SEARCH

SEARCH BY CITATION

Keywords:

  • RNA binding protein;
  • rbm47;
  • head development;
  • zebrafish;
  • gene regulation;
  • gene knockdown;
  • morpholino;
  • microarray

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Background: Vertebrate trunk induction requires inhibition of bone morphogenetic protein (BMP) signaling, whereas vertebrate head induction requires concerted inhibition of both Wnt and BMP signaling. RNA binding proteins play diverse roles in embryonic development and their roles in vertebrate head development remain to be elucidated. Results: We first characterized the human RBM47 as an RNA binding protein that specifically binds RNA but not single-stranded DNA. Next, we knocked down rbm47 gene function in zebrafish using morpholinos targeting the start codon and exon-1/intron-1 splice junction. Down-regulation of rbm47 resulted in headless and small head phenotypes, which can be rescued by a wnt8a blocking morpholino. To further reveal the mechanism of rbm47's role in head development, microarrays were performed to screen genes differentially expressed in normal and knockdown embryos. epcam and a2ml were identified as the most significantly up- and down-regulated genes, respectively. The microarrays also confirmed up-regulation of several genes involved in head development, including gsk3a, otx2, and chordin, which are important regulators of Wnt signaling. Conclusions: Altogether, our findings reveal that Rbm47 is a novel RNA-binding protein critical for head formation and embryonic patterning during zebrafish embryogenesis which may act through a Wnt8a signaling pathway. Developmental Dynamics 242:1395–1404, 2013. © 2013 Wiley Periodicals, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Formation of the head during vertebrate embryogenesis has been a hot topic in developmental biology since the discovery of the head organizer by Spemann and Mangold (Spemann, 1924). Embryological and genetic evidence indicates that vertebrate head induction requires the concerted inhibition of Nodal, Wnt, and bone morphogenetic protein (BMP) signaling (Piccolo et al., 1999). During anterior–posterior (AP) patterning, the Spemann organizer produces a group of factors that inhibit the posteriorizing effects of Wnt and BMP signaling (Glinka et al., 1997). This so-called “Two Inhibitor Model” proposes that inhibition of both pathways is responsible for the regional specification of vertebrate head induction.

RNA binding proteins (RBPs) are proteins containing one or more RNA binding domain, the most common being the RNA recognition motif (RRM) (Lunde et al., 2007). RBPs are gene regulators required throughout early vertebrate development. They exert their effects through interactions with gene transcripts, thus modulating their activity. There are a multitude of mechanisms through which RBPs can regulate gene expression (Colegrove-Otero et al., 2005). These effects may be exerted at all levels of posttranscriptional regulation: nonsense-mediated decay (e.g., UPF3; Ruiz-Echevarria et al., 1998), splicing (e.g., U2AF; Ruskin et al., 1988), and alternative splicing (e.g., hnRNPA1; Allemand et al., 2005), mRNA stability (e.g., HuD; Lazarova et al., 1999), RNA editing (e.g., ACF; Dance et al., 2002), RNA localization (e.g., HuR; Gallouzi et al., 2001), pre-rRNA complex formation (e.g., Nucleolin; Chen et al., 2012), and translation (e.g., PABP; Tarun and Sachs, 1996).

RBPs are important regulators during development of various organs, and tissues including germ cells, heart, and ear (Jiang et al., 1997; Beck et al., 1998; Gerber et al., 2002; Rowe et al., 2006). Several RBPs have been identified for their roles in neural development. For example, Vg1-RBP, expressed in embryonic and neoplastic cells, is required for the migration of cells forming the roof plate of the neural tube, and plays essential roles in neural crest migration (Yaniv et al., 2003). Quaking homolog, also known as KH domain RNA binding (QKI), regulates distinct mRNA targets to promote oligodendrocyte differentiation and myelin formation, which is associated with schizophrenia (Bockbrader and Feng, 2008). Depletion of cold-inducible RNA binding protein (CIRP), a Xenopus transcription factor 3 (XTcf-3)-specific target gene, by antisense morpholino oligonucleotide injection leads to an enlargement of the anterior neural plate (van Venrooy et al., 2008). There is emerging evidence to suggest the importance of RBPs in head development. For example, the putative RBP cellular nucleic acid binding protein (CNBP) controls neural crest cell expansion during rostral head development by affecting levels of cellular proliferation and apoptosis as well as fate determination (Weiner et al., 2011).

