Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Identification of Vascular and Hematopoietic Genes Downstream of etsrp by Deep Sequencing in Zebrafish

  • Gustavo Gomez,

    Affiliation Department of Molecular, Cell and Developmental Biology, University of California Los Angeles, Los Angeles, California, United States of America

  • Jae-Hyung Lee,

    Affiliation Department of Integrative Biology and Physiology, University of California Los Angeles, Los Angeles, California, United States of America

  • Matthew B. Veldman,

    Affiliation Department of Molecular, Cell and Developmental Biology, University of California Los Angeles, Los Angeles, California, United States of America

  • Jing Lu,

    Affiliation Department of Molecular, Cell and Developmental Biology, University of California Los Angeles, Los Angeles, California, United States of America

  • Xinshu Xiao,

    Affiliation Department of Integrative Biology and Physiology, University of California Los Angeles, Los Angeles, California, United States of America

  • Shuo Lin

    shuolin@ucla.edu

    Affiliation Department of Molecular, Cell and Developmental Biology, University of California Los Angeles, Los Angeles, California, United States of America

Abstract

The transcription factor etsrp/Er71/Etv2 is a master control gene for vasculogenesis in all species studied to date. It is also required for hematopoiesis in zebrafish and mice. Several novel genes expressed in vasculature have been identified through transcriptional profiling of zebrafish embryos overexpressing etsrp by microarrays. Here we re-examined this transcriptional profile by Illumina RNA-sequencing technology, revealing a substantially increased number of candidate genes regulated by etsrp. Expression studies of 50 selected candidate genes from this dataset resulted in the identification of 39 new genes that are expressed in vascular cells. Regulation of these genes by etsrp was confirmed by their ectopic induction in etsrp overexpressing and decreased expression in etsrp deficient embryos. Our studies demonstrate the effectiveness of the RNA-sequencing technology to identify biologically relevant genes in zebrfish and produced a comprehensive profile of genes previously unexplored in vascular endothelial cell biology.

Introduction

The cardiovasacular system, which includes the heart, vessels and blood, function together to deliver oxygen and nutrients to cells throughout the body and remove metabolic waste. Understanding the development of this system is instrumental to the advancement of both basic and clinical sciences. The zebrafish, Danio rerio, is an excellent model organism for such studies due to embryo transparency, high fecundity, and fast development of organogenesis. In particular, the cardiovascular system is formed within one day of birth [1], [2]. Through genetic, cellular and molecular studies in zebrafish, a great deal of knowledge regarding the molecular components and cellular events that establish this system has been obtained. It is notable that many of the key molecular players and events that drive organogenesis in zebrafish are evolutionarily and functionally conserved with other organisms, including mammals.

A critical transcription factor required for the development of vascular endothelial cells is ets related protein, etsrp, which was originally discovered through transcriptional profiling of the cardiovascular mutant line, cloche, that lacks blood and vasculature [3] and independently by a mutagenesis screen [4]. It was subsequently found that etsrp is both necessary and sufficient to induce both the vascular endothelial and primitive myelopoietic program in zebrafish [5], [6]. Its significance was further underscored by the identification and characterization of the functional homolog ets variant 2, etv2, in mammals [7], [8] and Xenopus [9], [10]. The etsrp transgenic fish line has been useful to reveal the cellular events that establish the cranial vasculature and myelopoiesis [11], and it is possible that etsrp might be associated with the initiation of the definitive hematopoietic program [12].

The zebrafish genome is now in its ninth version and remains incompletely annotated. Nonetheless bioinformatic approaches have been used to identify important genes encoding transcription factors containing the ETS box DNA binding domain in hematopoietic and endothelial development [13]. The use of microarrays by the zebrafish research community has increased the identification of genes expressed in the developing cardiovascular system [3], [14], [15], [16], [17], [18], [19]. The arrival of higher throughput next generation sequencing has expanded the possibilities to deepen our understanding and identification of novel genes, and has already proven its utility for studies in zebrafish. Although the majority of the published studies combining RNA-sequencing and zebrafish have focused on immunity [20], [21], [22], the use of high throughput sequencing to study other biological subjects in zebrafish has begun to increase [23], [24], [25]. The ability of this technological approach to examine the zebrafish transcriptome at greater depth than microarrays without an a priori bias prompted us to re-examine the transcriptional profile of embryos overexpressing etsrp. In this study we present a panel of 39 more genes that are expressed in the developing zebrafish vasculature that were identified by this approach.

Results

As previously demonstrated, the overexpression of etsrp results in the induction of vascular related genes during gastrulation stages of development, before the onset of angioblast specification [5]. In this study etsrp RNA was injected into one cell flk1-gfp transgenic embryos and their transcriptome profiles were examined at later stages of gastrulation, 70–90% epiboly, when ectopic induction of flk1-gfp is detected (Figure 1). Total RNA was extracted from pools of 125 embryos, and the induction of known hemangioblast markers, fli1a, scl and etsrp, was verified in the etsrp overexpressed group by RT-PCR before library construction and sequencing. Two separate sets of RNA-seq libraries were constructed from independent samples that were pre-validated using the same validation scheme as demonstrated in the workflow diagram on Figure 1.

thumbnail
Figure 1. Workflow diagram.

Embryos were either uninjected or injected with etsrp RNA at the one cell stage, then raised until the late gastrulation stages when the flk1-gfp transgenic reporter is induced ectopically. Equal paired groups of embryos were pooled for mRNA extraction, library construction, and solexa mRNA-sequencing. Samples were pre-validated prior to sequencing by RT-PCR with fli1A, the 3′UTR of etsrp, as well as scl, which were all preferentially induced as expected by etsrp overexpression (e) relative to uninjected controls (u).

https://doi.org/10.1371/journal.pone.0031658.g001

A total of 123 million single-end or paired-end reads were obtained and 35% of the reads were mapped onto the unigene transcript sequence database (Build #117) [26] using the Burrows Wheeler Aligner (BWA) alignment program [27]. Table S1 lists the number of reads in each control and estrp overexpressed (estrp oe) samples and the mapping results. About 43 million reads were uniquely mapped in total, which covered 77% of the unigene database (39,784 out of 51,481). The expression level difference between the control and estrp oe samples were compared, resulting in the upregulation of 849 unigene entries above a 1.9-fold cutoff value in the etsrp oe group with a corrected p-value<0.05, while 726 entries were downregulated using the same parameters. A sample of the top 35 hits of annotated upregulated genes obtained in this dataset is listed in Table 1.

