In most animals, primordial germ cell (PGC) specification and development depend on maternally provided cytoplasmic determinants that constitute the so-called germ plasm. Little is known about the role of germ plasm in vertebrate germ cell development, and its molecular mode of action remains elusive. While PGC specification in mammals occurs via different mechanisms, several germ plasm components required for early PGC development in lower organisms are expressed in mammalian germ cells after their migration to the gonad and are involved in gametogenesis. Here we show that the RNA of dead end, encoding a novel putative RNA binding protein, is a component of the germ plasm in zebrafish and is specifically expressed in PGCs throughout embryogenesis; Dead End protein is localized to perinuclear germ granules within PGCs. Knockdown of dead end blocks confinement of PGCs to the deep blastoderm shortly after their specification and results in failure of PGCs to exhibit motile behavior and to actively migrate thereafter. PGCs subsequently die, while somatic development is not effected. We have identified dead end orthologs in other vertebrates including Xenopus, mouse, and chick, where they are expressed in germ plasm and germ-line cells, suggesting a role in germ-line development in these organisms as well.

The molecular and cellular mechanisms governing cell motility and directed migration in response to the chemokine SDF-1 are largely unknown. Here, we demonstrate that zebrafish primordial germ cells whose migration is guided by SDF-1 generate bleb-like protrusions that are powered by cytoplasmic flow. Protrusions are formed at sites of higher levels of free calcium where activation of myosin contraction occurs. Separation of the acto-myosin cortex from the plasma membrane at these sites is followed by a flow of cytoplasm into the forming bleb. We propose that polarized activation of the receptor CXCR4 leads to a rise in free calcium that in turn activates myosin contraction in the part of the cell responding to higher levels of the ligand SDF-1. The biased formation of new protrusions in a particular region of the cell in response to SDF-1 defines the leading edge and the direction of cell migration.

The lozenge (lz) gene encodes a transcription factor involved in prepatterning photoreceptor precursors in the developing Drosophila eye. The central region of the predicted Lz protein product is homologous to AML1, a transcription factor associated with human leukemias, and to the Drosophila protein Runt. We show here that Lz plays a crucial role in governing the fate of two groups of cells that are born in a single round of mitosis in the larval eye disc. Lz helps define a subset of these cells as an equipotential group that is competent to respond to the Sevenless developmental signal. This is achieved by negative regulation of seven-up, a member of the steroid hormone receptor superfamily in these cells. In contrast, in a second group of cells, the Lz protein confers proper photoreceptor identity by positively regulating the homeo box gene Bar. Additionally, our genetic analysis suggests that Lz interacts with the Ras pathway to determine photoreceptor cell fate. This study suggests that the strategies involved in cell fate determination in the Drosophila eye are remarkably similar to those utilized during vertebrate hematopoietic development and require the coordinate action of growth factor and AML1-like pathways.

In the wild-type brain, the Drosophila classic cadherin DE-cadherin is expressed globally by postembryonic neuroblasts and their lineages ("secondary lineages"), as well as glial cells. To address the role of DE-cadherin in the larval brain, we took advantage of the dominant-negative DE-cad(ex) construct, the expression of which was directed to neurons, glial cells, or both. Global expression of DE-cad(ex) driven by a heat pulse during the early second instar resulted in a severe phenotype that included deficits in neural proliferation. Neuroblasts appeared in approximately normal numbers but had highly reduced mitotic activity. When the DE-cad(ex) construct was driven by the glial-specific driver gcm-Gal4, the effect of DE-cad(ex) on neuroblast proliferation could be replicated, which indicates that DE-cadherin acts in glial cells to promote proliferation of neuroblasts. Expression of DE-cad(ex) in neurons, cortex glia, or both results in abnormalities in cortex layering and in trajectories of secondary axons. In the wild-type brain, neuroblasts and neurons generated at different time points are arranged concentrically around the neuropile, with the DE-cadherin-positive neuroblasts and young secondary neurons at the surface, followed by older secondary neurons and primary neurons. Axons of secondary lineages follow a straight radial course toward the neuropile. Processes of glial cells located in the cortex form a scaffold, called trophospongium, that enwraps neuroblasts and neurons. Expression of DE-cad(ex) in neurons, cortex glia, or both disrupted the regular placement of neuroblasts and secondary neurons and resulted in abnormal trajectories of cell body fiber tracts. We conclude that DE-cadherin plays a pivotal role in larval brain proliferation, brain cortex morphogenesis, and axon growth.

