Figures
Abstract
Understanding the molecular interactions that lead to the establishment of the major body axes during embryogenesis is one of the main goals of developmental biology. Although the past two decades have revolutionized our knowledge about the genetic basis of these patterning processes, the list of genes involved in axis formation is unlikely to be complete. In order to identify new genes involved in the establishment of the dorsoventral (DV) axis during early stages of zebrafish embryonic development, we employed next generation sequencing for full transcriptome analysis of normal embryos and embryos lacking overt DV pattern. A combination of different statistical approaches yielded 41 differentially expressed candidate genes and we confirmed by in situ hybridization the early dorsal expression of 32 genes that are transcribed shortly after the onset of zygotic transcription. Although promoter analysis of the validated genes suggests no general enrichment for the binding sites of early acting transcription factors, most of these genes carry “bivalent” epigenetic histone modifications at the time when zygotic transcription is initiated, suggesting a “poised” transcriptional status. Our results reveal some new candidates of the dorsal gene regulatory network and suggest that a plurality of the earliest upregulated genes on the dorsal side have a role in the modulation of the canonical Wnt pathway.
Citation: Fodor E, Zsigmond Á, Horváth B, Molnár J, Nagy I, Tóth G, et al. (2013) Full Transcriptome Analysis of Early Dorsoventral Patterning in Zebrafish. PLoS ONE 8(7): e70053. https://doi.org/10.1371/journal.pone.0070053
Editor: Amit Singh, University of Dayton, United States of America
Received: March 16, 2013; Accepted: June 14, 2013; Published: July 29, 2013
Copyright: © 2013 Fodor et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research was supported by a Zoltán Magyary Postdoctoral Fellowship (T/213/2009) to MV and a BBSRC project grant to MV and SWW. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: At the time of manuscript preparation co-author János Molnár was employed by a commercial company, Biomi Ltd. This does not alter the authors’ adherence to all the PLOS ONE policies on sharing data and materials. All other authors declare no conflict of interest.
Introduction
The development of bilaterian embryos from a single-celled, fertilized egg into complex, multicellular, three-dimensional structure (the embryo) involves not only several rounds of cell divisions, but also a series of well-coordinated morphogenetic movements and patterning events, such as the establishment of the anteroposterior (AP) and dorsoventral (DV) axes.
Ever since the pioneering work of Hans Spemann and Hilde Mangold in the 1920s, it has been well known that during the formation of the DV-axis a specialized tissue with organizing properties is established in the future dorsal side of the embryo [1]. This tissue, called the Spemann organizer in amphibians, secretes signals that can instruct neighboring cells to form an axis. The molecular nature of these signals remained elusive for decades but with the introduction of molecular tools in developmental biology it became clear that several genes with organizing properties encode proteins that can antagonize BMP-signaling [2]–[7]. A parallel discovery of the “neural default model”, the property of BMP-signaling to suppress neural fate in ectodermal cells [8], [9] led to the recognition of the BMP-pathway as a major component of DV-axis patterning.
Zebrafish (Danio rerio) has become a widely used model organism, due to its fast development, high progeny number, transparency and ease of use. These advantages made it the subject of multiple extremely successful genetic screens, which have identified many key genes in diverse developmental processes, DV patterning being one of these [10]. From these screens, mutations that lead to DV patterning defects have often turned out to occur in genes that encode elements of the BMP signaling pathway (see [11] and references therein). However, some other key factors of early DV patterning have also been identified both by these screens and by other reverse genetic approaches [12]–[15]. The picture that emerges from these studies suggests that early zebrafish DV patterning is following a similar logic to that observed in the African clawed frog, Xenopus laevis [16].
A major similarity between the two anamniote model organisms is the pivotal early role of the canonical Wnt/β-catenin signaling pathway in dorsal determination. After fertilization, a dorsal determinant, Wnt11 in frogs [17], [18] and wnt8 mRNA in fish [19], is transferred from the vegetal part of the oocyte to the future dorsal side. In zebrafish this process is mediated by an active, microtubule-dependent process [19]–[21] and results in the activation of canonical Wnt/β-catenin signaling on the presumptive dorsal side. Consequently, the dorsal transcriptional network is activated, which will ultimately lead to the expression of BMP-antagonists and the formation of the BMP-signaling gradient across the DV-axis. This activity gradient will have powerful patterning effects across the ectoderm and the mesoderm [22], [23].
The recessive, maternal-effect ichabod (ich) mutation causes severe ventralization due to an impairment in the early Wnt/β-catenin signaling pathway in embryos derived from mothers homozygous for the mutation (ich embryos) [24]. This defect is the result of the decreased accumulation of maternal β-catenin-2 mRNA in the oocytes, caused most likely by a regulatory mutation [25].
Ventralized ich embryos lack a “shield” (the zebrafish equivalent of the Spemann organizer) during gastrulation and do not express any of the genes characteristic for organizer formation [24]. BMP-activity is also uniform across the ectoderm [26]. As the mutation can be completely rescued by the introduction of ectopic β-catenin-2 mRNA, ich embryos provide an excellent paradigm to study the formation of the early organizer and to identify genes involved in this process. Using this experimental logic, previously we were able to discern the epistatic relationship between the Fgf-, Nodal- and BMP-signaling pathways [27].
In order to obtain a genome-wide view on the transcriptional changes related to the specification of the dorsal side in the early zebrafish embryos and to identify novel components of early DV patterning, we have sequenced full transcriptomes of untreated and rescued ich embryos at sphere stage (shortly after the midblastula transition – MBT –, the onset of zygotic transcription [28]) using RNA-Seq.
After mapping and analyzing one hundred million tags per sample and validating our results by in situ hybridization, we identified 32 genes differentially upregulated in rescued embryos, many of them implicated by prior studies in DV patterning and/or axial mesoderm formation. Our analysis shows that the largest group of activated genes are modulators of canonical Wnt-signaling. This suggests the existence of a robust self-regulatory network, to coordinate the dynamic changes in canonical Wnt activity during early stages of development.
Results and Discussion
Identification of Genes Expressed during Early Dorsal Induction using RNA-Seq
The ability to completely rescue the severely ventralized ich embryos created the prospect to analyze the early induction of the dorsal organizer in a completely unbiased manner using next generation sequencing methods. As the earliest molecular manifestation of the future dorsal organizer occurs around MBT, we decided to compare rescued and untreated embryos at this early stage, so we can identify the most upstream components of the dorsal induction network (Figure 1A). In order to avoid false positive hits arising from forced expression of Wnt-target genes, we injected β-catenin-2 mRNA at concentrations that create overtly normal looking (non-dorsalized) embryos.