RBM proteins possess one or more RRMs, highly conserved RNA interaction motifs consisting of a four-stranded antiparallel β-sheet packed against two α-helices (Nagai et al., 1990). By regulating posttranscriptional processes, RBMs are capable of functioning through diverse mechanistic pathways. For example, RBM4, possessing two RRMs and a CCHC-type zinc finger, functions in several cellular processes including alternative splicing of pre-mRNA, translation, and RNA silencing (Lin and Tarn, 2005; Kar et al., 2006; Markus et al., 2006; Markus and Morris, 2006, 2009; Lin et al., 2007). RBM5, which contains 2 RRMs, is a modulator of apoptosis (Mourtada-Maarabouni and Williams, 2002). Some RRM domains are capable of protein–protein interaction, such as in the RBM protein heterogeneous ribonucleoprotein A1 (hnRNPA1), whose first RRM domain interacts with the cap region of topoisomerase I through a hydrophobic pocket on its β-surface, and thus may be involved in DNA relaxation (Trzcinska-Daneluti et al., 2007). Several RBMs appear to be important for vertebrate development. RBM19 is reported to play a role in digestive organ development in zebrafish (Mayer and Fishman, 2003) and preimplantation development in mice (Lorenzen et al., 2005; Borozdin et al., 2006; Zhang et al., 2008). RBM24a and b are involved in vasculogenesis, early angiogenesis, and vascular maintenance in the developing zebrafish (Maragh et al., 2011).

RNA Binding Motif Protein 47 (RBM47) (aka Ribonucleoprotein-47, NCBI Accession #AF262323) is an uncharacterized, putative RBP. In the current study, we have characterized human and zebrafish RBM47 (Rbm47), and explored its role in zebrafish embryonic development, demonstrating that it plays a pivotal role in head formation and early embryonic patterning through a pathway involving Wnt8a signaling.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Human RBM47 is a Novel RNA-Binding Protein

In searching for novel RNA interacting proteins, we identified human RBM47 on chromosome 4p14. RBM47 produces two transcripts resulting from alternative splicing events. Isoform a consists of seven exons, while isoform b has five exons. Isoform b lacks exons 5 and 7, but alternatively contains a shorter exon 4 in the mid-coding region. As shown in Figure 1A, isoform a and b encode two proteins that are 592 and 523 amino acids in length, respectively, with projected molecular weights of 64 kDa and 57 kDa. Structural analysis found that both protein isoforms contain three RRMs, suggesting RBP activity.

image

Figure 1. Characterization of human and zebrafish RBM47. A: Diagrammatic representation of human and zebrafish RBM47 (Rbm47) proteins (not to scale). Human RBM47 gene encodes two alternatively spliced transcripts, resulting in two protein isoforms a and b. The gray box indicates the nonhomology region resulting from alternative splicing. Zebrafish rbm47 gene encodes only one isoform. All proteins possess three RRM domains. B: Sequence alignment of highly conserved regions of human and zebrafish RBM47 (Rbm47) RRMs. The zebrafish Rbm47 RRM has 90% sequence identity compared with those of the human orthologue. C: HeLa cells transfected with RBM47-GFP fusion plasmid express RBM47 in the nucleus, compared with control-transfected cells that express green fluorescent protein (GFP) throughout the cell. D: The RNA binding assay demonstrates RBM47's ability to bind poly-A, -U, and -C RNA, while weakly binding poly-G RNA, and having no affinity for single-stranded DNA as shown in lane S. E: rbm47 is expressed ubiquitously during zebrafish embryogenesis, as shown by whole mount in situ hybridization. Scale bar = 200 μm.

