The Gal4-UAS technique has been used to misexpress a constitutively active Notch receptor variant (notch1a-intra) in the developing zebrafish retina. This is the first study to use this technique to misexpress genes and assess their function in neural development of the zebrafish. Expression of activated Notch1a either ubiquitously, driven by a heat- shock70 promoter, or in a spatially regulated manner, controlled by the deltaD promoter, causes a block in neuronal differentiation that affects all cell types. Developing cells take on either a glial fate or remain undifferentiated. A large number of cells eventually undergo apoptosis. These phenotypic effects of activated Notch1a are expressed cell autonomously. Cells within central regions of the retina adopt a glial fate if they express activated Notch1a in a time window that extends from 27 to 48 hours postfertilization. This period corresponds mainly to the time of origin of ganglion cells in the normal retina. Activation of notch1a at later stages results in defects in cell type specification that remain restricted to the ciliary marginal zone, whereas neuronal types are specified normally within the central region. These observations indicate that glial differentiation is initiated by Notch1a-intra expressing cells, which become postmitotic in the same time window. Our results strongly suggest that Notch1a instructs a certain cell population to enter gliogenesis, and keeps the remaining cells in an undifferentiated state. Some or all of these cells will eventually succumb to apoptosis.

The cells of the vertebrate retina differentiate into seven different major classes, one of which is glial (the Müller cells), while the other six are neuronal: the rod and cone photoreceptors, the horizontal, bipolar, amacrine and ganglion cells (Cajal, 1893; Dowling and Boycott, 1966). These seven major classes, however, comprise a large variety of different cell types, which can be classified on the basis of morphological and functional criteria. The cells of the retina develop from multipotential progenitor cells, giving rise to progeny that can differentiate into members of all the seven different cell classes (Turner and Cepko, 1987; Holt et al., 1988). However, at least some progenitor cells are biased toward the production of particular cell types. Thus, the VC1.1+ progenitor cells produce large numbers of amacrine and horizontal cells and, as development progresses, an increasing number of rods, but no cone cells (Alexiades and Cepko, 1997). Even though progenitor cells are very heterogeneous in their differentiation potential, and the time of cell birth is apparently not a major determinant of cell fate (Belecky Adams et al., 1996), a progressive restriction occurs, as later progenitor cells produce fewer different cell types (Alexiades and Cepko, 1997). A temporal sequence in the appearance of the different cell types has been described for several vertebrate species. Within a given region of the retina, ganglion cells, horizontal cells and amacrine cells, as well as cones, are specified first, followed by bipolar cells, rods and Müller cells (see review by Cepko et al., 1996).

A similar sequence for the differentiation of the retinal cell types has been observed in the zebrafish. The first cells to become postmitotic are ganglion cells, which are located in a ventronasal patch at about 28 hours postfertilization (hpf) (Nawrocki, 1985; Hu and Easter, 1999). At about 37 hpf, all ganglion cells of the central region of the retina are postmitotic (Hu and Easter, 1999). Cell differentiation within the remaining layers of the retina progresses in the same direction (Raymond et al., 1995; Schmitt and Dowling, 1996; Passini et al., 1997; Hu and Easter, 1999). Thus, the first postmitotic cells of the internal nuclear layer appear at 38 hpf, and the first photoreceptor cells at about 48 hpf (Larison and BreMiller, 1985; Nawrocki, 1985; Raymond et al., 1995; Hu and Easter, 1999). At about 60 hpf, approximately 90% of the neurones in the central region of the retina have differentiated and the various layers are clearly distinguishable (Malicki, 1999). No data on the time of origin of the Müller cells are available for the zebrafish.

The molecular basis of cell type diversification within the developing retina has been the object of several studies in different animal species (reviewed by Cepko et al., 1996; Harris, 1997). It is currently assumed that the competence of retinal progenitors to respond to environmental cues is modified in a sequential fashion. Progenitors thus become progressively restricted in their developmental potential and give rise to differentially specified progeny in response to specific signals.

One of those signals is mediated by the Notch pathway, which plays a prominent role in cell fate specification in the retina. Previous studies (Dorsky et al., 1995; Dorsky et al., 1997; Bao and Cepko, 1997; Henrique et al., 1997) used lipofection or replication-defective retroviruses to misexpress specific variants of Notch and Delta homologues in the developing retina. The results of most of these studies suggest that Notch signalling maintains cells in an undifferentiated state, thus contributing to specify the different cell types by making cells susceptible to regulatory signals at specific stages (Dorsky et al., 1995; Dorsky et al., 1997; Henrique et al., 1997). In still other studies, the histogenetic effects of misexpressing HES1, one of the target genes of Notch signalling, have been analysed. In such animals, retinal differentiation is blocked; conversely a null mutation in HES1 leads to premature differentiation of neuronal types that differentiate at earlier stages (Ishibashi et al., 1994; Ishibashi et al., 1995; Tomita et al., 1996). Recently, a number of papers have been published which imply that Notch signalling promotes glia cell development in rodents (Furukawa et al., 2000; Morrison et al., 2000; Gaiano et al., 2000). Similar results have been obtained by misexpressing Hes5, another target of Notch signalling, in the mouse retina (Hojo et al., 2000). In one of these papers (Morrison et al., 2000), an instructive function of Notch signalling in promoting glial development is proposed on the basis of in vitro studies.

We have used the Gal4-UAS system for targeted gene expression (Fisher et al., 1988; Brand and Perrimon, 1993) to express a constitutively active Notch receptor (notch1a-intra; Takke et al., 1999) in the developing zebrafish retina. This technique was recently adapted for the zebrafish (Scheer and Campos-Ortega, 1999) and the mouse (Rowitsch et al., 1999), and here we present, for the first time, results from a functional analysis of a neural developmental process using the Gal4- UAS technique in zebrafish. Our results indicate that Notch signalling in the developing retina blocks neuronal differentiation, causing cells either to enter gliogenesis or remain undifferentiated and/or die, instead of adopting a neuronal fate. These observations strongly suggest that Notch signalling has an instructive effect on retina cells, promoting glial development.

A fish line carrying one of the UAS:myc-notch-intra insertions described in Scheer and Campos-Ortega (1999) was used as effector. For the construction of the activator plasmids, the HindIII fragment of Gal4 from pGATB (Brand and Perrimon, 1993) was fused to the SV40 small intron and polyadenylation site, and either a 6 kb fragment from the deltaD promoter (S. H. and J. A. C.-O., unpublished), or a 1.5 kb fragment from the hsp70 promoter (Halloran et al., 2000), was cloned upstream of the translational start site of Gal4. Microinjection of DNA was as described previously (Scheer and Campos-Ortega, 1999).

Hybridization of digoxigenin-labelled Gal4 RNA probes to embryo whole mounts was as described previously (Bierkamp and Campos- Ortega, 1993). For this purpose the HindIII Gal4 fragment from pGATB was cloned into the HindIII site in pBluescript (KS), and digoxigenin-labelled Gal4 RNA probes were prepared using an RNA labelling kit (Boehringer Mannheim). Embryos and larvae of different genotypes were embedded in Araldite, 4-5 µ m transverse sections were made with an ultramicrotome (LKB) and stained with Methylene Blue.