In order to determine the biological relevance of the gene pool obtained here, a gene ontology analysis was performed with the Database for Annotation, Visualization and Integrated Discovery Gene Ontology (DAVID GO) program, as demonstrated in the identification of novel genes expressed in myeloid cells in zebrafish [28]. Of the 849 upregulated unigene hits examined by functional annotation clustering, 797 were queried successfully, but because some unigene entries are associated with more than one gene, and some genes are represented by more than one unigene identifier, a total of 787 entries were evaluated resulting in 148 clusters. The remaining 52 entries were not identified either because they represent expressed sequence tags not associated with a characterized gene, they are unique to zebrafish, or their orthologs were unidentifiable by DAVID GO gene conversion tools. Conversely, the 726 downregulated hits resulted in the successful query of 306 entries divided among 50 clusters, but relative to the induced clusters, these had rather low fold-enrichment scores.

The top two clusters from the upregulated data contained genes associated with “vascular development”, “angiogenesis,” and other terms related to vascular endothelial cell biology. These data were combined resulting in 68 genes after duplicate deletion, and examined individually for involvement in vascular development. 18 genes were excluded because although several are expressed in mesoderm or heart, their specific expression or function within endothelial cells remains to be demonstrated. We did examine one of these, xirp2a, which we found to have endothelial expression. These were noted as a subset of potential vascular genes in Table S2, which also includes 12 that were removed because there is no indication that they are associated with endothelial cells. By searching the list generated containing 849 unigene entries, 15 more vascular related genes were identified (Table 2), bringing the total number of genes associated with endothelial cells in this dataset to 53.

thumbnail
Table 2. Known vascular genes induced by etsrp in RNA-seq dataset.

https://doi.org/10.1371/journal.pone.0031658.t002

In addition to endothelial specific genes, there were 4 genes of the myeloid lineage that were preferentially induced by etsrp oe. They are lplastin, csf3r, cebp1, hsd3b7. This observation is consistent with what has been previously reported by our group [6]. Overall using the DAVID GO program and visual inspection of the RNA-seq dataset we observed that a substantial number of genes (6.2%) are associated with developing endothelial biology, confirming the biological relevance of the upregulated dataset. Contrarily, DAVID GO analysis of the downregulated genes did not result in a distinct biological category of genes, as genes associated with the development of all three germ layers including endoderm, mesoderm, and ectoderm were repressed.

In order to further biologically validate the data and identify new genes expressed in endothelial cells, several genes were selected from the upregulated dataset for analysis by RNA whole mount in situ hybridization (WISH) at the 80% epiboly stages, when they are predicted to be upregulated by overexpressing etsrp (Figure 2). For gene selection, the potential of the top 350 hits to be expressed in endothelial cells was evaluated thoroughly, while the remaining portion was assessed at random. Genes examined were selected based on a PUBMED literature search using the following criteria: 1. Not previously shown to be expressed in vascular endothelial cells. 2. Previously detected in endothelial cells but evidence is limited to in vitro studies. 3. Has not been examined in developing vasculature. One of the top ranked genes (highest fold-induction), similar to Src homology 2 domain containing E, she, was used as a positive control since its expression in the vasculature has been reported previously [19]. The gene set examined is listed on Table 3. This assay resulted in the clear ectopic induction by etsrp overexpression in 49 of the 50 genes examined (Figure 2). The high percentage of validated genes induced by etsrp overexpression by WISH supports the validity of both the deep sequencing and bioinformatic methods used to examine the transcriptional profiles of upregulated genes. The genes predicted to be suppressed by this approach was also evaluated by quantitative RT-PCR, but only 2of 8 genes examined were validated (Table S3). This suggests that this approach is more amenable to the examination of gene induction by etsrp. Nevertheless, the complete dataset containing downregulated unigene entries is listed in Table S8.

thumbnail
Figure 2. Verification of RNA-seq data.

Wild type and etsrp injected, etsrp oe, embryos were collected at late stages of gastrulation and processed by WISH. Ectopic induction was detected as random positively labeled cells in embryos injected with etsrp RNA at the one cell stage. Endogenous expression was observed in the presumptive forerunner cells (arrows) of wild-type uninjected embryos in (C) she and (E) spa17. Endogenous expression was also observed in (K) tmem119b (L) myl9 (AD′) sh3bp4 (AN′) hmha1 (AT′) aff3 and (AW′) acsbg2. Ectopic induction is clearly detected in 49 of the 50 genes examined, but not (AT′) aff3, which has a relatively high level of endogenous expression at this stage. Wild type uninjected embryos are positioned on the left column with their etsrp oe counterparts on the right for each gene, anterior is facing left where possible. Scale bar: 250 µm.

https://doi.org/10.1371/journal.pone.0031658.g002

thumbnail
Table 3. Genes examined from the RNA-seq etsrp overexpression dataset.

https://doi.org/10.1371/journal.pone.0031658.t003

To determine whether the 50 selected ectopically induced genes are expressed in the developing embryonic zebrafish vasculature, embryos were processed for WISH at a developmental stage when most of the primitive vasculature has formed, 24-hour post fertilization (24hpf). Marked expression in the vasculature was noted for 39 of the 50 genes (Figure 3). To confirm that these genes are functionally downstream of etsrp in the vasculature, we examined their expression in etsrp morphants, in which case the expression in the axial vasculature is prominently reduced as demonstrated in Figure 4. Sparse expression in myeloid cells was only detected for one of these genes, myo1F (Figure 3L). Although there is some ubiquitous expression of some of these genes, the more pronounced expression in the vasculature is clearly demonstrated in higher magnification images of the trunk regions (Figure S1), where the axial vascular expression in morphants is clearly reduced in all genes. Of the 39 genes with vascular expression only fhl3 (Figure 3 AD′ and Figure S1AD′) and acsbg2 (Figure 3AL′ and Figure S1AL′) are preferentially expressed in the axial venous system at this stage.

thumbnail
Figure 3. Genes with expression in vascular endothelial cells.