Download figure to PowerPoint

To determine subcellular localization, HeLa cells were transfected with an RBM47-GFP fusion protein, revealing nuclear localization (Fig. 1C). To validate our prediction that RBM47 is an RBP, an RNA binding assay was performed. Human RBM47 showed strong affinities to poly-A, -C, and -U RNAs, low affinity to poly-G RNA and no ability to interact with single-stranded DNA (ssDNA) (Fig. 1D).

Zebrafish rbm47 is Expressed During Early Embryogenesis

Zebrafish rbm47 encodes a protein of 599 amino acids, and contains three RRMs (Fig. 1A). Zebrafish Rbm47 protein shows 81.5% identity to human RBM47-a, with 90% similarity within the RRM sequences (Fig. 1B). No alternative splicing transcript was identified for zebrafish rbm47.

To reveal the spatiotemporal expression pattern of rbm47 during zebrafish embryogenesis, we carried out whole-mount in situ hybridization on embryos at different developmental stages using rbm47 riboprobes. As shown in Figure 1E, rbm47 is ubiquitously expressed from the one-cell stage up to 24 hr postfertilization (hpf). This expression pattern suggests that rbm47 plays an important role in early stages of zebrafish development.

rbm47 is Involved in Zebrafish Head Development

To determine the role of rbm47 in development, zebrafish rbm47 was knocked down in developing embryos using antisense morpholino oligonucleotides. Two blocking morpholinos were designed to target rbm47 RNA (Fig. 2A): MO-rbm47-ATG targets the start codon of rbm47 in exon 1, preventing translation initiation; MO-rbm47-E1I1 targets the exon-1/intron-1 boundary, interfering with pre-mRNA splicing. Both MOs were fluorescently labeled to achieve direct visualization upon microinjection. MO-rbm47-ATG was labeled with green fluorescein at the carboxyl terminal, whereas MO-rbm47-E1I1 was tagged with fluorescently red lissamine. A standard control oligo (MO-Ctrl), labeled with green fluorescein, was used as a control. Reverse transcriptase-polymerase chain reaction (RT-PCR) analysis of rbm47 mRNA from MO-injected embryos validated specific and effective splice-blocking of the rbm47 transcript in MO-rbm47-E1I1-injected embryos (Fig. 2B).

image

Figure 2. rbm47 knockdown in developing zebrafish embryos disrupts head formation. A: Two morpholino knockdown strategies were used to disrupt rbm47 function. MO-rbm47-ATG hybridizes to the start codon, preventing translation initiation. MO-rbm47-E1I1 hybridizes to the exon-1/intron-1 splice donor site, blocking splicing to exon 2. Schematic drawing is not to scale. B: Reverse transcriptase-polymerase chain reaction (RT-PCR) was used to show interrupted splicing of RNA extracted from MO-rbm47-E1I1 zebrafish, which amplifies a band of approximately 3.5 kb due to retention of intron 1. PCR cycles: lane 1 (34 cycles); lane 2 (30 cycles); lane 3 (30 cycles). C: Upper panels demonstrate typical loss of head phenotype in rbm47 knockdown zebrafish (b and c) and a control zebrafish injected with control morpholino (a). The lower panels depict typical small head phenotypes in MO-rbm47 injected fish (d and e). Scale bar = 200 μm. D: The incidence of total loss of head development in morpholino-injected zebrafish. Injection of morpholinos targeting rbm47 resulted in a 9–16% incidence of the headless phenotype. Co-injection of rbm47 mRNA or wnt8a blocking morpholino with rbm47 blocking morpholino resulted in a decreased incidence of headlessness (for detailed numbers and statistics, see Table 1).

