Gbx2 functions as a transcriptional repressor to regulate the specification and morphogenesis of the mid-hindbrain junction in a dosage- and stage-dependent manner.

The Gbx subfamily of homeodomain transcription factors is involved in the positioning of the isthmus, which patterns the midbrain and cerebellum in vertebrates. To uncover the details of Gbx functions, we first examined the dose dependency of its effects on brain formation in zebrafish and found that high-dose gbx2 mRNA injection affected the entire forebrain and midbrain, whereas low-dose mRNA specifically disrupted the isthmic folding at the midbrain-hindbrain boundary (MHB) but only weakly affected the expression of genes involved in MHB specification. Thus, isthmus morphogenesis, and not its early specification, is highly sensitive to gbx2. Transient induction of heat-inducible gbx2 using transgenic fish showed that MHB specification is most sensitive to gbx2 at the end of epiboly and further suggested that otx2 is the direct target gene. These together demonstrate that gbx2 regulates both specification and morphogenesis of the MHB/isthmus region. Deletion analyses showed that both the N- and C-terminal regions contribute to the suppressive activity of Gbx2 against the anterior brain and that the N-terminal core region, including the Eh1 and proline-rich sequences, is required for this Gbx2 activity. Comparison of the effects of activated and repressive forms with wild-type Gbx2 suggested that Gbx2 functions as a transcriptional repressor, which was further evidenced by a luciferase assay in which gbx2 repressed the MHB enhancer of fgf8a in mouse P19 cells.

Transgenesis in zebrafish with the tol2 transposon system.

The zebrafish (Danio rerio) is a useful model for genetic studies of vertebrate development. Its embryos are transparent and develop rapidly outside the mother, making it feasible to visualize and manipulate specific cell types in the living animal. Zebrafish is well suited for transgenic manipulation since it is relatively easy to collect large numbers of embryos from adult fish. Several approaches have been developed for introducing transgenes into the zebrafish germline, from the injection of naked DNA to transposon-mediated integration. In particular, the Tol2 transposable element has been shown to create insertions in the zebrafish genome very efficiently. By using Tol2, gene trap and enhancer trap vectors containing the GFP reporter gene or yeast transcription activator Gal4 gene have been developed. Here we outline methodology for creating transgenic zebrafish using Tol2 vectors, and their applications to visualization and manipulation of specific tissues or cells in vivo and for functional studies of vertebrate neural circuits.

Transient and stable transgenesis using tol2 transposon vectors.

Transgenesis is an important methodology for studying the function of genes and genomes in model plants and animals. For the model vertebrate zebrafish, methods using the Tol2 transposable element have been developed for this purpose. With these methods, the function of the transgene can be analyzed in both transient and stable transgenic fish. Recently, cis-sequences necessary for transposition of the Tol2 element were revealed. This enabled development of transposon vectors containing minimal DNA sequences that are easily manipulated. More recently, several transposon vectors containing the Gateway sequence were created and reported. These are useful because any foreign sequences can be cloned into a transposon vector fairly easily and rapidly. This chapter describes the features of these transposon vectors, and protocols to perform transgenesis in zebrafish.

Enhancer detection in zebrafish permits the identification of neuronal subtypes that express Hox4 paralogs.

Activity of zebrafish hoxb4a in the developing brain was analyzed in comparison to hoxa4a and hoxd4a using unique enhancer detection transgenes. Cytoplasmic YFP revealed shape and axonal projections of neurons in animals with insertions near the Hox4 genes and provided a means for the identification of neuronal subtypes. Despite an early activity of the genes in neuroepithelial cells and later in immature postmitotic neurons, we found reporter expression in distinct neuronal subtypes in the r7-r8-derived hindbrain. Most strikingly, hoxb4a neuronal subtypes projected through the vagus and into the pectoral fin while others formed symmetrically located fiber tracts innervating the cerebellum and the tectum, features that are partially shared by the other two paralogs. Collectively, our expression analysis indicates that hoxb4a in combination with its paralogs may play a significant role in the development of precerebellar, vagal, and pectoral fin neuronal subtypes.

meis1 regulates cyclin D1 and c-myc expression, and controls the proliferation of the multipotent cells in the early developing zebrafish eye.

