|Home | About | Journals | Submit | Contact Us | Français|
Physical interaction between the transmembrane proteins Delta and Notch allows only a subset of neural precursors to become neurons, as well as regulating other aspects of neural development. To examine the localization of Delta protein during neural development, we generated an antibody specific to zebrafish DeltaA (Dla). Here we describe for the first time the subcellular localization of Dla protein in distinct puncta at cell cortex and/or membrane, supporting the function of Dla in direct cell-cell communication. In situ RNA hybridization and immunohistochemistry revealed dynamic, coordinated expression patterns of dla mRNA and Dla protein in the developing and adult zebrafish nervous system. Dla expression is mostly excluded from differentiated neurons and is maintained in putative precursor cells at least until larval stages. In the adult brain, dla mRNA and Dla protein are expressed in proliferative zones normally associated with stem cells.
Notch ligands, including members of the Delta and Serrate families, are known to be involved in a sequential neuronal differentiation process, called lateral inhibition or lateral specification [see for reviews (Gaiano and Fishell, 2002; Chitnis, 2006; Louvi and Artavanis-Tsakonas, 2006)]. In a classical view, this process allows neighboring neural precursors to adopt distinct fates. For example, neural precursors expressing higher levels of Notch ligands may differentiate into neurons and at the same time inhibit differentiation of neighboring cells. These neighbors express lower levels of Notch ligand and may remain in an undifferentiated state, or they may adopt an alternative fate, such as differentiating as glia. The ability of Notch ligand-expressing cells to affect differentiation of their neighbors requires physical interaction of the transmembrane ligand, Delta or Serrate (also known as Jagged) with Notch or Notch-related transmembrane receptors, thus allowing selective cell-to-cell communication (Greenwald and Rubin, 1992; Shimizu et al., 2002; Yang et al., 2005; Yang et al., 2006). Upon ligand binding, Notch is cleaved and the intracellular domain enters the nucleus, where it regulates the expression of transcription factors.
Four delta genes and one delta-like gene 4 (dll4) are known to be expressed in the zebrafish embryo (Dornseifer et al., 1997; Appel and Eisen, 1998; Haddon et al., 1998; Leslie et al., 2007). These genes have partially overlapping but also unique expression patterns in the nervous system and gene-specific expression patterns in non-neural tissues. delta gene expression is found throughout all embryonic germ layers (ectoderm, mesoderm, and endoderm), where it appears to control cell fate through a conserved mechanism. For example, in endodermal derivatives, DeltaA (Dla) regulates the sequential segregation of pancreatic precursor cells from a common precursor pool (Zecchin et al., 2007). In non-neural ectodermal structures, Dla and DeltaD (Dld) were identified as regulators of choroid plexus formation (Bill et al., 2008). deltaC (dlc) shows high expression levels in the vascular system (Burns et al., 2005; Qian et al., 2007), controlling differentiation of hematopoetic precursor cells. In mesodermal tissues, dlc and dld are expressed in presomitic mesoderm and somites, and were shown to be required for somite segmentation (Holley et al., 2002; Jülich et al., 2005; Oates et al., 2005). dll4 is implicated as an important factor for normal vascular remodeling (Leslie et al., 2007). Interestingly, in the developing hindbrain, Dla function is implicated in maintaining rhombomere boundaries (Amoyel et al., 2005), suggesting a conserved function of Notch signaling in boundary formation. In the spinal cord, Delta proteins maintain a precursor pool by preventing their premature differentiation (Appel et al., 2001). Compromised Delta function results in a neurogenic phenotype, showing an increased number of early-differentiating neurons at the expense of later-developing neurons, and eventually leading to a premature depletion of the neuronal precursor pool (Appel and Eisen, 1998; Appel et al., 2001). Delta function also mediates the choice between the Rohon-Beard (RB) spinal sensory neuron fate and the neural crest fate (Cornell and Eisen, 2000). Although these studies demonstrate the importance of Delta function during development, particularly within the nervous system, the details of how precursor cells are chosen for a certain cell fate are still unclear. Learning the subcellular localization of Delta proteins is important for future studies to fully understand the mechanisms underlying Delta function.
