Development of the head skeleton involves reciprocal interactions between cranial neural crest cells (CNCCs) and the surrounding pharyngeal endoderm and ectoderm. Whereas elegant experiments in avians have shown a prominent role for the endoderm in facial skeleton development, the relative functions of the endoderm in growth versus regional identity of skeletal precursors have remained unclear. Here we describe novel craniofacial defects in zebrafish harboring mutations in the Sphingosine-1-phospate (S1P) type 2 receptor (s1pr2) or the S1P transporter Spinster 2 (spns2), and we show that S1P signaling functions in the endoderm for the proper growth and positioning of the jaw skeleton. Surprisingly, analysis of s1pr2 and spns2 mutants, as well as sox32 mutants that completely lack endoderm, reveals that the dorsal–ventral (DV) patterning of jaw skeletal precursors is largely unaffected even in the absence of endoderm. Instead, we observe reductions in the ectodermal expression of Fibroblast growth factor 8a (Fgf8a), and transgenic misexpression of Shha restores fgf8a expression and partially rescues the growth and differentiation of jaw skeletal precursors. Hence, we propose that the S1P-dependent anterior foregut endoderm functions primarily through Shh to regulate the growth but not DV patterning of zebrafish jaw precursors.

The Homeobox (Hox) and Paired box (Pax) gene families are key determinants of animal body plans and organ structure. In particular, they function within regulatory networks that control organogenesis. How these conserved genes elicit differences in organ form and function in response to evolutionary pressures is incompletely understood. We molecularly and functionally characterized one member of an evolutionarily dynamic gene family, plac8 onzin related protein 1 (ponzr1), in the zebrafish. ponzr1 mRNA is expressed early in the developing kidney and pharyngeal arches. Using ponzr1-targeting morpholinos, we show that ponzr1 is required for formation of the glomerulus. Loss of ponzr1 results in a nonfunctional glomerulus but retention of a functional pronephros, an arrangement similar to the aglomerular kidneys found in a subset of marine fish. ponzr1 is integrated into the pax2a pathway, with ponzr1 expression requiring pax2a gene function, and proper pax2a expression requiring normal ponzr1 expression. In addition to pronephric function, ponzr1 is required for pharyngeal arch formation. We functionally demonstrate that ponzr1 can act as a transcription factor or co-factor, providing the first molecular mode of action for this newly described gene family. Together, this work provides experimental evidence of an additional mechanism that incorporates evolutionarily dynamic, lineage-specific gene families into conserved regulatory gene networks to create functional organ diversity.

Using zinc-finger nuclease-mediated mutagenesis, we have generated mutant alleles of the zebrafish orthologue of the chicken talpid3 (ta3) gene, which encodes a centrosomal protein that is essential for ciliogenesis. Animals homozygous for these mutant alleles complete embryogenesis normally, but manifest a cystic kidney phenotype during the early larval stages and die within a month of hatching. Elimination of maternally derived Ta3 activity by germline replacement resulted in embryonic lethality of ta3 homozygotes. The phenotype of such maternal and zygotic (MZta3) mutant zebrafish showed strong similarities to that of chick ta3 mutants: absence of primary and motile cilia as well as aberrant Hedgehog (Hh) signalling, the latter manifest by the expanded domains of engrailed and ptc1 expression in the somites, reduction of nkx2.2 expression in the neural tube, symmetric pectoral fins, cyclopic eyes and an ectopic lens. GFP-tagged Gli2a localised to the basal bodies in the absence of the primary cilia and western blot analysis showed that Gli2a protein is aberrantly processed in MZta3 embryos. Zygotic expression of ta3 largely rescued the effects of maternal depletion, but the motile cilia of Kupffer’s vesicle remained aberrant, resulting in laterality defects. Our findings underline the importance of the primary cilium for Hh signaling in zebrafish and reveal the conservation of Ta3 function during vertebrate evolution.