Immunohistochemistry

Antibody staining was as described by Westerfield (Westerfield, 1993) with some modifications. The lenses of embryos and larvae older than 48 hpf were removed with sharp needles and the animals incubated for 30 minutes with 0.01% collagenase/PBT before bleaching in 1% H2O2/PBT overnight. The EliteABC kit (Vectastain) was used to enhance the sensitivity of the staining. All incubation steps were performed at 4°C overnight in 5% DMSO/5% horse serum/PBT. For myc antibody staining, the embryos were treated with a hybridoma supernatant containing a monoclonal anti-myc antibody (1:2). The 16A11, zn-8, FRet43, FRet11 and zrf-1 antibodies (kindly given to us by the Monoclonal Antibody Facility, University of Oregon) were diluted 1:50 for incubation. The antibody against phosphorylated histone H3 was from Upstate Biotechnology. The secondary antibody was in all cases a HRP-linked goat anti-mouse antiserum (1:500; Dianova). The stained specimens were embedded in Araldite and 4-5 µ m transverse sections were made with an ultramicrotome (LKB).

Cell transplantations

Donor embryos, derived from crosses between either deltaD:Gal4/+ or hsp70:Gal4/+ and UAS:myc-notch1a-intra/+ fish, were injected at the 1- to 2-cell stage with a mixture of Texas Red and biotin dextran (1:9; total concentration of 5% w/v in 0.2 M KCl) and allowed to develop until the 1000-cell/30% epiboly stage. Groups of 20-30 cells were then removed from the donors and isochronically transplanted into wild-type hosts. The genotype of individual donor embryos was established by PCR after the transplantation, and the host embryos were investigated at 24 hpf for the presence of Texas Red-labelled cells in the retina, fixed at 3-6 dpf, treated with streptavidin-HRP and stained with DAB. The animals were embedded in Araldite and transverse thick sections were made by hand with a razor blade.

As in other fish and in amphibians (Harris, 1997), the central region of the zebrafish retina is clearly distinct from the peripheral, ciliary marginal zone, where the stem cells are located and from which the retina grows during the lifetime of the animal (Harris and Perron, 1998). Cells within these two zones behave differently in response to Notch signalling at a given time point. Unless otherwise indicated, our observations refer exclusively to the central region.

Throughout this study we used as the effector gene a stably inserted UAS:notch1a-intra construct (in the figures abbreviated UAS:nic) that had been tagged with a tandem array of six copies of the Myc epitope (described by Takke et al., 1999; Scheer and Campos-Ortega, 1999). Analogous versions of this construct were tested in several animal species and invariably found to encode constitutively active Notch receptors (Coffman et al., 1993; Lieber et al.,1993; Rebay et al., 1993; Struhl et al., 1993; Austin et al., 1995; Dorsky et al., 1995; Takke et al., 1999). Transgenic insertions of two different Gal4 activator constructs were used to activate expression of notch1a-intra. The first activator, called deltaD:Gal4, is based on a genomic 6 kb fragment from the promoter region of the zebrafish deltaD gene (Dornseifer et al., 1997; Haddon et al., 1998), which contains all regulatory elements of this gene (S. H. and J. A. C.-O., unpublished). Staining with an anti-myc antibody of the progeny of crosses of deltaD:Gal4 fish with fish carrying a UAS:myc-notch1a- intra insertion shows that the topological distribution of reporter gene expression is essentially identical to that of the transcripts of the endogenous gene (not shown). Most cells in the central regions of the developing retina at 27 hpf, and all cells in the same region at 40 and 72 hpf, express the myc epitope (Fig. 1A,B). Therefore, deltaD:Gal4 activated notch1a-intra is expressed in mitotic as well as in postmitotic cells.

Fig. 1.

(A,B) Anti-myc antibody stainings of deltaD:Gal4;UAS:myc- notch1a-intra (nic) embryos at 40 hpf (A) and 72 hpf (B). Notice that the number of cells expressing the myc epitope increases with age, although all central retina cells contain the epitope, and that the ciliary marginal zone is devoid of myc expression (arrows in A and B). (C) An anti-myc antibody staining of a hsp70:Gal4;UAS:myc- notch1a-intra animal, which had been heat-shocked at 24 hpf and fixed at 54 hpf. (D) A western blot with protein extract from 48 hpf embryos, which had been heat-shocked at 24 hpf and probed with anti-myc antibody. Lane 1, wild-type; lane 2, hsp70:Gal4;UAS:myc- notch1a-intra embryos. The right lane shows the SeeBlueTMPlus2 marker (Novex). Dorsal is to the left in all pictures.

Fig. 1.

(A,B) Anti-myc antibody stainings of deltaD:Gal4;UAS:myc- notch1a-intra (nic) embryos at 40 hpf (A) and 72 hpf (B). Notice that the number of cells expressing the myc epitope increases with age, although all central retina cells contain the epitope, and that the ciliary marginal zone is devoid of myc expression (arrows in A and B). (C) An anti-myc antibody staining of a hsp70:Gal4;UAS:myc- notch1a-intra animal, which had been heat-shocked at 24 hpf and fixed at 54 hpf. (D) A western blot with protein extract from 48 hpf embryos, which had been heat-shocked at 24 hpf and probed with anti-myc antibody. Lane 1, wild-type; lane 2, hsp70:Gal4;UAS:myc- notch1a-intra embryos. The right lane shows the SeeBlueTMPlus2 marker (Novex). Dorsal is to the left in all pictures.

The second activator construct, hsp70:Gal4, carries a zebrafish heat-shock70 promoter (Halloran et al., 2000), which activates transcription of UAS:myc-notch1a-intra ubiquitously immediately after heat shock (for 30 minutes at 40°C, not shown). The translation product can be detected with the anti- myc antibody as early as 1 hour after heat shock in all cells of the body (not shown). To assess the persistence of the Notch1a- intra protein, 27 hpf embryos were heat shocked for 30 minutes at 40°C and stained with the anti-myc antibody 24 and 48 hours later. Staining was very strong 24 hours after heat-shock in several organs, such as the retina (Fig. 1C), the neural tube, the notochord and, at much lower intensity, the somites (not shown). To test whether the epitope detected by the anti-myc antibody corresponds to the chimeric myc-notch1a-intra protein or to degradation products, a western blot was prepared with proteins from hsp70:Gal4;UAS:myc-notch1a-intra embryos 24 hours after a 30 minutes heat-shock (40°C) at 24 hpf, which was then probed with the anti-myc antibody. The result shows a band of about 91 kDa, the size expected for the chimeric protein (Fig. 1D). The retina still contains detectable amounts of the myc epitope 48 hours after heat shock. However, no attempt was made to detect the chimeric protein by western blotting in these animals.

Individual progeny of crosses between either deltaD:Gal4 or hsp70:Gal4 fish with UAS:myc-notch1a-intra fish were genotyped by PCR. Mutant progeny of the former cross are in fairly good general physiological conditions and live for a few days after hatching (until 6-7 dpf; days postfertilization), whereas those of the latter cross are more strongly affected and die earlier (at 4-5 dpf). Striking pattern defects were found in the retinas of deltaD:Gal4;UAS:myc-notch1a-intra individuals (Fig. 2B), and even stronger defects were seen in those of hsp70:Gal4;UAS:myc-notch1a-intra larvae following heat shock at 27 hpf (30 minutes at 40°C) and fixation at 3 dpf (Fig. 3B,C). Sections of eyes from these animals stained with methylene blue demonstrate that the normal layered organisation of the retina is hardly distinguishable. Neither the characteristic palisade-like arrangement of outer segments of the photoreceptor cells nor plexiform layers can be distinguished, although isolated islands of neuropile are visible among clusters of cell bodies.

