Functional development of the zebrafish pineal gland: light-induced expression of period2 is required for onset of the circadian clock.

In zebrafish, the pineal gland is a photoreceptive organ that contains an intrinsic circadian oscillator and exhibits rhythmic arylalkylamine-N-acetyltransferase (zfaanat2) mRNA expression. In the present study, we investigated the role of light and of a clock gene, zperiod2 (zper2), in the development of this rhythm. Analysis of zfaanat2 mRNA expression in the pineal gland of 3-day-old zebrafish embryos after exposure to different photoperiodic regimes indicated that light is required for proper development of the circadian clock-controlled rhythmic expression of zfaanat2, and that a 1-h light pulse is sufficient to initiate this rhythm. Analysis of zper2 mRNA expression in zebrafish embryos exposed to different photoperiodic regimes indicated that zper2 expression is transiently up-regulated by light but is not regulated by the circadian oscillator. To establish the association between light-induced zper2 expression and light-induced clock-controlled zfaanat2 rhythm, zPer2 knock-down experiments were performed. The zfaanat2 mRNA rhythm, induced by a 1-h light pulse, was abolished in zPer2 knock-down embryos. These experiments indicated that light-induced zper2 expression is crucial for establishment of the clock-controlled zfaanat2 rhythm in the zebrafish pineal gland.

A gene expression screen in zebrafish embryogenesis.

A screen for developmentally regulated genes was conducted in the zebrafish, a system offering substantial advantages for the study of the molecular genetics of vertebrate embryogenesis. Clones from a normalized cDNA library from early somitogenesis stages were picked randomly and tested by high-throughput in situ hybridization for restricted expression in at least one of four stages of development. Among 2765 clones that were screened, a total of 347 genes with patterns judged to be restricted were selected. These clones were subjected to partial sequence analysis, allowing recognition of functional motifs in 163 among them. In addition, a portion of the clones were mapped with the aid of the LN54 radiation hybrid panel. The usefulness of the in situ hybridization screening approach is illustrated by describing several new markers for the characteristic structure in the fish embryo named the yolk syncytial layer, and for different regions of the developing brain.

The IA-2 gene family: homologs in Caenorhabditis elegans, Drosophila and zebrafish.

AIMS/HYPOTHESIS: IA-2 and IA-2beta are major autoantigens in Type I (insulin-dependent) diabetes mellitus and are expressed in neuroendocrine tissues including the brain and pancreatic islets of Langerhans. Based on sequence analysis, IA-2 and IA-2beta are transmembrane protein tyrosine phosphatases but lack phosphatase activity because of critical amino acid substitutions in the catalytic domain. We studied the evolutionary conservation of IA-2 and IA-2beta genes and searched for homologs in non-mammalian vertebrates and invertebrates. METHODS: IA-2 from various species was identified from EST sequences or cloned from cDNA libraries or both. Expression in tissues was determined by transfection and in situ hybridization. RESULTS: We identified homologs of IA-2 in C. elegans, Drosophila, and zebrafish which showed 46, 58 and 82 % identity and 60, 65 and 87 % similarity, respectively, to the amino acids of the intracellular domain of human IA-2. Further studies showed that IA-2 was expressed in the neural tissues of the three species. Comparison of the genomic structure of the intracellular domain of human IA-2 with that of human IA-2beta showed that they were nearly identical and comparison of the intron-exon boundaries of Drosophila IA-2 with human IA-2 and IA-2beta showed a high degree of relatedness. CONCLUSION/INTERPRETATION: Based on these findings and sequence analysis of IA-2 homologs in mammals, we conclude that there is an IA-2 gene family which is a part of the larger protein tyrosine phosphatase superfamily. The IA-2 and IA-2beta genes represent two distinct subgroups within the IA-2 family which originated over 500 million years ago, long before the development of the pancreatic islets of Langerhans.

Zebrafish serotonin N-acetyltransferase-2: marker for development of pineal photoreceptors and circadian clock function.

