Notch signaling plays an essential role in early brain development including, but not limited to, cell fate determination (Dontu et al.,2004), pattern formation (Lewis et al.,2009), and neurogenesis and gliogenesis (Julich et al.,2005; Taylor et al.,2007; Wheeler et al.,2008; Xiao et al.,2009). Membrane-bound Notch is proteolyzed by tumor necrosis factor alpha (TNFα) -converting enzyme, metalloproteases, and presenilin into an active form that is translocated to the nucleus where it forms a complex with CSL (CBF1/RBP-Jκ, Su(H), Lag-1; Jarriault et al.,1995; Lu and Lux,1996). This complex is a transcriptional activator triggering expression of downstream target genes such as Hairy/Enhancer-of-split (Hes), which inhibit basic helix–loop–helix (bHLH) transcription factors with additional roles in neurodevelopment (Bailey and Posakony,1995; Lecourtois and Schweisguth,1998; Mumm and Kopan,2000; Hirata et al.,2002; Holley et al.,2002; Oates and Ho,2002).
Mutagenesis screens in zebrafish have identified several mutants of the Notch signaling pathway: after eight (aei) (delta D), deadly seven (des) (notch 1a), beamter (bea) (delta C), and mind bomb (mib; Jiang et al.,1996; van Eeden et al.,1996). Loss-of-function mutations in zebrafish display a so-called “neurogenic” phenotype, marked by overproduction of primary neurons early in development and a later reduction in secondary neurons. The mutation identified in mibhi904 zebrafish disrupts a conserved putative E3 ubiquitin ligase regulating Notch signaling (Golling et al.2002; Chen and Corliss,2004). Ubiquitin ligases which contain substrate-binding domains providing critical elements for specificity of target proteins or catalytic domains (such as HECT and RING fingers), comprise one of the largest families of enzymes in human cells, and have been linked to multiple human diseases. For example, the human neurogenetic disorder, Angelman syndrome (AS), can be caused by loss-of-function mutations in E3 ubiquitin ligase (ube3a; Kishino et al.,1997). Phenotypic hallmarks of AS include motor dysfunction leading to an ataxic gait, frequent severe seizures, profound learning disability, absent speech, and characteristic happy demeanor (Lossie et al.,2001). Mouse models of AS exhibit an increased incidence of seizures, poor performance on Rotarod assays, and defects in long-term potentiation (Jiang et al.,1998; Miura et al.,2002). We recently showed that a mibhi904 E3 ubiquitin ligase mutant zebrafish exhibits electrographic and behavioral seizure activity (Hortopan et al.,2010). A transcriptome analysis of mibhi904 zebrafish, at 3 days postfertilization (dpf), identified differential expression of GABA signaling pathway and neurodevelopmental (Notch signaling pathway related) genes (Hortopan et al.,2010). In the previous manuscript, we focused on analysis of GABA signaling genes as these may directly underlie the observed epileptic phenotype. Here, we examined expression of a group of microarray-identified neurodevelopment genes in the central nervous system of mibhi904 zebrafish at 3 dpf, e.g., a larval time point when these mutants exhibit a severe neurogenic phenotype and epilepsy. Confocal analysis of early brain development using live green fluorescent protein (GFP) reporter fish (Gfap:GFP and Dlx5/6:GFP) was also performed.
A transcriptome analysis using an Affymetrix zebrafish microarray containing 16,416 genes was recently performed on mibhi904 mutants (homozygotes) and age-matched sibling controls at 3 dpf (Hortopan et al.,2010). Microarray data was validated using conventional reverse transcriptase-polymerase chain reaction (RT-PCR) and real-time quantitative PCR (qPCR) for nine genes with putative roles in neurodevelopment (Table 1; Fig. 1). To further define the expression of these genes, we completed in situ hybridization studies using whole-mount larvae and cryostat sections through the central nervous system. Results of these studies are presented in detail below.