During eye development, retinal progenitors are drawn from a multipotent, proliferative cell population. In Drosophila the maintenance of this cell population requires the function of the TALE-homeodomain transcription factor Hth, although its mechanisms of action are still unknown. Here we investigate whether members of the Meis gene family, the vertebrate homologs of hth, are also involved in early stages of eye development in the zebrafish. We show that meis1 is initially expressed throughout the eye primordium. Later, meis1 becomes repressed as neurogenesis is initiated, and its expression is confined to the ciliary margin, where the retinal stem population resides. Knocking down meis1 function through morpholino injection causes a delay in the G1-to-S phase transition of the eye cells, and results in severely reduced eyes. This role in cell cycle control is mediated by meis1 regulating cyclin D1 and c-myc transcription. The forced maintenance of meis1 expression in cell clones is incompatible with the normal differentiation of the meis1-expressing cells, which in turn tend to reside in undifferentiated regions of the retinal neuroepithelium, such as the ciliary margin. Together, these results implicate meis1 as a positive cell cycle regulator in early retinal cells, and provide evidence of an evolutionary conserved function for Hth/Meis genes in the maintenance of the proliferative, multipotent cell state during early eye development.

Identification and real-time imaging of a myc-expressing neutrophil population involved in inflammation and mycobacterial granuloma formation in zebrafish.

By enhancer trap screening we identified a transgenic zebrafish line showing leukocyte-specific YFP expression during late embryo and early larval development. Its enhancer detection insertion was mapped near a novel member of the myc proto-oncogene family, encoding transcription factors known to be important for regulating human myelopoiesis. Characterization of the zebrafish myc family showed that only this particular myc gene is strongly expressed in leukocytes. To identify the myc/YFP-expressing cell type, we re-examined specificity of described myeloid markers by multiplex fluorescent in situ hybridization, showing that lcp1 can be considered as a general leukocyte marker, csf1r as a macrophage-specific marker, and mpx and lyz as neutrophil-specific markers. Subsequent colocalization analysis defined the YFP-positive cells as a subset of the neutrophil population. Using real-time confocal imaging we demonstrate that these cells migrate to sites of inflammation and are involved in innate immune responses towards infections, including Mycobacterium marinum-induced granuloma formation.

Retroviral enhancer detection insertions in zebrafish combined with comparative genomics reveal genomic regulatory blocks - a fundamental feature of vertebrate genomes.

A large-scale enhancer detection screen was performed in the zebrafish using a retroviral vector carrying a basal promoter and a fluorescent protein reporter cassette. Analysis of insertional hotspots uncovered areas around developmental regulatory genes in which an insertion results in the same global expression pattern, irrespective of exact position. These areas coincide with vertebrate chromosomal segments containing identical gene order; a phenomenon known as conserved synteny and thought to be a vestige of evolution. Genomic comparative studies have found large numbers of highly conserved noncoding elements (HCNEs) spanning these and other loci. HCNEs are thought to act as transcriptional enhancers based on the finding that many of those that have been tested direct tissue specific expression in transient or transgenic assays. Although gene order in hox and other gene clusters has long been known to be conserved because of shared regulatory sequences or overlapping transcriptional units, the chromosomal areas found through insertional hotspots contain only one or a few developmental regulatory genes as well as phylogenetically unrelated genes. We have termed these regions genomic regulatory blocks (GRBs), and show that they underlie the phenomenon of conserved synteny through all sequenced vertebrate genomes. After teleost whole genome duplication, a subset of GRBs were retained in two copies, underwent degenerative changes compared with tetrapod loci that exist as single copy, and that therefore can be viewed as representing the ancestral form. We discuss these findings in light of evolution of vertebrate chromosomal architecture and the identification of human disease mutations.

Genomic regulatory blocks encompass multiple neighboring genes and maintain conserved synteny in vertebrates.

