No description available.

In this thesis I describe a novel role for the Endoplasmic Reticulum (ER) in spindle positioning and stability in the early <i>C. elegans</i> embryo. Within the embryo the ER is intimately associated with the microtubule cytoskeleton and I show that the disruption of this association leads to instability of nuclear and spindle positioning. I first describe RNA interference screens that I performed to identify new genes involved in the regulation of spindle positioning. This led to the identification of three ER resident proteins that comprise <i>C. elegans</i> homologs of the Oligosaccharyl Transferase (OST) complex. Through characterisation of the spindle phenotypes induced by knockdown of OST components, I show that these proteins are similar to a diverse collection of mutants affecting the ER, such as the trans-membrane ER protein, <i>ooc-3</i> and the <i>npp</i> (nuclear pore protein) genes. I found that identical nuclear and spindle instability phenotypes could be replicated by using the drug Brefeldin A to rapidly disrupt the ER, suggesting the ER might have a structural role in stabilizing nuclear and spindle position. I also provide evidence suggesting that the association between the ER and the microtubule cytoskeleton involves Clathrin heavy-chain (CHC). CHC may form part of the link between the ER and microtubule networks. Overall my results describe a role for the reticular ER network in stabilizing the assembling mitotic spindle and a potential new role for clathrin in linking the ER and microtubules. I addition, I show that Early Endosomes in the <i>C. elegans</i> embryo are polarized and asymmetrically segregated with the actin-myosin cytoskeleton, and present work optimizing the creation of transgenic lines by microparticle bombardment.

Morphogen gradients have long been proposed as a mechanism of developmental patterning. This mechanism requires an individual cell to be able to measure at least two concentration thresholds of a single signalling molecule and to make at least three different cell fate decisions. The mechanism by which cells are able to make these quantitative decisions is not at all well understood. One example of a morphogen response is that of the induction of mesoderm in <I>Xenopus</I> animal cap cells by activin. The grade of mesoderm induced is dependent on the concentration of activin. In this dissertation the induction of mesoderm in animal cap cells by activin is used as a model system for the investigation of how cells determine their response to a morphogen gradient. First, the <I>in vivo</I> role of activin is demonstrated with the use of a novel dominant negative receptor. Second, overexpression of cloned activin receptors is used to show that cells can use a single affinity receptor to generate the different responses to a morphogen, rather than by using different affinity receptors. Third, a new assay has been developed in order to make quantitative measurements for how cells respond to a morphogen. Using this assay it has been shown that cells first respond to activin when only 2% of their receptors are occupied and can switch response when 6% of their receptors are filled. It has also been shown that cells measure the absolute number of their occupied receptors and not the ratio of occupied to unoccupied. Fourth, potential signal transduction components, downstream of the activin receptor, are tested for their involvement in activin signal transduction from receptors to the nucleus. Fifth, the involvement of another signalling pathway, the Wnt pathway, is investigated in the response to activin. It is shown that there is cooperativity between activin signalling and Wnt signalling for the induction of some, but not all, responses to activin and that this cooperativity is reciprocal.

The imprinting centre 2 (IC2) on mouse distal chromosome 7 is part of a cluster of genes of which ten are currently known to be monoallelically expressed. All genes in the cluster are paternally silenced, with the exception of the non-coding RNA gene <i>Kcnqlot1</i>, which is maternally silenced. In this study I have been interested to understand how and when allelic expression and allelic histone marks are established in the IC2 cluster, and how they are erased in germ cells. I have used trophoblast stem (TS) and trophoblast giant cells as a model system of placental development to determine lineage specific histone modifications and gene expression. Using this model system, I was able to show that placental allele specific histone modifications are established after the blastocyst stage. Using a mutant for an oocyte-specific form of Dnmt1 (Dnmtlo), I have analysed imprinted expression and levels of transcription of genes in the cluster. This has allowed me to identify a particular time point during development from which expression of <i>Kcnqlot1</i> can target gene silencing and repressive histone marks in <i>cis </i>to the imprinted genes in the cluster. Finally, I was interested to see if histone marks associated with imprinting are removed during the process of imprint erasure in primordial germ cells (PGCs). To this end I have developed a sensitive chromatin immunoprecipitation (ChIP) technique that can be used in combination with single strand conformational polymorphism (SSCP). This method enabled me to investigate allele specific histone modifications in small numbers of cells (10³-10⁴). This has allowed me to show that some allele specific histone modifications remain in primordial germ cells (PGCs) at the IC2 cluster after DNA methylation has been erased at E13.5.