Download figure to PowerPoint

Zebrafish embryos (n = 500) at the one- or two-cell stage were microinjected with 5 ng of MO-rbm47-ATG, MO-rbm47-E1I1 or two MOs combined. Twenty-four hours later, similar phenotypes were observed in both MO-injected groups, characterized by defects in anterior head development. Approximately 9–16% of MO-injected embryos lacked heads, demonstrating a headless phenotype (Fig. 2C–b & c; Table 1), and 20–30% had reduced head development (Fig. 2C–d & e; Table 1), while the embryos injected with MO-Ctrl developed normally (Fig. 2C; Table 1).

Table 1. Phenotype Summary of Zebrafish Embryos from rbm47 Morpholino Knockdown
 Phenotype classes
  smallabnormal trunk & tail  P value 
Morpholinosheadlessheadnormaln=headlessversus
  1. Denotes standard error.

Standard control0±0%06%94%240  
MO-ATG9±0.87%a30%14%47%5001.62E-6MO-CTRL
MO-E1I112±2.29%a20%15%53%5002.16E-8MO-CTRL
MO-ATG+MO-E1I116±1.28%a30%15%39%1501.59E-10MO-CTRL
MO-E1I1+ RNA2±1.35%a9%6%83%1502.83E-4MO-E1I1
MO-E1I1+MO-wnt82±1.10%a12%6%80%1502.83E-4MO-E1I1

Next, we performed rescue experiments to test whether the headless phenotype is a specific effect of rbm47 knockdown. Five pg rbm47 mRNA was co-injected with 10 ng MO-rbm47-E1I1 into zebrafish embryos. Examining the development of embryos at 2-hr intervals, the incidence of the headless phenotype was reduced to ∼2% (Fig. 2D; Table 1). These data suggest that rbm47 expression is required for head formation.

rbm47 Acts on Head Development Through the Wnt8a Signaling Pathway

To reveal the molecular mechanism of rbm47 in regulation of zebrafish head development, microarray gene expression analysis was performed on MO-rbm47 embryos. RNA was isolated from MO-injected embryos (n = 200 for each group) at 75% epiboly. The 75% epiboly stage is characterized by anterior axial hypoblast development as the prechordal plate reaches the animal pole. This occurs 2 hr before head-mesoderm formation, and we anticipate that genes critical for head induction are expressed at this time point.

The expression levels of 15,619 genes were probed and compared across groups. A minimum four-fold decrease in expression was detected for 26 genes in both the MO-rbm47-ATG and MO-rbm47-E1I1 groups compared with control MO injected embryos. Meanwhile, 20 genes had a minimum four-fold increase in expression as a result of rbm47 knockdown (Fig. 3A; Table 2). Of interest, the microarray analysis revealed that wnt8a expression was increased 1.4-fold and 2.3-fold in MO-rbm47-E1I1 and MO-rbm47-ATG fish, respectively, suggesting the involvement of rbm47 in this major pathway of head development. Furthermore, gene expression profiling detected up-regulation of other genes involved in Wnt as well as BMP signaling, including gsk3a, otx2, and chordin (Table 3).

image

Figure 3. RNA microarray to identify candidate rbm47 target genes. A: Twenty genes were found to be up-regulated and 26 genes were found to be down-regulated by at least four-fold as shown on the left and right side of the graph, respectively. B,C: Confirmation of the most significantly up- and down-regulated genes by real-time reverse transcriptase-polymerase chain reaction (epcam and a2ml).