We report evidence for a mechanism for the maintenance of long-range conserved synteny across vertebrate genomes. We found the largest mammal-teleost conserved chromosomal segments to be spanned by highly conserved noncoding elements (HCNEs), their developmental regulatory target genes, and phylogenetically and functionally unrelated "bystander" genes. Bystander genes are not specifically under the control of the regulatory elements that drive the target genes and are expressed in patterns that are different from those of the target genes. Reporter insertions distal to zebrafish developmental regulatory genes pax6.1/2, rx3, id1, and fgf8 and miRNA genes mirn9-1 and mirn9-5 recapitulate the expression patterns of these genes even if located inside or beyond bystander genes, suggesting that the regulatory domain of a developmental regulatory gene can extend into and beyond adjacent transcriptional units. We termed these chromosomal segments genomic regulatory blocks (GRBs). After whole genome duplication in teleosts, GRBs, including HCNEs and target genes, were often maintained in both copies, while bystander genes were typically lost from one GRB, strongly suggesting that evolutionary pressure acts to keep the single-copy GRBs of higher vertebrates intact. We show that loss of bystander genes and other mutational events suffered by duplicated GRBs in teleost genomes permits target gene identification and HCNE/target gene assignment. These findings explain the absence of evolutionary breakpoints from large vertebrate chromosomal segments and will aid in the recognition of position effect mutations within human GRBs.

Three enhancer regions regulate gbx2 gene expression in the isthmic region during zebrafish development.

In vertebrate embryos, positioning of the boundary between the midbrain and hindbrain (MHB) and subsequent isthmus formation are dependent upon the interaction between the Otx2 and Gbx genes. In zebrafish, sequential expression of gbx1 and gbx2 in the anterior hindbrain contributes to this process, whereas in mouse embryos, a single Gbx gene (Gbx2) is responsible for MHB development. In the present study, to investigate the regulatory mechanism of gbx2 in the MHB/isthmic region of zebrafish embryos, we cloned the gene and showed that its organization is conserved among different vertebrates. Promoter analyses revealed three enhancers that direct reporter gene expression after the end of epiboly in the anterior-most hindbrain, which is a feature of the zebrafish gbx2 gene. One of the enhancers is located upstream of gbx2 (AMH1), while the other two enhancers are located downstream of gbx2 (AMH2 and AMH3). Detailed analysis of the AMH1 enhancer showed that it directs expression in the rhombomere 1 (r1) region and the dorsal thalamus, as has been shown for gbx2, whereas no expression was induced by the AMH1 enhancer in other embryonic regions in which gbx2 is expressed. The AMH1 enhancer is composed of multiple regulatory subregions that share the same spatial specificity. The most active of the regulatory subregions is a 291-bp region that contains at least two Pax2-binding sites, both of which are necessary for the function of the main component (PB1-A region) of the AMH1 enhancer. In accordance with these results, enhancer activity in the PB1-A region, as well as gbx2 expression in r1, was missing in no isthmus mutant embryos that lacked functional pax2a. In addition, we identified an upstream conserved sequence of 227bp that suppresses the enhancer activity of AMH1. Taken together, these findings suggest that gbx2 expression during the somitogenesis stage in zebrafish is regulated by a complex mechanism involving Pax2 as well as activators and suppressors in the regions flanking the gene.

Enhancer detection in the zebrafish using pseudotyped murine retroviruses.

Vectors based on murine retroviruses are among the most efficient means to insert reporter constructs into the context of a vertebrate chromosome with the aim to visualize cis-regulatory information available to a basal promoter at the site of insertion. In combination with using the zebrafish embryo as a readout for the activity of regulatory elements, enhancer detection becomes a powerful technique for gene discovery and for the mapping of the extent of regulatory domains in a vertebrate genome. Our laboratory has performed the only large-scale enhancer detection screen to date in any vertebrate and we describe in this paper the methods we developed to generate viral particles, to insert reporter constructs into the zebrafish germ line, the screening of detection events in heterozygous F1 embryos, and the isolation of genomic sequence flanking the inserted vector for the purpose of genomic mapping. Given sufficient scale, the technology described here can be used to obtain cis-regulatory information across the entire zebrafish genome for any given basal promoter.

