stab2) homologs (45.2 and 46.6% identity; 61.4 and 63.5% similarity, respectively); significant similarity is present over the entire protein sequence (Supp. Fig. S7). Stabilin proteins feature seven fasciclin-like adhesion domains, 18–20 epidermal-growth factor (EGF) domains, one X-link domain and 3 to 6 hyaluronan-binding motifs and a transmembrane domain (Politz et al.,2002). Mouse stab2 is expressed within sinusoidal endothelial cells in the liver, lymph node, spleen, and bone marrow as well as heart valve mesenchyme, brain, eye, and kidney (Falkowski et al.,2003). Stab2 functions as the endocytic hyaluronan receptor of hepatic sinusoidal endothelial cells (Zhou et al.,1999,2000,2002). Zebrafish Stabilin homologs have not been previously characterized.

As analyzed by in situ hybridization, zebrafish stab2 was expressed in the ALPM and PLPM regions in the presumptive endothelial cell precursors as early as the 10-somite stage (Fig. 7A). At the 22-somite stage, stab2 was expressed strongly in the precursors of dorsal aorta and cardinal vein, and weaker in the head vessels (Fig. 7B,C). At 24 hpf, intense stab2 staining was observed within axial and venous head vessels and the posterior cardinal vein tail plexus region while intersegmental vessels stained only weakly (Fig. 7D). Notably, stab2 exhibited stronger expression in the cardinal vein than in the dorsal aorta. Endothelial stab2 expression in the axial vessels, branchial arch vessels, and venous head vessels persisted at 36 hpf and 48 hpf (Fig. 6E,F).

thumbnail image

Figure 7. Expression of stabilin2 (stab2). A: Dorsal view, anterior to the left. B: Anterior view, dorsal is up. CF: Lateral view, anterior to the left. A: A flat-mounted embryo at the 10-somite stage. B–F: Whole-mount embryos at 22-somite (B,C), 24 hours postfertilization (hpf; D; inset in D, larger magnification of the trunk region), 36 hpf (E), 48 hpf (F) stage embryo, tail region. A: Note stab2 expression in the endothelial cell precursors in the anterior (arrowheads) and posterior (arrows) lateral plate mesoderm. B–F: Stab2 is expressed strongly in axial and venous head vessels including the cardinal vein tail plexus region, intersegmental and aortic arch vessels. D: Note that stab2 expression in the posterior cardinal vein is more intense than in the dorsal aorta. aa, aortic arches; ccv, common cardinal vein; cv, cardinal vein; hv, head vessels; da, dorsal aorta; isv, intersegmental vessels; mcev, middle cerebral vein; pcv, posterior cardinal vein; phbc, primordial hindbrain channel; pmbc, primordial midbrain channel.

Download figure to PowerPoint

Expression in Etsrp Morphants

To confirm that Etsrp regulates vascular expression of the novel genes described, we analyzed their expression in Etsrp morphants at 22–24 hpf. Expression of agtr2, esam, stab2, she, yrk was strongly down-regulated in Etsrp morphants with only a few scattered endothelial cells remaining in the trunk and tail regions (Fig. 8A–J), while mrc1 expression could not be detected in Etsrp morphants (Fig. 8K,L). As described in previous paragraphs, endothelial-specific fog2b (Fig. 5E,F) and myeloid-specific ker18 (Fig. 6D,E) expression domains were also absent in Etsrp morphants. The results from Etsrp knockdown and overexpression experiments argue that expression of the described novel vascular-specific genes during embryogenesis is dependent on Etsrp function.

thumbnail image

Figure 8. AL: Expression of agtr2 (A,B), esam (C,D), stab2 (E,F), she (G,H), yrk (I,J), and mrc1 (K,L) in control uninjected embryos (A,C,E,G,I,K) and etsrp morphants (B,D,F,H,J,L). A–L: At 24 hours postfertilization (hpf), tail region (A–J); at 22 hpf (K,L). Note that expression of vascular endothelial markers is strongly down-regulated in etsrp morphants with only a few endothelial cells within the axial vessels and the tail plexus region present.

Download figure to PowerPoint

Overall, 8 of 14 novel genes analyzed displayed specific expression patterns in vascular endothelial or myeloid cells. Of interest, expression patterns are different between multiple genes up-regulated by Etsrp. These results indicate that endothelial cells up-regulate distinct gene expression programs even during early stages of vasculogenesis. Our results will be important for further functional studies, which will greatly expand our knowledge of vasculature formation.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  8. Supporting Information