Vertebrate hematopoiesis is a complex developmental process that is controlled by genes in diverse pathways. To identify novel genes involved in early hematopoiesis, we conducted an ENU (N-ethyl-N-nitrosourea) mutagenesis screen in zebrafish. The mummy (mmy) line was investigated because of its multiple hematopoietic defects. Homozygous mmy embryos lacked circulating blood cells types and were dead by 30 hours post-fertilization (hpf). The mmy mutants did not express myeloid markers and had significantly decreased expression of progenitor and erythroid markers in primitive hematopoiesis. Through positional cloning, we identified a truncation mutation in dhx8 in the mmy fish. dhx8 is the zebrafish ortholog of the yeast splicing factor prp22, which is a DEAH-box RNA helicase. Mmy mutants had splicing defects in many genes, including several hematopoietic genes. Mmy embryos also showed cell division defects as characterized by disorganized mitotic spindles and formation of multiple spindle poles in mitotic cells. These cell division defects were confirmed by DHX8 knockdown in HeLa cells. Together, our results confirm that dhx8 is involved in mRNA splicing and suggest that it is also important for cell division during mitosis. This is the first vertebrate model for dhx8, whose function is essential for primitive hematopoiesis in developing embryos.

The cranial vasculature is essential for the survival and development of the central nervous system and is important in stroke and other brain pathologies. Cranial vessels form in a reproducible and evolutionarily conserved manner, but the process by which these vessels assemble and acquire their stereotypic patterning remains unclear. Here, we examine the stepwise assembly and patterning of the vascular network of the zebrafish hindbrain. The major artery supplying the hindbrain, the basilar artery, runs along the ventral keel of the hindbrain in all vertebrates. We show that this artery forms by a novel process of medial sprouting and migration of endothelial cells from a bilateral pair of primitive veins, the primordial hindbrain channels. Subsequently, a second wave of dorsal sprouting from the primordial hindbrain channels gives rise to angiogenic central arteries that penetrate into and innervate the hindbrain. The chemokine receptor cxcr4a is expressed in migrating endothelial cells of the primordial hindbrain channels, whereas its ligand cxcl12b is expressed in the hindbrain neural keel immediately adjacent to the assembling basilar artery. Knockdown of either cxcl12b or cxcr4a results in defects in basilar artery formation, showing that the assembly and patterning of this crucial artery depends on chemokine signaling.

Here, we show that a novel Rspo1-Wnt-Vegfc-Vegfr3 signaling pathway plays an essential role in developmental angiogenesis. A mutation in R-spondin1 (rspo1), a Wnt signaling regulator, was uncovered during a forward-genetic screen for angiogenesis-deficient mutants in the zebrafish. Embryos lacking rspo1 or the proposed rspo1 receptor kremen form primary vessels by vasculogenesis, but are defective in subsequent angiogenesis. Endothelial cell-autonomous inhibition of canonical Wnt signaling also blocks angiogenesis in vivo. The pro-angiogenic effects of Rspo1/Wnt signaling are mediated by Vegfc/Vegfr3(Flt4) signaling. Vegfc expression is dependent on Rspo1 and Wnt, and Vegfc and Vegfr3 are necessary to promote angiogenesis downstream from Rspo1-Wnt. As all of these molecules are expressed by the endothelium during sprouting stages, these results suggest that Rspo1-Wnt-VegfC-Vegfr3 signaling plays a crucial role as an endothelial-autonomous permissive cue for developmental angiogenesis.

Mef2 transcription factors have been strongly linked with early heart development. D-mef2 is required for heart formation in Drosophila, but whether Mef2 is essential for vertebrate cardiomyocyte (CM) differentiation is unclear. In mice, although Mef2c is expressed in all CMs, targeted deletion of Mef2c causes lethal loss of second heart field (SHF) derivatives and failure of cardiac looping, but first heart field CMs can differentiate. Here we examine Mef2 function in early heart development in zebrafish. Two Mef2c genes exist in zebrafish, mef2ca and mef2cb. Both are expressed similarly in the bilateral heart fields but mef2cb is strongly expressed in the heart poles at the primitive heart tube stage. By using fish mutants for mef2ca and mef2cb and antisense morpholinos to knock down either or both Mef2cs, we show that Mef2ca and Mef2cb have essential but redundant roles in myocardial differentiation. Loss of both Mef2ca and Mef2cb function does not interfere with early cardiogenic markers such as nkx2.5, gata4 and hand2 but results in a dramatic loss of expression of sarcomeric genes and myocardial markers such as bmp4, nppa, smyd1b and late nkx2.5 mRNA. Rare residual CMs observed in mef2ca;mef2cb double mutants are ablated by a morpholino capable of knocking down other Mef2s. Mef2cb over-expression activates bmp4 within the cardiogenic region, but no ectopic CMs are formed. Surprisingly, anterior mesoderm and other tissues become skeletal muscle. Mef2ca single mutants have delayed heart development, but form an apparently normal heart. Mef2cb single mutants have a functional heart and are viable adults. Our results show that the key role of Mef2c in myocardial differentiation is conserved throughout the vertebrate heart.