Table 1. Summary of Data Extracted From Microarrays
aStatistical analysis (Ttest, GeneSpring GX 7.3.1) and quantitative polymerase chain reaction (qPCR) analysis (SPSS, two-tailed Student's t test) are shown. GenBank ID, fold changes, and P values.
hairy and enhancer of split 5
Basic helix-loop-helix domain containing, class B, 5
homeo box A5a
homeobox protein (hoxb5b) gene
diencephalon/mesencephalon homeobox 1a
developing brain homeobox 1a
glial fibrillary acidic protein
Hairy-related 4.2 (her4.2) and hairy/enhancer of split 5 (hes5), Notch responsive bHLH genes (Takke and Campos-Ortega,1999; Kawamura et al.,2005; Bae et al.,2005; Hegde et al.,2008), show a significant four-fold microarray and three-fold real-time qPCR decrease in mibhi904 zebrafish consistent with prior observations in mibta52b mutants (Hwang et al.,2009). A lateral view of the head, in whole-mount in situ hybridization (WISH), for wild-type (WT) control zebrafish revealed prominent her4.2 mRNA expression in forebrain, midbrain and hindbrain (Fig. 2A). Transverse cryostat sections showed a more detailed expression pattern of her4.2 in telencephalon, especially in the large pallial domain and continued posteriorly into subpallium through the anterior commissure (Fig. 2A1). In diencephalon, strongly homogenous clusters of her4.2 cells can be found along the midline in dorsal (Fig. 2A2) and ventral thalamus (Fig. 2A3). Caudal to pallium, her4.2 expression appears in the eminentia thalami and preoptic region (Fig. 2A2,A3). Metencephalon (midbrain) displays strong and extensive her4.2 expression in the ventral region of posterior tuberculum and in the intermediate and caudal hypothalamus (Fig. 2A4,A5). Still at midbrain level, more posteriorly in the mesencephalon, her4.2 expression is prominent down the midline and as a band along the optic tectum and tegmentum boundary (Figs. 2A4,A5). In contrast, her4.2 WISH expression is dramatically reduced in age-matched mibhi904 mutants and detectable only in sparsely distributed clusters throughout the central nervous system (CNS). A lateral WISH view, shows only a few small clusters of her4.2 cells in subpallium and pallium (compare Figs. 2B and 2A). Transverse cryostat sections of the forebrain clearly demonstrate the presence of only small clusters of her4.2 expression in pretectum, dorsal region of the posterior tuberculum, and anterior commissure (Fig. 2B1). Diencephalon displays diffuse small clusters of her2.4 expression in dorsal thalamus and in the lateral forebrain bundle (Fig. 2B2,B3). In optic tectum, medial, basal, and lateral small clusters of her2.4 can be seen throughout the three transverse cryostat sections (Fig. 2B3–B5), together with small her2.4 expressing cell clusters located in the ventral region of posterior tuberculum (Fig. 2B4,B5) and hypothalamus (Fig. 2B4). Midline expression was strongly reduced in mibhi904 mutants and no detectable expression was noted in the retina.
A similar, although somewhat less prominent, pattern of WISH expression was found for hes5. Of interest in WT controls, a distinct pattern of hes5 expression, with a predominant accumulation in forebrain and midbrain, was observed (Fig. 3A). In cryostat sections through telencephalon, pallium, and subpallium, hes5 expression appears near the ventricular surface (Fig. 3A1). In diencephalon, hes5 signal continues into the optic tectum (Fig. 3A2–A4), eminentia thalami and preoptic region (Figs. 3A1,A2) as well as the ventral portion of posterior tuberculum and intermediate hypothalamus (Fig. 3A3). Strong hes5 expression was noted near the tectal ventricle, torus semicircularis, and rostral region of medulla oblongata (Fig. 3A4). “Blobs” of hes5 expression in caudal hypothalamus are also present. As described above for her4.2, strong hes5 expression is seen in the retina of WT zebrafish (Fig. 3A1) but not age-matched mibhi904 mutants; WISH also revealed a cluster of hes5 expression primarily restricted to midbrain (Fig. 3B). Transverse sections show hes5-expressing aggregates corresponding to early migrating cells in the pre-tectum (Fig. 3B1,B2) that fail to extend down into thalamus and hypothalamus (Fig. 3B3,B4).