Fig. 2.

Neuronal development is impaired following expression of activated notch1a. A and B show Methylene Blue-stained sections through the retina of 5 dpf wild-type (A)and deltaD:Gal4; UAS:myc-notch1a-intra (nic)(B)larvae. Note the perturbation of the normal layered organisation in the mutant. (C-J) Neuronal development is strongly suppressed in deltaD:Gal4; UAS:myc-notch1a-intra animals (D,F,H,J), as compared to wild-type(C,E,G,I). (C,D) Sections from retinas stained with the 16A11 antibody. (E,F) Whole-mounts stained with zn-8 (48 hpf embryos, white arrows point to the optic chiasm). (G,H) Sections stained with FRet43 (5 dpf), (I,J) Sections stained with FRet11 (5 dpf). Notice that the number of cells expressing the corresponding marker is strongly reduced in the mutants. These cells are situated in outer layers of the retina.

Fig. 2.

Neuronal development is impaired following expression of activated notch1a. A and B show Methylene Blue-stained sections through the retina of 5 dpf wild-type (A)and deltaD:Gal4; UAS:myc-notch1a-intra (nic)(B)larvae. Note the perturbation of the normal layered organisation in the mutant. (C-J) Neuronal development is strongly suppressed in deltaD:Gal4; UAS:myc-notch1a-intra animals (D,F,H,J), as compared to wild-type(C,E,G,I). (C,D) Sections from retinas stained with the 16A11 antibody. (E,F) Whole-mounts stained with zn-8 (48 hpf embryos, white arrows point to the optic chiasm). (G,H) Sections stained with FRet43 (5 dpf), (I,J) Sections stained with FRet11 (5 dpf). Notice that the number of cells expressing the corresponding marker is strongly reduced in the mutants. These cells are situated in outer layers of the retina.

Fig. 3.

Neuronal development is impaired following expression of activated Notch1a. Temporal evolution of retinal defects. (A- C) Methylene Blue-stained sections of wild-type (A) and hsp70:Gal4; UAS:myc-notch1a-intra (nic). (B,C) 3 dpf larvae that had been heat-shocked at 27 (B) and 48 hpf (C). Nuclei of differentiated ganglion cells (white arrows) and a rudimentary optic nerve (black arrow) can be seen in C. (D) wild- type and (E,F) hsp70:Gal4; UAS:myc-notch1a-intra 3 dpf larvae that had been heat-shocked at 27 hpf (E) and 48 hpf (F), and stained with the 16A11 and FRet43 antibodies. Whereas activation of notch1a-intra expression at 27 hpf completely blocks neuronal differentiation (E), patterning and immunoreactivity of neuronal markers is less affected after heat shocks at 48 hpf. The arrows in F point to cells expressing the neuronal markers 16A11 and FRet43. (G) Wild-type and hsp70:Gal4;UAS:myc-notch1a- intra 3 dpf animals that had been heat-shocked at 48 hpf, and stained with FRet11. Presumptive rods are organised in palisades (arrow in H), but do not express the FRet11 antigen. (I) A 6 dpf hsp70:Gal4;UAS:myc-notch1a-intra larva that had been heat-shocked at 3 dpf and stained with Methylene Blue. Defects are restricted to the ciliary marginal zone (arrow).

Fig. 3.

Neuronal development is impaired following expression of activated Notch1a. Temporal evolution of retinal defects. (A- C) Methylene Blue-stained sections of wild-type (A) and hsp70:Gal4; UAS:myc-notch1a-intra (nic). (B,C) 3 dpf larvae that had been heat-shocked at 27 (B) and 48 hpf (C). Nuclei of differentiated ganglion cells (white arrows) and a rudimentary optic nerve (black arrow) can be seen in C. (D) wild- type and (E,F) hsp70:Gal4; UAS:myc-notch1a-intra 3 dpf larvae that had been heat-shocked at 27 hpf (E) and 48 hpf (F), and stained with the 16A11 and FRet43 antibodies. Whereas activation of notch1a-intra expression at 27 hpf completely blocks neuronal differentiation (E), patterning and immunoreactivity of neuronal markers is less affected after heat shocks at 48 hpf. The arrows in F point to cells expressing the neuronal markers 16A11 and FRet43. (G) Wild-type and hsp70:Gal4;UAS:myc-notch1a- intra 3 dpf animals that had been heat-shocked at 48 hpf, and stained with FRet11. Presumptive rods are organised in palisades (arrow in H), but do not express the FRet11 antigen. (I) A 6 dpf hsp70:Gal4;UAS:myc-notch1a-intra larva that had been heat-shocked at 3 dpf and stained with Methylene Blue. Defects are restricted to the ciliary marginal zone (arrow).

The use of specific antibodies allows one to further dissect the patterning defects. The 16A11 antibody recognises an antigen expressed by amacrine and ganglion cells (Link et al., 2000), whereas FRet43 (Larison and BreMiller, 1990) and FRet11 (Schmitt and Dowling, 1996) stain double cones and rods, respectively. The number of cells expressing the corresponding antigens is strongly reduced in deltaD:Gal4;UAS:myc-notch1a-intra (Fig. 2C,D,G-J), and such cells are absent altogether in hsp70:Gal4;UAS:myc- notch1a-intra 3 dpf retinas that had been heat-shocked with 40°C for 30 minutes at 27 hpf (Fig. 3E and not shown). Similar observations were made using the zn-8 antibody, which stains the axons of the ganglion cells (Trevarrow et al., 1990). Although a rudimentary optic nerve is occasionally present, the number of axons in these nerves is always strongly reduced (Fig. 2E,F).

zrf-1 recognises radial glia cells (Trevarrow et al., 1990). In the wild-type retina, the earliest signs of immunoreactivity to zrf- 1 are found at approximately 5 dpf, in fibres that traverse the plexiform and nuclear layers (Fig. 4A-C), corresponding to the Müller cells. However, following deltaD:Gal4-mediated activation of notch1a-intra, zrf-1-positive cells can already be distinguished at 2 dpf, that is to say several days prior to the time at which cells expressing this marker appear in the wild type (Fig. 4D). The number of zrf-1-positive cells increases, and at 6 dpf apparently all cells of the inner layers of the retina express the antigen strongly and have differentiated fibres traversing the retina (Fig. 4E-F).

Fig. 4.