(A) Total RNA samples were isolated from untreated (ventralized) ich embryos, and from “rescued” (β-catenin mRNA injected) counterparts. The samples were used for RNA-Seq, and the resulting datasets were compared to uncover genes upregulated during early dorsal specification. (B) Two independent in silico approaches were used to identify differentially expressed genes: DESeq and Cufflinks. The two scripts resulted in overlapping, but not identical gene-sets, which were combined for further analysis. (Grey demarks genes where differential dorsal expression could not be confirmed– see text for details.).
Using sequencing-by-ligation on the SOLiD V4 platform, we sequenced over one hundred million tags per sample. We first filtered and mapped the reads to the zebrafish genome (Zv9– Ensembl) using Genomics Workbench ver4.6. Next, in order to identify genes showing significantly altered expression level in “rescued” over untreated mutant (ich) sample we used a dual approach. On the one hand, we employed DESeq [29], a package developed to analyse count data from high-throughput sequencing (Figure S1A). For positive hits the alignments were manually analyzed in order to filter out false positive results. This conservative approach identified 32 genes, 17 of which have been implicated in organizer and/or axial mesoderm formation by previous studies.
In parallel we also used TopHat [30] splice junction mapper and Cufflinks [30], a package that assembles transcripts, estimates their abundances, and tests for differential expression in RNA-Seq samples (Figure S1B). After manual curation this approach resulted in 25 hits, with 12 of them previously known for their role in dorsal mesoderm formation.
The fact that established “dorsal” genes were overrepresented in our datasets underscored the strength of this approach and suggested that other genes on the list would prove to be bona fide components of the early DV patterning network.
The two datasets (DESeq and Cufflinks) only partially overlapped: just 15 genes were found in both lists (Figure 1B), with prominent dorsally-expressed genes missing from both (e.g. gsc in the case of DESeq and dharma for Cufflinks). As both datasets produced only conservative estimates, this was not surprising and showed the complementarity of the two approaches. Consequently for further analysis we used the combination of the two datasets.
To further test if our results consist of truly zygotic, early expressed genes, we compared our results against an independent dataset. Using data from a recent study of maternal-to-zygotic transition in zebrafish [31] (GEO database, accession number GSE22830), we found that most of the identified genes are indeed upregulated during MBT (Figure S2). Two genes, map2k6 and tmem68 had to be excluded from the combined dataset, as they showed, respectively, decreased and unchanged expression after MBT.
Validation of the Results with in situ Hybridization
A further validation of our in silico dataset came from a series of in situ hybridization experiments, where we tested the expression pattern of the examined genes in both wild type and ich embryos at sphere stage and at 30% epiboly stage. In our previous studies we already performed such comparative expression studies for a small subset of key organizer-specific genes, such as dharma, gsc and chd [25]. We were able to obtain specific probes for 36 genes out of our 39 tested genes (the only exceptions being arrdc3b, clec19a and rbl2).
On the basis of the expression data, we observed two major phenotypic classes. Most genes showed early dorsal expression in wild type embryos, followed by circumferential upregulation in the germ ring at 30% epiboly stage (Figure 2). In ich embryos, while the early dorsal wave of expression was absent, the later circumferential expression was present (Figure 2). Other genes, similar to the dorsal-specific gsc, kept an “all dorsal” expression profile in wild type controls and showed restricted expression even at 30% epiboly (Figure 3). Such genes, as expected, were absent in ich embryos even at the onset of gastrulation. An interesting exception was rasl11b, which had an early dorsal expression domain (Figure 3Fi), but by mid-gastrula stages it was restricted to the ventral side of wild type embryos (Figure 3Fiii). In ventralized ich embryos, the early expression domain was missing (Figure 3Fii), while in later stage embryos the gene was expressed ubiquitously (Figure 3Fiv).
The majority of the identified genes show early dorsal expression in wild type embryos and lack thereof in ich controls. This is followed by a second wave of circumferential, mesodermal expression, which could be detected in both genetic backgrounds. An interesting exception is flh, which shows a somewhat counterintuitive expression dynamics: a very strong circumferential expression (complemented by strong dorsal signal in wild type embryos) can be detected as early as late sphere stage (Ei, Eii), but this expression becomes significantly weaker by 30% epiboly, except in the dorsal side of wild type embryos (Eiii, Eiv). Sphere stage embryos are shown from a lateral view, whereas ∼30% epiboly stage embryos are presented from an animal view (dorsal to the right in both cases).
The majority of genes in this category can be characterized by a “dorsal-only” expression pattern: they are expressed both at sphere and at 30% epiboly stages in the dorsal side of wild type controls and are absent from ich embryos. The exception is rasl11b which shows a strong dorsal, and a fainter ubiquitous expression at sphere stage in wild types (Fi), but by 30% epiboly it is absent from the dorsal side and expression can be detected only ventrally (Fiii). Accordingly, in ich embryos lacking dorsal induction, faint ubiquitous expression (Fii) becomes progressively stronger during development (Fiv). Sphere stage embryos are shown from a lateral view, whereas ∼30% epiboly stage embryos are presented from an animal view (dorsal to the right in both cases).
Two genes, LOC100334443 and dnase1l3 were upregulated only at MBT, but no significant difference between expression in wild type and ich embryos could be detected for either of them prior to late stages of gastrulation, when both were found to be specifically expressed in the dorsal forerunner cells (DFCs) of control embryos (Figure 4 and not shown).
(A) ier2 expression starts early in the future dorsal side, persists until early mid-gastrulation, then fades away, and becomes detectable only during somitogenesis in a stripe in the hindbrain area, and in the tailbud. (B) rnd1l expression is weakly upregulated in all the blastomeres after MBT, and it is expressed at higher levels in a few dorsal precursors. Later the expression becomes restricted in the mesoderm (both axial and non-axial), where it is progressively downregulated. During somitogenesis two prominent neural expression domains appear: in the eyefield and in the hindbrain. (C) ppp1r3ca expression starts at the dorsal edge of the embryo, and later it is upregulated in a narrow stripe around the germring. Expression weakens during late gastrula stages, but it will be upregulated during somitogenesis in the anterior forebrain and in the posterior tailbud regions. (D) LOC100334443 expression is strongly upregulated after the onset of zygotic transcription. During gastrulation it becomes restricted to the ectoderm, where it is gradually downregulated, except in the dorsal forerunner cells. During somitogenesis expression is also prominent in the neurectoderm, especially in the hindbrain. All embryos are shown from lateral view, dorsal to the right.
Other genes, such as aplnrb and isg15 completely lacked dorsal expression during the earliest stages of post-MBT development, and were present right after MBT in ich embryos, too. Both of these genes showed an apparently randomized expression before 30% epiboly, and later became restricted to the germring (Figure S3A-N and [32], [33]). As the expression pattern after MBT for these two genes ranged from a few cells to larger clusters, it is likely that the expression level differences detected by DESeq and Cufflinks were genuine, yet they reflected a stochastic difference only, thus these genes could be classified clearly as false positives.