Class B5 (bhlhb5), a bHLH domain containing transcription factor (Brunelli et al.,2003), exhibited a five-fold down-regulation in microarray and almost three-fold down-regulation in qPCR in mibhi904 mutants compared with controls. Lateral and dorsal WISH views in WT controls revealed prominent bhlhb5 expression at the forebrain and midbrain–hindbrain boundary (Fig. 4 A,B). Strong forebrain expression is also present in the olfactory pits, as noted in transverse and coronal sections (Figs. 4A1,Ci). Although only faint diencephalic bhlhb5 expression appears in transverse sections (Fig. 4A2), bhlhb5 can be seen clearly in dorsal views and coronal sections (Fig. 4C,Ci). More posteriorly in optic tectum, approximately at the level of the tectal ventricle, bhlhb5 expression becomes even stronger (Fig. 4A3). In addition, bhlhb5 is expressed in the WT eye where it is normally distributed into retinal nuclear layers. It has been shown that targeted deletion of Bhlhb5 causes loss of GABA-producing amacrine and Type 2 OFF-cone bipolar cells (Feng et al.,2006; Dulin et al.,2007). Considering the crucial role of Bhlhb5 in the specification of both cell subtypes, this could explain why expression is more diffuse in the mibhi904 mutant eye where the retinal layers do not appear to differentiate (Fig. 4B2,Di). Also, in mind bomb mutants, there is no detectable bhlhb5 expression in the olfactory pits (Fig. 4B,B1,D) and only faint or weak expression in diencephalon (dorsal thalamus; Fig. 4B2,B3).
Hoxa5a, hoxb5b, and dmbx1a, homeobox genes with roles in embryonic hindbrain patterning and brain development (Holland and Takahashi,2005), exhibit a nearly two-fold down-regulation in mibhi904 mutants compared with controls. WISH revealed remarkable WT expression patterns at rhombomere boundaries and lower expression in pharyngeal arches 6 and 7 for hoxa5a and hoxb5b genes (Figs. 5A,B,E,F), in agreement with previous studies (McGinnis and Krumlauf,1992; Krumlauf,1994, Davis and Stellwag,2010). However, in mibhi904 mutants there is no detectable hoxa5a expression in hindbrain (Fig. 5C,D) and only small collections of hoxb5b at rhombomere boundaries and in pharyngeal arches (Fig. 5G,H). In WT, dmbx1a was expressed robustly throughout midbrain and hindbrain (Fig. 6A). Transverse sections revealed prominent clusters of dmbx1a expression in optic tectum and retina (Figs. 6A1,A2). In midbrain, dmbx1a clusters were noted at the level of the cerebellum, in the cerebellar commissure and torus semicircularis (Fig. 6A3) and posteriorly reaching the medulla oblongata (Fig. 6A). In contrast, dmbx1a expression is barely detectable in mibhi904 mutants (Fig. 6B); only few dmbx1a-expressing cell clusters are visible in optic tectum, retina (Fig. 6B1,B3) and the midbrain–hindbrain boundary (Fig. 6B).
Another related homeobox gene (Fjose et al.,1994), dbx1a, exhibited two-fold up-regulation in microarray and almost two-fold up-regulation in qPCR in mibhi904 mutants. In WT, WISH revealed dbx1a expression in forebrain, continuing through midbrain and ending at the hindbrain–medulla oblongata boundary (Fig. 7A). Transverse sections show discrete domains of dbx1a in the pallium (Fig. 7A1) and similar clusters of expression in dorsal thalamus (Fig. 7A2), optic tectum, and pretectum (Fig. 7A2,A3). Caudal to the pallium, dbx1a expression appears in the eminentia thalami (Fig. 7A3); a small number of dbx1a-expressing cells can be seen in rostral hypothalamus (Fig. 7A4) and cerebellum (Fig. 7A5). WISH in mibhi904 mutant revealed a greater degree of dbx1a expression following the same distribution profile seen in WT siblings (Fig. 7B). In transverse sections, a dense strip of pallial dbx1a expression was observed in diencephalon and a very prominent expression pattern was noted in optic tectum, pretectum, and dorsal thalamus (Fig. 7B1) extending posteriorly (Fig. 7B2,B3). In diencephalon, some dbx1a-expressing cells were observed in the eminentia thalami (Fig. 7B3). More caudally, dbx1a expression extends dorsally reaching cerebellum and posteriorly into the medulla oblongata (Fig. 7B4,B5). There is no detectable dbx1a expression in hypothalamus of mibhi904 mutants.