The following genes are restricted to the vasculature and are expressed in the cranial vasculature, dorsal aorta, caudal hemtopoietic tail region, posterior cardinal vein, and intersegmental vessels: (E) cd93; (F) gpr183; (S) tmem88a; (AF′) hmha1, while the following are expressed in the same structures but exclude the posterior cardinal vein: (A) lyg2 is also expressed in the posterior yolk extension; (AC′) tagap; (AK′) si:ch211-250g4.3. These genes are expressed in the cranial vasculature, dorsal aorta, caudal hematopoietic tail region, posterior cardinal vein and other tissues as noted: (B) cdc42gap1, is also expressed in the hypochord, hatching gland, and sparsely in spinal cord; (D) klhl4 is also expressed in the heart, intersegmental vessels, and epiphysis; (G) xirp2a is also expressed in the heart, somite boundaries and pronephric duct; (N) rgl2, is also expressed in the intersegmental vessels, forebrain and primitive erythorcytes; (AA′) myof is also expressed in the fin fold, and tailbud; (AM′) arhgap27 is also expressed in the intersegmental vessels, otic vesicles, pronephric duct, and hatching gland. The following genes are expressed in the cranial vasculature, dorsal aorta, caudal hematopoietic tail region, and intersegmental vessels as well as (C) spa17, the olfactory bulb, epiphysis, otoliths, pronephric duct, and caudal notocord; (H) rassf4 was not detected in the intersegmental vessels but is observed in the forebrain, floor plate of neural tube, and spinal cord neurons; (I) tmem119b is also expressed in spinal cord, and tailbud; (AJ′) aff3 is also expressed in the forebrain, spinal cord, and caudal notocord. (AD′) fhl3 is expressed in the cranial vasculature, posterior cardinal vein, caudal vein, fin fold, lateral line primordium, and lightly in caudal somites. (AL′) acsbg2 is expressed in the posterior cardinal vein, caudal vein, spinal cord neurons, optic stalk, forebrain, and midbrain-hindbrain boundary. The following genes have basal levels of ubiquitous expression with prominent labeling of the cranial vasculature, dorsal aorta, caudal hematopoietic tail region and other structures and tissues where indicated: (J) myl9 is also expressed in the floor plate; (K) cntn4, is also expressed in the intersegmental vessels, olfactory bulb, and spinal cord neurons; (L) myo1f is also expressed in primitive myeloid cells; (M) iclp2, also has slight expression in anterior intersegmental vessels; (Q) dgki; (R) ifit5-like; (U) samd10 is also expressed in the caudal somites, and neurons of the anterior spinal cord; (V) cald1 is also expressed in the intersegmental vessels and pronephric duct; (X) zgc:171494 is also expressed in the floor plate, and pronephric duct; (Y) ankdd1a is also expressed in the spinal cord neurons; (Z) sh3bp4 is also expressed in the posterior cardinal vein, and intersegmental vessels; (AB′) grtp1b is also expressed in the posterior cardinal vein; (AE′) irf9; (AG′) ankrd58; (AH′) ccdc135 is also expressed in the posterior cardinal vein, pronephric duct, posterior neural tube, and slightly in the hatching gland; (AI′) tmem151a is also expressed in the posterior cardinal vein, spinal cord neurons and intersegmental vessels (see Figure S1AI′). Basal ubiquitous expression with darker staining of the dorsal aorta and caudal hematopoietic tail region is noted in: (O) capn8 which includes expression in the forebrain, hatching gland, and lateral line primordium; (P) mhc1uea; (T) rasa4 is also expressed in the yolk extension; (W) sept5b is also expressed in the forebrain and somites. Embryos are positioned laterally with the anterior facing left. Abbreviations: cht, caudal hematopoietic tail region; cn, caudal notochord; cs, caudal somites; cv, cranial vasculature; da, dorsal aorta; e, epiphysis; ff, fin fold; fp, floor plate; h, heart; hg, hatching gland; isv, intersegmental vessels; llp, lateral line primordium; mhb, midbrain hindbrain boundary; o, otoliths; ob, olfactory bulb; os, optic stalk; ov, otic vesicle; pd, pronephric duct; pe, primitive erythroid cells; pm, primitive myeloid cells; py, posterior yolk extension; sb, somite boundaries; sc, spinal cord; scn, spinal cord neurons; ss, somites; tb, tailbud; ye, yolk extension. Scale bar: 250 µm.

https://doi.org/10.1371/journal.pone.0031658.g003

thumbnail
Figure 4. Analysis of gene expression in etsrp morphants.

Flk1-gfp embryos were injected at the one cell stage with a mixture of translation blocking morpholinos for etsrp and analyzed by WISH at the 24hpf stage. Marked reduction is apparent in all genes examined in the axial vasculature, which includes the dorsal aorta and posterior cardinal vein, and is marked with a downward facing arrow in all images. The primitive myeloid cells stained in (L) myo1f wild-type controls are absent in their etsrp morphant counterparts. The staining in the axial trunk region of rgl2 in etsrp morphants marks primitive erythrocytes that are trapped due to lack of circulation. The expression of non-vascular structures is not affected otherwise in etsrp morphants. Embryos are positioned laterally with the anterior facing left. Abbreviations: cht, caudal hematopoietic tail region; cv, cranial vasculature; da, dorsal aorta. Scale bar: 250 µm.

https://doi.org/10.1371/journal.pone.0031658.g004

In the course of gene selection and WISH probe generation the Ensembl databases were utilized together with those at NCBI to examine gene structure and evolutionary ontology. Several of the genes probed for were noted to have redundant gene names as of the current genome build, Zv9 in Ensembl. These paralogs have been noted in their description below with an asterisk and further depicted in Table S4 with chromosomal location and identity. Following is a brief description of 10 selected genes in the order that they are presented in Figures 3 and 4. Descriptions for the remaining 29 genes are continued in Text S1.