Serotonin N-acetyltransferase (AANAT), the penultimate enzyme in melatonin synthesis, is typically found only at significant levels in the pineal gland and retina. Large changes in the activity of this enzyme drive the circadian rhythm in circulating melatonin seen in all vertebrates. In this study, we examined the utility of using AANAT messenger RNA (mRNA) as a marker to monitor the very early development of pineal photoreceptors and circadian clock function in zebrafish. Zebrafish AANAT-2 (zfAANAT-2) cDNA was isolated and used for in situ hybridization. In the adult, zfAANAT-2 mRNA is expressed exclusively in pineal cells and retinal photoreceptors. Developmental analysis, using whole mount in situ hybridization, indicated that pineal zfAANAT-2 mRNA expression is first detected at 22 h post fertilization. Retinal zfAANAT-2 mRNA was first detected on day 3 post fertilization and appears to be associated with development of the retinal photoreceptors. Time-of-day analysis of 2- to 5-day-old zebrafish larvae indicated that zfAANAT-2 mRNA abundance exhibits a dramatic 24-h rhythm in a 14-h light, 10-h dark cycle, with high levels at night. This rhythm persists in constant darkness, indicating that the zfAANAT-2 mRNA rhythm is driven by a circadian clock at this stage. The techniques described in this report were also used to determine that zfAANAT-2 expression is altered in two well characterized genetic mutants, mindbomb and floating head. The observations described here suggest that zfAANAT-2 mRNA may be a useful marker to study development of the pineal gland and of circadian clock mechanisms in zebrafish.

Gastrulation in zebrafish: what mutants teach us.

A major approach to the study of development is to compare the phenotypes of normal and mutant individuals for a given genetic locus. Understanding the development of a complex metazoan therefore requires examination of many mutants. Relatively few organisms are being studied this way, and zebrafish is currently the best example of a vertebrate for which large-scale mutagenesis screens have successfully been carried out. The number of genes mutated in zebrafish that have been cloned expands rapidly, bringing new insights into a number of developmental pathways operating in vertebrates. Here, we discuss work on zebrafish mutants affecting gastrulation and patterning of the early embryo. Gastrulation is orchestrated by the dorsal organizer, which forms in a region where maternally derived beta-catenin signaling is active. Mutation in the zygotic homeobox gene bozozok disrupts the organizer genetic program and leads to severe axial deficiencies, indicating that this gene is a functional target of beta-catenin signaling. Once established, the organizer releases inhibitors of ventralizing signals, such as BMPs, and promotes dorsoanterior fates within all germ layers. In zebrafish, several mutations affecting dorsal-ventral (D/V) patterning inactivate genes functioning in the BMP pathway, stressing the central role of this pathway in the gastrula embryo. Cells derived from the organizer differentiate into several axial structures, such as notochord and prechordal mesoderm, which are thought to induce various fates in adjacent tissues, such as the floor plate, after the completion of gastrulation. Studies with mutants in nodal-related genes, in one-eyed pinhead, which is required for nodal signaling, and in the Notch pathway reveal that midline cell fate specification is, in fact, initiated during gastrulation. Furthermore, the organizer coordinates morphogenetic movements, and zebrafish mutants in T-box mesoderm-specific genes help clarify the mechanism of convergence movements required for the formation of axial and paraxial mesoderm.

Isolation, characterization and mutation analysis of PEX13-defective Chinese hamster ovary cell mutants.

We isolated peroxisome biogenesis mutants ZP128 and ZP150 from rat PEX2 -transformed Chinese hamster ovary (CHO) cells, by the 9-(1'-pyrene)nonanol/ultraviolet method. The mutants lacked morphologically recognizable peroxisomes and showed a typical peroxisome assembly-defective phenotype such as a high sensitivity to 12-(1'-pyrene)dodecanoic acid/UV treatment. By means of PEX cDNA transfection and cell fusion, ZP128 and ZP150 were found to belong to a recently identified complementation group H. Expression of human PEX13 cDNA restored peroxisome assembly in ZP128 and ZP150. CHO cell PEX13 was isolated; its deduced sequence comprises 405 amino acids with 93% identity to human Pex13p. Mutation in PEX13 of mutant ZP150 was determined by RT-PCR: G to A transition resulted in one amino acid substitution, Ser319Asn, in one allele and truncation of a 42 amino acid sequence from Asp265 to Lys306 in another allele. Therefore, ZP128 and ZP150 are CHO cell lines with a phenotype of impaired PEX13.