We know little about how motor circuits are assembled to produce appropriately coordinated behaviours. <i>Drosophila</i> is an excellent system in which to investigate this topic. I therefore developed an imaging technique which enables contraction-relaxation cycles in individual muscles in intact, freely behaving embryos and larvae, to be visualised and documented for quantitative analysis. I have used this technique to document the ventral longitudinal muscle contractions in the embryo. I found that movement begins two thirds of the way through embryonic development. Embryos then go through a sequence of distinct phases, characterised by particular patterns of muscle contractions, before the first sequences resembling larval crawling emerge, some four hours after the first movements. Interestingly, the earliest movements in <i>Drosophila</i> embryos occur before neurons become capable of firing action potentials. I was able to prove that these movements are myogenic in origin. I next pinpointed when the transition to neurogenic movement occurs. Then, by disabling the sensory systems, I demonstrated that the onset of neurogenic movement is not a reflex response to sensory stimulation. On the other hand, I found evidence that spontaneous firing in the nervous system, as neurons acquire their electrical properties, could be responsible for the first neurally-driven muscle contractions. The first motor output in embryos is a disorganised burst of activity, which bears no resemblance to larval crawling. Over the next hour, I find that motor output occurs as repeated bursts, separated by log period of inactivity. Within these bursts increasingly coordinated patterns of muscle contractions are seen, eventually culminating in the first complete waves of peristaltic crawling. These bursts of activity, like the first motor output, do not require sensory input to occur. Instead, I find evidence that they are the consequence of a network which is spontaneously active and subject to activity-dependent depression. Intrinsic episodic activity of this kind is also a feature of maturing networks in vertebrates.

To determine the role of BMP signalling in neural development I chose to make use of the REMI transgenic technique. I cloned the <I>Xenopus laevis</I> Pax-6 promoter and used it as a tool to drive the expression of BMP4 and the ΔXBMPR (a dominant negative BMP receptor) in the anterior neural plate, from the end of gastrulation. The main conclusion I draw from the misexpression of BMP4 in transgenic <I>Xenopus </I>F0 embryos is that developing neural tissue is still competent to respond to BMP4 signalling at neural plate stages. Furthermore, BMP4 continues to function as an inhibitor of the majority of neural genes expressed into the neural plate stages of development, though not all neural genes are affected by the ectopic expression of BMP4 equally. In addition, I present evidence to suggest that the inhibition of BMP signalling by the misexpression of a truncated form of the BMP receptor in transgenic F0 embryos perturbs neural development. I also present evidence that BMP4 signalling is required for specifying the anterior border of the neural plate in <I>Xenopus. </I>I show that BMP4 and BMP7 are present in the prechordal mesoderm, a region underlying the midline of the anterior neural plate, which could receive the signalling molecules. In sum, I suggest that BMP signalling is involved in the fine-tuning of patterning and development of the neural plate. To further our understanding of the role BMPs play in neural development and to advance our technical abilities to misexpress factors, I have developed the Gal4-UAS system in <I>Xenopus. </I>I show that given factors can be transactivated in a precise manner, by the cross-fertilisation of transgenic founder lines. Additionally, I show that transactivation by Gal4 can result in a predicted phenotype for the effector gene in question.

No description available.