A. lyg2

The goose type lysozyme ortholog lyg2 was originally identified in goose egg white, where it functions in the catalysis of bacterial cell wall break down, and is expressed ubiquitously in flounder [29]. In chicken, it is expressed in lung and non-adherent fraction of bone marrow cells, possibly reflecting expression within myelocytes where it exhibits an innate immune function [30]. Contrary to mammalian lysozyme G, the lysozyme G2 in flounder and zebrafish lack a signal peptide and a conserved cystein catalytic triad, suggesting an intracellular role with a potentially distinct function [31]. There are 2 other paralogs for this gene. lyg1 (zgc:92608), is expressed in macrophages [32], while the expression of zgc:162941 has not been reported.

B. cdc42GAP1/ARHGAP1

Originally cloned in mice, cdc42GAP1 catalyzes the hydrolysis of GTP bound by the Rho GTPases cdc42 and rac1, inactivating them [33]. In mice and humans it is expressed ubiquitously but exhibits marked expression in heart and lungs [34]. Mice knockouts result in generally reduced organ size relative to wild-type siblings due to increased rates of basal apoptosis, and display reduced viability [35]. The few survivors experience premature aging associated with heightened genomic instability [36]. In vitro evidence suggests cdc42GAP1 negatively modulates angiogenesis [37]. Although it is expressed ubiquitously in mammals, it is predominantly expressed in the developing vasculature and heart in zebrafish at the stage examined here (Figure 3B).

C. spa17/zgc:153721

The protein is homologous with the human sperm protein antigen 17, spa17, and both contain a cAMP dependent protein kinase regulatory subunit domain and an IQ motif that binds proteins with EF-hand motifs, but zgc:153721/spa17 of zebrafish has a glutamine rich region between these motifs that is absent in spa17 in humans and mice. Although in mice it is expressed ubiquitously and is expressed in embryos, in rabbits, it is most highly expressed in the tissue from which it was originally identified, testis [38], [39]. Increased expression of spa17 has been detected in esophageal, ovarian, and cervical cancers and has been used as a clinical marker for cancer in these tissues [40], [41], [42]. In zebrafish, spa17 is also expressed in the pronephros (Figure 3C).

D. klhl4

kelch like homolog like 4, klhl4 is a member of the kelch motif containing family of proteins [43]. The founding member, kelch, was cloned from Drosophila, and mediates cytoplasmic streaming through ring canals in the developing oocyte [44]. In human fetal tissue, klhl4 shows ubiquitous expression, and the protein is structurally similar to the founding member, kelch1 [45]. As in humans, the klhl4 of zebrafish has a BTB/POZ like domain and a BACK domain at the N-terminus, but while human klhl4 has 6 kelch domains at the carboxy terminus, zebrafish has 3. The function of this protein has not been characterized.

E. cd93

Other aliases of cd93 include complement component C1q receptor like (C1qrl), lymphocyte antigen 68 (ly68), and AA4.1. cd93 is a widely conserved member of the c type lectin family whose expression in endothelial and other cell types has been documented in mice and humans [46], [47]. In previous reports the name complement receptor C1qR-like (crl) has been used for a different gene in zebrafish, which is the likely homolog of c-type lectin domain family 14, member a (clec14a) [3], [18], [19]. Although both of these genes are members of the c-type lectin family, they are paralogs of each other that deserve distinction, and we note that both are expressed in the vasculature, since clec14a [3] and cd93 shown here, are vascular specific at the stage examined (Figure 3E).

F. gpr183*

Epstein bar induced 2, ebi2, or g protein coupled receptor 183, gpr183, is an orphan receptor that was originally identified as a gene induced by Epstein Bar Virus infection in Buritt lymphoma cells [48]. Knockout mice are viable without any notable gross phenotype, however they experience defects in B-cell mobility within lymphoid tissues, and consequently have difficulty mounting rapid antibody responses [49], [50]. Northern blot evaluation of its expression was originally limited to peripheral blood mononuclear cells (PBMCs), lymphocytic tissues and lungs [48]. However, in separate assays it has been detected in the aorta, atria, gastrointestinal tract, but not in the hematopoietic bone marrow or fetal liver [51]. At the stage examined, gpr183 is clearly expressed in the developing vasculature of zebrafish embryos (Figure 3F), but not in the heart.

G. xirp2a

xin actin binding repeat containing 2a, xirp2a, is one of three orthologs for the mammalian counterpart, xirp1β [52]. The paralogs in zebrafish are xirp2b, which is restricted to the myotome borders during the larval stages of development (ZFIN), and xirp2c, which has not been characterized. Functionally, xin repeat motifs cross-link f-actin with β-catenin at adherens junctions [53]. Xirp2a expression in the myoseptum, pronephric duct, heart, otic vesicle, head mesoderm, and liver has been demonstrated in developing zebrafish larvae [32]. In this study the expression in cranial and axial vasculature in the trunk becomes more evident. The most likely explanation for the differences observed might be a longer in situ color developing time used here.

H. rassf4

ras association (RalGDS/AF-6) domain family member 4, rassf4, is a member of Ras association domain family of proteins, which contain a Ras association domain and in the case of rassf4, a SARAH domain (reviewed in [54]). rassf4 is expressed ubiquitously in humans, but is downregulated in tumorous cells due to the hypermethylation of its promoter [55]. The idea that rassf4 is a tumor suppressor is supported by the apoptotic death of tumor cells following forced expression of rassf4 [55]. Besides expression in the vasculature of 1 dpf zebrafish embryos, rassf4 is also highly expressed in spinal cord neurons (Figure 3H, and Figure S1H).