Nonsense and temperature-sensitive mutations in PEX13 are the cause of complementation group H of peroxisome biogenesis disorders.

Peroxisome biogenesis disorders, including Zellweger syndrome (ZS), neonatal adrenoleukodystrophy (NALD) and infantile Refsum disease, are lethal hereditary diseases caused by abnormalities in peroxisomal assembly. To date, 12 genotypes have been identified. We now have evidence that the complete human cDNA encoding Pex13p, an SH3 protein of a docking factor for the peroxisome targeting signal 1 receptor (Pex5p), rescues peroxisomal matrix protein import and its assembly in fibroblasts from PBD patients of complementation group H. In addition, we detected mutations on the human PEX13 cDNA in two patients of group H. A severe phenotype of a ZS patient (H-02) was homozygous for a nonsense mutation, W234ter, which results in the loss of not only the SH3 domain but also the putative transmembrane domain of Pex13p. A more mildly affected NALD patient (H-01), whose fibroblasts showed the temperature-sensitive (TS) phenotype, was homozygous for a missense mutation in the SH3 domain of Pex13p, I326T. This mutant PEX13 cDNA expression in a PEX13-defective CHO mutant showed I326T to be a TS mutation and thus suggested that Pex13p with the I326T mutation in the SH3 domain is stable at 30 degrees C but is somewhat unstable at 37 degrees C.

Newly identified Chinese hamster ovary cell mutants defective in peroxisome assembly represent complementation group A of human peroxisome biogenesis disorders and one novel group in mammals.

We isolated peroxisome biogenesis-defective mutants from rat PEX2-transformed Chinese hamster ovary (CHO) cells, using the 9-(1'-pyrene)nonanol/ultraviolet method. A total of 18 mutant cell clones showing cytosolic localization of catalase were isolated. By complementation group (CG) analysis by means of PEX cDNA transfection and cell fusion, cell mutants, ZP124 and ZP126, were found to belong to two novel CGs of CHO mutants. Mutants, ZP135 and ZP167, were also classified to the same CG as ZP124. Further cell fusion analysis using 12 CGs fibroblasts from patients with peroxisome deficiency disorders such as Zellweger syndrome revealed that ZP124 belonged to human CG-A, the same group as CG-VIII in the United States. ZP126 could not be classified to any of human and CHO CGs. These mutants also showed typical peroxisome assembly-defective phenotypes such as severe loss of catalase latency and impaired biogenesis of peroxisomal enzymes. Collectively, ZP124 represents CG-A, and ZP126 is in a newly identified CG distinct from the 14 mammalian CGs previously characterized.

cyclops encodes a nodal-related factor involved in midline signaling.

Ventral structures in the central nervous system are patterned by signals emanating from the underlying mesoderm as well as originating within the neuroectoderm. Mutations in the zebrafish, Danio rerio, are proving instrumental in dissecting these midline signals. The cyclops mutation leads to a loss of medial floor plate and to severe deficits in ventral forebrain development, leading to cyclopia. Here, we report that the cyclops locus encodes the nodal-related protein Ndr2, a member of the transforming growth factor type beta superfamily of factors. The evidence includes identification of a missense mutation in the initiation codon and rescue of the cyclops phenotype by expression of ndr2 RNA, here renamed "cyclops." Thus, in interaction with other molecules implicated in these processes such as sonic hedgehog and one-eyed-pinhead, cyclops is required for ventral midline patterning of the embryonic central nervous system.

Zebrafish nodal-related genes are implicated in axial patterning and establishing left-right asymmetry.

Nodal-related 1 (ndr1) and nodal-related 2 (ndr2) genes in zebrafish encode members of the nodal subgroup of the transforming growth factor-beta superfamily. We report the expression patterns and functional characteristics of these factors, implicating them in the establishment of dorsal-ventral polarity and left-right asymmetry. Ndr1 is expressed maternally, and ndr1 and ndr2 are expressed during blastula stage in the blastoderm margin. During gastrulation, ndr expression subdivides the shield into two domains: a small group of noninvoluting cells, the dorsal forerunner cells, express ndr1, while ndr2 RNA is found in the hypoblast layer of the shield and later in notochord, prechordal plate, and overlying anterior neurectoderm. During somitogenesis, ndr2 is expressed asymmetrically in the lateral plate as are nodal-related genes of other organisms, and in a small domain in the left diencephalon, providing the first observation of asymmetric gene expression in the embryonic forebrain. RNA injections into Xenopus animal caps showed that Ndr1 acts as a mesoderm inducer, whereas Ndr2 is an efficient neural but very inefficient mesoderm inducer. We suggest that Ndr1 has a role in mesoderm induction, while Ndr2 is involved in subsequent specification and patterning of the nervous system and establishment of laterality.