Download figure to PowerPoint

Table 2. Significantly Changes in Gene Expression Identified by Microarray Analysis
Gene name access no.Fold change
MO-rbm47-E1I1MO-rbm47-ATG
Epithelial cell adhesion molecule (epcam)BQ262802255.93211.44
Wingless homolog (Drosophila) (wls)BM775264182.53388.67
Kelch repeat and BTB (POZ) domain containing 10bAI64154243.456.3
Calpain 8AW23361647.143.9
Methionine-R-sulfoxide reductase B1AA49721916.338.76
Hedgehog acyltransferase-like,bAI62634810.7215.12
Alpha-2-macroglobulin-like (a2ml)AI793675−53.5−95.4
HORMA domain-containing proteinBM316732−9.6−106.8
Solute carrier family 20, member 1aAW077636−27.4−34.3
Heat shock factor binding protein 1-likeBM530302−13.6−10.9
Apolipoprotein A-IVAI545593−5.2−6.8
Annexin A2bBC153582−3.0−8.9
LIM domain only 7aBM279822−4.5−6.3
Table 3. Microarray Profiling of BMP and Wnt Pathway Genes Following rbm47 Knockdown
Gene nameMO-E1I1MO-ATG
  1. a

    Denotes non-changed.

  2. b

    “I” denotes that the gene is increased, expressed as a fold-change value.

Goosecoida
Sonic hedgehog
ChordinI,1.2
Follistatin
Noggin
Tcf3
Dickkopf-1
Pax2, pax6
Otx2I,1.7b
Lmo-1
Gsk3aI,1.9I,2.6
Gsk3b
Wnt8aI,1.4I,2.3
Wnt8b
Wnt-1
Wnt2
Wnt4
Wnt5
Wnt-10
Wnt-11

To examine the involvement of rbm47 in Wnt8a signaling, MO-rbm47 was co-injected with MO-wnt8 to see whether wnt8 knockdown is capable of rescuing the morphant phenotype. Co-injection dramatically decreased the incidence of the headless and small head phenotypes (Fig. 2D; Table 1), suggesting that Rbm47 does indeed function through the Wnt signaling pathway in regulating head development and early embryonic patterning.

Among genes with altered expression levels, epithelial cell adhesion molecule (epcam) had the highest up-regulation, while alpha-2-macroglobulin-like (a2ml) was the most severely down-regulated (Table 2). epcam is a Wnt/β-catenin signaling target gene in hepatocellular carcinoma cells (Yamashita et al., 2007). Human A2M is reportedly associated with Wnt/β-catenin signaling (Lindner et al., 2010). These genes were chosen for quantitative RT-PCR (qRT-PCR) analysis to verify Rbm47's effect on gene expression. epcam expression was elevated ∼four-fold following rbm47 knockdown (Fig. 3B), while the expression of a2ml was reduced to ∼23% normal levels compared with the control MO group (Fig. 3C).

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES

In the present study, we have characterized human and zebrafish RBM47 (Rbm47), and investigated its role as a putative RNA binding protein involved in zebrafish embryogenesis. RBPs mediate their effects by altering posttranscriptional events of specific gene transcripts. RBPs interact with target transcripts through RNA binding domains, such as RNA Recognition Motifs. Zebrafish Rbm47 possesses three RRMs with high homology to those found in the human orthologue. We found that human rbm47 strongly interacts with poly-A, -C, and -U RNAs, while binding with poly-G RNA occurs with low affinity, demonstrating its ability to bind to RNA. rbm47 is not capable of interaction with ssDNA. Additionally, we created an rbm47-GFP fusion protein to determine its subcellular localization, which was found to be within the HeLa cell nuclei. Based on this information, we propose that rbm47 is a novel RNA Binding Protein.

To characterize the spatiotemporal expression of rbm47, whole mount in situ hybridization for rbm47 mRNA was performed on developing zebrafish embryos. rbm47 is expressed ubiquitously throughout early embryonic development. To study its function during zebrafish development, we used a morpholino-based knockdown approach to target either the rbm47 translation start codon or exon-1/intron-1 splicing. We demonstrated that a high percentage of rbm47 knockdown embryos had incomplete head formation or total loss of head development. This striking phenotype was rescued upon co-injection of rbm47 mRNA, supporting our conclusion that defective head development is a consequence of rbm47 knockdown, and that rbm47 expression is required for normal head development.