I. tmem119b

Neither the protein encoded by tmem119b nor its paralog, tmem119a, have been characterized. tmem119b encodes a single pass type I transmembrane protein while tmem119a encodes a multipass transmembrane protein with two transmembrane domains. As a conserved gene in vertebrates, it is currently unclear whether tmem119b or tmem119a is the ortholog of their counterpart in mice, osteoblast induction factor, obif, which promotes osteoblast differentiation [56]. Although obif and tmem119b share a structural topological profile, tmem119a has a negligible higher amount of amino acid conservation with obif. It is also currently unknown whether tmem119/obif plays the same roles in humans.

J. myl9

myosin light chain 9 regulatory, myl9, is a calcium-regulated protein that regulates the contraction of myosin heavy chains in non-skeletal muscle cells that was originally cloned from human umbilical artery [57]. Currently there is evidence that the expression of myl9 may be directly regulated by junb [58], and runx1 [59]. In zebrafish there is a close paralog, myosin light polypeptide 9-like (zgc: 77916), with 95% gene identity to myl9 that also remains to be characterized.

Discussion

In this report we provide further support for the utility of combining the relatively unbiased approach of high throughput next generation RNA-sequencing to analyze whole embryo transcriptomes in zebrafish. The high correlation between genes induced by etsrp as predicted by RNA-seq and their confirmation by WISH reinforces the effectiveness of the methods used to interrogate transcriptional profiles in zebrafish. Of the genes selected from the dataset obtained, we identified a very high percentage of genes expressed in the vasculature of embryos, validating this approach to identify biologically relevant genes. However there may be non-vascular genes in this data that are induced due to the induction of off-target genes by the artificially elevated level of etsrp. This issue can be addressed by WISH using wild type embryos at a proper embryonic stage. Additionally, this method does not discriminate direct and indirect targets induced by etsrp. Further analysis by ChIP-Seq should reveal those genes that are transcriptionally regulated by etsrp.

Both the ectopic induction of these genes by etsrp overexpression and down-regulation by etsrp deficiency in vascular endothelial and primitive myeloid cells highlights their genetic epistatic relationship to etsrp and show that etsrp is a potent transcriptional activator whose expression and activity must be tightly regulated. Although several genes exhibit basal levels of ubiquitous expression their level in the vasculature is still higher. These genes deserve as much attention as those only specific to the vascular lineage, as several studies have shown that the down-regulation of such genes resulted in the disruption of vascular development. For instance, knockdown of nrarp-a and nrarp-b [60], [61], lpa1 [62], and dep1a/dep1b [63] respectively, resulted in defects in intersegmental vessels, lymphangiogenesis, and arterial specification. We obtained approximately 850 genes that were induced 1.9 fold or more by etsrp expression. Based on previous annotation and studies we can classify 6.2% of them as vascular genes. Experimental studies identified 39 more vascular genes and further increased the percentage to 10.8%. This leaves over 700 unigene entries in the dataset, some of which may be expressed and actively involved in the development of the zebrafish vasculature, and are potentially worth examining. The complete dataset is available in Table S5 and we hope this will benefit research on the development of the vascular system. The panel of genes found here remain to be characterized further and will require subsequent functional studies.

Materials and Methods

Embryo injection

Flk1-gfp embryos were injected with 75 pg etsrp RNA at the one cell stage as described [5], and collected at the late gastrulation stages, 70–90% epiboly, when ectopic flk1-gfp expression is notable in injected embryos. Total RNA was extracted with the PerfectPure RNA Tissue kit (5 Prime). For pre-sequencing validation, cDNA was made with Oligo(dT) primers and Superscript II reverse transcriptase (Invitrogen). The following primers pairs were used for verification by RT-PCR: β-actin: 5′ -TGTTTTCCCCTCCATTGTTG-3′, 5′-ACATACATGGCAGGGGTGTT-3′; etsrp 3′UTR: 5′-GAGGAATTCTCGAAGGATTGG-3′;TGGTTTTCTAAAGGCACCTAG-3′; Fli1a: 5′-CCGAGGTCCTGCTCTCACAT-3′;5′-GGGACTGGTCAGCGTGAGAT-3′; scl: 5′-GGAGATGCGGAACAGTATGG-3′, 5′-GAAGGCACCGTTCACATTCT-3′. To test gene expression in etsrp morphants, flk1-gfp embryos were injected with a mixture of 4 ng each of translation blocking etsrp morpholinos as described in [5] at the one cell stage, and embryos were harvested at 24hpf for WISH.

Cloning and Whole Mount In Situ Hybridization (WISH)

Different segments of the genes selected for evaluation were cloned into the pCRII Topo vector with the TOPO TA cloning kit (dual promoter) according to the manufacturer's instructions (Invitrogen). Amplicons were obtained from an etsrp overexpression cDNA library at late stages of gastrulation (70%–90% epiboly). Primers used for amplification are noted in Table S5. Positive plasmids were confirmed by sequencing. WISH was performed as described [64]. DIG labeled RNA probes were generated by linearizing TOPO-cloned genes with restriction endonucleases (New England Biolabs), and transcribing with SP6 or T7 RNA polymerase (Promega).

Image acquisition and processing

In situ stained embryos were further processed by a serial dehydration in ethanol, followed by rehydration into 1XPBS. Embryos were imaged in 2% methylcellulose in depression microscope slides. Images were captured with a color digital CCD camera (Axiocam, Zeiss) mounted on an upright microscope (Axioskop2 plus, Zeiss) with Openlab 4.0.2 software (Improvision, Lexington, MA). Serial images were combined, merged and processed with Adobe Photoshop CS5.

Deep sequencing library construction and bioinformatics

mRNA-seq libraries were constructed with the Tru-Seq RNA prep kit (Illumina) according to manufacturer's instructions. After this project was initiated, the Illumina sequencer has gone through a series of improvements. As a result, we obtained both single-end and paired-end reads with different lengths (76, 51, and 46 bp) using the Illumina GA II sequencer and its earlier versions. The obtained reads were mapped to 51,481 Unigene transcript sequences (Build #117, October 2009) [26] using BWA [27] allowing up to 4 mismatches. In case of the paired-end reads, we determined whether the mapped results support correct pairing of the reads according to the unigene annotation. If one read in a pair is mapped to a gene, the other one should be mapped to the same gene. Since each read in a pair may map to multiple locations in the genome, all possible combinations of their mappings were examined for correct pairing. The pair of reads is considered uniquely mapped only if one unique pair of mapped locations was identified.