Overexpression of the forebrain-specific homeobox gene six3 induces rostral forebrain enlargement in zebrafish.

The Drosophila homeobox gene sine oculis is expressed in the rostral region of the embryo in early development and is essential for eye and brain formation. Its murine homolog, Six3, is expressed in the anterior neural plate and eye anlage, and may have crucial functions in eye and brain development. In this study, we describe the cloning and expression of zebrafish six3, the apparent ortholog of the mouse Six3 gene. Zebrafish six3 transcripts are first seen in hypoblast cells in early gastrula embryos and are found in the anterior axial mesendoderm through gastrulation. six3 expression in the head ectoderm begins at late gastrula. Throughout the segmentation period, six3 is expressed in the rostral region of the prospective forebrain. Overexpression of six3 in zebrafish embryos induced enlargement of the rostral forebrain, enhanced expression of pax2 in the optic stalk and led to a general disorganization of the brain. Disruption of either the Six domain or the homeodomain abolish these effects, implying that these domains are essential for six3 gene function. Our results suggest that the vertebrate Six3 genes are involved in the formation of the rostral forebrain.

Expression of LIM-domain binding protein (ldb) genes during zebrafish embryogenesis.

LIM homeodomain proteins are developmental regulators whose functions depend on synergism with LIM domain binding proteins (Ldb proteins). We have isolated four members of the ldb gene family from the zebrafish, Danio rerio. Ldb1, Ldb2 and Ldb3 share 95%, 73% and 62% amino acid identity with mouse Ldb1, respectively. In overlay assays, Ldb proteins bind LIM homeodomain proteins and LMO1, but not zyxin or MLP. Whole mount in situ hybridization showed that zebrafish ldb1 is expressed ubiquitously from gastrulation onward. Ldb2 is ubiquitous at gastrulation, and later is found in many but not all tissues, especially the anterior central nervous system (CNS) and vasculature. Ldb3 mRNA was expressed primarily in the anterior CNS.

Human PEX1 cloned by functional complementation on a CHO cell mutant is responsible for peroxisome-deficient Zellweger syndrome of complementation group I.

The peroxisome biogenesis disorders (PBDs), including Zellweger syndrome (ZS) and neonatal adrenoleukodystrophy (NALD), are autosomal recessive diseases caused by defects in peroxisome assembly, for which at least 10 complementation groups have been reported. We have isolated a human PEX1 cDNA (HsPEX1) by functional complementation of peroxisome deficiency of a mutant Chinese hamster ovary (CHO) cell line, ZP107, transformed with peroxisome targeting signal type 1-tagged "enhanced" green fluorescent protein. This cDNA encodes a hydrophilic protein (Pex1p) comprising 1,283 amino acids, with high homology to the AAA-type ATPase family. A stable transformant of ZP107 with HsPEX1 was morphologically and biochemically restored for peroxisome biogenesis. HsPEX1 expression restored peroxisomal protein import in fibroblasts from three patients with ZS and NALD of complementation group I (CG-I), which is the highest-incidence PBD. A CG-I ZS patient (PBDE-04) possessed compound heterozygous, inactivating mutations: a missense point mutation resulting in Leu-664 --> Pro and a deletion of the sequence from Gly-634 to His-690 presumably caused by missplicing (splice site mutation). Both PBDE-04 PEX1 cDNAs were defective in peroxisome-restoring activity when expressed in the patient fibroblasts as well as in ZP107 cells. These results demonstrate that PEX1 is the causative gene for CG-I peroxisomal disorders.

LIM domains: multiple roles as adapters and functional modifiers in protein interactions.