Vertebrate head induction requires the concerted inhibition of both Wnt and BMP signaling pathways (Glinka et al., 1997; Piccolo et al., 1999). Indeed, the headless phenotype as a consequence of single gene knockdown is an important observation that has only be seen as a result of altering the regulation of a small set of master regulatory genes involved in early vertebrate development, including foxA3 and gsc (Yao and Kessler, 2001; Seiliez et al., 2006), tcf3 (Kim et al., 2000), and dkk1 (Glinka et al., 1998). Wnt8 is the key transcriptional motivator to act on the anterior neuroectoderm from the lateral mesoderm to produce the AP regional patterning of the central nervous system (Erter et al., 2001). The graded Wnt8 activity mediates overall neuroectodermal posteriorization and thus determines the location of the midbrain–hindbrain boundary organizer (Rhinn et al., 2005). Wnt8 expression is inhibited within the organizer, but is found in the lateral margin of the zebrafish gastrula (Kim et al., 2000). Thus, excess Wnt8 activity due to overexpression or loss of inhibition leads to loss of anterior structure. Our observation that morpholino knockdown of rbm47 causes headless and reduced head phenotypes suggests that it may act through a pathway involving Wnt8.

In investigating Rbm47's mechanistic pathway, we used microarray analysis to screen the expression levels of 15,619 zebrafish genes from rbm47 MO-knockdown embryos at 75% epiboly. We found 92 genes with increased expression in both splice- and translation-blocked knockdown groups, compared with MO-control embryos, with 20 genes having a minimum four-fold increase. epcam was identified as the most up-regulated gene by this screen. epcam regulates cell adhesion, integrity, plasticity and morphogenesis as a partner of E-cadherin during zebrafish epiboly and skin development (Slanchev et al., 2009). Two Tcf-binding elements were identified in the epcam promoter and epcam was found to be a Wnt-β-catenin target gene in hepatocellular carcinoma cells (Yamashita et al., 2007). These findings support the idea that rbm47's effect on head development occurs through the canonical Wnt8 signaling pathway (Lu et al., 2011).

In accordance with this, rescue experiments demonstrated that a wnt8a-blocking morpholino can partially rescue the rbm47 knockdown phenotype (Fig. 2D). In addition to epcam, several other genes involved in Wnt signaling were up-regulated, including gsk3a, otx2, and chordin. This further supports our conclusion that the effect of rbm47 knockdown on head development occurs through an overactive Wnt pathway.

Meanwhile, of the genes with decreased expression, 26 exhibited a minimum four-fold expression reduction by microarray analysis, with a2ml being the most severely affected. The qRT-PCR results confirmed reduced gene expression of a2ml in rbm47 knockdown embryos. A2M, the human homologue of zebrafish a2ml, is a plasma protease inhibitor, cytokine carrier, and ligand for cell-signaling receptors (Roberts, 1985). A2M in the human and rat brain is an acute-phase protein synthesized primarily by astrocytes, and is associated with Alzheimer's disease due to its ability to mediate the clearance and degradation of amyloid β (Cavus et al., 1996; Kovacs, 2000). The activated forms of A2M can bind to neurotrophic factors and directly inhibit neurotrophic factor-receptor signal transduction to repress neurite outgrowth of central neurons (Koo and Liebl, 1992; Liebl and Koo, 1993; Koo et al., 1994; Hu and Koo, 1998). Most importantly, human A2M is reported to regulate β-catenin signaling though the Wnt inhibitory co-receptor low-density lipoprotein receptor-related protein-1 (LRP1) (Lindner et al., 2010). An A2M conformational intermediate is capable of regulating peripheral nerve injury response by a mechanism that requires LRP1 (Arandjelovic et al., 2007). These previous studies have demonstrated that A2M plays an essential role in neurogenesis. In this study, zebrafish a2ml's down-regulation by rbm47 knockdown provides an important insight into the mechanism of rbm47 on development, suggesting that it is also involved in neural development. However, to identify the RNA binding partners of rbm47, detailed mechanistic evaluation is required in future investigations.