Matching of neurons and glia is necessary for central nervous system development. I aimed to elucidate the pattern of longitudinal glial (LG) proliferation and its potential control by neurons. I characterised the LG cell lineage by wide-field and confocal microscopy using immunohistochemistry to monitor the distribution of phosphoHistone H3 ad Cyclins A, B and E. This lineage divides first into four and then six, finally resulting in nine to eleven glia. Neuronal ablation using the Ga14/UAS system resulted in an early reduction and a late increase in the proliferation of the subset of LG which express the gene <i>prospero</i>. Despite the symmetric cell divisions of the lineage wild-type and mutant analysis revealed that Miranda segregates Prospero and Numb asymmetrically during early LG divisions. The unequal distribution of Prospero segregates glial fate, causing the activation of the MAPKinase pathway by Vein in two of four glia, to coordinate glial cell division with axon guidance. In fact, <i>prospero</i> mutations also affect cell division. Normally the first divisions are rapid and the LG pause at four cells. At this point Prospero regulates Cyclin E introducing the first G1. Prospero regulates Notch in the LG and Notch also regulates Prospero in antagonism to Numb. This feed back loop results in the segregation of Prospero to a subset of six of the final LG. I show that the subset of six LG expressing <i>Notch</i> and <i>prospero</i> are arrested in G1 as opposed to G0. Target expression of <i>cyclin E</i> to the LG can only induce entry into S phase, monitored with BrdU, of the LG that express <i>prospero</i>. Non-Prospero longitudinal glia express <i>dacapo</i> triggering cell cycle exit and terminal differentiation. Ectopic Dacapo and Prospero demonstrate that these two proteins antagonise each other.

In this study, I have shown that XStau 1 and 2 are expressed throughout oogenesis, embryogenesis and in the adult <i>Xenopus</i>. Immunostaining and confocal microscopy has demonstrated that both XStaul 1 and 2 localise to the vegetal cortex during oogenesis and colocalise with the ER but not the MC, at a time consistent with a role in the vegetal localisation of late pathway mRNAs. Biochemical analysis has shown that on meiotic maturation XStau 1 and 2 become phosphorylated and that this phosphorylation coincides with their rearrangement at the vegetal cortex in mature eggs, where they no longer co-localise with ER. Using gel filtration separation, I have shown that XStaufen 1 and 2 are part of large RNP complexes with a minor proportion present in smaller particles. RNA-independent ligands have been identified by co-immunoprecipitation with XStaul antibody, and have been sequenced by mass spectrometry and their identities confirmed by Western blot analysis and in <i>in vitro</i> binding assays. Binding patterns identified by immunoprecipitation did not include proteins with a role in mRNA localisation, but instead suggest a role for XStau1 in translational control and included CPEB, Xp54, p100, ePABP and XUpf1. This was confirmed in MS2 tethering assays where XStau1 but not XStau2 was shown to repress the translation, but not stability, of a firefly luciferase reporter mRNA 1.5 to 2 fold. These studies extend our understanding of the vertebrate Staufen proteins in the control of gene expression and indicate that in <i>Xenopus</i> oocytes, XStau1 mediates translational repression.

To understand vertebrate brain development we must determine how the cells of the simple, two-dimensional anterior neural plate are signalled to undergo complex reorganisation in to the elaborately structured, three-dimensional forebrain. To address this, I have collected time-lapse 3D movies of brain development in the optically transparent zebrafish embryo. I used these movies to accurately track the paths of many hundreds of cells representing the important cell territories of the brain, and so constructed dynamic, high-resolution fate maps of their transformations. My findings in wild-type embryos challenge the accepted model of forebrain folding. The resolution of bilateral eyes is not, as believed, achieved by the anterior-ward progression of the ventral diencephalon (hypothalamus) in the form of a neural keel, through the initially monocular eye field. The neural keel is actually formed by the hypothalamus moving deep than anterior-wards, subducting but not reshaping, more anterior progenitors of the dorsal forebrain (telencephalon) and medial eye field. In a novel observation, I can show that the separation of bilateral eyes occurs subsequent to this, coincident with the anterior-ward movement of dorsal diencephalic cells along the neural tube. I then used this new detailed model of normal development to address the causes of cyclopia – one consequence of the common human neural tube defect, holoprosencephaly (HPE). My analysis of time-lapse movies of animals with reduced <i>Nodal </i>or <i>Wnt11</i> signalling – modelling HPE, identified two distinct morphogenetic events, whose failure can cause cyclopia. Under conditions of reduced <i>Nodal</i> signalling, cyclopia occurs because the cells moving deep to form the ventral midline are erroneously derived from eye territories anterior to the normal presumptive hypothalamus. In contrast, in mutant embryos with reduced <i>Wnt11</i> signalling, cyclopia<i> </i>results from a reduced anterior-ward extension and movement of the future dorsal diencephalons, causing incomplete evagination of the optic lobes.