To determine expression levels of genes and exons, we used the variable RPKM (reads per kilobase of exon per million mapped reads) defined by Mortazavi et al. 2008 [65]. Analysis of differential gene expression was carried out to obtain the differentially expressed genes between control sample and estrp oe sample. The number of uniquely mapped read-pairs for each gene in each sample was stored. The total number of mapped reads in each lane was normalized using the total mapped reads in each lane. Fisher's exact test was then performed using the above read counts for each gene. The resulting p-values were corrected via the Benjamini and Hochberg method as implemented in R. Finally, differentially expressed genes were defined as those with changes of at least 1.9-fold between a pair of samples at a false discovery rate (FDR) of 5%. Table S7 lists all 849 upregulated unigene hits in the estrp oe sample, while the 726 downregulated hits are tabulated in Table S8.

Quantitative PCR (qPCR)

qPCR was performed and analyzed exactly as stated in [12], with three independent biological replicates of either uninjected or etsrp oe groups of embryos harvested at late gastrulation stages. Primers used are listed in Table S6.

Supporting Information

Figure S1.

Higher magnification of Figure 4. The axial trunk vasculature of embryos displayed in figure 4 were imaged at higher magnification to highlight the changes observed. Wild-type embryos are on the left half of each column with their etsrp morphant counterparts on the right for each gene. Embryos were positioned with anterior facing left.

https://doi.org/10.1371/journal.pone.0031658.s001

(TIF)

Table S1.

Total RNA-Seq reads obtained in control and estrp oe samples and mapping results.

https://doi.org/10.1371/journal.pone.0031658.s002

(XLS)

Table S2.

DAVID GO derived gene subset without clear vascular expression or not expressed in developing vasculature.

https://doi.org/10.1371/journal.pone.0031658.s003

(XLS)

Table S3.

Quantitative PCR of unigene hits predicted to be suppressed by etsrp oe via RNA-seq.

https://doi.org/10.1371/journal.pone.0031658.s004

(XLS)

Table S4.

Genes with double entries in Ensembl.

https://doi.org/10.1371/journal.pone.0031658.s005

(XLS)

Table S5.

Primers used for cloning to make WISH probes.

https://doi.org/10.1371/journal.pone.0031658.s006

(XLS)

Table S6.

Primers used to test downregulated gene set following etsrp overexpression by qPCR.

https://doi.org/10.1371/journal.pone.0031658.s007

(XLS)

Table S7.

Complete etsrp mRNA-seq dataset (induced greater than or equal to 1.9 fold by etsrp overexpression).

https://doi.org/10.1371/journal.pone.0031658.s008

(XLS)

Table S8.

Downregulated mRNA-seq dataset (reduced greater than or equal to 1.9 fold by overexpression of etsrp ).

https://doi.org/10.1371/journal.pone.0031658.s009

(XLS)

Text S1.

Supplementary Text. A continuation of the results section that describes information regarding the genes examined in this study.

https://doi.org/10.1371/journal.pone.0031658.s010

(DOC)

Acknowledgments

We thank Scott Doherty for excellent technical support and Anqi Liu for help with animal husbandry.

Author Contributions

Conceived and designed the experiments: GAG SL. Performed the experiments: GAG MBV. Analyzed the data: GAG J-HL MBV JL. Contributed reagents/materials/analysis tools: XX. Wrote the paper: GAG.