The LIM domain is a specialized double-zinc finger motif found in a variety of proteins, in association with domains of divergent functions or forming proteins composed primarily of LIM domains. LIM domains interact specifically with other LIM domains and with many different protein domains. LIM domains are thought to function as protein interaction modules, mediating specific contacts between members of functional complexes and modulating the activity of some of the constituent proteins. Nucleic acid binding by LIM domains, while suggested by structural considerations, remains an unproven possibility. LIM-domain proteins can be nuclear, cytoplasmic, or can shuttle between compartments. Several important LIM proteins are associated with the cytoskeleton, having a role in adhesion-plaque and actin-microfilament organization. Among nuclear LIM proteins, the LIM homeodomain proteins form a major subfamily with important functions in cell lineage determination and pattern formation during animal development.

lim6, a novel LIM homeobox gene in the zebrafish: comparison of its expression pattern with lim1.

A novel LIM class homeobox gene, lim6, was isolated from a zebrafish embryonic cDNA library. The encoded protein shares a high degree of sequence similarity with the previously described Lim1 and Lim5 proteins. This study compares the spatial and temporal expression pattern of the closely related lim6 and lim1 genes during early embryogenesis. Generally, lim6 mRNA was found at rather low amounts compared to lim1 mRNA. At the shield stage, lim6 mRNA, similar to lim1 mRNA, was predominantly expressed in the shield. Lim6 was transiently expressed in a restricted region of the anterior neural plate at the bud stage, distinct from the expression of lim1 in the notochord and the pronephros and pronephric ducts. During the segmentation period, the lim6 gene started to be expressed in single cells in the spinal cord, followed by a gradually increasing wide-spread expression throughout the CNS. During this stage, lim1 mRNA disappeared in the notochord and pronephric ducts and was found in the pronephroi and single cells in the CNS. In 24 hr embryos, lim6 and lim1 were expressed in the fore-, mid-, and hindbrain and the spinal cord, except that lim1 mRNA was limited to two small domains in the telencephalon, whereas lim6 mRNA was widely expressed in this region. A comparison of expression of lim1 and lim6 and of the previously characterized lim5 show that, in spite of close sequence similarity, distinct expression patterns imply nonredundant functions for each member of this group of genes.

Retinoid X receptor (RXR) within the RXR-retinoic acid receptor heterodimer binds its ligand and enhances retinoid-dependent gene expression.

Retinoic acid receptor (RAR) and retinoid X receptor (RXR) form heterodimers and regulate retinoid-mediated gene expression. We studied binding of RXR- and RAR-selective ligands to the RXR-RAR heterodimer and subsequent transcription. In limited proteolysis analyses, both RXR and RAR in the heterodimer bound their respective ligands and underwent a conformational change in the presence of a retinoic acid-responsive element. In reporter analyses, the RAR ligand (but not the RXR ligand), when added singly, activated transcription, but coaddition of the two ligands led to synergistic activation of transcription. This activation required the AF-2 domain of both RXR and RAR. Genomic footprinting analysis was performed with P19 embryonal carcinoma cells, in which transcription of the RARbeta gene is induced upon retinoid addition. Paralleling the reporter activation data, only the RAR ligand induced in vivo occupancy of the RARbeta2 promoter when added singly. However, at suboptimal concentrations of RAR ligand, coaddition of the RXR ligand increased the stability of promoter occupancy. Thus, liganded RXR and RAR both participate in transcription. Finally, when these ligands were tested for teratogenic effects on zebra fish and Xenopus embryos, we found that coadministration of the RXR and RAR ligands caused more severe abnormalities in these embryos than either ligand alone, providing biological support for the synergistic action of the two ligands.

Oncocytic hepatocellular carcinoma with numerous globular hyaline bodies.