We focused our analysis on the distribution and localization of dla mRNA and Dla protein in the developing and adult zebrafish nervous system. We show that dla mRNA and Dla protein are present at the same time in the same cell populations and that Dla protein is localized in puncta at the cell cortex and/or membrane. Thus, Dla is located in the right place at the right time to interact with Notch during cell-to-cell communication to determine neural cell fate.
We generated monoclonal antibodies recognizing the last 167 amino acids of the C-terminal portion of the zebrafish Dla protein (ZDB-GENE-980526-29 on LG1), which is the most divergent region between all Delta family members. We isolated two monoclonal antibodies (14A10 and 18D2) that resulted in identical labeling patterns in whole-mount zebrafish embryos; we refer to both of these as Dla antibody. For this study, we used mostly 18D2. However, we used 14A10 for the colocalization experiments in Fig. 7, because 18D2 is the same isotype as the zdD2 Dld antibody (Crosnier et al., 2005; Matsuda and Chitnis, 2009).
Overlap between dla mRNA expression and labeling of both of our Dla antibodies (Fig. 1) suggested that they are both specific for Dla. We compared the expression of dla mRNA and Dla protein in stage-matched sibling embryos. At 24 hours postfertilization (hpf), dla transcripts are expressed in the developing forebrain, midbrain, and hindbrain (Fig. 1A). In the forebrain dla expression appears strongest in the ventral telencephalon, located above the optic recess, and in two longitudinal stripes in the diencephalon. Although we detected dla in the tectum, expression in the tegmentum appeared stronger. We found that dla transcript is depleted from a zone separating the midbrain and the hindbrain (Fig. 1A), described as the intervening zone (Geling et al., 2004). At 24 hpf this zone is located at the midbrain-hindbrain boundary, a region which is kept in an undifferentiated state by pathways independent of Notch signaling (Geling et al., 2004). Dla protein (Fig. 1B) labeling reflected dla mRNA distribution. A section of the head gives a better resolution of Dla protein staining, showing Dla in several neuronal clusters throughout the developing brain and in a ventrally located string of cells along the basal midbrain and hindbrain (Fig. 1S).
A dorsal view of the head (Fig. 1E) revealed dla mRNA expression in the olfactory bulbs. Within the brain, dla transcript was restricted to some groups of cells. More laterally, we detected expression in the ventrally located cranial ganglia. High levels of dla mRNA were expressed in the cerebellum and in a striped pattern in all rhombomeres, adjacent to but not within the boundaries, as has been previously reported (Cheng et al., 2004; Amoyel et al., 2005). Dla protein labeling (Fig. 1F) appeared in similar regions as dla mRNA. Note that dla mRNA (Fig. 1E) and Dla protein expression levels (Fig. 1F) appeared weaker in the ventricular zone at this stage. This is in contrast to the expression pattern observed at 48 hpf, when dla mRNA is expressed in the dorsal forebrain, tectum, cerebellum and rhombomeres (Fig. 1I,M), with high levels in proliferative zones near the midline. Further, dla expression formed a distinctive pattern in the rhombomere borders and in bilateral strings of cell clusters (Fig. (Fig.1M1M and and4F).4F). This intriguing hindbrain expression is described in more detail later. dla mRNA was also prominent in the trigeminal ganglia (Fig. 1I).Regional Dla protein distribution was similar to dla mRNA expression.
In the spinal cord, labeling of Dla protein at 24 hpf (Fig. 1D,H) and 48 hpf (Fig. 1L,P) reflected the expression pattern of dla mRNA at both stages (Fig. 1C,G and Fig. 1K,O, respectively). We found Dla protein expression in scattered cells throughout the spinal cord. A dorsal view of the spinal cord (Fig. 1P) revealed a confined pattern of Dla protein location within the developing spinal cord at 48 hpf. Interestingly, Dla appeared more punctate in dorsal views (Fig. 1H) than in lateral views (Fig. 1D). We describe this pattern in more detail later.