Excessive and premature differentiation of glial cells following expression of activated notch1a. (A-C) Wild-type and (D-F) deltaD:Gal4; UAS:myc-notch1a-intra (nic) animals at 2 dpf (A,D), 3 dpf (B,E) and 5-6 dpf (C,F), all stained with the zrf-1 antibody. Note that, in the mutants, reactivity to the antibody is already detectable in central regions of the retina at 2 dpf, and increases considerably until 6 dpf, whereas immunoreactivity first appears in the wild type around 4-5 dpf. (G-I) 3-dpf hsp70:Gal4; UAS:myc-notch1a-intra (nic) retinas, heat-shocked at 27 hpf (G), 35 hpf (H), and 48 hpf (I). Note that the most severe gliosis results from expression of activated Notch1a at earlier stages (G), that gliosis is less intense after activation of Notch1a at 35 hpf (H), and that pulses at 48 hpf bring about excessive glial development only outside the central retina (arrows in I). (J) A hsp70:Gal4; UAS:myc-notch1a- intra 48 hpf embryo that had been heat-shocked at 27 hpf. Note that the severity of gliosis is comparable to that in G, despite the age difference. (K,L) 3 dpf hsp70:Gal4; UAS:myc-notch1a-intra larvae after a single heat shock(K)at 30 hpf, or repeated heat shocks at 30, 48 and 56 hpf(L)to show that activated Notch1a is expressed at the same time as the glial phenotype. The severity of glial hyperplasia is comparable in both cases despite the use of repeated heat shocks in L.

Fig. 4.

Excessive and premature differentiation of glial cells following expression of activated notch1a. (A-C) Wild-type and (D-F) deltaD:Gal4; UAS:myc-notch1a-intra (nic) animals at 2 dpf (A,D), 3 dpf (B,E) and 5-6 dpf (C,F), all stained with the zrf-1 antibody. Note that, in the mutants, reactivity to the antibody is already detectable in central regions of the retina at 2 dpf, and increases considerably until 6 dpf, whereas immunoreactivity first appears in the wild type around 4-5 dpf. (G-I) 3-dpf hsp70:Gal4; UAS:myc-notch1a-intra (nic) retinas, heat-shocked at 27 hpf (G), 35 hpf (H), and 48 hpf (I). Note that the most severe gliosis results from expression of activated Notch1a at earlier stages (G), that gliosis is less intense after activation of Notch1a at 35 hpf (H), and that pulses at 48 hpf bring about excessive glial development only outside the central retina (arrows in I). (J) A hsp70:Gal4; UAS:myc-notch1a- intra 48 hpf embryo that had been heat-shocked at 27 hpf. Note that the severity of gliosis is comparable to that in G, despite the age difference. (K,L) 3 dpf hsp70:Gal4; UAS:myc-notch1a-intra larvae after a single heat shock(K)at 30 hpf, or repeated heat shocks at 30, 48 and 56 hpf(L)to show that activated Notch1a is expressed at the same time as the glial phenotype. The severity of glial hyperplasia is comparable in both cases despite the use of repeated heat shocks in L.

Heat-shock-mediated activation of notch1a-intra at early developmental stages, e.g. tailbud or early somite stages (10-12 hpf), results in anophthalmic animals (not shown). Heat- shock pulses at later stages allow eye development and lead to gliosis, the strength and topological location of which clearly depends on the stage at which notch1a-intra is expressed.

Following a 30 minute 40°C pulse at 27 hpf, the eyes develop but 16A11-, FRet11- and FRet43-positive cells are absent at 3 dpf (Fig. 3D-E), whereas the number of zrf-1-positive cells is high (Fig. 4G). It is remarkable that, although all cells of the retina expressed myc-notch1a-intra, as judged by the myc antibody, immunoreactivity to zrf-1 was present only within central regions of the retina, with the highest activity concentrated in the innermost layers. However, zrf-1-positive cells were less well developed, with shorter fibres, than following activation of notch1a-intra with deltaD:Gal4 (compare Fig. 4F with 4G). Heat-shock treatments later, at 35 hpf, resulted in fewer zrf-1-positive cells at 3 dpf, but their distribution was similar to that in the previous case (Fig. 4H). Patterning in the retina at 3 dpf is still strongly affected by heat shocks at 48 hpf (Fig. 3C), but ganglion cells, amacrine cells and double cones expressed their characteristic markers, although very weakly (Fig. 3F). Presumptive rods did not express the FRet11 antigen in these animals, but were arranged in characteristic palisades (Fig. 3G-H). The defects in gliogenesis following heat-shock at 48 hpf are virtually absent from central regions of these retinas, as judged from the number of zrf-1 cells (Fig. 4I). However, obvious defects are visible internal to the ciliary marginal zone, where newly originated and differentiating cells are located (Fig. 4I). Even later heat-shock pulses, at 3 dpf, result in pattern defects exclusively within the marginal region of the 6 dpf larva, whereas the central regions are apparently normal (Fig. 3I).

Therefore, premature and excessive glial development can be elicited in central regions of the retina only between 27 and 48 hpf, i.e. during the time at which mainly ganglion cells become postmitotic in untreated embryos (Hu and Easter, 1999). Activation of Notch1a at 48 hpf leads to pattern defects without accompanying gliosis. Activation at still later stages does not affect the central retina.

Simultaneous expression of activated Notch1a and glial markers

The premature expression of glial markers suggests that gliogenesis is initiated shortly after activation of notch1a-intra transcription by Gal4. However, it is also conceivable that Notch1a-intra exerts a permissive function in this process and that specification of a given cell as a Müller glia cell occurs only when Notch1a-intra is no longer present. This does not seem to be the case, as hsp70:Gal4; UAS:myc- notch1a-intra embryos (heat-shocked at 27 hpf and fixed 24 hours later) express the myc and zrf-1 epitopes within central regions of their retinas (Figs 1C, 5D). For technical reasons (both antibodies had been raised in mouse) double staining could not be done on the same specimens. However, it seems highly probable that both markers are simultaneously expressed by the same cells, as the anti-myc antibody stains all cells within the central region of the developing retina, thus including the zrf-1-positive cells, and the Myc-Notch1a-intra protein persists for at least 24 hours (see above). Nevertheless,to ensure that the cells do indeed express Notch1a-intra and glial epitopes at the same time, repeated heat-shocks were applied. hsp70:Gal4;UAS:myc-notch1a-intra embryos were heat-shocked at 30, 48 and 56 hpf, and fixed and processed with the zrf-1 antibody at 3 dpf (Fig. 4L). The glial phenotype of these retinas is as severe as that of embryos of the same age that had received a single heat shock at 30 hpf (compare Fig. 4K with 4L). All these data strongly suggest that the presence of Notch1a-intra is causally related to the deficit in neuronal differentiation and the gliosis.

notch1a-intra expression leads to gliogenesis cell autonomously

Cell transplantations following the technique described by Link et al. (Link et al., 2000) were used to test whether the gliogenesis defects are expressed cell autonomously. Following isochronic transplantation of either deltaD:Gal4;UAS:myc-notch1a-intra or hsp70:Gal4;UAS:myc-notch1a-intra cells into wild-type hosts, the mosaic distribution of labelled cells in the retina of the wild-type hosts was similar in all cases, irrespective of the genotype of the donors (Fig. 5). That is to say, patches of labelled cells alternated with unlabelled cells; in some cases, single donor cells were completely surrounded by host cells. However, the phenotype of the transplanted cells clearly depended on the genotype of the donor. Cells from either wild- type donors or from those carrying only one of the two transgenes differentiated into any of the different cell types of the retina, as judged from the morphology of the labelled cells (Fig. 5A). Thus, labelled cells when surrounded by unlabelled cells were distributed in columns extending from the photoreceptor to the ganglion cell layer, with perikarya located in all layers. In contrast to this, cells transplanted from embryos carrying both transgenes either adopted a glial phenotype or remained undifferentiated, that is, with a round appearance and devoid of cell processes (Fig. 5B-D). When differentiated, the labelled cells were also grouped in columns, but the processes never extended beyond the level of the inner segments of the photoreceptors. Within small cell groups, the perikarya of labelled cells were located exclusively in the outer nuclear layer; in large, compact groups of deltaD:Gal4; UAS:myc- notch1a-intra cells, perikarya were occasionally found in other layers of the retina. When progeny of transplanted cells grew as isolated single cells or patches, their glial morphology could be observed unambiguously (Fig. 5C). In addition, biotin could also be detected in cells without processes and in dead or dying cells distributed throughout the retina (not shown).