Plexin D1, a receptor for the semaphorin family of ligands with a crucial role in regulating axonal pathfinding, neuronal patterning (Tamagnone and Comoglio,2000), and patterning of developing blood vessels (Torres-Vazquez et al.,2004), exhibited a three-fold up-regulation in microarray and nearly two-fold in qPCR. WISH shows faint plxnd1 expressed only at the level of the branchial arches in WT siblings (Fig. 8A). In sharp contrast, mibhi904 mutants show an aberrant increase in plxnd1 expression concentrated to the forebrain, optic tectum, and branchial arches (Fig. 8B). In contrast, neurexophilin 1 (nxph1), a family of neuropeptide-like secreted glycoproteins (Petrenko et al.,1996), appeared to have the opposite pattern of expression to Plexin D1 in mibhi904 mutants. WISH revealed clusters of nxph1 expression in forebrain, midbrain, and hindbrain in WT (Fig. 9A). In transverse sections, much of this nxph1 expression is restricted to the pallium region (telencephalon) and some glomerular structures within the habenula and dorsal thalamus (Fig. 9A1). Caudal to pallium, nxph1 expression appears in eminentia thalami (Fig. 9A2), ventral thalamus, through the posterior tuberculum (Fig. 9A2) and rostral hypothalamus (Fig. 9A3). Nxph1 expression appears heterogeneous in optic tectum and smaller scattered clusters were noted in medulla oblongata (Fig. 9A); nxph1 is barely detected around the branchial arches and medulla oblongata in mibhi904 mutants at 3 dpf (Fig. 9B,B1–B3).
GFP Expression in Live Zebrafish
Glial fibrillary acidic protein (Gfap) is a member of the intermediate filament family of proteins found in astroglial cells and early neuronal progenitors. Consistent with previous studies (Hegde et al.,2008), Gfap expression appeared to be six-fold down-regulated in microarray and almost four-fold in qPCR in mibhi904 mutants. To further assess in vivo expression using confocal microscopy, a transgenic line of zebrafish expressing a green fluorescent protein driven by Gfap specific regulatory elements (Gfap:GFP tg) was crossed into the mibhi904 background. In confocal images from anesthetized and immobilized WT siblings, Gfap drove expression (shown as green fluorescence) in telencephalon, retina and midbrain and was prominent in hindbrain and spinal cord (Fig. 10A,B). A distinctly different and more diffuse expression pattern was observed in mibhi904 mutants, where reduced Gfap-GFP fluorescence was observed in optic tectum and retina. Although more GFP fluorescence was spread posteriorly in hindbrain and medulla oblongata (compared with optic tectum), overall GFP fluorescence in mibhi904 mutants was diffuse and notably less than in WT controls (Fig. 10C,D).
The eight distal-less (Dlx) genes present in zebrafish (Ekker et al.,1992; Akimenko et al.,1994) encode a family of transcription factors involved in the formation of the forebrain, branchial arches, pharyngeal dentition, sensory organs, and limbs (Zerucha et al.,2000; Park et al.,2004; Borday-Birraux et al.,2006; Burton,2008; MacDonald et al.,2010a). Anesthetized and immobilized, 3 dpf WT larvae show GFP expression in the telencephalon, diencephalon, and optic tectum, but also in some groups of cells in the cerebellar region (Fig. 10E,F) consistent with previous reports in transgenic zebrafish at 5 dpf (Mione et al.,2008). Only few cells in diencephalon and cerebellum (Fig. 10G,H) express Dlx5a-6a in mibhi904 mutants.