Conserved co-regulation and promoter sharing of hoxb3a and hoxb4a in zebrafish.

The expression of zebrafish hoxb3a and hoxb4a has been found to be mediated through five transcripts, hoxb3a transcripts I-III and hoxb4a transcripts I-II, driven by four promoters. A "master" promoter, located about 2 kb downstream of hoxb5a, controls transcription of a pre-mRNA comprising exon sequences of both genes. This unique gene structure is proposed to provide a novel mechanism to ensure overlapping, tissue-specific expression of both genes in the posterior hindbrain and spinal cord. Transgenic approaches were used to analyze the functions of zebrafish hoxb3a/hoxb4a promoters and enhancer sequences containing regions of homology that were previously identified by comparative genomics. Two neural enhancers were shown to establish specific anterior expression borders within the hindbrain and mediate expression in defined neuronal populations derived from hindbrain rhombomeres (r) 5 to 8, suggesting a late role of the genes in neuronal cell lineage specification. Species comparison showed that the zebrafish hoxb3a r5 and r6 enhancer corresponded to a sequence within the mouse HoxA cluster controlling activity of Hoxa3 in r5 and r6, whereas a homologous region within the HoxB cluster activated Hoxb3 expression but limited to r5. We conclude that the similarity of hoxb3a/Hoxa3 regulatory mechanisms reflect the shared descent of both genes from a single ancestral paralog group 3 gene.

Segregation of telencephalic and eye-field identities inside the zebrafish forebrain territory is controlled by Rx3.

Anteroposterior patterning of the vertebrate forebrain during gastrulation involves graded Wnt signaling, which segregates anterior fields (telencephalon and eye) from the diencephalon. How the telencephalic and retinal primordia are subsequently subdivided remains largely unknown. We demonstrate that at late gastrulation the Paired-like homeodomain transcription factor Rx3 biases cell specification choices towards the retinal fate within a population of bipotential precursors of the anterior forebrain: direct cell tracing demonstrates that retinal precursors acquire a telencephalic fate in embryos homozygous for the rx3-null allele ckh(ne2611), characterized by an enlarged telencephalon and a lack of eyes. Chimera analyses further indicate that this function of Rx3 is cell autonomous. Transfating of the eye field in the absence of Rx3 function correlates with a substantial posterior expansion of expression of the Wnt antagonist Tlc and the winged-helix transcription factor Foxg1. These results suggest that the process segregating the telencephalic and eye fields is isolated from diencephalic patterning, and is mediated by Rx3.

Large-scale enhancer detection in the zebrafish genome.

Murine retroviral vectors carrying an enhancer detection cassette were used to generate 95 transgenic lines of fish in which reporter expression is observed in distinct patterns during embryonic development. We mapped 65 insertion sites to the as yet unfinished zebrafish genome sequence. Many integrations map close to previously known developmental genes, including transcription factors of the Pax, Hox, Sox, Pou, Otx, Emx, zinc-finger and bHLH gene families. In most cases, the activated provirus is located in, or within a 15 kb interval around, the corresponding transcriptional unit. The exceptions include four insertions into a gene desert on chromosome 20 upstream of sox11b, and an insertion upstream of otx1. In these cases, the activated insertions are found at a distance of between 32 kb and 132 kb from the coding region. These as well as seven other insertions described here identify genes that have recently been associated with ultra conserved non-coding elements found in all vertebrate genomes.

gbx2 Homeobox gene is required for the maintenance of the isthmic region in the zebrafish embryonic brain.