Fig. 5.

Glial development is induced cell autonomously.(A) An example of transplantation of wild-type cells into a wild-type host. Note that labelled perykarya are distributed throughout all layers of the retina, and that the cells can differentiate into any of the normal cell types of the retina. (B,C) deltaD:Gal4; UAS:myc-notch1a-intra (nic) and hsp70:Gal4; UAS:myc-notch1a-intra cells, which had grown in wild-type hosts. Note that all labelled cells differentiate characteristics of Müller cells, i.e. perykarya within the outer nuclear layer (asterisks) and extending from the basal membrane but not extending beyond the inner segments of the photoreceptors (arrows in B and D). These characteristics can be clearly seen in the two cells in D, which grew isolated from other mutant cells.

Fig. 5.

Glial development is induced cell autonomously.(A) An example of transplantation of wild-type cells into a wild-type host. Note that labelled perykarya are distributed throughout all layers of the retina, and that the cells can differentiate into any of the normal cell types of the retina. (B,C) deltaD:Gal4; UAS:myc-notch1a-intra (nic) and hsp70:Gal4; UAS:myc-notch1a-intra cells, which had grown in wild-type hosts. Note that all labelled cells differentiate characteristics of Müller cells, i.e. perykarya within the outer nuclear layer (asterisks) and extending from the basal membrane but not extending beyond the inner segments of the photoreceptors (arrows in B and D). These characteristics can be clearly seen in the two cells in D, which grew isolated from other mutant cells.

Activation of notch1a-intra strongly reduces the number of mitoses

To test the hypothesis that Notch signalling affects the proliferation of the cells in the developing retina (Ishibashi et al., 1995; Henrique et al, 1997; Furukawa et al., 2000), we estimated the number of mitoses in 2- and 3-dpf embryos, which were either wild type, deltaD:Gal4; UAS:myc-notch1a- intra or hsp70:Gal4; UAS:myc-notch1a-intra, that had been heat-shocked for 30 minutes at 30 hpf. Following staining with the mitosis marker phosphohistone H3 the numbers of labelled cells observed were consistently lower in the mutants than in the wild-type, and the effect was severe after activation of notch1a-intra expression with hsp70:Gal4 (Table 1 and Fig. 6A-C). Mitotic activity was apparently normal within marginal zones, but strongly reduced within central regions.

Table 1.

Effect of activated notch1a expression on the number of mitoses

Effect of activated notch1a expression on the number of mitoses
Effect of activated notch1a expression on the number of mitoses
Fig. 6.

Expression of activated Notch1a reduces the mitotic activity in the retina. (A-C) H3 antibody staining. A 47 hpf wild-type, and B,C are hsp70:Gal4; UAS:myc-notch1a-intra (nic), at 47 hpf and 3 dpf, respectively, all three heat-shocked at 30 hpf. Mitotic nuclei are located in external layers and are much more abundant in the wild-type than in the mutant retinas (see Table 1).

Fig. 6.

Expression of activated Notch1a reduces the mitotic activity in the retina. (A-C) H3 antibody staining. A 47 hpf wild-type, and B,C are hsp70:Gal4; UAS:myc-notch1a-intra (nic), at 47 hpf and 3 dpf, respectively, all three heat-shocked at 30 hpf. Mitotic nuclei are located in external layers and are much more abundant in the wild-type than in the mutant retinas (see Table 1).

Activated Notch 1a causes retinal cells to enter apoptosis

As mentioned above, a relatively large number of retina cells of notch1a-intra-expressing embryos and larvae do not develop either a glial or a neuronal phenotype. This is particularly striking when notch1a-intra is activated by early heat-shocks, as neuronal markers are completely absent. We assume that these cells remain undifferentiated as they do not exhibit any process nor do they express neural markers (neuronal or glial). To assess the role that cell death may play in the fate of these non-glial and non-neuronal cells, we performed Tunel staining in wild-type and transgenic fish (hsp70:Gal4;UAS-myc- notch1a-intra and deltaD:Gal4;UAS-myc-notch1a-intra) of different ages, following the protocols described by Abdelilah et al. (Abdelilah et al., 1996) and Daly and Sandell (Daly and Sandell, 2000).

In the wild type, a few apoptotic cells are present in the retina from about 24 hpf on; the frequency remains similar at 48 and 60 hpf, 3 dpf and 3.5 dpf, and has decreased at 5 dpf (Fig. 7A,C). The frequency of cell death in the two mutant genotypes is several times higher than in the wild-type at all stages (Fig. 7B,D). The frequency is highest in 48 hpf embryos of both genotypes, at least relatively, as the total number of cells increases in the older retinas. Unfortunately, individual variation makes it very difficult to give precise absolute values for the frequency of cell death in the mutants.

Fig. 7.

The frequency of apoptosis is considerably increased following expression of activated Notch1a. Tunel staining of wild- type (A,C) and hsp70:Gal4; UAS:myc-notch1a-intra (B,D) animals at 2 dpf (A,B) and 5 dpf (C,D). Dead cells are blue (arrows in A).

Fig. 7.

The frequency of apoptosis is considerably increased following expression of activated Notch1a. Tunel staining of wild- type (A,C) and hsp70:Gal4; UAS:myc-notch1a-intra (B,D) animals at 2 dpf (A,B) and 5 dpf (C,D). Dead cells are blue (arrows in A).

Both types of activator transgenes used here, deltaD:Gal4 and hsp70:Gal4, can drive strong Gal4 expression in the retina, and both cause similar defects by activating notch1a-intra, i.e. a strong reduction in numbers, or even complete disappearance, of cells expressing neuronal markers, and strong glial hyperplasia, which is initiated at very early stages and eventually leads to severe morphogenetic defects. hsp70:Gal4 activates ubiquitous expression. However, as expression is inducible by temperature, it can be restricted to a given time point. deltaD-Gal4 activates notch1a-intra in cells that normally express deltaD (Dornseifer et al., 1997; Haddon et al., 1998). It seems likely that all cells of the central regions of the retina express deltaD during the critical developmental stages that we are considering, i.e. between 24 and 72 hpf, as in the deltaD:Gal4;UAS:myc-notch1a-intra retinas all central cells contain detectable amounts of the myc epitope. Accordingly, Notch1a-intra protein is present in mitotic as well as in postmitotic cells irrespective of whether it was activated by deltaD:Gal4 or hsp70:Gal4.