Zebrafish mind bomb mutants are characterized by a severe neurogenic phenotype with defects somite, neural crest, and vasculature development. These defects have been interpreted as a consequence of abnormal Notch signaling (Jiang et al.,1996; Schier et al.,1996; van Eeden et al.,1996; Haddon et al.,1998; Riley et al.,1999; Lawson et al.,2001). For example, early in development, Notch target genes such as basic Helix–Loop–Helix (bhlh) or Hairy/Enhancer of Split (hes) are up-regulated in critical areas of the central nervous system (Bray and Furriols,2001) and these, in turn, suppress transcription of proneural genes such as neurogenin, which prevents neighboring cells from adopting a neuronal fate in a process called “lateral inhibition” (Camposortega,1995). Disruption of E3 ubiquitin ligase activity in mind bomb mutants leads to a failure in Notch signaling, resulting in a down-regulation of bhlh and hes genes, an excess of early differentiating neurons, a deficit of late differentiating neurons, impaired lateral inhibition (Schier et al.,1996; Jiang et al.,1996; Itoh et al.,2003; Park and Appel,2003; Yeo and Chitnis,2007) and a striking disorganization of all regions of the CNS (Golling et al.,2002). Decreased Notch activity is also suggested by reports of reduced expression of the transcriptional repressor her4 gene and increased expression of neurogenin-1 (Chen and Corliss,2004; Hegde et al.,2008; Hwang et al.,2009). Here, down-regulation in her4 and two additional Notch pathway genes (hes5 and bhlhb5) were confirmed in telencephalon, metencephalon, and optic tectum of mibhi904 mutants at 3 dpf. Hairy-related and hairy/enhancer of split genes were previously shown to be down-regulated in microarray studies on mibta52b mutants (Hwang et al.,2009), but these studies did not examine spatial expression patterns in the CNS. That these Notch signaling genes show decreased expression along the midline, a proliferative zone in the larval zebrafish brain (Mueller and Wullimann,2005), further suggests a critical role for E3 ubiquitin ligase in mediating neuronal differentiation.
Expression of a proneural bHLH gene (bhlhb5) required for forebrain organization and co-expressed in regions along the midline where Dlx genes (i.e., transcription factors required for the tangential migration of GABAergic interneurons during brain development; Anderson et al.,1997) have been reported (Mueller and Wullimann,2005) was nearly below detectable levels in mibhi904 mutants. Not surprisingly, loss of neurodevelopmental gene expression in this region of the developing brain ultimately leads to a disruption of forebrain cytoarchitecture and marked reductions in both GABAergic interneuron markers (Hortopan et al.,2010) and Dlx5/6-expressing interneurons (Fig. 10). At a functional level, neuronal disorganization and reduced interneuron density are likely contributors to the observed epileptic phenotype in mibhi904 mutants. Interestingly, expression of multiple members of the bhlh family of transcription factors (Lee,1997; Bramblet et al.,2002; McLellan et al.,2002; Xu et al.,2002) are also differentially regulated following chemically induced status epilepticus in rats (Elliott et al.,2001). Given that bhlh genes show similar expression changes in an epileptic mibhi904 mutant and in zebrafish CNS structures shown to generate abnormal electrical activity, e.g., optic tectum and telencephalon (Hortopan et al.2010), suggests they might be functionally related to epileptogenesis.