We also observed foxo3b expression showing a similar, stochastic dynamic during blastula and gastrula stages (Figure S3O–H’). This is in contrast with the quasi-ubiquitous expression described before [34], however, repeated experiments yielded the same results. Of note, though expression seemed stochastic both in wild type and ich embryos, the overall expression appeared higher in wild types during early stages of development.
All three genes manifesting stochastic expression were previously associated with cellular and physiological stress [35]–[37], thus their collective post-MBT expression suggests that the blastomeres are particularly prone to stress at this stage.
Promoter Analysis and Histone Modifications in Dorsal Genes
Previous work has established the core of the gene regulatory network (GRN) responsible for establishing the DV axis after the onset of the zygotic transcription [38]. While on the dorsal side canonical Wnt/β-catenin signaling, directly or indirectly, induces the expression of typical dorsal genes, on the ventral side maternal and then zygotic Pou5f1/Spg (the zebrafish homolog of mammalian Oct4), in combination with the homeobox repressors Vox, Vent and Ved inhibits the activation of the same dorsal genes (Figure 5A).
(A) Previous work has identified the gene regulatory network (GRN) involved in the DV patterning of the zebrafish embryo. While in the dorsal side canonical Wnt signaling activates directly or indirectly the genes involved in the induction of the organizer, ventrally Vox/Vent/Ved and Pou5f1 represses the expression of these genes. (B) We tested the upstream 4 kb genomic region of the candidate genes for a possible enrichment in the binding sites of key transcription factors. There was no significant difference between the validated genes and a random gene set. (As expected, significant difference in the number of TcfLef binding sites could be detected between the promoter region of the confirmed Wnt-target subset of our gene-set, and the randomized set of genes (p<0.05).) (C) Using a previously published dataset we were able to demonstrate that the histones associated to our candidate genes often show a “bivalent status”, carrying activating (H3K4me3) and repressing (H3K27me3) methylation marks at the same time. This is in stark contrast with the histone-methylation patterns of randomly selected genes.
We wanted to test if the promoter regions of the candidate genes show enrichment for the binding sites of the aforementioned ventral transcription factors, or for recognition motifs of TcfLef, the transcriptional effectors of canonical Wnt-signaling. Therefore, using the Transcription Element Search System (TESS) script [39] we counted the number of such binding sites in the putative regulatory genomic regions of the analyzed genes. As transgenic reporter lines mimicking the expression of early transcribed developmental regulators often use 1–3 kb of the upstream genomic sequence [40]–[43], we decided to concentrate our efforts to the upstream 4 kb region.
Compared with a random set of zebrafish genes, we could not detect a statistically significant difference in the number of Pou5f1 and TcfLef binding sites (Figure 5B). When examining only the promoter regions of previously confirmed Wnt target genes in our gene-set, we did observe a significant enrichment in TcfLef sites, suggesting that our approach was sound. It appears then that the majority of dorsal genes detected by our transcriptome comparison are not direct targets of canonical Wnt-signaling. We also note that a tendency for enrichment for XVent1 binding sites can also be seen in the data, yet this difference was not large enough to be statistically significant (Figure 5B).
Of note, the enrichment of the TcfLef sites could not observed when we analyzed genomic regions that covered the first introns, or the −8 to −4 kb upstream regions or the downstream 4 kb genomic sequences of the validated and random genes (Figure S4). Interestingly, the analysis of the downstream 4 kb regions suggested a statistically significant difference in the number of Pou5f1 binding sites between the validated set and the random set of genes (Figure S4). This result, however, should be interpreted cautiously, as when the respective genomic regions of the validated genes known to be regulated by Pou5f1 [44] were compared with the rest of the validated dataset, no difference could be observed in the number of the Pou5f1 binding sites (not shown).
Recent results suggest that the nucleosomes at the promoters of developmentally important genes often acquire activating trimethylation marks on the lysine at position four (K4) in histone 3 (H3), or a “bivalent” mark of both activating K4 and inhibiting H3 lysine 27 (K27) trimethylation during transcriptional activation in zebrafish [45]–[47]. Using the recently published data of Lindeman and coworkers [46], we tested if our validated genes show similar patterns of epigenetic modifications.
Our validated “dorsal” gene set show an increase in the activating H3 K4 trimethylation marks in the promoter region, compared with the random set of genes, and an overabundance of “bivalent” histone-marks at the onset of zygotic transcription (Figure 5C), showing that dorsal genes are epigenetically marked prior to transcriptional activation. (Of note, according to the Lindeman et al. dataset [46] dharma, one of the earliest transcribed genes in the zebrafish genome [48], carries only H3K27 trimethylation during the time of its transcription. This could be an error in the dataset, but could also suggest a peculiar epigenetic regulatory mechanism at this genomic locus.)
Many of the Dorsal Genes are Modulators of Wnt-signaling
We performed a manual literature search to assign biological function to the validated dataset. The results showed that the single biggest class of genes (11/32) was associated with modulation of Wnt-signaling, followed by regulators of cell motility (5/32) and modulators of Fgf-signaling (4/32) (Table 1).
The enrichment in Wnt modulators is not surprising, as previous studies already demonstrated that several direct targets of Wnt-signaling are at the same time feed-back regulators of the pathway [49], [50]. Indeed, given the importance of Wnts during early zebrafish development a tight regulation can be expected. The close relationship between the pathway and its regulators is also underscored by an apparent enrichment in TcfLef binding sites in the promoter regions of the respective genes (Figure 5B). Furthermore, the “bivalent” epigenetic status observed at the majority of the Wnt-modulators (Figure 6) also suggests a tight and efficient regulation of this regulatory module.
A close examination revealed that in our validated dataset genes that modulate Wnt-signaling are overrepresented. Interestingly, genes that are solely expressed in the dorsal side act usually as transcriptional repressors, while those genes that are also expressed in the ventrolateral mesodermal domains interact with the Wnt signaling pathway post-translationally. (Although foxa3 and mst1 both show circumferential expression, the dorsal expression domain is much more prominent for both genes, therefore we treated them as intermediates.) The most likely reason for this difference is that while dorsally Wnt-signaling has to be completely abolished to avoid its later ventralizing effects, in other domains the role of feedback repressors is just to ensure that Wnt-signaling is not overactivated. The majority of the Wnt-modulators carries “bivalent” H3K4/H3K27 histone trimethylation marks, suggesting they are in a “poised” status during MBT, which could also enhance the tight temporal and spatial control of the pathway. (In the case of zic2.1 the “bivalent” status could be detected only prior MBT, and not during MBT, indicated by the lighter orange color.).
Regulation of Wnt-signaling can happen at the transcriptional level (by inhibiting/activating the expression of specific components of the canonical pathway) or at the translational level (by modulating the function of proteins). Interestingly, Wnt-modulators associated with transcriptional repression showed an elevated dorsal (foxa3) or dorsal-only (dharma, gsc) expression pattern, while the majority of post-translational repressors (dkk1b, notum1a, zic2.1, zic3) were upregulated circumferentially in the germ-ring at the onset of gastrulation (Figure 6).