References

  1. 1. Stainier DY, Fishman MC (1994) The zebrafish as a model system to study cardiovascular development. Trends Cardiovasc Med 4: 207–212.
  2. 2. Weinstein BM (2002) Plumbing the mysteries of vascular development using the zebrafish. Semin Cell Dev Biol 13: 515–522.
  3. 3. Sumanas S, Jorniak T, Lin S (2005) Identification of novel vascular endothelial-specific genes by the microarray analysis of the zebrafish cloche mutants. Blood 106: 534–541.
  4. 4. Pham VN, Lawson ND, Mugford JW, Dye L, Castranova D, et al. (2007) Combinatorial function of ETS transcription factors in the developing vasculature. Dev Biol 303: 772–783.
  5. 5. Sumanas S, Lin S (2006) Ets1-related protein is a key regulator of vasculogenesis in zebrafish. PLoS Biol 4: e10.
  6. 6. Sumanas S, Gomez G, Zhao Y, Park C, Choi K, et al. (2008) Interplay among Etsrp/ER71, Scl, and Alk8 signaling controls endothelial and myeloid cell formation. Blood 111: 4500–4510.
  7. 7. Lee D, Park C, Lee H, Lugus JJ, Kim SH, et al. (2008) ER71 acts downstream of BMP, Notch, and Wnt signaling in blood and vessel progenitor specification. Cell Stem Cell 2: 497–507.
  8. 8. Kataoka H, Hayashi M, Nakagawa R, Tanaka Y, Izumi N, et al. (2011) Etv2/ER71 induces vascular mesoderm from Flk1+PDGFR{alpha}+ primitive mesoderm. Blood.
  9. 9. Neuhaus H, Müller F, Hollemann T (2010) Xenopus er71 is involved in vascular development. Dev Dyn 239: 3436–3445.
  10. 10. Salanga MC, Meadows SM, Myers CT, Krieg PA (2010) ETS family protein ETV2 is required for initiation of the endothelial lineage but not the hematopoietic lineage in the Xenopus embryo. Dev Dyn 239: 1178–1187.
  11. 11. Proulx K, Lu A, Sumanas S (2010) Cranial vasculature in zebrafish forms by angioblast cluster-derived angiogenesis. Dev Biol 348: 34–46.
  12. 12. Ren X, Gomez GA, Zhang B, Lin S (2010) Scl isoforms act downstream of etsrp to specify angioblasts and definitive hematopoietic stem cells. Blood 115: 5338–5346.
  13. 13. Liu F, Patient R (2008) Genome-wide analysis of the zebrafish ETS family identifies three genes required for hemangioblast differentiation or angiogenesis. Circulation Research 103: 1147–1154.
  14. 14. Mathavan S, Lee SG, Mak A, Miller LD, Murthy KR, et al. (2005) Transcriptome analysis of zebrafish embryogenesis using microarrays. PLoS Genet 1: 260–276.
  15. 15. Covassin L, Amigo JD, Suzuki K, Teplyuk V, Straubhaar J, et al. (2006) Global analysis of hematopoietic and vascular endothelial gene expression by tissue specific microarray profiling in zebrafish. Dev Biol 299: 551–562.
  16. 16. Qian F, Zhen F, Ong C, Jin S-W, Meng Soo H, et al. (2005) Microarray analysis of zebrafish cloche mutant using amplified cDNA and identification of potential downstream target genes. Dev Dyn 233: 1163–1172.
  17. 17. Weber GJ, Choe SE, Dooley KA, Paffett-Lugassy NN, Zhou Y, et al. (2005) Mutant-specific gene programs in the zebrafish. Blood 106: 521–530.
  18. 18. Gomez GA, Veldman MB, Zhao Y, Burgess S, Lin S (2009) Discovery and characterization of novel vascular and hematopoietic genes downstream of etsrp in zebrafish. PLoS ONE 4: e4994.
  19. 19. Wong KS, Proulx K, Rost MS, Sumanas S (2009) Identification of vasculature-specific genes by microarray analysis of Etsrp/Etv2 overexpressing zebrafish embryos. Dev Dyn 238: 1836–1850.
  20. 20. Hegedus Z, Zakrzewska A, Agoston VC, Ordas A, Rácz P, et al. (2009) Deep sequencing of the zebrafish transcriptome response to mycobacterium infection. Mol Immunol 46: 2918–2930.
  21. 21. Ordas A, Hegedus Z, Henkel CV, Stockhammer OW, Butler D, et al. (2010) Deep sequencing of the innate immune transcriptomic response of zebrafish embryos to Salmonella infection. Fish Shellfish Immunol.
  22. 22. Stockhammer OW, Rauwerda H, Wittink FR, Breit TM, Meijer AH, et al. (2010) Transcriptome analysis of Traf6 function in the innate immune response of zebrafish embryos. Mol Immunol 48: 179–190.
  23. 23. Aanes H, Winata CL, Lin CH, Chen JP, Srinivasan KG, et al. (2011) Zebrafish mRNA sequencing deciphers novelties in transcriptome dynamics during maternal to zygotic transition. Genome research.
  24. 24. Soares AR, Pereira PM, Santos B, Egas C, Gomes AC, et al. (2009) Parallel DNA pyrosequencing unveils new zebrafish microRNAs. BMC Genomics 10: 195.
  25. 25. Aday AW, Zhu LJ, Lakshmanan A, Wang J, Lawson ND (2011) Identification of cis regulatory features in the embryonic zebrafish genome through large-scale profiling of H3K4me1 and H3K4me3 binding sites. Dev Biol.
  26. 26. Wheeler DL, Church DM, Federhen S, Lash AE, Madden TL, et al. (2003) Database resources of the National Center for Biotechnology. Nucleic Acids Res 31: 28–33.
  27. 27. Li H, Durbin R (2009) Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25: 1754–1760.
  28. 28. Zakrzewska A, Cui C, Stockhammer OW, Benard EL, Spaink HP, et al. (2010) Macrophage-specific gene functions in Spi1-directed innate immunity. Blood 116: e1–11.
  29. 29. Hikima J, Minagawa S, Hirono I, Aoki T (2001) Molecular cloning, expression and evolution of the Japanese flounder goose-type lysozyme gene, and the lytic activity of its recombinant protein. Biochim Biophys Acta 1520: 35–44.
  30. 30. Nakano T, Graf T (1991) Goose-type lysozyme gene of the chicken: sequence, genomic organization and expression reveals major differences to chicken-type lysozyme gene. Biochim Biophys Acta 1090: 273–276.
  31. 31. Irwin DM, Gong Z (2003) Molecular evolution of vertebrate goose-type lysozyme genes. J Mol Evol 56: 234–242.
  32. 32. Thisse B, Heyer V, Lux A, Alunni V, Degrave A, et al. (2004) Spatial and temporal expression of the zebrafish genome by large-scale in situ hybridization screening. Methods Cell Biol 77: 505–519.
  33. 33. Lamarche-Vane N, Hall A (1998) CdGAP, a novel proline-rich GTPase-activating protein for Cdc42 and Rac. J Biol Chem 273: 29172–29177.
  34. 34. Tcherkezian J, Triki I, Stenne R, Danek EI, Lamarche-Vane N (2006) The human orthologue of CdGAP is a phosphoprotein and a GTPase-activating protein for Cdc42 and Rac1 but not RhoA. Biol Cell 98: 445–456.