Activation of Notch1a blocks neuronal development Previous

work indicated that expression of activated Notch variants blocks neuronal development in various parts of the developing CNS (Coffman et al., 1993; Henrique et al., 1995; Chitnis et al., 1995; Dorsky et al., 1995; Tomita et al., 1996; Henrique et al., 1997; Appel and Eisen, 1998; Takke et al., 1999; Furukawa et al., 2000; Morrison et al., 2000; Gaiano et al., 2000). We confirm and extend these observations. Expression of notch1a-intra at the critical stage of 27-48 hpf leads in 3 dpf larvae to a strong reduction, in the case of activation by deltaD:Gal4, or to complete absence, when activated by hsp70:Gal4, of cells expressing markers of ganglion cells, amacrine cells, rods and cones. The difference in intensity of the effects caused by deltaD:Gal4 and hsp70:Gal4 is most probably related to the amount of Gal4 protein expressed from each activator and, consequently, the amount of Notch1a-intra protein expressed from the effector. This is supported by the observation that different insertions of the same activator induce different amounts of Gal4 mRNA, and that the phenotypic defects in crosses with effector fish do indeed correlate with the amount of Gal4 mRNA (not shown). The temporal evolution of the effects indicates that neuronal development is blocked very early. As assessed in the retina of the 3 dpf larva, a single 30-minute heat shock at 27 hpf abolishes all neuronal development. Following a single heat- shock at 48 hpf there is a relatively large number of 16A11- and FRet43-positive cells, although these cells express the antigens weakly, and general patterning is less affected. If the heat shock is given even later, at 3 dpf, the defects are restricted to the peripheral regions of the retina, and layering and differentiation of the retinal cell types are essentially normal. That is to say, specification of cells that had become postmitotic prior to notch1a-intra expression is not perturbed. However, all developing neuronal types that we have been able to assess are sensitive to activated notch1a.

Notch signalling has an instructive role in retina development

The developmental fate of the cells that express activated Notch and do not adopt a neuronal fate is a controversial question. Following studies in zebrafish, Xenopus, chicken and mouse, some authors proposed that cells expressing activated Notch remain in an undifferentiated state (Coffman et al., 1990; Henrique et al., 1995; Chitnis et al., 1995; Dorsky et al., 1995; Tomita et al., 1996; Henrique et al., 1997; Appel et al., 1998; see a discussion in Harris, 1997). Other authors, however, reported that activation of Notch receptors within the developing CNS of mammals leads to glial development (Furukawa et al., 2000; Morrison et al., 2000; Gaiano et al., 2000).

We find that Gal4-mediated activation of notch1a-intra causes a large proportion of retinal cells to enter glial development in the zebrafish; this is a cell-autonomous effect, as a glial phenotype develops even when individual notch1a- intra-expressing cells are surrounded by wild-type cells. However, the remaining retina cells seem indeed to remain undifferentiated, as they do not express either glial or neuronal phenotypes. Repeated heat shocks and the relatively long persistence of the notch1a-intra protein ensure that activated Notch1a is still present in the cells developing a glial phenotype (see also Furukawa et al., 2000; Gaiano et al., 2000). Therefore, our results suggest that activated Notch1a has at least two different functions in the developing retina of the zebrafish. The most obvious function is the suppression of neuronal development associated with a diversion of a large excess of cells into glial development. Glial development is initiated as early as 48 hpf, as shown when notch1a-intra is activated by deltaD:Gal4. The temporal evolution of the gliogenic effect suggests that retinal cells of the central region are competent to react to Notch signalling by developing as glial cells only between 27 and 48 hpf, whereas activation at later stages allows cell-type specification in this region. As this time window corresponds chiefly to that of the origin of ganglion cells (Hu and Easter, 1999), we propose that these cells are the main contributors to the glial hyperplasia. The number of glial cells is higher after activation of Notch1a-intra by deltaD-Gal4 than by heat-shock. This difference suggests that the gliogenic effects may cumulate over a period of time longer than that uncovered by the heat-shocks and corresponding to the origin of ganglion cells.

The second apparent function of activated Notch1a is to keep other retinal cells undifferentiated, at least temporarily. However, activated Notch1a also causes many cells to enter apoptosis, beginning around 48 hpf and continuing for as long as the animal survives. It also remains unclear whether apoptosis is indirect consequence of the lack of neuronal development, that is, of inductive influences normally derived from the neurones, or rather elicited by Notch1a-intra directly. The possibility that the apparently undifferentiated cells correspond to those undergoing apoptosis is an attractive hypothesis. However, as we have been unable to quantify the undifferentiated and apoptotic cells, we cannot exclude the possibiliy that they represent distinct populations of cells.

Relationship between Notch activation and the cell cycle

A relationship between the Delta/Notch signalling pathway and the cell cycle has been proposed by several authors (see Ohnuma et al., 1999). However, the basis for this interrelationship is largely unknown. The observations described here show a high degree of similarity to the effects of misexpressing p27Xic1, a member of the Cip/Kip family of Cdk (cyclin-dependent kinase) inhibitors, in Xenopus (Ohnuma et al., 1999). Misexpression of p27Xic1 leads to premature differentiation of retinal glia cells and blocks the differentiation of at least one neuronal cell type, the bipolar cells. As a Cdk inhibitor, p27Xic1 causes the cell cycle to arrest, thus impeding the division of the retinal cells. However, in order for the cells to enter gliogenesis, interruption of the cell cycle as such is not sufficient; expression of p27Xic1 itself is required. This gliogenic function of the Cdk inhibitor is enhanced by the simultaneous expression of activated Notch. Thus, Ohnuma et al. speculate that the function of p27Xic1 in promoting glial development in the retina is intimately related to the Delta/Notch signal pathway (Ohnuma et al., 1999). Thus, in an indirect manner, our results support this possibility.

Our results do not support the contention, as proposed by others in mouse and chicken (Ishibashi et al., 1995; Henrique et al., 1997; Furukawa et al., 2000), that activation of Notch leads to increased cell proliferation. On the contrary, the present results indicate, first, that mitotic activity is strongly reduced, indeed almost totally suppressed, as shown by the immunoreactivity to the H3 antibody, in notch1a-intra- expressing retinas. The reduction in the number of mitoses may be either a direct consequence of activated notch1a, or reflect the increased apoptotic activity – or both.

On the function of Delta/Notch signalling and cellular diversity in the developing retina

The results reported above indicate that, as a consequence of Notch1a activation, retinal cells either acquire a glial fate or remain undifferentiated and/or enter apoptosis. The premature expression of glial markers by cells that simultaneously express activated Notch1a strongly suggests an instructive rather than a permissive role for Notch signalling. In this context it is noteworthy that in Xenopus and rats only differentiating Müller glia cells continue to express the Notch receptor, whereas other differentiating cells of the retina apparently do not express Notch (Ohnuma et al., 1999; Furukawa et al., 2000). Under the conditions of our experiment, many other cells, besides the normally developing Müller cells, would continue to express activated Notch1a and thus be forced to initiate glial development. An instructive function for Notch signalling in promoting glial development has recently been proposed on the basis of in vitro studies using mammalian cells (Morrison et al., 2000). Our results represent the first in vivo evidence that Notch signalling may perform such an instructive function in retinal development in the zebrafish.