Human PITX2 mutations are associated with Axenfeld-Rieger syndrome, an autosomal-dominant developmental disorder that involves ocular anterior segment defects, dental hypoplasia, craniofacial dysmorphism and umbilical abnormalities. Characterization of the PITX2 pathway and identification of the mechanisms underlying the anomalies associated with PITX2 deficiency is important for better understanding of normal development and disease; studies of pitx2 function in animal models can facilitate these analyses. A knockdown of pitx2 in zebrafish was generated using a morpholino that targeted all known alternative transcripts of the pitx2 gene; morphant embryos generated with the pitx2^ex4/5 splicing-blocking oligomer produced abnormal transcripts predicted to encode truncated pitx2 proteins lacking the third (recognition) helix of the DNA-binding homeodomain. The morphological phenotype of pitx2^ex4/5 morphants included small head and eyes, jaw abnormalities and pericardial edema; lethality was observed at ~6–8-dpf. Cartilage staining revealed a reduction in size and an abnormal shape/position of the elements of the mandibular and hyoid pharyngeal arches; the ceratobranchial arches were also decreased in size. Histological and marker analyses of the misshapen eyes of the pitx2^ex4/5 morphants identified anterior segment dysgenesis and disordered hyaloid vasculature. In summary, we demonstrate that pitx2 is essential for proper eye and craniofacial development in zebrafish and, therefore, that PITX2/pitx2 function is conserved in vertebrates.

The vertebrate body forms in an anterior-to-posterior progression, driven by a population of undifferentiated cells at the posterior-most end of the embryo. Recent studies have demonstrated that these undifferentiated cells are multipotent stem cells, suggesting that local signaling factors specify cell fate. However, the mechanism of cell fate specification during this process is unknown. Using a combination of single cell transplantation and newly developed cell-autonomous inducible Wnt inhibitor and activator transgenic zebrafish lines, we show that canonical Wnt signaling is continuously necessary and sufficient to specify mesoderm from a bipotential neural/mesodermal precursor. Surprisingly, we also find that Wnt signaling functions subsequently within the mesoderm to specify somites instead of posterior vascular endothelium. Our results demonstrate that dynamic local Wnt signaling cues specify germ layer contribution and mesodermal tissue type specification of multipotent stem cells throughout the formation of the early vertebrate embryonic body.

The creation of molecular tools able to unravel in vivo spatiotemporal activation of specific cell signaling events during cell migration, differentiation and morphogenesis is of great relevance to developmental cell biology. Here, we describe the generation, validation and applications of two transgenic reporter lines for Wnt/β-catenin signaling, named TCFsiam, and show that they are reliable and sensitive Wnt biosensors for in vivo studies. We demonstrate that these lines sensitively detect Wnt/β-catenin pathway activity in several cellular contexts, from sensory organs to cardiac valve patterning. We provide evidence that Wnt/β-catenin activity is involved in the formation and maintenance of the zebrafish CNS blood vessel network, on which sox10 neural crest-derived cells migrate and proliferate. We finally show that these transgenic lines allow for screening of Wnt signaling modifying compounds, tissue regeneration assessment as well as evaluation of potential Wnt/β-catenin genetic modulators.

The neural crest is a migratory, multipotent cell lineage that contributes to myriad tissues, including sensory neurons and glia of the dorsal root ganglia (DRG). To identify genes affecting cell fate specification in neural crest, we performed a forward genetic screen for mutations causing DRG deficiencies in zebrafish. This screen yielded a mutant lacking all DRG, which we named sensory deprived (sdp). We identified a total of four alleles of sdp, all of which possess lesions in the gene coding for reversion-inducing cysteine-rich protein containing Kazal motifs (Reck). Reck is an inhibitor of metalloproteinases previously shown to regulate cell motility. We found reck function to be both necessary for DRG formation and sufficient to rescue the sdp phenotype. reck is expressed in neural crest cells and is required in a cell-autonomous fashion for appropriate sensory neuron formation. In the absence of reck function, sensory neuron precursors fail to migrate to the position of the DRG, suggesting that this molecule is crucial for proper migration and differentiation.