The spatiotemporal Dla protein distribution during development is very dynamic. This distribution pattern is consistent with the function of the Notch signaling pathway as a regulator of cell fate and cell maturation state, and with possible oscillatory expression to regulate maintenance of progenitor cells, as has been recently demonstrated for Delta-like1 (Shimojo et al., 2008). Early in development (14 hpf), Dla protein was widely expressed throughout the dorsoventral extent of the neural plate (Fig. 1Q). At 18 hpf, we found high expression levels of Dla protein in the dorsolateral region of the hindbrain and spinal cord (Fig. 1R). We noticed some accumulation of Dla puncta in the medial region of the spinal cord. Restriction of Dla labeling to groups of cells in the developing brain becomes visible in a confocal section (Fig. 1S). In contrast to the high number of Dla-positive cells in early stages, later in development Dla was restricted to fewer cells. For example, in 5 days postfertilization (dpf) larvae, we found Dla protein in cells located in a medial region of the spinal cord, in close proximity to the ventricle (Fig. 1T). This is in agreement with the observation that later in development, Dla is located within or in proximity to identified proliferative zones (Wullimann, 2005; Chapouton et al., 2006). Consistent with the role of Delta-Notch signaling in the choice between neural cell fates, as well as in other processes later in development, Dla protein in the spinal cord did not appear to be localized to specific cell types. For example, we tested antibodies to markers for motoneurons and RB neurons (Islet1/2) (Korzh et al., 1993), and for interneurons (GABA), and did not find any correlation between Dla expression and expression of these markers (not shown).
To further verify the specificity of our Dla antibody, we tested it using a protein knock-out approach. The zebrafish line dlahi781Tg (Chen et al., 2002; Golling et al., 2002) carries a viral insertion disrupting the dla locus in the second exon of the coding region, leading to early termination. We identified adult fish carrying the viral insertion using dla-specific external primers and a viral insertion-specific internal primer. These primers distinguish between unaffected (no insertion) and affected (insertion) dla gene loci, respectively. In a cross of two heterozygous carriers we found about 25% homozygous mutants, confirming Mendelian inheritance of the insertion. Disruption of dla function leads to a premature maturation of early differentiating primary motoneurons and RB neurons at the expense of later differentiating neurons and other cell types (Appel and Eisen, 1998; Cornell and Eisen, 2000; Lutolf et al., 2002; Park and Appel, 2003; Shin et al., 2007). Consistent with previous reports, the number of RB neurons appeared higher in homozygous carriers (dlahi781/hi781) of the dla viral insertion (Fig. 2A,B), suggesting that Dla function is depleted. In dlahi781/hi781 embryos, the insertion leads to a premature stop codon early in the N-terminal part of Dla, suggesting to result in a null allele. Importantly, the C-terminal part, which contains the epitope recognized by the Dla antibody, is absent. Indeed, only dlahi781/hi781 mutants homozygous for the viral insertion show an absence of Dla labeling (Fig. 2B), suggesting that Dla antibody is specific to Dla and does not recognize other Delta proteins. Consistent with this idea, Dla labeling is normal in embryos homozygous for a mutation in deltaD (data not shown). We genotyped single embryos after analyzing their Dla protein level to confirm the presence of an uninterrupted dla locus (Fig. 2C) and/or the presence of the viral insertion perturbing the dla locus (Fig. 2D). The Dla staining and the PCR genotype correlated in all embryos tested (n=18).
To further characterize the Dla expression profile in neuronal cells, we decided to use the well-described Elavl antibody (previously known as anti-HuC/D) (Marusich et al., 1994) as a marker for differentiating and mature neurons (Okano and Darnell, 1997; Wakamatsu and Weston, 1997). In the spinal cord, Dla and Elavl were expressed in domains that appeared mostly exclusive of one another (Fig. 3). In general, we detected very few cells co-labeled for Dla and Elavl proteins (not shown). Cells that did express both proteins might have been caught in a transition state, in which the cell had begun to express Elavl before fully down-regulating expression of Dla. In a few instances we saw cells of the same type that exhibited differences in Dla and Elavl labeling. For example, in a pair of bilaterally-positioned developing cells in the dorsal spinal cord, one cell labeled with Elavl, identifying it as a committed RB neuron (Fig. 3F). In contrast, the cell situated on the opposite side of the dorsal spinal cord labeled with Dla, suggesting that it had not yet committed to a neuronal fate. Thus, protein expression in these two cells reveals two different maturation stages.
To gain a better appreciation of the spatial localization of Dla and Elavl proteins, we examined their distributions at different locations within the spinal cord. We examined sections taken very laterally, close to the overlying muscle, as well as sections taken more medially, closer to the spinal cord midline (Fig. 3G). By 24 hpf, the number of cells expressing either Dla or Elavl protein appeared to be different in lateral (Fig. 3A,A’) versus medial spinal cord regions (Fig. 3C,C’). The difference in the ratios of Dla and Elavl expressing cells became more striking at 48 hpf; the majority of cells in the lateral region expressed Elavl (Fig. 3B,B’) and only a few cells expressed Dla. In contrast, in the more medial region, we found a high number of Dla-positive cells and few Elavl-positive cells (Fig. 3D,D’). To confirm this pattern, we examined transverse sections of the spinal cord (Fig. 3E-F”). Consistent with the observations described above, we found Dla protein in cells closer to the midline, whereas Elavl was expressed in cells located more laterally (Fig. 3E-E” and not shown); these Dla-expressing cells were often in direct contact with Elavl-expressing cells. Elavl expression identifies cells that have already committed to a neuronal fate, and in some cases have differentiated as neurons, consistent with their location at the periphery of the developing spinal cord. In contrast, Dla is typically expressed in cells that are in the process of making a fate commitment, and have not yet begun to differentiate. We found that cells adjacent to the midline rarely expressed Dla. Our interpretation is that these cells are within the proliferative zone and are maintained in an undifferentiated state by activation of Notch through Dla expressed by their more lateral neighbors. Our results are consistent with the developmental pattern of the spinal cord, in which progenitor cells are located medially, next to an intermediate zone in which cells are undergoing commitment, which in turn is next to a lateral zone in which committed neurons are differentiating.
The division into zones representing different differentiation stages was very pronounced in the developing brain, shown in 48 hpf old embryos (Fig. 4). In the forebrain, ElavI protein was located in two bilateral stripes along the anteroposterior axis of the brain, whereas Dla was expressed in cells in closer proximity to the midline (Fig. 4A-A”, see Fig. 4G for orientation). A more dorsal and caudal section showed a similar Dla-positive domain, with 1 or 2 cell rows on either side of the midline expressing Dla (Fig. 4B-B”). In addition, Dla-expressing cells were aligned in a striped pattern, revealing rhombomere borders. At this medial level, we detected Elavl-positive cells arranged in a similar striped pattern, but not overlapping with Dla-positive cells. Elavl labeling was more robust at a slightly more dorsally located level (Fig. 4C-C”).
Compared to Dla, the Elavl expression pattern was less orderly and we noticed that the intensity of the Elavl signal was variable. We conclude that, as has been reported for dla RNA at 18 hpf (Amoyel et al., 2005), Dla protein is located adjacent to rhombomere boundaries where Notch signaling helps shape the rhombomeres by lateral inhibition (Fig. 4D-D”). Neuronal precursor cells located within the rhombomeres are able to differentiate (Trevarrow et al., 1990) and hence are positive for Elavl (Fig. 4E-E”). In contrast, Dla-positive cells located near the rhombomere boundary are thought to allow cells to adopt a neural fate and, in addition, to contribute to suppression of boundary cell formation (Amoyel et al., 2005). Consistent with Dla protein distribution, we observe a highly ordered pattern in dla mRNA expression, visible as strings of cells arranged along the anteroposterior and mediolateral axes (Fig. 4F), with either one or two cell-wide gaps between dla-positive cells. As in the spinal cord, hindbrain neurons are generated sequentially from precursors over an extended period during development (Kintner, 2002). Thus, this pattern might be important to allow controlled growth of the rhombomeres in the developing hindbrain, using a mechanism common to various growing brain regions (Fish et al., 2008).
Precursor cells located in the ventricular zone along the neuraxis can expand, self-renew, and generate postmitotic neurons and/or glia (García-Verdugo et al., 2002). The precursor cells that give rise to the neurons generated in the adult teleost brain are generally located in the walls of the brain ventricles [for reviews see (Stigloher et al., 2008; Zupanc, 2008)]. From these proliferative regions, neuronal precursors migrate toward their final targets where they differentiate (Alvarez-Buylla and Lois, 1995; García-Verdugo et al., 2002; Kintner, 2002)(Alvarez-Buylla and Lois, 1995; García-Verdugo et al., 2002; Kintner, 2002). In the adult zebrafish brain, we found Dla protein expression in one or two cell rows delineating the walls of the mesencephalic ventricle (Fig. 5A, see Fig. 5F for orientation) and also in the walls of more anterior brain regions (not shown). In the spinal cord dla is expressed primarily in post-mitotic neurons, although it is also expressed in a subset of BrdU-positive cells that are still in the process of dividing (Appel et al., 2001)(Appel et al., 2001). The Dla-positive cells within the wall of the mesencephalic ventricle were interspersed with single cells that expressed the neuronal marker Elavl; there were also Elavl-positive cells at various distances from the ventricle wall. In addition, there are cells in the mesencephalic ventricle wall that express neither Dla or Elavl. We interpret the Elavl-positive cells interspersed between Dla-positive cells as committed neurons; the Elavl-positive cells located in close proximity to the ventricle might be newly differentiating neurons caught en route, as they are migrating to their final destinations. We also found large domains of Elavl-expressing neurons adjacent to the ventricle, which is not surprising as Elavl was reported to be expressed in postmitotic neurons throughout development of the CNS (Sakakibara and Okano, 1997). We investigated in more detail whether Dla and Elavl are co-expressed by examining a region that had a neuronal cluster expressing Elavl located adjacent to the ventricular zone delineated by Dla-expressing precursors (Fig. 5B-B”). The majority of cells clearly expressed either Dla or Elavl. Only a few cells were labeled for both proteins, suggesting that these cells might be down-regulating Dla as they start to differentiate into neurons, as we have previously shown for spinal motoneurons (Appel and Eisen, 1998).
In addition to the ventricular zone, we found scattered cells expressing Dla in regions adjacent to, but more distant from, the ventricle (Fig. 5C-C”). Again, Elavl expression was excluded from cells positive for Dla protein. Consistent with Dla protein labeling, we found dla mRNA expression in the walls of the brain ventricle (Fig. 5D). We examined whether Dla protein was expressed in cells positive for the glial fibrillary acidic protein (GFAP), a marker for the neural precursor cells called astrocytes in the adult brain (Doetsch, 2003). As expected, we found Dla protein and GFAP co-expressed in cells near the ventricular zone, confirming Dla protein in neural precursor cells in the adult zebrafish brain.
After analyzing the overall distribution of Dla protein, we were interested in localizing Dla to a subcellular structure. To accomplish this, we performed double labeling of Dla protein and the cortical marker, Beta-catenin (Nathke et al., 1994) (Fig. 6A-D) or the membrane marker, GFP in the transgenic line Tg(Bactin:HRAS-EGFP) in which GFP is in the plasma membrane (Cooper et al., 2005) (Fig. 6E). Both experiments suggest that in the embryonic spinal cord, Dla protein is localized to puncta at the cell membrane and/or cortex; widespread Dla localization in the cytoplasm is not supported by our data (Fig. 6). Interestingly, the apparent accumulation of Dla protein in clusters did not show a specific cellular pattern (Fig. 6A). One interpretation is that cell-cell interactions are occurring between many different cell types at this stage (26 hpf) of spinal cord development. These Dla accumulations were typically localized to one region of the cell cortex and/or membrane, suggesting that, if these Dla puncta are involved in intercellular communication via Delta-Notch signaling, this signaling may be restricted to only one adjacent cell (Fig. 6B). Occasionally, Dla clusters were found adjacent to two or more other cells (Fig. 6C); this could be interpreted to mean that interactions can also occur with more than one cell at a time. Another possibility is based on recent work that suggests that Delta and Notch are localized to the region underlying the zonula adherens, and that this localization is important for Delta-Notch signaling in polarized epithelial cells and potentially during vertebrate neurogenesis (Krahn and Wodarz, 2009; Ossipova et al., 2009). If this is also the case in the neural cells in developing zebrafish, then the puncta could appear very different when viewed in differently oriented embryos. Although the spatial localization of Dla protein in the membrane and/or cortex tended to be punctate during early development (Fig. 6A-E,G), in adult brain Dla protein appeared to be more evenly distributed around the cell (Fig. 6F). The interesting differences in Dla protein localization we report raise questions about possible differences in Dla function at different stages and in different central nervous system regions. These questions will need further investigation to be answered.
Our observation that Dla protein was mostly restricted to the cell membrane and/or cortex at the stages we examined contrasts with a recent report that Dld protein is primarily localized in cytoplasmic puncta in the developing nervous system at both early tail bud stages and later in the hindbrain, and is not colocalized with Beta-catenin protein (Matsuda and Chitnis, 2009). However, in the hindbrain there does appear to be some colocalization of Dld protein and Beta-catenin [compare Fig. 1F’ of (Matsuda and Chitnis, 2009) with Fig. 6A and Fig. 7 of this paper]. Double labeling of Dla and Dld proteins (Fig. 7) suggests that Dla and Dld differ somewhat in localization. Dla protein appears to be restricted to the cell membrane and/or cortex. Dld protein is also localized to the cell membrane and/or cortex. However, in addition, Dld protein also shows clear localization to the cytoplasm, as previoiusly reported (Matsuda and Chitnis, 2009). The expression similarities and differences between Dla and Dld raise questions about the specific functions of the two different Notch ligands, as the genes encoding both proteins share mRNA expression patterns in at least some parts of the nervous system.
In conclusion, we show that dla mRNA and Dla protein are expressed in a similar pattern in the developing and adult nervous system, and that embryos that do not express dla mRNA also do not label with our Dla antibody. We found that Dla was broadly expressed in undifferentiated cells within the central nervous system, probably precursor cells, and was not expressed in identified neurons. Dla protein clusters are located in the right place to promote selective and specific cell-to-cell communication via Delta-Notch signaling.
Embryos were collected from natural crosses of wild-type (AB or AB/TU), Tg(Bactin:HRAS-EGFP) (allele vu119) (Cooper et al., 2005), or dlahi781Tg/+ (Chen et al., 2002; Golling et al., 2002; Amsterdam and Hopkins, 2004) fish in the University of Oregon Zebrafish Facility, and were staged hourly at 28.5°C according to Kimmel et al. (Kimmel et al., 1994).
Fish were anaesthetized in tricaine, placed in ice water for 5-10 minutes, decapitated, their jaws removed, and the heads fixed for 5 hours at 4°C in 2% paraformaldehyde fixation buffer. Brains were dissected in PBS, rinsed in 30% sucrose and embedded in sucrose/agarose. Embryos for in situ hybridization on sections were fixed over night at 4°C in 4% paraformaldehyde fixation buffer, rinsed in PBS, put in 30% sucrose and embedded in sucrose/agarose. Embryos for imaging antibody distribution were sectioned after the Immunohistochemistry procedure. Embryos were rinsed in PBS, equilibrated in 30% sucrose solution and embedded in sucrose/agarose. Brain tissue and embryos were cut into 16 μm thick sections.
We generated antibodies against the C-terminal region of Dla protein, encoded by the gene dla (ZFIN:ZDB-GENE-980526-29 on LG 1). We amplified a fragment of about 500 bp (corresponding to amino acid positions 636-802) and cloned it into the expression vector pET28-C(+) using the following primers (sequences binding to dla DNA are underlined):
We purified Dla fusion protein using Ni-NTA columns under denaturing conditions (QIAexpressionist). Mice were immunized with about 2-4 ng Dla protein in the University of Oregon Monoclonal Antibody Facility and splenocyte myeloma fusions were generated (Marusich, 1988). We screened for Dla-specific antibodies using ELISA, Western blot analyses using a DeltaA fusion protein, and finally by using antibody serum on whole-mount zebrafish embryos. We isolated two different hybridoma cell clones, derived from the same mouse, that produced Dla-specific antibodies; cell clone 18D2 was identified as an IgG1 isotype, 14A10 as an IgG2a isotype. We found that antibodies from both cell clones produce the same pattern of cell labeling. Dla antibodies were kept in cell culture serum. The 18D2 Dla antibody is available from the Zebrafish International Resource Center (ZIRC; http://zebrafish.org).
RNA in situ hybridization and immunohistochemistry on whole embryos were carried out according to standard protocols (Hauptmann and Gerster, 1994). In situ hybridization on cryostat sections were performed using a modified protocol from Jensen et al. (Jensen et al., 2001). For RNA in situ hybridization we used the previously described probe dla (Appel and Eisen, 1998; Haddon et al., 1998). Whole embryo and adult brain section immunohistochemistry were performed with following primary antibodies: Dla (IgG1, see above) was used at 1:7; Dla 14A10 (IgG2a, see above) was used at 1:5; anti-Elavl (formerly known as anti-HuC/D 16A11; IgG2b; University of Oregon) was used at 1:2000; rabbit anti-Beta-catenin (Sigma) was used at 1:1000; anti-Islet (IgG2a; 39.4D5 Developmental Studies Hybridoma Bank) was used at 1:200; anti-GFAP (IgG1; Chemicon) was used at 1:1000 and anti-Dld (zdD2) was used at 1:50 Primary antibodies were revealed using secondary antibodies coupled to Alexa Fluor488 [goat anti-mouse IgG1, IgG2a, IgG2b and (H+L), Molecular Probes], Alexa Fluor546 [goat anti-mouse IgG1, IgG2a, IgG2b and (H+L), Molecular Probes] and AlexaFluor568 (goat anti-mouse IgG2b) were used at 1:750; AlexaFluor546 [goat anti-rabbit, IgG (H+L) Molecular Probes] was used at 1:500 and Goat Anti-rabbit Cy5 [IgG (H+L), Jackson ImmunoResearch Laboratiories] was used 1:750.
Heterozygous fish and single embryos were genotyped for the presence of the viral insertion perturbing the dla locus. We performed PCR on single embryos after immunohistochemistry to confirm the identity of homozygous carriers. As a control we used genomic DNA extracted from an unrelated ABC fish line. PCR primers were chosen to detect the viral insertion in the dla coding sequence (exon2) or to detect wild-type dla coding sequence. We expected to see a 276 bp product for the dla locus carrying the viral insertion, amplified with the primer pair dla-1 and dla-2. The undisturbed wild type dla locus was amplified using primer pair dla-1 and dla-3, resulting in a 450 bp PCR product. Primers dla-1 and dla-3 bind to dla coding sequence flanking the viral insertion, producing a putative PCR product of more than 3 kb. However, with the PCR conditions used, we did not amplify a specific PCR product in embryos homozygous for the viral insertion. We used following oligonucleotides:
Embryos were scored and photographed with a Zeiss Axioplan microscope and photographed using a Nikon Coolpix 995 digital camera or a Zeiss Axiocam MRc5 live camera. Embryos were imaged using a Zeiss LSM5 confocal microscope or an inverted Nikon TU-2000 microscope with an EZ-C1 confocal system (Nikon). The following objectives were used for confocal imaging: 40X water immersion objective, 60X water immersion objective, 100X oil immersion objective.
The dlahi781Tg insertion line was generated by Nancy Hopkins’ lab (Amsterdam and Hopkins, 2004) and is available from the Zebrafish International Resource Center (ZIRC). The Tg(Bactin:HRAS-EGFP) (allele vu119) was generated by J. Topczewski and L. Solnica-Krezel (Cooper et al., 2005). We thank Mike Marusich and the University of Oregon Monoclonal Antibody Facility for help generating the Dla antibody, Wenbiao Chen, Julian Lewis and Karen Guillemin for reagents, Jonathan Leslie for technical advice, the University of Oregon Zebrafish Facility staff for animal care, the University of Oregon Histology Facility for preparing sections, and Jacob Lewis for help with fish husbandry. We also thank Philip Washbourne and Prisca Chapouton for helpful comments on the manuscript. Supported by NIH grant NS23915, HD22486 and AHA postdoctoral fellowship 0420027Z.
Grant sponsors: NIH NS23915 (JSE), NIH HD22486 (JSE) and AHA postdoctoral fellowship 0420027Z (ATallafuss).