Background: Zebrafish germ cells contain granular-like structures, organized around the cell nucleus. These structures share common features with polar granules in Drosophila, germinal granules in Xenopus and chromatoid bodies in mice germ cells, such as the localization of the zebrafish Vasa, Piwi and Nanos proteins, among others. Little is known about the structure of these granules as well as their segregation in mitosis during early germ-cell development.

During their migration, zebrafish primordial germ cells (PGCs) rely on directional cues provided by the chemokine SDF-1a, whose receptor is CXCR4b. The molecular mechanisms whereby CXCR4b activation is interpreted intracellularly into directional migration are not known. Here we investigate the role of two important biochemical pathways - G-protein-dependent and phosphoinositide 3-kinase (PI3K)-dependent signaling - in directing PGC migration in zebrafish. We show that G proteins of the Gi family are essential for directional migration but not for PGC motility. Inhibition of PI3K signaling in PGCs slows down their migration and leads to abnormal cell morphology as well as to reduced stability of filopodia. Invariably, during directed PGC migration, the distribution of the products of PI3K activity - phosphatidylinositol (3,4,5)-trisphosphate [PtdIns(3,4,5)P3] and/or phosphatidylinositol (3,4)bisphosphate [PtdIns(3,4)P2] - is not polarized, and reducing the level of these 3-phosphoinositides does not affect the ability of PGCs to migrate directionally. We therefore conclude that Gi-dependent signaling is essential for directional migration, whereas the PI3K pathway is important for the actual motility of PGCs.

The Drosophila E-cadherin homolog, DE-cadherin, is expressed postembryonically by brain neuroblasts and their lineages of neurons ("secondary lineages"). DE-cadherin appears in neuroblasts as soon as they can be identified by their increase in size and then remains expressed uninterruptedly throughout larval life. DE-cadherin remains transiently expressed in the cell bodies and axons of neurons produced by neuroblast proliferation. In general, axons of neurons belonging to one lineage form tight bundles. The trajectories of these bundles are correlated with the location of the neuronal lineages to which they belong. Thus, axon bundles of lineages that are neighbors in the cortex travel parallel to each other and reach the neuropile at similar positions. It is, therefore, possible to assign coherent groups of neuroblasts and their lineages to the individual neuropile compartments and long axon tracts introduced in the accompanying articles (Nassif et al. [2003] J Comp Neurol 455:417-434; Younossi-Hartenstein et al. [2003] J Comp Neurol 455:435-450). In this study, we have reconstructed the pattern of secondary lineages and their projection in relationship to the compartments and Fasciclin II-positive long axon tracts. Based on topology and axonal trajectory, the lineages of the central brain can be subdivided into 11 groups that can be followed throughout successive larval stages. The map of larval lineages and their axonal projection will be important for future studies on postembryonic neurogenesis in Drosophila. It also lays a groundwork for investigating the role of DE-cadherin in larval brain development.

The Drosophila E-cadherin homolog, DE-cadherin, is expressed and required in all epithelial tissues throughout embryogenesis. Due to a strong maternal component of DE-cadherin, its early function during embryogenesis has remained elusive. The expression of a dominant negative DE-cadherin construct (UAS-DE-cad(ex)) using maternally active driver lines allowed us to analyze the requirements for DE-cadherin during this early phase of development. Maternally expressed DE-cad(ex) result in phenotype with variable expressivity. Most severely affected embryos have abnormalities in epithelialization of the blastoderm, resulting in loss of the blastodermal cells' apico-basal polarity and monolayered structure. Another phenotypic class forms a rather normal blastoderm, but shows abnormalities in proliferation and morphogenetic movements during gastrulation and neurulation. Mitosis of the mesoderm occurs prematurely before invagination, and proliferation in the ectoderm, normally a highly ordered process, occurs in a random pattern. Mitotic spindles of ectodermal cells, normally aligned horizontally, frequently occurred vertically or at an oblique angle. This finding further supports recent findings indicating that, in the wild-type ectoderm, the zonula adherens is required for the horizontal orientation of mitotic spindles. Proliferation defects in DE-cad(ex)-expressing embryos are accompanied by the loss of epithelial structure of ectoderm and neuroectoderm. These germ layers form irregular double or triple layers of rounded cells that lack zonula adherens. In the multilayered neuroectoderm, epidermal precursors, neuroblasts and ganglion mother cells occurred intermingled, attesting to the pivotal role of DE-cadherin in delamination and polarized division of neuroblasts. By contrast, the overall number and spacing of neuroblasts was grossly normal, indicating that DE-cadherin-mediated adhesion is less important for cell-cell interaction controlling the ratio of epidermal vs. neural progenitors.