These differences can be explained in the context of the changing role of Wnt-signaling during early development. While the early activation of the canonical Wnt/β-catenin pathway is indispensable for the formation of the organizer and the specification of dorsal cell fates, at the onset of gastrulation Wnt (specifically Wnt8) becomes a potent morphogen with ventralizing and posteriorizing effects [11]. Due to this later role, pathway activation has to be completely excluded from the dorsal side.
The second wave of Wnt-signaling is driven by the activation of wnt8 in the germring, by Nodal-signaling emanating from the yolk syncitial layer (YSL) [51]–[53]. As this signal is circumferential, dorsal transcriptional repressors counteract its effects by inhibiting the activation of wnt8 in this region.
When Wnt-signaling becomes a potent posteriorizer during gastrulation, its levels have to be tightly regulated. Too much Wnt activity results in the expansion of posterior tissues at the expense of anterior ones, while reduced Wnt-signaling has the opposite effect. To achieve such a fine-tuning of Wnt activity, post-translational modulation is an ideal tool. While not interfering with the source of Wnt-signaling, the circumferentially expressed Wnt-suppressors can set the level of Wnt activity through well characterized negative feed-back mechanisms.
Conclusions
Using an unbiased full transcriptome sequencing approach we generated a comprehensive list of post-MBT dorsally induced genes in the zebrafish embryo (Figure 1B). Our results suggest that the majority of these genes are epigenetically marked during the activation of zygotic transcription, often with “bivalent” H3 K4/K27 trimethylation (Figure 5C). Given that in Xenopus dorsal β-catenin has a role in establishing a “poised“ chromatin state prior to MBT [54], we propose that ich embryos, due to insufficient β-catenin-2 levels [25], lack these chromatin modifications and consequently fail to induce their dorsal developmental programs.
We also show that many of the dorsally activated genes are modulators of the Wnt-pathway, but components of the Fgf- and Nodal-signaling pathways are also induced together with genes important for cell motility and adhesion (Table 1).
Of note is the absence of sqt/ndr1 from our list, as it was previously shown to be an early essential component of the dorsal induction pathway [27]. This finding is most likely explained by the maternal component of ndr1, also observed in ich embryos [55], wich will buffer the changes occurring during MBT at the dorsal side. The activation of the Fgf-signaling pathway, itself dependent on Nodal-signaling [27] is also indicative that the dorsal activation of ndr1 in fact occurs.
It is also noteworthy that several genes on our list are not direct targets of Wnt-signaling. As mentioned above, the transcription of Fgf ligands requires prior activation of Nodal-signaling. Furthermore, other genes, such as the Fgf-signaling feedback inhibitors spry2 and spry4, and the BMP-antagonist chd are dependent on the activation of Fgf-pathway [27], [56], [57]. Taken together, this suggests that several levels of the dorsal GRN are activated promptly and quickly after MBT.
Finally, although most of the examined genes showed an early dorsal activity, followed by a later, pan-mesodermal expression in the germ-ring, a couple of genes were restricted to the dorsal side. It will be interesting to understand the transcriptional logic regulating the expression of the latter genes, as it might reveal how context-dependent gene activation can be encoded on genomic level.
Our analysis also determined the expression pattern during early development for several, previously uncharacterized genes, such as ier2, rnd1l, ppp1r3ca and LOC100334443 (Figure 4). Although not all of them show restricted dorsal expression during blastula and early gastrula stages, for those that do, it will be interesting to determine how they fit into the established genetic network of dorsal induction in zebrafish embryos.
Materials and Methods
Fish Care
Wild-type ekwill (ekw) and mutant ichp1/p1 fish stocks were maintained in the animal facility of Eötvös Loránd University. All protocols used in this study were approved by the Hungarian National Food Chain Safety Office (Permit Number: XIV-I-001/515-4/2012).
RNA Preparation
Rescue experiments were performed as described before [27]. Total RNA was isolated from 50–50 uninjected and β-catenin-2 mRNA injected embryos with TRIZOL (Invitrogen), using the manufacturer’s protocol. An extra round of ethanol precipitation was applied in order to eliminate residual TRIZOL contaminations.
Whole Transcriptome Sequencing
RNA quality and quantity measurements were performed on Bioanalyzer (Agilent Technologies) and Qubit (Life Technologies). High quality (RIN >8.5) total RNA samples from three biological replicates were pooled and processed using the SOLiD total RNA-Seq Kit (Life Technologies), according to the manufacturers suggestions. Briefly, 5 µg of pooled RNA was DNaseI treated and the ribosomal RNA depleted using Eukaryote RiboMinus rRNA Removal Kit (Life Technologies). The leftover was fragmented using RNaseIII, the 50–200 nt fraction size-selected, sequencing adaptors ligated and the templates reverse transcribed using ArrayScript RT. The cDNA library was purified with Qiagen MinElute PCR Purification Kit (Qiagen) and size-selected on a 6% TBE-Urea denaturing polyacrylamide gel. The 150–250 nt cDNA fraction was amplified using AmpliTaq polymerase and purified by AmPureXP Beads (Agencourt). Concentration of each library was determined using the SOLiD Library TaqMan Quantitation Kit (Life Technologies). Each library was clonally amplified on SOLiD P1 DNA Beads by emulsion PCR (ePCR). Emulsions were broken with butanol, and ePCR beads enriched for template-positive beads by hybridization with magnetic enrichment beads. Template-enriched beads were extended at the 3′ end in the presence of terminal transferase and 3′ bead linker. Beads with the clonally amplified DNA were deposited onto sequencing slide and sequenced on SOLiD V4 Instrument using the 50-base sequencing chemistry.
Data Availability
Short-read data of the two sequenced transcriptomes were deposited in NCBI’s Short Read Archive at http://www.ncbi.nlm.nih.gov/sra/under accession SRA075737.
Bioinformatics
Bioinformatic analysis of the whole transcriptome sequencing was performed in color space using Genomics Workbench ver4.6 (CLC Bio). Raw sequencing data were trimmed by removal of low quality, short sequences so that only 50 nucleotide long sequences were used in further analysis. Sequences were mapped in a strand specific way onto the Danio rerio Zv9 (Ensembl) reference genome, using default parameters. In order to avoid possible false positive hits, genes where the mapped reads showed a highly skewed distribution in the “rescued” dataset were removed from further analysis. DESeq results were manually curated, to remove (further) false positive hits, which showed highly skewed mapping of reads. Only genes where a 1.5 fold upregulation was detected after normalization were considered for further analysis.