As Rbm47 is ubiquitously expressed during zebrafish embryonic development, the finding that its knockdown results in a tissue-specific phenotype requires explanation. We hypothesize that the rbm47 target gene(s) and/or its binding partner(s) are tissue-specific regulators. Our preliminary study indeed demonstrates that a2ml is expressed in the anterior head region during embryonic patterning by RNA in situ hybridization (data not shown).

In summary, the present study demonstrates that Rbm47 is an RNA binding protein that plays an important role in head development during zebrafish embryogenesis.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES

RNA and DNA Constructs

RNA was extracted from zebrafish using the Illustra RNAspin mini isolation kit (GE Healthcare, Little Chalfont, Buckinghamshire, UK). RNase-free DNase digestion was performed to eliminate genomic DNA. RNA was reverse transcribed into cDNA using the Roche Reverse Transcription kit (Roche Applied Science, Laval, Quebec, Canada). The full-length cDNA of zebrafish rbm47 was amplified by PCR from zebrafish cDNA and then subcloned into the pBluescriptII vector.

Determination of the Cellular Localization of Human RBM47

A pRMB47-GFP plasmid was built to express the RBM47-GFP fusion protein in cultured human cell lines using pEGFP-C1 (Clontech, Mountain View, CA) as the cloning vector. HeLa cells were transfected with pRMB47-GFP or the control plasmid, pEGFP-C1, which expresses GFP from the CMV promoter, using lipofectamine. The transfected cells were examined under a fluorescent microscope (Leica-model DM IRB, Deerfield, IL) 48 hr posttransfection.

RNA Binding Assay

Individual RNA-Sepharose Beads (Sigma, St. Louis, MO), poly (A), poly (G), poly (C), poly (U), or single-stranded DNA-sepharose beads were washed in a buffer containing 20 mM HEPES pH 8.0, 20% glycerol, 100 mM KCl, 0.2 mM ethylenediaminetetraacetic acid, 1 mM DDT, 0.5 mM phenylmethylsulfonyl fluoride, and RNA-guard at 38 units/ml, and packed in mini-columns. RBM47 produced from E. coli was loaded in each column and washed with the same buffer. Proteins retained in the columns were separated by polyacrylamide gel electrophoresis.

Morpholino Embryo Injections

Morpholinos were purchased from Gene Tools, LLC (Philomath, OR). Two experimental morpholino oligos were designed to target the translation start codon and the first exon–intron boundary of zebrafish rbm47 pre-mRNA. The experimental and control morpholino sequences are as follows:

MO-rbm47-ATG: 5′ CGGAGTCTTCTGCTGTCATTCTGAA 3′-carboxyfluorescein

MO-rbm47-E1I1: 5′ TGATTGTAACTAAGATTAACCTGAA 3′-lissamine

MO-wnt8a: 5′ ACGCAAAAATCTGGCAAGGGTTCAT 3′

Standard control: 5′ CTCTTACCTCAGTTACAATTTATA 3′–carboxyfluorescein

Morpholino oligonucleotides were solubilized in water at 1 mM. The resulting stock solution was heated at 60°C for 10 min and then diluted to working concentrations in sterile water before injection. The yolks of embryos at the one- to two-cell stage were microinjected with a volume of 5 nl of MO (200 embryos/μl). Effective doses were determined separately for each morpholino.

mRNA Injections

pBluescript II SK(+)-RBM47 was linearized with Acc65I (Promega, Madison, WI). Capped mRNA was transcribed in vitro using T3 RNA polymerase (Roche Applied Science, Laval, Quebec, Canada). Synthesized RBM47-RNA was purified by centrifugation through the Ambion NucAway column (Ambion, Austin, TX). The mRNA was co-injected into 1–2 cell stage embryos with the indicated morpholino. Siblings from the same batch served as the internal control for these experiments.