Dramatically reduced mibhi904 mutant expression of homeobox (hoxa5a and hoxb5b) genes was also noted. Hox genes act as super-regulators of development and are often simultaneously expressed in tissue where they activate or repress transcription of multiple downstream target genes involved in morphogenesis, segmental specification and neurodevelopment (Gilbert,2000; Kawahara et al.,2002). Both hoxa5a and hoxb5b show almost negligible expression in forebrain structures sub-serving higher brain functions such as the telencephalon and optic tectum; dmbx1a expression was robust throughout midbrain and hindbrain of control fish but nearly absent from mibhi904 mutants. In contrast, one up-regulated homeobox gene (dbx1a) was heavily overexpressed in the ventral forebrain and hindbrain. Recent studies in dbx1 mutant mice indicate that homeodomain transcription factors act upstream of Notch signaling, and it is believed that homeodomain proteins control spatial distribution of Notch ligands and proteins (Marklund et al.,2010). Using software that predicts interaction partners for one protein within a specific species and dbx1 as an example, we also noted a strong association between hes5, dmbx1a, and dbx1a. COGs (Clusters of Orthologous Groups of proteins) are very powerful and can help identify direct (physical) and indirect (functional) associations between genes. These in silica analyses support our qPCR findings, where we found a high Pearson correlation between the corresponding genes (Supp. Table S1; which is available online). To better illustrate these correlations and interactions, 3D graphs were created using relative expression values from qPCR. One example is shown in Supp. Fig. S1, where dbx1a expression, shown to be up-regulated in the microarray assay in mibhi904 mutants (B), is plotted with two other down-regulated genes, her4.2 and nxph1. The pattern observed in WT zebrafish (A) is opposite; these last two genes, her4.2 and nxph1, increase their expression while the dbx1a drops down (inversely correlated). A different pattern of response was observed when the same gene, dbx1a, was plotted with other two genes, plxnd1 and hes5, in WT (C) and in mibhi904 mutants (D). This would be the first step for the definition of a model to predict co-variation gene expression patterns and for the use of some of these genes (molecular markers) as a potential diagnostic tool.
Notch signaling is also involved in vascular development (Jakobsson et al.,2009; Roca and Adams,2007) and vascular defects in the trunk of three different Notch zebrafish mutants were recently described (Therapontos and Vargesson,2010). Although these studies focused on trunk vascularization, expression of plxnd1 in the head of these mutants was also reported and is similar to the pattern of expression we observed in mibhi904 mutants. Of interest, nxph1, expressed in discrete clusters in the habenula, pallium, and ventral thalamus of control fish (but nearly absent in the forebrain and midbrain of mibhi904 mutants) functions as an endogenous ligand for α-neurexins (Missler and Sudhof,1998). Neurexins, together with neuroligins, are thought to play an essential role in synaptic transmission, particularly at GABAergic synapses (Craig and Kang,2007). This is consistent with our recent demonstration (Hortopan et al.,2010) that GABA signaling may be reduced in mind bomb mutants presumably contributing to defective inhibitory synaptic transmission and epilepsy. Finally, neurexin-ligand interactions are also important for development and/or maturation of synaptic connections (Clarris et al.,2002) and implicated in the pathophysiology of neurodevelopmental disorders, e.g., neurexin-1 was recently associated with autism (Ching et al.,2010) a co-morbidity noted in children with Angelman syndrome.
When mibhi904 mutants were crossed with transgenic reporter lines (Gfap:GFP or Dlx5/6:GFP) a strong general down-regulation in fluorescence was noted. For glial fibrillary acidic protein, this is consistent with microarray and qPCR data (Hortopan et al.,2010) and previous independent analysis of the spinal cord (Song et al.,2010). Mutations in human gfap have been described in association with a severe childhood brain disorder, i.e., Alexander disorder (Brenner et al.,2001; Quinlan et al.,2007) characterized by enlarged brain and head size, seizures, stiffness in the arms/legs, intellectual disability, and developmental delay. Dlx5a-6a:GFP transgenic zebrafish provide a means to further study GABAergic interneurons (Mione et al.,2008; MacDonald et al.,2010b). Confocal images of mibhi904 zebrafish mutants crossed into this reporter line show a clear decrease of cells in forebrain and midbrain, suggestive of a failure in the differentiation or migration of early born interneurons. Although deficits in cell density could be a contributing factor to gene expression patterns seen throughout this manuscript, general brain morphology as indicated by these in vivo GFP imaging studies suggests that major CNS structures are largely intact in mibhi904 mutants.
In conclusion, altered expression levels and patterns in mibhi904 mutants of genes critical to early stages of neurodevelopment support the view that an ubiquitin E3 ligase is involved in Notch signaling. In contrast to earlier studies of mind bomb zebrafish mutants focused on development of the pituitary gland (Dutta et al.,2008), hindbrain (Bingham et al.,2003), spinal cord (Itoh et al.,2003), or mesoderm (Hwang et al.,2009), our studies focused on gene expression patterns in critical regions of the zebrafish telencephalon and metencephalon. Failure or reduction in ligase activity in these regions can, as shown here, lead to down-regulation of several Notch signaling genes that are required for normal neuronal development. Loss of this signaling led to dramatic alterations in how the brain develops with potentially catastrophic consequences for inhibitory synapse formation, cognitive function, and survival (mib mutants do not survive past 4 dpf). Although this study only represents a relatively small subset of neurodevelopmental genes that may be disturbed, it is possible that mibhi904 mutants could help to characterize mechanisms that underlie symptoms of disorders that require proper development of synaptic circuits.
Animals and Maintenance
Heterozygote zebrafish (mind bomb, line #hi904) were obtained from the Zebrafish International Resource Center (Eugene, OR; http://zebrafish.org/zirc/fish/lineAll.php). The following transgenic lines were also used in this study: Gfap:GFP (Tg(gfap:GFP)mi2001/+; Chen et al.,2010) and Dlx5a-6a:GFP (Zerucha et al.,2000). Adult zebrafish were maintained according to standard procedures (Westerfield,1993), and following guidelines approved by the University of California, San Francisco Institutional Animal Care and Use Committee. Zebrafish embryos and larvae were maintained in egg water (0.03% Instant Ocean).
At 3 dpf, larvae were sorted by morphology and used for RNA isolation. Total RNA was isolated from 14 pools of larvae (4 fish/pool): mib mutants (n = 7) and age-matched WT sibling controls (n = 7). Fish were treated with Trizol Reagent (Invitrogen, Carlsbad, CA), RNase-free DNase to remove possible genomic DNA contamination, and quantified with GeneQuant spectrophotometer.
PCR, Cloning, and Sequencing
cDNA was generated using a mix of random primers and oligo(dT)20 in a reverse transcription kit (SuperScript III First-Strand Synthesis System, Invitrogen) according to the manufacturer's protocol. Primers pairs, forward and reverse, were specifically designed using Primer 3 web software (http://frodo.wi.mit.edu/primer3/) for each investigated gene (Supp. Table S2). The most conserved regions were identified by sequence alignment (ClustalW, Thompson et al.,1994) of all available gene sequences from GeneBank including other fish species. We used BLAST software to investigate primer cross-specificity and Mfold software (Zuker,2003) to check for secondary structure of the entire DNA sequence. Each reaction cycle (32 loops) consisted of incubations at 94°C (30 sec), 60°C (30 sec), and 72°C (60 sec) with Taq DNA Polymerase (Taq PCR Core kit, Qiagen). A 2% agarose gel electrophoresis stained with ethidium bromide was used to separate PCR products which were further cloned in pCRII-TOPO plasmid vector (TOPO TA Cloning System, Invitrogen) according to the manufacturer's specifications. DNA sequencing was performed by Elim Biopharmaceuticals, Inc. (Hayward, CA).
Quantitative Real-Time PCR (qPCR)
Gene expression levels were determined by real-time qPCR using SybrGreen fluorescent master mix on an ABI Prism 7700 Sequence Detection System driven by ABI prism SDS v9.1 software (Applied Biosystems). The cDNA templates were diluted 1:2 with DEPC (diethyl pyrocarbonate) sterile water before qPCR applications to minimize the presence of potential inhibitors. Primer Express v3.0 software (Applied Biosystems) was used to design all primers on our own sequenced cDNA to produce amplicons ranging in size between 71 bp and 125 bp (Supp. Table S3) and then synthesized by Invitrogen. Samples were run in triplicate in 10 μL of 1× SYBR green master mix containing 100 nM of each primer and RNAse free water. Samples without reverse transcriptase and samples without RNAs were run for each reaction as negative controls. Cycling parameters were as follows: 50°C × 2 min, 95°C × 10 min, then 45 cycles of the following 95°C × 15 sec, 60°C × 1 min. For each sample, a dissociation step was performed at 95°C × 15 sec, 60°C × 20 sec, and 95°C × 15 sec. Dissociation (melting) curve analysis showed no sign of primer-dimers or other nonspecific reaction products.