Two well circumscribed tumors, oncocytic and non-oncocytic, were removed from the non-cirrhotic liver of a 67 year old male. The large oncocytic tumor (OCT), occupying the entire left lobe, was multilobulated with focal coagulation necrosis and areas of hemorrhage. Light microscopy revealed that it consisted of exclusively large, granular oxyphilic cells with moderate nuclear atypia and occasional mitotic figures, which were trabecular and/or pseudoglandular in structure, but no lamellar fibrosis was seen. Characteristically, the OCT cells included numerous globular hyaline bodies (GHB) of various sizes which were stained red with acid fuchsin and deep blue or magenta with phosphotungstic acid hematoxylin (PTAH), but negative for periodic acid Schiff (PAS), orcein, rhodamine and Grimelius methods. Immunohistochemically, alpha-fetoprotein (AFP), alpha-1-antitrypsin, alpha-1-antichymotrypsin, fibrinogen and ferritin were all negative. On ultrastructural examination, tumor cells were mitochondria-rich, including electron dense, ovoid or polyhedral inclusions, with the delineated membrane identical with that of the GHB. In contrast, the small tumor in the right lobe (Segment 7) was a solid adenoma with no oncocytic transition. Based on these findings, it was postulated that OCT consists of heterogenous proliferation of mitochondria-rich hepatocytes which tend to induce lysosomal GHB closely associated with mitochondrial abnormalities.

Retinoid X receptor-selective ligands produce malformations in Xenopus embryos.

Retinoids exert pleiotropic effects on the development of vertebrates through the action of retinoic acid receptors (RAR) and retinoid X receptors (RXR). We have investigated the effect of synthetic retinoids selective for RXR and RAR on the development of Xenopus and zebrafish embryos. In Xenopus, both ligands selective for RAR and RXR caused striking malformations along the anterior-posterior axis, whereas in zebrafish only ligands specific for RAR caused embryonic malformations. In Xenopus, RAR- and RXR-selective ligands regulated the expression of the Xlim-1, gsc, and HoxA1 genes similarly as all-trans-retinoic acid. Nevertheless, RXR-selective ligands activated only an RXR responsive reporter but not an RAR responsive reporter introduced by microinjection into the Xenopus embryo, consistent with our failure to detect conversion of an RXR-selective ligand to different derivatives in the embryo. These results suggest that Xenopus embryos possess a unique response pathway in which liganded RXR can control gene expression. Our observations further illustrate the divergence in retinoid responsiveness between different vertebrate species.

The LIM class homeobox gene lim5: implied role in CNS patterning in Xenopus and zebrafish.

LIM homeobox genes are characterized by encoding proteins in which two cysteine-rich LIM domains are associated with a homeodomain. We report the isolation of a gene, named Xlim-5 in Xenopus and lim5 in the zebrafish, that is highly similar in sequence but quite distinct in expression pattern from the previously described Xlim-1/lim1 gene. In both species studied the lim5 gene is expressed in the entire ectoderm in the early gastrula embryo. The Xlim-5 gene is activated in a cell autonomous manner in ectodermal cells, and this activation is suppressed by the mesoderm inducer activin. During neurulation, expression of the lim5 gene in both the frog and fish embryo is rapidly restricted to an anterior region in the developing neural plate/keel. In the 2-day Xenopus and 24-hr zebrafish embryo, this region becomes more sharply defined, forming a strongly lim5-expressing domain in the diencephalon anterior to the midbrain-forebrain boundary. In addition, regions of less intense lim5 expression are seen in the zebrafish embryo in parts of the telencephalon, in the anterior diencephalon coincident with the postoptic commissure, and in restricted regions of the midbrain, hindbrain, and spinal cord. Expression in ventral forebrain is abolished from the 5-somite stage onward in cyclops mutant fish. These results imply a role for lim5 in the patterning of the nervous system, in particular in the early specification of the diencephalon.

LIM domain proteins.

The LIM domain is a cysteine-rich domain composed of 2 special zinc fingers that are joined by a 2-amino acid spacer. Some proteins are constituted by LIM domains only while others contain a variety of different functional domains. LIM proteins form a diverse group which includes transcription factors and cytoskeletal proteins. The primary role of LIM domains appears to be in protein-protein interaction, through the formation of dimers with identical or different LIM domains or by binding distinct proteins. In LIM homeodomain proteins, LIM domains seem to function as negative regulatory domains. LIM homeodomain proteins are involved in the control of cell lineage determination and the regulation of differentiation, and LIM-only proteins may have similar roles. LIM-only proteins are also implicated in the control of cell proliferation since several genes encoding such proteins are associated with oncogenic chromosome translocations. In analyzing sequence relationships between LIM domains we suggest that they may be arranged into 5 groups which appear to correlate with the structural and functional properties of the proteins containing these domains.