We isolated cDNA clones for the zebrafish gbx2 gene, which is implicated in the establishment of the midbrain-hindbrain boundary (MHB) in other vertebrates. Spatially localized expression of gbx2 was observed at the MHB from 90% epiboly through to the hatching stage. Comparisons with the expression of otx2, wnt1, and krox20 showed that gbx2 is expressed in the anterior hindbrain. Ectopic expression of gbx2 by mRNA injection caused cyclopia or truncation of the fore- and midbrain and severely affected isthmic and cerebellar structures, while hindbrain formation was not significantly affected. At the molecular level, gbx2 suppressed the expression of otx2 in the fore/midbrain, six3 in the anterior forebrain, and MHB-specific genes such as eng2 and wnt1. In contrast, gbx2 did not down-regulate the expression of the hindbrain marker genes. Therefore, gbx2 specifically suppressed the formation of the entire fore/midbrain. Meanwhile, misexpression of otx2 suppressed the expression of gbx2 in the embryonic brain. Abrogation of gbx2 expression with an antisense morpholino oligonucleotide disrupted the midbrain/anterior hindbrain region, and these loss-of-function effects were rescued by activating the Gbx2 protein immediately after the end of gastrulation. Taken together, these results suggest that the zebrafish gbx2 gene is essential for the maintenance of MHB and/or the formation of the isthmic structure during somitogenesis, rather than for the MHB establishment during gastrulation. We also suggest that other factors, including gbx1, is required for the establishment of the MHB during gastrulation.

Expression of the FGF receptor 2 gene (fgfr2) during embryogenesis in the zebrafish Danio rerio.

We isolated a full-length cDNA clone for the zebrafish homologue of fibroblast growth factor receptor (FGFR) 2. The deduced protein sequence is typical of vertebrate FGFRs in that it has three Ig-like domains in the extracellular region. The expression of fgfr2 is initiated during epiboly in the paraxial mesoderm. During early somitogenesis, fgfr2 expression was noted in the anterior neural plate as well as in newly formed somites. Whereas fgfr2 expression in somites is transient, it increases in the central nervous system (CNS), i.e. in the ventral telencephalon, anterior diencephalon, midbrain, and respective rhombomeres of the hindbrain, from the mid-somitogenesis stage. The dorsal telencephalon and the region around the midbrain-hindbrain boundary are devoid of fgfr2 expression. Essentially the same expression pattern is observed until 48 h post-fertilization in the CNS, although rhombomeric expression in the hindbrain is progressively confined to narrower stripes. After somitogenesis, fgfr2 expression was also observed in the lens, hypochord, endoderm, and fin mesenchyme. We compared the expression of the four fgfr genes (fgfr1-4) in the CNS of zebrafish embryos and show that fgfr1 is the only fgfr gene that is expressed in the dorsal telencephalon and isthmic region from which expression of fgfr2-4 is absent.

Expression of the FGF receptor 2 gene (fgfr2) during embryogenesis in the zebrafish Danio rerio.

We isolated a full-length cDNA clone for the zebrafish homologue of fibroblast growth factor receptor (FGFR) 2. The deduced protein sequence is typical of vertebrate FGFRs in that it has three Ig-like domains in the extracellular region. The expression of fgfr2 is initiated during epiboly in the paraxial mesoderm. During early somitogenesis, fgfr2 expression was noted in the anterior neural plate as well as in newly formed somites. Whereas fgfr2 expression in somites is transient, it increases in the central nervous system (CNS), i.e. in the ventral telencephalon, anterior diencephalon, midbrain, and respective rhombomeres of the hindbrain, from the mid-somitogenesis stage. The dorsal telencephalon and the region around the midbrain-hindbrain boundary are devoid of fgfr2 expression. Essentially the same expression pattern is observed until 48 h post-fertilization in the CNS, although rhombomeric expression in the hindbrain is progressively confined to narrower stripes. After somitogenesis, fgfr2 expression was also observed in the lens, hypochord, endoderm, and fin mesenchyme. We compared the expression of the four fgfr genes (fgfr1-4) in the CNS of zebrafish embryos and show that fgfr1 is the only fgfr gene that is expressed in the dorsal telencephalon and isthmic region from which expression of fgfr2-4 is absent.