We are grateful to John Kuwada for the gift of the hsp70 promoter prior to publication, to Paul Hardy for comments on the manuscript, Iris Riedl and Christel Schenkel for expert technical assistance, and our colleagues for discussions. This work was supported by grants of the Deutsche Forschungsgemeinschaft (DFG, SFB 243).

Abdelilah
,
S.
,
Mountcastle-Shah
,
E.
,
Harvey
,
M.
,
Solnica-Krezel
,
L.
,
Schier
,
A. F.
,
Stemple
,
D. L.
,
Malicki
,
J.
,
Neuhauss
,
S. C.
,
Zwartkruis
,
F.
,
Stainier
,
D. Y.
,
Rangini
,
Z.
and
Driever
,
W.
(
1996
).
Mutations affecting neural survival in the zebrafish Danio rerio
.
Development
123
,
217
227
.
Alexiades
,
M. R.
and
Cepko
,
C. L.
(
1997
).
Subsets of retinal progenitors display temporally regulated and distinct biases in the fates of their progeny
.
Development
124
,
1119
1131
.
Appel
,
B.
and
Eisen
,
J. S.
(
1998
).
Regulation of neuronal specification in the zebrafish spinal cord by Delta function
.
Development
125
,
371
380
.
Austin
,
C. P.
,
Feldman
,
D. E.
,
Ida
,
J. J.
and
Cepko
,
C. L.
(
1995
).
Vertebrate retinal ganglion cells are selected from competent progenitors by the action of Notch
.
Development
121
,
3637
3650
.
Bao
,
Z.-Z.
and
Cepko
,
C. L.
(
1997
).
The expression and function of Notch pathway genes in the developing rat eye
.
J. Neurosci
.
17
,
1425
1434
.
Belecky Adams
T.
,
Cook
,
B.
and
Adler
,
R.
(
1996
).
Correlations between terminal mitosis and differentiated fate of retinal precursor cells in vivo and in vitro: analysis with the “window labeling” technique
.
Dev. Biol
.
178
,
304
315
.
Bierkamp
,
C.
and
Campos-Ortega
,
J. A.
(
1993
).
A zebrafisch homologue of the Drosophila neurogenic gene Notch and its pattern of transcription during early embryogenesis
.
Mech. Dev
.
43
,
87
100
.
Brand
,
A.
and
Perrimon
,
N.
(
1993
).
Targeted gene expression as a means of altering cell fates and generating dominant phenotypes
.
Development
118
,
401
415
.
Cajal
,
S. R.
(
1893
).
La retina des vertebres
.
La Cellule
9
,
17
257
.
Cepko
,
C. L.
,
Austin
,
C. P.
,
Yang
,
X.
,
Alexiades
,
M.
and
Ezzeddine
,
D.
(
1996
).
Cell fate determination in the vertebrate retina
.
Proc. Natl. Acad. Sci. USA
93
,
589
595
.
Chitnis
,
A.
,
Henrique
,
D.
,
Lewis
,
J.
,
Ish-Horowicz
,
D.
and
Kintner
,
C.
(
1995
).
Primary neurogenesis in Xenopus embryos regulated by a homologue of the Drosophila neurogenic gene Delta
.
Nature
375
,
761
766
.
Coffman
,
C.
,
Harris
,
W.
and
Kintner
,
C.
(
1990
).
tXotch, the Xenopus homolog of Drosophila notch
.
Science
249
,
1438
1441
.
Coffman
,
C. R.
,
Skoglund
,
P.
,
Harris
,
W. A.
and
Kintner
,
C. R.
(
1993
).
Expression of an extracellular deletion of Xotch diverts cell fate in Xenopus embryos
.
Cell
73
,
659
671
.
Daly
,
F. J.
and
Sandell
,
J. H.
(
2000
).
Inherited retinal degeneration and apoptosis in mutant zebrafish
.
Anat. Rec
.
258
,
145
155
.
Doetsch
,
F.
,
Caille
,
I.
,
Lim
,
D. A.
,
Garcia-Verdugo
,
J. M.
and
Alvarez-Buylla
,
A.
(
1999
).
Subventricular zone astrocytes are neural stem cells in the adult mammalian brain
.
Cell
97
,
703
716
.
Dornseifer
,
P.
,
Takke
,
C.
and
Campos-Ortega
,
J. A.
(
1997
).
Overexpression of a zebrafish homologue of the Drosophila neurogenic gene Delta perturbs differentiation of primary neurons and somite development
.
Mech. Dev
.
63
,
159
171
.
Dorsky
,
R. I.
,
Chang
,
W. S.
,
Rapaport
,
D. H.
and
Harris
,
W. A.
(
1997
).
Regulation of neuronal diversity in the Xenopus retina by Delta signalling
.
Nature
385
,
67
70
.
Dorsky
,
R. I.
,
Rapaport
,
D. H.
and
Harris
,
W. A.
(
1995
).
Xotch inhibits cell differentiation in the Xenopus retina
.
Neuron
14
,
487
496
.
Dowling
,
J. E.
and
Boycott
,
B. B.
(
1966
).
The organization of the primate retina: electron microscopy
.
Proc. R. Soc. Lond. B Biol. Sci
.
166
,
80
111
Fischer
,
J.
,
Giniger
,
E.
,
Maniatis
,
T.
and
Ptashne
,
M.
(
1988
).
GAL4 activates transcription in Drosophila
.
Nature
332
,
853
856
.
Furukawa
,
T.
,
Mukherjee
,
S.
,
Bao
,
Z.-Z.
,
Morrow
,
E. M.
and
Cepko
,
C. L.
(
2000
).
rax, Hes1, and notch1 promote the formation of Müller glia by postnatal retinal progenitor cells
.
Neuron
26
,
383
394
.
Gaiano
,
N.
,
Nye
,
J. S.
and
Fishell
,
G.
(
2000
).
Radial glial identity is promoted by Notch1 signaling in the murine forebrain
.
Neuron
26
,
395
404
.
Haddon
,
C.
,
Smithers
,
L.
,
Schneider-Maunoury
,
S.
,
Coche
,
T.
,
Henrique
,
D.
and
Lewis
,
J.
(
1998
).
Multiple delta genes and lateral inhibition in zebrafish primary neurogenesis
.
Development
125
,
359
370
.
Halloran
,
M. C.
,
Sato-Maeda
,
M.
,
Warren
,
J. T.
,
Su
,
F.
,
Lele
,
Z.
,
Krone
,
P. H.
,
Kuwada
,
J. Y.
and
Shoji
,
W.
(
2000
).
Laser-induced gene expression in specific cells of transgenic zebrafish
.
Development
127
,
1953
1960
.
Harris
,
W. A.
(
1997
)
Cellular diversification in the vertebrate retina
.
Curr. Opin. Genet. Dev
.
7
,
651
658
.
Harris
,
W. A.
and
Perron
,
M.
(
1998
)
Molecular recapitulation: the growth of the vertebrate retina
.
Int. J. Dev. Biol
.
42
,
299
304
.
Henrique
,
D.
,
Adam
,
J.
,
Myat
,
A.
,
Chitnis
,
A.
,
Lewis
,
J.
and
Ish-Horowicz
,
D.
(
1995
).
Expression of a Delta homologue in prospective neurons in the chick [see comments]
.
Nature