Global Expression Analysis

Etsrp RNA was synthesized as previously described (Sumanas and Lin,2006). Approximately 50 pg of etsrp RNA was microinjected into zebrafish embryos at the one- to four-cell stages. Three separate batches of approximately 150 microinjected embryos and uninjected controls were frozen on dry ice at the tail bud stage. Total RNA was purified using the RNAquous-4PCR kit (Ambion). Approximately 10–20 μg of RNA from each batch was sent to the NimbleGen facility, where it was processed, labeled, and hybridized to standard NimbleGen 385K gene expression array. Briefly, RNA was converted into cDNA using the SuperScript II cDNA Conversion Kit (Invitrogen, Carlsbad, CA). Double-stranded cDNA was random-prime labeled with Cy3-nonamers and hybridized to the microarrays for 16 hr at 42°C. The zebrafish microarray was based on Zv6 genome assembly and contained approximately 32,000 genes with approximately 12 probes of 60mer oligonucleotides per each gene. The arrays were washed, dried, and scanned at 5-μm resolution using a GenePix 4000B microarray scanner (Molecular Devices, Sunnyvale, CA). Data were extracted from scanned images using NimbleScan software (Roche NimbleGen). Quantile normalization was performed across replicate arrays, and RMA (Robust Multichip Average) analysis was performed to generate gene expression values (Irizarry et al.,2003). Average ratios of expression values of etsrp-injected vs. control were calculated from three experiments. Results were sorted by the fold change from the highest to the lowest. Genes that displayed values of P > 0.05 or showed expression levels less than twofold above the background level in both control and experimental samples were eliminated from further analysis.

Real-Time RT-PCR Analysis

Embryos were injected at the one- to two-cell stage with 25–50 pg of circular etsrp-XeX or hEts1-pSG5neo plasmid (Stratagene, La Jolla, CA; kindly provided by R. Forough, Texas A&M University, College Station, TX) as previously described (Sumanas et al.,2008). Batches of 50 injected and control uninjected embryos were frozen on dry ice at the tail bud stage. Total RNA was purified using the RNAquous-4PCR kit (Ambion). cDNA was synthesized using Superscript III reverse transcriptase and oligo-dT primer (Invitrogen). Real-time PCR was performed using Chromo4 thermal cycler (Bio-Rad) and iQ SYBR Green Supermix (Bio-Rad). The following PCR profile was used: 95°C 5 min; 95°C 1 min, 58°C 1 min, 72°C 1 min, detection at 82°C for 10 sec; steps 2 through 5 repeated 44 times. Primer sequences used are listed in Table S1. Relative cDNA amounts for genes klhl4, neurogranin-similar, she, sytVI were calculated using the iCycler software (Bio-Rad) and normalized to the expression of elongation factor 1α (EF1α). For the genes arhgef9, MHC class I, lyg, ker18, mrc1, agtr2, fog2b, esam, yrk, PCR amplification of which resulted in minor amounts of nonspecific products, the amount of specific product was calculated using ImageJ software (NIH) from the intensity values of an image of an ethidium bromide-stained agarose gel obtained with Kodak gel imaging system.

cDNA Analysis, Probe Synthesis, and In Situ Hybridization

cDNA clones with catalog numbers EDR1052-464677 (corresponds to agtr2), MDR1738-8979828 (klhl4), MDR1738-8993193 (she), EDR1052-8871873 (mrc1), EDR1052-4299875 (esam), EDR4422-98297746 (arhgef9), EDR1052-5563160 (neurogranin-similar), EDR4422-98267804 (lyg), MDR1734-9509671 (yrk), EDR4422-98215784 (fog2b), EDR1052-5562874 (sytVI), EDR1052-99605490 (ker18), EDR1052-99566219 (MHC), and EDR1052-99605248 (stab2) were obtained from Open Biosystems and verified by single-pass sequencing. Clones corresponding to agtr2, she, mrc1, esam, arhgef9, yrk, and stab2 were sequenced completely. Homology trees and protein alignments were constructed using MacVector 9.0 software. Probes for in situ hybridization were synthesized as shown in the Table S2. Probes were labeled with DIG-UTP (Roche) and synthesized using T3, T7, or SP6 RNA polymerases (Ambion/Promega). In situ hybridization was performed as described (Jowett,1999).

Image Capture and Processing

Stained embryos younger than 22 hpf were mounted on the agarose-coated Petri dishes. Images were captured with Sony DSC-H9 digital camera mounted on Zeiss SV8 microscope. Different focal planes were manually merged using Adobe Photoshop CS2 to yield the maximum clarity image. Embryos 22 hpf and older were mounted on glass slides in 2% methylcellulose. Images were captured with using ×5 or ×10 objective on AxioImager Z1 (Zeiss) compound microscope with Axiocam ICC3 color camera (Zeiss). Images in different focal planes were combined using Extended Focus module within Axiovision software (Zeiss). Image levels were adjusted using Adobe Photoshop CS2 to increase the contrast.

Protein Sequence Analysis

Protein sequence alignments were performed using MacVector 9.0 software and CLUSTALW algorithm, BLOSUM matrix. Phylogenetic trees have been calculated using neighbor joining method. Uncorrected “p” distances between two sequences are shown for each branch. Conserved domains were detected using BLAST NCBI Conserved Domain Database (CDD) search (v2.14)(Marchler-Bauer et al.,2007).


  1. Top of page
  2. Abstract
  6. Acknowledgements
  8. Supporting Information

We thank K. Yutzey for providing critical comments. This research was supported by CCHMC Trustee Award to S.S.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  6. Acknowledgements
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

DVDY_21990_sm_SuppFigS1.eps1856KFigure S1. Zebrafish Angiotensin Receptor 2 sequence and homology analysis. (A) Zebrafish Agtr2 contains 7-transmembrane domains as revealed by NCBI Conserved Domain Database (CDD) search (v2.14)(Marchler-Bauer et al., 2007). (B) Homology tree of agtr1 and agtr2 subfamilies displaying homologs from human, mouse, frog Xenopus laevis, chick and zebrafish. (C) Alignment of zebrafish Agtr2 protein sequence with its human and mouse Agtr2 homologs as well as with more distant human Agtr1. Identical amino acids are shaded in dark grey while similar amino acids are shaded in medium grey.
DVDY_21990_sm_SuppFigS2.eps2840KFigure S2. Zebrafish Mannose Receptor C1 sequence and homology analysis. (A) N-terminal Ricin-type beta-trefoil, fibronectin type II and C-type lectin-like domains are predicted by CDD search. (B) Homology tree of Mrc1 and Mrc2 subfamilies from mouse, human, chick and zebrafish. Two Mrc1 homologs are predicted within zebrafish genome. (C) Alignment of protein sequences from the zebrafish, human and mouse Mrc1 homologs. Identical amino acids are shaded in dark grey while similar amino acids are shaded in medium grey. The first 333 a.a. of zebrafish Mrc1a sequence are based on a prediction from genomic sequence while the rest have been confirmed by cDNA sequencing (Genbank Accession Number FJ205703).
DVDY_21990_sm_SuppFigS3.eps1904KFigure S3. Zebrafish Src homology domain 2 containing E (She) sequence and homology analysis. (A) She contains C-terminal SH2 domain as predicted by CDD search. (B) Homology tree of she, shd, shf and shb subfamilies from zebrafish, human, mouse and chick. (C) Alignment of predicted zebrafish, human and mouse She protein sequences. Identical amino acids are shaded in grey.
DVDY_21990_sm_SuppFigS4.eps1988KFigure S4. Zebrafish Esam sequence and homology analysis. (A) Zebrafish Esam contains two immunoglobulin domains, including a V-set domain as predicted by CDD search. Predicted transmembrane and cytoplasmic domains are not shown. (B) Homology tree of related Esam and Immunoglobulin superfamily, member 11 (IGSF11) subfamilies from zebrafish, chick, mouse and humans. (C) Alignment of zebrafish ESAMa and ESAMb protein sequences with its closest human and mouse homologs. Identical amino acids are shaded in dark grey, while similar amino acids are shaded in medium grey. ESAMa cDNA sequence has been submitted to GenBank with accession number FJ205704.
DVDY_21990_sm_SuppFigS5.eps2107KFigure S5. Zebrafish Yrk sequence and homology analysis. (A) Zebrafish Yrk contains Src homology 2 and protein tyrosine kinase catalytic domains based on CDD homology analysis. (B) Homology tree of Yrk, Fyn, Yes, Src and Fgr protein kinases from human, mouse, chick, frog Xenopus laevis and zebrafish. (C) Alignment of two zebrafish Yrk/Fyn homologs with their closest homologs from mouse and human. Zebrafish yrk cDNA sequence has been submitted to GenBank with accession number FJ205701.
DVDY_21990_sm_SuppFigS6.tif53KFigure S6. Confirmation of fog2b downregulation in etsrp morphants by RT-PCR analysis. Total RNA from control and morphant embryos at the 16-somite stage was purified and used for cDNA synthesis. EF1? expression was used as a control reference. Primer sequences are listed in Table S1.
DVDY_21990_sm_SuppFigS7.eps4515KFigure S7. Zebrafish Stabilin 2 sequence and homology analysis. (A) Stab2 is predicted to contain 7 fasciclin-like adhesion domains, an X-link domain and putative hyaluronan binding sites based on CDD analysis. In addition, Stab2 proteins contain 18-20 EGF domains and a transmembrane domain (not shown). (B) Homology tree of Stab1 and Stab2 subfamilies from zebrafish, chick, mouse and humans. (C) Alignment of mouse, human and zebrafish Stab2 sequences. Zebrafish stab2 cDNA sequence has been submitted to GenBank with an Accession Number FJ945321.
DVDY_21990_sm_SuppTableS1.doc39KTable S1. Primer sequences used for the confirmation of the microarray results by the real-time RT-PCR.
DVDY_21990_sm_SuppTableS2.doc43KTable S2. Vector information and probe synthesis methods.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.