As an alternative approach, the 50 nucleotide long color-space RNA-Seq data were mapped onto the Danio rerio Zv9 reference using TopHat v1.3.2 [58], which, by allowing for recognition of splice junctions, facilitates correct mapping of sequencing reads that span multiple exons. Assembly of transcripts and estimation of their relative abundance between the two samples was carried out using Cufflinks v1.1.0 [59]. For candidate genes the alignments in the rescued dataset were manually checked to avoid false positive arising from uneven distribution of reads. (In a typical false positive sample more than half of the reads mapped to a short, 20–30 bp long segment.).
For promoter analysis the indicated genomic sequences were downloaded from Ensembl both for our validated gene set and a randomized set of genes, and analyzed using TESS (http://www.cbil.upenn.edu/tess) [39] for TcfLef, XVent1 and Pou5f1 binding sites. Only hits with a log-odd score better than 8.0 were counted. The number of such binding sites for each gene was recorded (Table S1) and analyzed with the R program package (http://www.r-project.org).
Histone modification status for the validated and random genes was determined using the dataset of Lindeman and coworkers [46]. Data visualization was performed with the ggplot2 package (http://ggplot2.org).
The random gene-set was generated using the corresponding application of the RSA-tools package (http://rsat.ulb.ac.be/rsat/random-genes_form.cgi) and can be found in Table S1.
In situ Hybridization
Whole mount in situ hybridization stainings were performed as described before [25]. The plasmids used for probe synthesis were as follows: pCS2+aplnrb [32], pZL-dkk1 [60], pCS2+fgf8 [61], pcDNA3.0-fgf17b [62], pBS-flh [63], pBS-foxa3 [64], pBS-foxb1.2 [64], pBS-foxd3 [64], pGEMT-foxd5 [65], pSPORT1-ism1 [66], pBS-lft1 [67], pGEMT-mixer [68], pBS-ntl [69], osr1 [70], pCRII-TOPO-sp5l [66], pBS-zspry2 [71], pBS-spry4 [56], pBS-zic2.1 [72], zic3 [73]. The remainder genes were cloned through nested PCRs and cloned into pCRII-TOPO-Blunt (Invitrogen) or pGEM-T-Easy (Promega) vectors. PCR primers were designed based on the available annotations in the Ensembl database, and their sequences can be found in the Methods S1 file.
Supporting Information
Figure S1.
Identification of candidate genes. Genes showing a significant (more than 1.5 fold in the case of DESeq) upregulation after normalization were considered for further analysis. Two complementary approaches, DESeq ([29] in the Main Text) (A) and Cufflinks ([30] in the Main Text) (B) packages were used to identify differentially regulated genes. Positive hits are colored red.
https://doi.org/10.1371/journal.pone.0070053.s001
(TIF)
Figure S2.
Majority of the candidate genes are upregulated after MBT. Using a previously published dataset ([31] in Main Text), we tested whether the expression of our candidate genes is upregulated at MBT, as expected. Our results show that indeed, this is the case for all, except two genes: map2k6 and tmem68. The former showed a weak maternal expression, which was downregulated after MBT, whereas the latter showed no change in expression during and after MBT.
https://doi.org/10.1371/journal.pone.0070053.s002
(TIF)
Figure S3.
Genes with stochastic early expression. Genes belonging to this class show stochastic expression at the onset of the zygotic transcription. We could detect clusters of cells both in wild type and ich embryos that upregulated these particular genes, but no clear pattern emerged. For aplnrb (A–F) and isg15 (G–N), the expression pattern later became more coherent, and localized around the germring. Interestingly, although no clear expression pattern emerged, foxo3b expression in general appeared higher in wild type embryos (O–V). Later stages of foxo3b expression in wild type embryos also showed stochastic and varying expression patterns (A’–H’). All sphere and 30% epiboly stage embryos are shown from an animal view. Shield, 70% epiboly and bud stage embryos are shown from a lateral view, with dorsal to the right.
https://doi.org/10.1371/journal.pone.0070053.s003
(TIF)
Figure S4.
Extended genomic region analysis of validated genes. (A) For multi-exon genes, statistical analysis of the first introns for enrichment in putative TcfLef, XVent1 or Pou5f1 binding sites yielded negative results. (B) In the 4 kb downstream genomic regions, we could not detect significant differences between the random and validated gene-sets in the number of TcfLef and XVent1 sites. The validated gene-set was enriched however in putative Pou5f1 sites (p<0.05). However, this result has to be interpreted with caution, as proven Pou5f1-targets [44] within the validated set have about the same number of potential Pou5f1 binding sites as non-target counterparts (not shown). (C) Analysing the −8 kb to −4 kb upstream genomc region no differences were observed in the number of TcfLef, XVent1 and Pou5f1 sites between our validated and random data-sets.
https://doi.org/10.1371/journal.pone.0070053.s004
(TIF)
Table S1.
In silico binding site analysis for candidate and randomly selected genes.
https://doi.org/10.1371/journal.pone.0070053.s005
(XLS)
Methods S1.
Sequence of PCR primers used for the nested PCR reactions.
https://doi.org/10.1371/journal.pone.0070053.s006
(DOC)
Acknowledgments
We would like to express gratitude to all the colleagues who provided us plasmids to help our work. We are grateful to the Sequencing Platform of the Biological Research Centre, Hungarian Academy of Sciences for RNA sequencing. The supercomputing infrastructure provided by the Hungarian National Information Infrastructure Development (NIIF) Program was used for parallel computing during Cufflinks and TopHat analysis. MV is a János Bolyai Fellow of the Hungarian Academy of Sciences.
Author Contributions
Conceived and designed the experiments: MV. Performed the experiments: EF ÁZ BH MV. Analyzed the data: BH JM GT IN MV. Contributed reagents/materials/analysis tools: SWW. Wrote the paper: MV.
References
- 1. Spemann H, Mangold H (2001) Induction of embryonic primordia by implantation of organizers from a different species. 1923. Int J Dev Biol 45: 13–38.
- 2. Smith WC, Harland RM (1992) Expression cloning of noggin, a new dorsalizing factor localized to the Spemann organizer in Xenopus embryos. Cell 70: 829–840.
- 3. Sasai Y, Lu B, Steinbeisser H, Geissert D, Gont LK, et al. (1994) Xenopus chordin: a novel dorsalizing factor activated by organizer-specific homeobox genes. Cell 79: 779–790.
- 4. Piccolo S, Sasai Y, Lu B, De Robertis EM (1996) Dorsoventral patterning in Xenopus: inhibition of ventral signals by direct binding of chordin to BMP-4. Cell 86: 589–598.
- 5. Zimmerman LB, De Jesus-Escobar JM, Harland RM (1996) The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell 86: 599–606.
- 6. Fainsod A, Deissler K, Yelin R, Marom K, Epstein M, et al. (1997) The dorsalizing and neural inducing gene follistatin is an antagonist of BMP-4. Mech Dev 63: 39–50.
- 7. Hemmati-Brivanlou A, Kelly OG, Melton DA (1994) Follistatin, an antagonist of activin, is expressed in the Spemann organizer and displays direct neuralizing activity. Cell 77: 283–295.