375
,
787
790
.
Henrique
,
D.
,
Hirsinger
,
E.
,
Adam
,
U.
,
Le Roux
,
I.
,
Pourquié
,
O.
,
Ish-Horowicz
,
D.
and
J.
Lewis
(
1997
).
Maintenance of neuroepithelial progenitor cells by Delta-Notch signaling in the embryonic chick retina
.
Curr. Biol
.
7
,
661
670
.
Hojo
,
M.
,
Ohtsuka
,
T.
,
Hashimoto
,
N.
,
Gradwohl
,
G.
,
Guillemot
,
F.
and
Kageyama
,
R.
(
2000
).
Glial cell fate specification modulated by the bHLH gene Hes5 in mouse retina
.
Development
127
,
2515
2522
.
Holt
,
C.
,
Bertsch
,
T.
,
Ellis
,
H.
and
Harris
,
W. A.
(
1988
).
Cellular determination in the Xenopus retina is independent of lineage and birth date
.
Neuron
1
,
15
26
.
Hu
,
M.
and
Easter
,
S. S.
(
1999
).
Retinal neurogenesis: the formation of the initial central patch of postmitotic cells
.
Dev. Biol
.
207
,
309
321
.
Ishibashi
,
M.
,
Moriyoshi
,
K.
,
Sasai
,
Y.
,
Shiota
,
K.
,
Nakanishi
,
S.
and
Kageyama
,
R.
(
1994
).
Persistent expression of helix-loop-helix factor HES-1 prevents mammalian neural differentiation in the central nervous system
.
EMBO J
.
13
,
1799
1805
.
Ishibashi
,
M.
,
Ang
,
S. L.
,
Shiota
,
K.
,
Nakanishi
,
S.
,
Kageyama
,
R.
and
Guillemot
,
F.
(
1995
).
Targeted disruption of mammalian hairy and Enhancer of split homolog-1 (HES-1) leads to up-regulation of neural helix-loop-helix factors, premature neurogenesis, and severe neural tube defects
.
Genes Dev
.
9
,
3136
3148
.
Larison
,
K. D.
and
BreMiller
,
R.
(
1990
).
Early onset of phenotype and cell patterning in the embryonic zebrafish retina
.
Development
109
,
567
576
.
Lieber
,
T.
,
Kidd
,
S.
,
Alcamo
,
E.
,
Corbin
V.
and
Young
,
M. W.
(
1993
)
Antineurogenic phenotypes induced by truncated Notch proteins indicate a role in signal transduction and may point to a novel function for Notch in the nuclei
.
Genes Dev
.
7
,
1949
1965
.
Link
,
B. A.
,
Fadool
,
J. M.
,
Malicki
,
J.
and
Dowling
,
J. E.
(
2000
).
The zebrafish young mutation acts non-autonomously to uncouple differentiation from specification for all retinal cells
.
Development
127
,
2177
2188
.
Kimmel
,
C. B.
,
Ballard
,
W. W.
,
Kimmel
,
S. R.
,
Ullmann
,
B.
and
Schilling
,
T. F.
(
1995
)
Stages of embryonic development of the zebrafish
.
Dev. Dyn
.
203
,
253
310
.
Malicki
,
J.
(
1999
).
Development of the retina
.
Methods Cell Biol
.
59
,
273
299
.
Morrison
,
S. J.
,
Perez
,
S. E.
,
Qiao
,
Z.
,
Verdi
,
J. M.
,
Hicks
,
C.
,
Weinmaster
,
G.
and
Anderson
,
D. J.
(
2000
).
Transient Notch activation initiates an irreversible switch from neurogenesis to gliogenesis by neural crest stem cells
.
Cell
101
,
499
510
.
Nawrocki
,
W.
(
1985
).
Development of the neural retina in zebrafish, Brachydanio rerio
.
University of Oregon
,
Eugene, OR, USA
Ohnuma
,
S.
,
Philpott
,
A.
,
Wang
,
K.
,
Holt
,
C. E.
and
Harris
,
W. A.
(
1999
).
P27Xic, a Cdk inhibitor promotes the determination of glial cells in the Xenopus retina
.
Cell
99
,
499
510
.
Passini
,
M. A.
,
Levine
,
E. M.
,
Canger
,
A. K.
,
Raymond
,
P. A.
and
Schechter
,
N.
(
1997
).
Vsx-1 and Vsx-2: differential expression of two Paired-like homeobox genes during zebrafish and goldfish retinogenesis
.
J. Comp. Neurol
.
388
,
495
505
.
Raymond
,
P.
,
Barthel
,
L.
and
Curran
,
G.
(
1995
).
Developmental patterning of rod and cone photoreceptors in embryonic zebrafish
.
J. Comp. Neurol
.
359
,
537
550
.
Rebay
,
I.
,
Fehon
,
R. G.
and
Artavanis-Tsakonas
,
S.
(
1993
),
Specific truncations of Drosophila Notch define dominant activated and dominant negative forms of the receptor
.
Cell
74
,
319
329
.
Rowitsch
,
D. H.
,
St-Jacques
,
B.
,
Lee
,
S. M. K.
,
Flax
,
J. D.
,
Snyder
,
E. Y.
and
McMahon
,
A. P.
(
1999
).
Sonic hedgehog regulates proliferation and inhibits differentiation of CNS precursor cells
.
J. Neurosci
.
19
,
8954
8965
.
Scheer
,
N.
and
Campos-Ortega
,
J. A.
(
1999
).
Use of the GAL4-UAS technique for targeted gene expression in the zebrafish
.
Mech. Dev
.
80
,
153
158
.
Schmitt
,
E. A.
and
Dowling
,
J. E.
(
1996
).
Comparison of topographical patterns of ganglion and photoreceptor cell differentiation in the retina of the zebrafish, Danio rerio
.
J. Comp. Neurol
.
371
,
222
234
.
Struhl
,
G.
,
Fitzgerald
,
K.
and
Greenwald
,
I.
(
1993
)
Intrinsic activity of the Lin-12 and Notch intracellular domains in vivo
.
Cell
74
,
331
345
.
Takke
,
C.
,
Dornseifer
,
P.
, von
Weizäcker
,
E.
and
Campos-Ortega
,
J. A.
(
1999
).
her4, a zebrafish homologue of the Drosophila neurogenic gene E(spl), is a target of NOTCH signalling
.
Development
126
,
1811
1821
.
Tomita
,
K.
,
Ishibashi
,
M.
,
Nakahara
,
K.
,
S.
L.
,
Nakanishi
,
S.
,
Guillemot
,
F.
, and
Kageyama
,
R.
(
1996
).
Mammalian hairy and Enhancer of split homolg 1 regulates differentiation of retinal neurons and is essential for eye morphogenesis
.
Neuron
16
,
723
734
.
Trevarrow
,
B.
,
Marks
,
D. L.
and
Kimmel
,
C. B.
(
1990
).
Organization of hindbrain segments in the zebrafish embryo
.
Neuron
4
,
669
679
.
Turner
,
D. L.
and
Cepko
,
C. L.
(
1987
).
A common progenitor for neurons and glia persists in rat retina late in development
.
Nature
328
,
131
136
.
Westerfield
,
M.
(
1993
). A guide for laboratory use of zebrafish (Brachydanio rerio).
The Zebrafisch Book
.
Oregon University Press
,
Oregon
.