  35. 35. Wang L, Yang L, Burns K, Kuan C-Y, Zheng Y (2005) Cdc42GAP regulates c-Jun N-terminal kinase (JNK)-mediated apoptosis and cell number during mammalian perinatal growth. Proc Natl Acad Sci USA 102: 13484–13489.
  36. 36. Wang L, Yang L, Debidda M, Witte D, Zheng Y (2007) Cdc42 GTPase-activating protein deficiency promotes genomic instability and premature aging-like phenotypes. Proc Natl Acad Sci USA 104: 1248–1253.
  37. 37. Engelse MA, Laurens N, Verloop RE, Koolwijk P, van Hinsbergh VWM (2008) Differential gene expression analysis of tubule forming and non-tubule forming endothelial cells: CDC42GAP as a counter-regulator in tubule formation. Angiogenesis 11: 153–167.
  38. 38. Richardson RT, Yamasaki N, O'Rand MG (1994) Sequence of a rabbit sperm zona pellucida binding protein and localization during the acrosome reaction. Dev Biol 165: 688–701.
  39. 39. Wen Y, Richardson RT, Widgren EE, O'Rand MG (2001) Characterization of Sp17: a ubiquitous three domain protein that binds heparin. Biochem J 357: 25–31.
  40. 40. Gupta G, Sharma R, Chattopadhyay TK, Gupta SD, Ralhan R (2007) Clinical significance of sperm protein 17 expression and immunogenicity in esophageal cancer. Int J Cancer 120: 1739–1747.
  41. 41. Li FQ, Liu Q, Han YL, Wu B, Yin HL (2010) Sperm protein 17 is highly expressed in endometrial and cervical cancers. BMC Cancer 10: 429.
  42. 42. Straughn JM, Shaw DR, Guerrero A, Bhoola SM, Racelis A, et al. (2004) Expression of sperm protein 17 (Sp17) in ovarian cancer. Int J Cancer 108: 805–811.
  43. 43. Adams J, Kelso R, Cooley L (2000) The kelch repeat superfamily of proteins: propellers of cell function. Trends Cell Biol 10: 17–24.
  44. 44. Xue F, Cooley L (1993) kelch encodes a component of intercellular bridges in Drosophila egg chambers. Cell 72: 681–693.
  45. 45. Braybrook C, Warry G, Howell G, Arnason A, Bjornsson A, et al. (2001) Identification and characterization of KLHL4, a novel human homologue of the Drosophila Kelch gene that maps within the X-linked cleft palate and Ankyloglossia (CPX) critical region. Genomics 72: 128–136.
  46. 46. Fonseca MI, Carpenter PM, Park M, Palmarini G, Nelson EL, et al. (2001) C1qR(P), a myeloid cell receptor in blood, is predominantly expressed on endothelial cells in human tissue. J Leukoc Biol 70: 793–800.
  47. 47. Petrenko O, Beavis A, Klaine M, Kittappa R, Godin I, et al. (1999) The molecular characterization of the fetal stem cell marker AA4. Immunity 10: 691–700.
  48. 48. Birkenbach M, Josefsen K, Yalamanchili R, Lenoir G, Kieff E (1993) Epstein-Barr virus-induced genes: first lymphocyte-specific G protein-coupled peptide receptors. J Virol 67: 2209–2220.
  49. 49. Gatto D, Paus D, Basten A, Mackay CR, Brink R (2009) Guidance of B cells by the orphan G protein-coupled receptor EBI2 shapes humoral immune responses. Immunity 31: 259–269.
  50. 50. Pereira JP, Kelly LM, Xu Y, Cyster JG (2009) EBI2 mediates B cell segregation between the outer and centre follicle. Nature 460: 1122–1126.
  51. 51. Rosenkilde MM, Benned-Jensen T, Andersen H, Holst PJ, Kledal TN, et al. (2006) Molecular pharmacological phenotyping of EBI2. An orphan seven-transmembrane receptor with constitutive activity. J Biol Chem 281: 13199–13208.
  52. 52. Grosskurth SE, Bhattacharya D, Wang Q, Lin JJ-C (2008) Emergence of Xin demarcates a key innovation in heart evolution. PLoS ONE 3: e2857.
  53. 53. Choi S, Gustafson-Wagner EA, Wang Q, Harlan SM, Sinn HW, et al. (2007) The intercalated disk protein, mXinalpha, is capable of interacting with beta-catenin and bundling actin filaments [corrected]. J Biol Chem 282: 36024–36036.
  54. 54. Avruch J, Xavier R, Bardeesy N, Zhang X-F, Praskova M, et al. (2009) Rassf family of tumor suppressor polypeptides. J Biol Chem 284: 11001–11005.
  55. 55. Eckfeld K, Hesson L, Vos MD, Bieche I, Latif F, et al. (2004) RASSF4/AD037 is a potential ras effector/tumor suppressor of the RASSF family. Cancer Res 64: 8688–8693.
  56. 56. Kanamoto T, Mizuhashi K, Terada K, Minami T, Yoshikawa H, et al. (2009) Isolation and characterization of a novel plasma membrane protein, osteoblast induction factor (obif), associated with osteoblast differentiation. BMC Dev Biol 9: 70.
  57. 57. Kumar CC, Mohan SR, Zavodny PJ, Narula SK, Leibowitz PJ (1989) Characterization and differential expression of human vascular smooth muscle myosin light chain 2 isoform in nonmuscle cells. Biochemistry 28: 4027–4035.
  58. 58. Licht AH, Nübel T, Feldner A, Jurisch-Yaksi N, Marcello M, et al. (2010) Junb regulates arterial contraction capacity, cellular contractility, and motility via its target Myl9 in mice. J Clin Invest 120: 2307–2318.
  59. 59. Jalagadugula G, Mao G, Kaur G, Goldfinger LE, Dhanasekaran DN, et al. (2010) Regulation of platelet myosin light chain (MYL9) by RUNX1: implications for thrombocytopenia and platelet dysfunction in RUNX1 haplodeficiency. Blood 116: 6037–6045.
  60. 60. Topczewska JM, Topczewski J, Szostak A, Solnica-Krezel L, Hogan BLM (2003) Developmentally regulated expression of two members of the Nrarp family in zebrafish. Gene Expr Patterns 3: 169–171.
  61. 61. Phng LK, Potente M, Leslie JD, Babbage J, Nyqvist D, et al. (2009) Nrarp coordinates endothelial Notch and Wnt signaling to control vessel density in angiogenesis. Dev Cell 16: 70–82.
  62. 62. Lee SJ, Chan TH, Chen TC, Liao BK, Hwang PP, et al. (2008) LPA1 is essential for lymphatic vessel development in zebrafish. FASEB J 22: 3706–3715.
  63. 63. Rodriguez F, Vacaru A, Overvoorde J, den Hertog J (2008) The receptor protein-tyrosine phosphatase, Dep1, acts in arterial/venous cell fate decisions in zebrafish development. Dev Biol 324: 122–130.
  64. 64. Thisse C, Thisse B (2008) High-resolution in situ hybridization to whole-mount zebrafish embryos. Nat Protoc 3: 59–69.
  65. 65. Mortazavi A, Williams BA, McCue K, Schaeffer L, Wold B (2008) Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat Methods 5: 621–628.