For qPCR data, significant differences were considered at P value ≤ 0.05 (Student's t-test). Relative quantification of the target gene transcript with β-actin reference gene transcript (Hortopan et al.,2010) was made following both the Comparative ΔΔCT (Livak and Schmittgen,2001) and the Efficiency Based (Pfaffl,2001) methods using qCalculator software (programmed by Ralf Gilsbach, Institute of Pharmacology and Toxicology, University of Bonn, Germany), which also estimates qPCR efficiency E = 10(−1/slope). Similar results were obtained with both types of analyses. Standard curves for all nine genes, to estimate qPCR efficiencies, were constructed using a four-fold serial dilution of pooled cDNA; 5 standards assayed in triplicate: 1/1; 1/4; 1/16; 1/64; 1/256 (the efficiencies, slope of the curves and the correlation coefficient are summarized in Supp. Table S3).
Whole-Mount In Situ Hybridization (WISH)
Antisense and sense RNA probes were generated from plasmids corresponding to each of the nine selected genes using specific restriction enzymes for linearization (New England Biolabs, UK). Linearized DNA template (1 μg) was purified (QIAquick, Qiagen) and incubated for 3 hr at 37°C in a solution containing 10× transcription buffer, dithiothreitol (DTT; 100 mM), 10× Dig NTP Mix (Roche), RNAse inhibitor (20 U/μl), and RNA polymerase (20 U/μl) T7 or SP6. After digestion of the DNA template with DNase (10 U/μl) for 15 min at 37°C and incubation, the product was purified using a mix of RNAse-free water and LiCl (30 μl, 1:2) and left overnight at −20°C. After centrifugation at 4°C and washing with 70% ethanol (RNAse free), the pellet was dried and stored in hybridization mix solution at −20°C until use.
Three days postfertilization embryos, mib mutants (n = 8), and WT controls (n = 8) for each antisense and sense RNA probes, were sorted and fixed in 4% paraformaldehyde (PFA) then stored in 100% methanol at −20°C. Following storage at −20°C, fixed larvae were rehydrated in a series of methanol and phosphate buffered saline-0.1%Tween20 (PBST) washes. WISH was performed as previously described (Hauptmann and Gerster,1994). Larvae were fixed in 4% PFA, washed in PBST, and processed for cryo-sectioning using quick-frozen samples mounted in O.C.T. compound (Tissue TEK; slice thickness: 15–20 μm).
Transgenic Gfap:GFP zebrafish (Tg(gfap:GFP)mi2001/+) founder lines were obtained from the Zebrafish International Research Center (http://zebrafish.org/zirc/home/guide.php). Adult GFP founder lines on an AB background were crossed with heterozygote mind bomb (mibhi904Tg/+) founders on an AB background. F1 fish were sorted by fluorescence as embryos, raised to adulthood and crossed to obtain mib:Gfap:GFP larvae. 30 fish larvae were sorted at 3 dpf (mib; n = 15 and WT; n = 15) and anesthetized in a cocktail containing 0.02% Tricaine and α-bungarotoxin (1 mg/ml) or curare (4.5 mM), then immersed in 1.2% low melting point agarose to immobilize and orientate the embryos for imaging. Visual assessment of GFP expression was performed using a Leica SP5 confocal scanning fluorescence microscope. Confocal images were reconstructed using z-stack projections produced from serial scanning every 4 μm.
Microscopy and Imaging
Tissue sections of larval zebrafish were chosen to be representative of gene expression and matched, as well as possible, with respect to location and cutting angle. WISH figures are annotated to show the approximate location and cutting angle for each section. Pictures of whole-mount in situ hybridization embryos mounted in 70% glycerol and slide-mounted cryosections were taken using a Zeiss Axioskop microscope equipped with a computer-controlled Optronics MicroFire camera system. Raw images were imported into Adobe Photoshop and slightly adjusted for contrast and sharpness.
Neuroanatomical designations were taken from Mueller and Wullimann (2005) in consultation with Thomas Mueller.
We thank Matthew Dinday and Thomas Mueller for their valuable contributions to this manuscript. S.C.B. was funded by funds from the National Institutes of Health.