- 8. Weinstein DC, Hemmati-Brivanlou A (1997) Neural induction in Xenopus laevis: evidence for the default model. Curr Opin Neurobiol 7: 7–12.
- 9. Wilson PA, Hemmati-Brivanlou A (1995) Induction of epidermis and inhibition of neural fate by Bmp-4. Nature 376: 331–333.
- 10. Mullins MC, Hammerschmidt M, Kane DA, Odenthal J, Brand M, et al. (1996) Genes establishing dorsoventral pattern formation in the zebrafish embryo: the ventral specifying genes. Development 123: 81–93.
- 11. Langdon YG, Mullins MC (2011) Maternal and zygotic control of zebrafish dorsoventral axial patterning. Annu Rev Genet 45: 357–377.
- 12. Reim G, Brand M (2006) Maternal control of vertebrate dorsoventral axis formation and epiboly by the POU domain protein Spg/Pou2/Oct4. Development 133: 2757–2770.
- 13. Fekany K, Yamanaka Y, Leung T, Sirotkin HI, Topczewski J, et al. (1999) The zebrafish bozozok locus encodes Dharma, a homeodomain protein essential for induction of gastrula organizer and dorsoanterior embryonic structures. Development 126: 1427–1438.
- 14. Imai Y, Gates MA, Melby AE, Kimelman D, Schier AF, et al. (2001) The homeobox genes vox and vent are redundant repressors of dorsal fates in zebrafish. Development 128: 2407–2420.
- 15. Shimizu T, Yamanaka Y, Nojima H, Yabe T, Hibi M, et al. (2002) A novel repressor-type homeobox gene, ved, is involved in dharma/bozozok-mediated dorsal organizer formation in zebrafish. Mech Dev 118: 125–138.
- 16. De Robertis EM, Kuroda H (2004) Dorsal-ventral patterning and neural induction in Xenopus embryos. Annu Rev Cell Dev Biol 20: 285–308.
- 17. Kofron M, Birsoy B, Houston D, Tao Q, Wylie C, et al. (2007) Wnt11/beta-catenin signaling in both oocytes and early embryos acts through LRP6-mediated regulation of axin. Development 134: 503–513.
- 18. Tao Q, Yokota C, Puck H, Kofron M, Birsoy B, et al. (2005) Maternal wnt11 activates the canonical wnt signaling pathway required for axis formation in Xenopus embryos. Cell 120: 857–871.
- 19. Lu FI, Thisse C, Thisse B (2011) Identification and mechanism of regulation of the zebrafish dorsal determinant. Proc Natl Acad Sci U S A 108: 15876–15880.
- 20. Nojima H, Rothhamel S, Shimizu T, Kim CH, Yonemura S, et al. (2010) Syntabulin, a motor protein linker, controls dorsal determination. Development 137: 923–933.
- 21. Tran LD, Hino H, Quach H, Lim S, Shindo A, et al. (2012) Dynamic microtubules at the vegetal cortex predict the embryonic axis in zebrafish. Development 139: 3644–3652.
- 22. Barth KA, Kishimoto Y, Rohr KB, Seydler C, Schulte-Merker S, et al. (1999) Bmp activity establishes a gradient of positional information throughout the entire neural plate. Development 126: 4977–4987.
- 23. von der Hardt S, Bakkers J, Inbal A, Carvalho L, Solnica-Krezel L, et al. (2007) The Bmp gradient of the zebrafish gastrula guides migrating lateral cells by regulating cell-cell adhesion. Curr Biol 17: 475–487.
- 24. Kelly C, Chin AJ, Leatherman JL, Kozlowski DJ, Weinberg ES (2000) Maternally controlled (beta)-catenin-mediated signaling is required for organizer formation in the zebrafish. Development 127: 3899–3911.
- 25. Bellipanni G, Varga M, Maegawa S, Imai Y, Kelly C, et al. (2006) Essential and opposing roles of zebrafish beta-catenins in the formation of dorsal axial structures and neurectoderm. Development 133: 1299–1309.
- 26. Varga M, Maegawa S, Weinberg ES (2011) Correct anteroposterior patterning of the zebrafish neurectoderm in the absence of the early dorsal organizer. BMC Dev Biol 11: 26.
- 27. Maegawa S, Varga M, Weinberg ES (2006) FGF signaling is required for {beta}-catenin-mediated induction of the zebrafish organizer. Development 133: 3265–3276.
- 28. Kane DA, Kimmel CB (1993) The zebrafish midblastula transition. Development 119: 447–456.
- 29. Anders S, Huber W (2010) Differential expression analysis for sequence count data. Genome Biol 11: R106.
- 30. Trapnell C, Roberts A, Goff L, Pertea G, Kim D, et al. (2012) Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat Protoc 7: 562–578.
- 31. 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 Res 21: 1328–1338.
- 32. Scott IC, Masri B, D’Amico LA, Jin SW, Jungblut B, et al. (2007) The g protein-coupled receptor agtrl1b regulates early development of myocardial progenitors. Dev Cell 12: 403–413.
- 33. Zeng XX, Wilm TP, Sepich DS, Solnica-Krezel L (2007) Apelin and its receptor control heart field formation during zebrafish gastrulation. Dev Cell 12: 391–402.
- 34. Xie XW, Liu JX, Hu B, Xiao W (2012) Zebrafish foxo3b negatively regulates canonical Wnt signaling to affect early embryogenesis. PLoS One 6: e24469.
- 35. Andersen JB, Hassel BA (2006) The interferon regulated ubiquitin-like protein, ISG15, in tumorigenesis: friend or foe? Cytokine Growth Factor Rev 17: 411–421.
- 36. Eijkelenboom A, Burgering BM (2013) FOXOs: signalling integrators for homeostasis maintenance. Nat Rev Mol Cell Biol 14: 83–97.
- 37. O’Carroll AM, Don AL, Lolait SJ (2003) APJ receptor mRNA expression in the rat hypothalamic paraventricular nucleus: regulation by stress and glucocorticoids. J Neuroendocrinol 15: 1095–1101.
- 38. Chan TM, Longabaugh W, Bolouri H, Chen HL, Tseng WF, et al. (2009) Developmental gene regulatory networks in the zebrafish embryo. Biochim Biophys Acta 1789: 279–298.
- 39.
Schug J (2008) Using TESS to predict transcription factor binding sites in DNA sequence. Curr Protoc Bioinformatics Chapter 2: Unit 2 6.
- 40. Eberhart JK, Swartz ME, Crump JG, Kimmel CB (2006) Early Hedgehog signaling from neural to oral epithelium organizes anterior craniofacial development. Development 133: 1069–1077.
- 41. Harvey SA, Tumpel S, Dubrulle J, Schier AF, Smith JC (2010) no tail integrates two modes of mesoderm induction. Development 137: 1127–1135.
- 42. Concha ML, Russell C, Regan JC, Tawk M, Sidi S, et al. (2003) Local tissue interactions across the dorsal midline of the forebrain establish CNS laterality. Neuron 39: 423–438.
- 43. Ryu SL, Fujii R, Yamanaka Y, Shimizu T, Yabe T, et al. (2001) Regulation of dharma/bozozok by the Wnt pathway. Dev Biol 231: 397–409.
- 44. Onichtchouk D, Geier F, Polok B, Messerschmidt DM, Mossner R, et al. (2010) Zebrafish Pou5f1-dependent transcriptional networks in temporal control of early development. Mol Syst Biol 6: 354.
- 45. Vastenhouw NL, Zhang Y, Woods IG, Imam F, Regev A, et al. (2010) Chromatin signature of embryonic pluripotency is established during genome activation. Nature 464: 922–926.
- 46. Lindeman LC, Andersen IS, Reiner AH, Li N, Aanes H, et al. (2011) Prepatterning of developmental gene expression by modified histones before zygotic genome activation. Dev Cell 21: 993–1004.
- 47. Vastenhouw NL, Schier AF (2012) Bivalent histone modifications in early embryogenesis. Curr Opin Cell Biol 24: 374–386.
- 48. Yamanaka Y, Mizuno T, Sasai Y, Kishi M, Takeda H, et al. (1998) A novel homeobox gene, dharma, can induce the organizer in a non-cell-autonomous manner. Genes Dev 12: 2345–2353.
- 49. Flowers GP, Topczewska JM, Topczewski J (2012) A zebrafish Notum homolog specifically blocks the Wnt/beta-catenin signaling pathway. Development 139: 2416–2425.
- 50. Shinya M, Eschbach C, Clark M, Lehrach H, Furutani-Seiki M (2000) Zebrafish Dkk1, induced by the pre-MBT Wnt signaling, is secreted from the prechordal plate and patterns the anterior neural plate. Mech Dev 98: 3–17.
- 51. Erter CE, Solnica-Krezel L, Wright CV (1998) Zebrafish nodal-related 2 encodes an early mesendodermal inducer signaling from the extraembryonic yolk syncytial layer. Dev Biol 204: 361–372.
- 52. Erter CE, Wilm TP, Basler N, Wright CV, Solnica-Krezel L (2001) Wnt8 is required in lateral mesendodermal precursors for neural posteriorization in vivo. Development 128: 3571–3583.
- 53. Varga M, Maegawa S, Bellipanni G, Weinberg ES (2007) Chordin expression, mediated by Nodal and FGF signaling, is restricted by redundant function of two beta-catenins in the zebrafish embryo. Mech Dev 124: 775–791.
- 54. Blythe SA, Cha SW, Tadjuidje E, Heasman J, Klein PS (2010) beta-Catenin primes organizer gene expression by recruiting a histone H3 arginine 8 methyltransferase, Prmt2. Dev Cell 19: 220–231.
- 55. Gore AV, Maegawa S, Cheong A, Gilligan PC, Weinberg ES, et al. (2005) The zebrafish dorsal axis is apparent at the four-cell stage. Nature 438: 1030–1035.
- 56. Furthauer M, Reifers F, Brand M, Thisse B, Thisse C (2001) sprouty4 acts in vivo as a feedback-induced antagonist of FGF signaling in zebrafish. Development 128: 2175–2186.
- 57. Furthauer M, Van Celst J, Thisse C, Thisse B (2004) Fgf signalling controls the dorsoventral patterning of the zebrafish embryo. Development 131: 2853–2864.
- 58. Trapnell C, Pachter L, Salzberg SL (2009) TopHat: discovering splice junctions with RNA-Seq. Bioinformatics 25: 1105–1111.
- 59. Trapnell C, Williams BA, Pertea G, Mortazavi A, Kwan G, et al. (2010) Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol 28: 511–515.
- 60. Hashimoto H, Itoh M, Yamanaka Y, Yamashita S, Shimizu T, et al. (2000) Zebrafish Dkk1 functions in forebrain specification and axial mesendoderm formation. Dev Biol 217: 138–152.
- 61. Furthauer M, Thisse C, Thisse B (1997) A role for FGF-8 in the dorsoventral patterning of the zebrafish gastrula. Development 124: 4253–4264.
- 62. Cao Y, Zhao J, Sun Z, Zhao Z, Postlethwait J, et al. (2004) fgf17b, a novel member of Fgf family, helps patterning zebrafish embryos. Dev Biol 271: 130–143.
- 63. Talbot WS, Trevarrow B, Halpern ME, Melby AE, Farr G, et al. (1995) A homeobox gene essential for zebrafish notochord development. Nature 378: 150–157.
- 64. Odenthal J, Nusslein-Volhard C (1998) fork head domain genes in zebrafish. Dev Genes Evol 208: 245–258.
- 65. Lee HC, Tseng WA, Lo FY, Liu TM, Tsai HJ (2009) FoxD5 mediates anterior-posterior polarity through upstream modulator Fgf signaling during zebrafish somitogenesis. Dev Biol 336: 232–245.
- 66. Weidinger G, Thorpe CJ, Wuennenberg-Stapleton K, Ngai J, Moon RT (2005) The Sp1-related transcription factors sp5 and sp5-like act downstream of Wnt/beta-catenin signaling in mesoderm and neuroectoderm patterning. Curr Biol 15: 489–500.
- 67. Thisse C, Thisse B (1999) Antivin, a novel and divergent member of the TGFbeta superfamily, negatively regulates mesoderm induction. Development 126: 229–240.
- 68. Alexander J, Rothenberg M, Henry GL, Stainier DY (1999) casanova plays an early and essential role in endoderm formation in zebrafish. Dev Biol 215: 343–357.
- 69. Schulte-Merker S, Ho RK, Herrmann BG, Nusslein-Volhard C (1992) The protein product of the zebrafish homologue of the mouse T gene is expressed in nuclei of the germ ring and the notochord of the early embryo. Development 116: 1021–1032.
- 70. Swartz ME, Sheehan-Rooney K, Dixon MJ, Eberhart JK (2011) Examination of a palatogenic gene program in zebrafish. Dev Dyn 240: 2204–2220.
- 71. Labalette C, Bouchoucha YX, Wassef MA, Gongal PA, Le Men J, et al. (2011) Hindbrain patterning requires fine-tuning of early krox20 transcription by Sprouty 4. Development 138: 317–326.
- 72. Toyama R, Gomez DM, Mana MD, Dawid IB (2004) Sequence relationships and expression patterns of zebrafish zic2 and zic5 genes. Gene Expr Patterns 4: 345–350.
- 73. Grinblat Y, Sive H (2001) zic Gene expression marks anteroposterior pattern in the presumptive neurectoderm of the zebrafish gastrula. Dev Dyn 222: 688–693.
- 74. Fekany-Lee K, Gonzalez E, Miller-Bertoglio V, Solnica-Krezel L (2000) The homeobox gene bozozok promotes anterior neuroectoderm formation in zebrafish through negative regulation of BMP2/4 and Wnt pathways. Development 127: 2333–2345.
- 75. Glinka A, Wu W, Delius H, Monaghan AP, Blumenstock C, et al. (1998) Dickkopf-1 is a member of a new family of secreted proteins and functions in head induction. Nature 391: 357–362.
- 76. Semenov MV, Tamai K, Brott BK, Kuhl M, Sokol S, et al. (2001) Head inducer Dickkopf-1 is a ligand for Wnt coreceptor LRP6. Curr Biol 11: 951–961.
- 77. Seiliez I, Thisse B, Thisse C (2006) FoxA3 and goosecoid promote anterior neural fate through inhibition of Wnt8a activity before the onset of gastrulation. Dev Biol 290: 152–163.
- 78. Pan D (2010) The hippo signaling pathway in development and cancer. Dev Cell 19: 491–505.
- 79. Varelas X, Miller BW, Sopko R, Song S, Gregorieff A, et al. (2010) The Hippo pathway regulates Wnt/beta-catenin signaling. Dev Cell 18: 579–591.
- 80. Wang WD, Melville DB, Montero-Balaguer M, Hatzopoulos AK, Knapik EW (2011) Tfap2a and Foxd3 regulate early steps in the development of the neural crest progenitor population. Dev Biol 360: 173–185.
- 81. Yao J, Kessler DS (2001) Goosecoid promotes head organizer activity by direct repression of Xwnt8 in Spemann’s organizer. Development 128: 2975–2987.
- 82. Giraldez AJ, Copley RR, Cohen SM (2002) HSPG modification by the secreted enzyme Notum shapes the Wingless morphogen gradient. Dev Cell 2: 667–676.
- 83. Martin BL, Kimelman D (2008) Regulation of canonical Wnt signaling by Brachyury is essential for posterior mesoderm formation. Dev Cell 15: 121–133.
- 84. Pourebrahim R, Houtmeyers R, Ghogomu S, Janssens S, Thelie A, et al. (2011) Transcription factor Zic2 inhibits Wnt/beta-catenin protein signaling. J Biol Chem 286: 37732–37740.
- 85. Fujimi TJ, Hatayama M, Aruga J (2012) Xenopus Zic3 controls notochord and organizer development through suppression of the Wnt/beta-catenin signaling pathway. Dev Biol 361: 220–231.
- 86. Reifers F, Bohli H, Walsh EC, Crossley PH, Stainier DY, et al. (1998) Fgf8 is mutated in zebrafish acerebellar (ace) mutants and is required for maintenance of midbrain-hindbrain boundary development and somitogenesis. Development 125: 2381–2395.
- 87. Cabrita MA, Christofori G (2008) Sprouty proteins, masterminds of receptor tyrosine kinase signaling. Angiogenesis 11: 53–62.
- 88. Lin W, Jing N, Basson MA, Dierich A, Licht J, et al. (2005) Synergistic activity of Sef and Sprouty proteins in regulating the expression of Gbx2 in the mid-hindbrain region. Genesis 41: 110–115.
- 89. Kunwar PS, Zimmerman S, Bennett JT, Chen Y, Whitman M, et al. (2003) Mixer/Bon and FoxH1/Sur have overlapping and divergent roles in Nodal signaling and mesendoderm induction. Development 130: 5589–5599.
- 90. Meno C, Gritsman K, Ohishi S, Ohfuji Y, Heckscher E, et al. (1999) Mouse Lefty2 and zebrafish antivin are feedback inhibitors of nodal signaling during vertebrate gastrulation. Mol Cell 4: 287–298.
- 91. Bisgrove BW, Essner JJ, Yost HJ (1999) Regulation of midline development by antagonism of lefty and nodal signaling. Development 126: 3253–3262.
- 92. Patel M, Chiang TC, Tran V, Lee FJ, Cote JF (2011) The Arf family GTPase Arl4A complexes with ELMO proteins to promote actin cytoskeleton remodeling and reveals a versatile Ras-binding domain in the ELMO proteins family. J Biol Chem 286: 38969–38979.
- 93. Waxman JS, Hocking AM, Stoick CL, Moon RT (2004) Zebrafish Dapper1 and Dapper2 play distinct roles in Wnt-mediated developmental processes. Development 131: 5909–5921.
- 94. Hong SK, Dawid IB (2009) FGF-dependent left-right asymmetry patterning in zebrafish is mediated by Ier2 and Fibp1. Proc Natl Acad Sci U S A 106: 2230–2235.
- 95. Hong SK, Tanegashima K, Dawid IB (2011) XIer2 is required for convergent extension movements during Xenopus development. Int J Dev Biol 55: 917–921.
- 96. Pezeron G, Lambert G, Dickmeis T, Strahle U, Rosa FM, et al. (2008) Rasl11b knock down in zebrafish suppresses one-eyed-pinhead mutant phenotype. PLoS One 3: e1434.
- 97. Ogata S, Morokuma J, Hayata T, Kolle G, Niehrs C, et al. (2007) TGF-beta signaling-mediated morphogenesis: modulation of cell adhesion via cadherin endocytosis. Genes Dev 21: 1817–1831.
- 98. Schulte-Merker S, Lee KJ, McMahon AP, Hammerschmidt M (1997) The zebrafish organizer requires chordino. Nature. 387: 862–863.
- 99. Yan B, Neilson KM, Moody SA (2009) foxD5 plays a critical upstream role in regulating neural ectodermal fate and the onset of neural differentiation. Dev Biol 329: 80–95.
- 100. Song J, Kim HJ, Gong Z, Liu NA, Lin S (2007) Vhnf1 acts downstream of Bmp, Fgf, and RA signals to regulate endocrine beta cell development in zebrafish. Dev Biol 303: 561–575.
- 101. Hernandez RE, Rikhof HA, Bachmann R, Moens CB (2004) vhnf1 integrates global RA patterning and local FGF signals to direct posterior hindbrain development in zebrafish. Development 131: 4511–4520.
- 102. Xiang W, Ke Z, Zhang Y, Cheng GH, Irwan ID, et al. (2011) Isthmin is a novel secreted angiogenesis inhibitor that inhibits tumour growth in mice. J Cell Mol Med 15: 359–374.
- 103. Delpire E, Gagnon KB (2008) SPAK and OSR1: STE20 kinases involved in the regulation of ion homoeostasis and volume control in mammalian cells. Biochem J 409: 321–331.
- 104. Thorpe CJ, Weidinger G, Moon RT (2005) Wnt/beta-catenin regulation of the Sp1-related transcription factor sp5l promotes tail development in zebrafish. Development 132: 1763–1772.