In Situ Hybridization

Plasmid pBluescript II SK(+)-RBM47 was digested with BstXI (Fermentas, Ottawa, ON, Canada). DIG-labeled RNA antisense RBM47 probes were synthesized in vitro using T7 RNA polymerase (New England BioLabs, Ipswich, MA) and digoxigenin-labeled UTP (Roche Applied Science, Laval, Quebec, Canada). The sense probe was transcribed using T3 RNA polymerase (Roche Applied Science, Laval, Quebec, Canada) and used as a negative control. The sizes of antisense and sense probes are 821 bp and 988 bp, respectively. Whole-mount in situ hybridization was performed as previously described (Jowett and Lettice, 1994; Jowett and Yan, 1996) .

Zebrafish Microarray

Following microinjection, developing embryos were observed under the dissecting microscope, and dead embryos were removed every 2 hr. At approximately 8 hr postfertilization, we carefully checked and harvested the embryos at 75% epiboly. 200 embryos of each group were collected for RNA extraction. RNA was isolated using 1 ml of Trizol (Invitrogen, Carlsbad, CA) per 100 embryos, according to the manufacturer's protocol, followed by further purification using the Illustra RNAspin mini isolation kit (GE Healthcare, Little Chalfont, Buckinghamshire, UK). Before microarray analysis, RNA quality was determined on an Agilent 2100 RNA Bioanalyzer (Agilent Technologies, Santa Clara, CA). Finally, 1 μg of RNA from each group was subjected to microarray analysis at The Center for Applied Genomics (Hospital for Sick Children, Toronto, ON, Canada).

RT-PCR and Real-Time PCR

RNA was isolated using Trizol (Invitrogen) method, followed by Illustra RNAspin purification. Total RNA (1 μg) was reverse transcribed using random hexamer primers and SuperScript II reverse transcriptase (Invitrogen), following the manufacturer's instructions. RNase-free DNase digestion was performed to eliminate genomic DNA. A reverse-transcriptase negative control was also used to exclude genomic DNA contamination. To confirm disruption of mRNA maturation using the splice-blocking morpholino MO-rbm47-E1I1, two oligo primers across intron 1 were designed for RT-PCR. The primer sequences are as follows:

Forward primer (F): 5′ ATGACAGCAGAAGACTCCGCCT 3′

Reverse primer (R): 5′ TCAGTAGGTCTGGTATACATCA 3′

For real-time PCR, 20-ng template cDNA was sequence using the ABI Prism 7700 sequence detection system (Applied Biosystems, Foster City, CA). For relative quantification analysis, values of RNA expression were compared between groups after normalization with β-actin expression. All measurements were performed as previously described (Livak and Schmittgen, 2001). Primers were designed by the Prism 7700 system and synthesized at The Center for Applied Genomics (Hospital for Sick Children, Toronto, ON, Canada).

The primer sequences are as follows:

  • z-epcam forward: 5′ CAAGACGAGCCATAACTTTATTTCAT 3′
  • z-epcam reverse: 5′ CAAACAAGGCAACTAAAACCTTCA 3′
  • z-a2ml forward: 5′ GGATCTGGGAGCTTGCTGAA 3′
  • z-a2ml reverse: 5′ CAAGTCGTGATGGTGTCAGGAA 3′
  • β-actin forward: 5′ CGAGCAGGAGATGGGAACC 3′
  • β-actin reverse: 5′ CAACGGAAACGCTCATTGC 3′

Statistical Analysis

Microinjected embryos were divided into four groups: headless, small head, abnormal trunk and tail, and normal. Data were analyzed using Pearson's chi-squared test, comparing the number of headless fish to all other phenotypes (normal, abnormal trunk and tail, and small headed) across morpholino-injected groups.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES

We thank Drs. Ashley Bruce and Vince Tropepe of the Department of Cell and Systems Biology, University of Toronto for helpful discussions of the project. Thanks to Dr. Bruce for reading the manuscript with helpful comments and suggestions. We acknowledge the Canada Foundation for Innovation (CFI) for infrastructural funding in support of the Zebrafish Core Facility at St. Michael's Hospital.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES