The induction of senescence in response to persistent stress induces major phenotypic changes in senescent cells, including the secretion of a host of inflammatory factors and reactive oxygen species. Recent evidence has implicated senescent cells in the diseases of ageing and cancer; however, the mechanism by which this occurs is still unknown. This thesis uses a reporter cell line with cells expressing a fluorescent conjugate that allows real time live cell imaging of a sub set of cells within a co-culture, to provide the first evidence that senescent cells can induce a DNA damage response in healthy cells, and thus implicates a potential mechanism by which senescent cells could non-autonomously contribute to the ageing process. The use of specific inhibitors, stimulation, and targeted repression indicate that gap junctions, reactive oxygen species, p38, mTOR and NF-κB all play a key role in this observed bystander effect of senescent cells, and offer potential targets for therapies designed to reduce the damaging effects of senescent cells.

The pharyngeal apparatus is comprised of a series of pharyngeal arches that are defined by the evagination of endodermal pouches toward the ectoderm. Whilst the transcription factor T‐box 1 (Tbx1) and the Fibroblast growth factor 8 (Fgf8) are both required for caudal pouch formation, a role for these factors in the endoderm during pouch morphogenesis has not been confirmed. The observation that Fgf8 expression is lost in Tbx1 homozygous null mutant embryos has lead to the hypothesis that FGF8 functions directly downstream of TBX1 during pouch formation. To test this hypothesis the Sox17‐2A­‐icre line was used to delete Fgf8, Fibroblast growth factor receptor 1 (Fgfr1) and Fibroblast growth factor receptor 2 (Fgfr2) and Tbx1 in the endoderm, and analyse the development of the pouches and their derivatives. In all embryos with an endoderm specific deletion of Tbx1, the caudal PA remained unsegmented as caudal pouches 3 and 4 do not form. The deletion of endodermal Tbx1 severely reduced the expression of Fgf8 in the endoderm; however, the expression of Fibroblast Growth Factor (FGF) signalling readouts was maintained. All rostral and caudal pouches are present in embryos with an endoderm specific deletion of Fgf8, indicating that the loss of Fgf8 alone from this epithelium is not sufficient to disrupt the process of pouch evagination. The compound deletion of Fgfr1 and Fgfr2 from the endoderm did not prevent pouch evagination but did cause 3rd pouch hypoplasia and the rostral pouches to fuse. These data indicate that FGF signalling downstream of Fgfr1 and Fgfr2 and Tbx1 are important for pouch formation but that they are likely to function in parallel pathways within the endoderm. Further data suggests that TBX1 may be acting in an FGF independent manner to control the polarity of actin within the endodermal epithelia. Overall this data reveals that both cell signalling and actin polarity must be tightly regulated within the endoderm for the pharyngeal pouches to form correctly.

The mammalian Rab family consists of between 60-70 members, making it the largest sub family of the Ras superfamily. Rabs are responsible for vesicle trafficking within cells, acting as molecular switches cycling between the GDP inactive and GTP bound active forms. Far from being just cellular housekeeping genes, these genes have been shown to have specific functions which, when disrupted, can lead to clinical disorders and interesting developmental defects. This thesis therefore seeks to investigate this interesting family of genes and their roles in zebrafish development. Using antisense morpholino oligonuleotides in a loss of function screen, this thesis identifies the function of 13 zebrafish rabs. Three of these, rabla3, rab3cl and rab28 have specific and interesting phenotypes, with pigmentation defects seen in rabla3 and rab3cl and behavioural defects seen in rab28. In particular, the pigmentation defect in rab3cl resulted in the discovery that the embryos were blind. This thesis also shows an essential role for rab5a2 in zebrafish development and Nodal signalling. Disruption of this rab causes a dramatic early phenotype, with 100% mortality in embryos prior to 24 hours post fertilization. rab5a2 morpholino injected embryos show no visable organizer and have reduced nodal target gene expression. Overexpression of rab5a2 shows embryos with additional expression of Nodal target genes no tail and goosecoid in the animal pole of the embryos but not the dorsal marker chd. Microarray analysis of rab5a2 morpholino injected embryos showed reduction and upregulation of expression of many genes involved in dorsal ventral patterning. This suggests a complex role for rab5a2 in patterning the early embryo, as both dorsalizing and ventralizing genes such as chd, bmp4 and wnt8 were down regulated while, ventralizing genes such as bmp2b were upregulated.

In 1924, Spemann and Mangold (Spemann and Mangold, 1924) demonstrated the instructive effect of the vertebrate organiser, located in the dorsal lip of the blastopore in amphibia, in the acquisition of neural character during early development. This discovery led to much interest in the process of neural induction. More recently, a "default model" was proposed to account for neural induction, which postulates that ectoderm cells acquire a neural fate autonomously, if they receive no signal. However, this neural fate is inhibited by BMPs, which induce epidermis. The organiser in turn secretes BMP antagonists, which release neighbouring cells from inhibition. However, other results suggested that neural induction is more complex, comprising several sequential and/or parallel signals, which cooperate to induce neural fates. One such finding was that chick epiblast cells do not respond to BMP antagonists unless they are first exposed to an organiser for at least 5 hours, suggesting that other signals are required upstream of BMP inhibition. To identify the differences between cells that have or have not received such signals, a differential screen was performed. Among 15 genes isolated, only 3 correspond to previously known genes: Defender Against Cell Death (Dad1), Polyubiquitin II (Ubll) and Heavy chain ferritin (hcf). All 3 of these had been implicated in Programmed Cell Death (PCD). This project was designed to investigate whether PCD is important in neural induction and to study the role of these three proteins in both processes. First, the distribution of apoptotic cells was examined at different stages of development using TUNEL staining. It was found that PCD is random at first, but by the time the neural plate starts to be established PCD becomes concentrated at its lateral and anterior border. The expression patterns of the three genes were studied: all three are expressed in the neural plate around the time of neural induction, however they are not exclusive markers of the neural plate. The ability of the organiser (Hensen's node) to induce their expression within 5 hours was confirmed, and the ability of FGF signals to mimic this was also tested. The effects of the overexpression of the genes outside their normal expression domain were then examined, scoring both for changes in the distribution of PCD and for the expression of markers for neural, epidermal and border (prospective neural crest) territories. This revealed that overexpression of hcf leads to an increase in the number of cells undergoing PCD, but this effect is not cell autonomous. Finally, the expression patterns of the pro-apoptotic gene Cas9 and of the death effector Cas3 were tested and compared to the apoptotic patterns. An increase in PCD was observed in the region where expression of Cas3 and Cas9 overlaps and which is free from Dad1, consistent with the notion of an amplification loop between Caspase 3 and Caspase9, and with a possible involvement of DAD1 in the regulation of this process.

The coordinated growth of the conceptus is sustained by reciprocal signalling between the epiblast and the extraembryonic ectoderm (ExE). In mice, FGF4 produced by the epiblast promotes Cdx2 expression in the trophoblast stem cell (TSC) niche located in the ExE. Cdx2, Eomes and EIf-5 expression in the ExE constitute an auto-regulatory circuit that maintains the mouse TSC gene network. Cells in the ExE, in turn secrete BMP4, which is a critical mesoderm determinant. The ExE however, is absent in non-rodent embryos, raising the question whether the cross-talk between embryonic and extraembryonic domains is conserved in mammals. Pig embryos represent "mammotypical" embryos in that the flat epiblast is surrounded by trophectoderm (TE). The pig TE undergoes a remarkable elongation, however it is not known whether the epiblast regulates this process and whether a TSC niche supports this remarkable growth.

Periconceptional deficiencies in one-carbon (l-C) metabolites (e.g. folate, vitamin Bl2 and methionine) influence parental fertility and offspring health. Knowledge of l-C metabolism is limited to the liver where species-specific differences have been reported. Functionality in the ovary, germ cells and preimplantation embryo is poorly understood. The multitude of allosteric interactions involved in l-C metabolism makes it challenging to understand the principles by which this pathway is regulated. Mathematical models of this cycle, based on published enzyme kinetics and regulatory mechanisms within the liver, exist and can address this problem in part. We recently found that transcripts for specific l-C metabolism enzymes were either absent or poorly expressed in the bovine ovary and preimplantation embryo. Extending these initial observations, to include the sheep, pig, rat and human hepatocyte and granulosa cell lines, this thesis sought to increase the understanding of the regulation of l-C metabolism in the ovary and preimplantation embryo by combining experimental data with theoretical models. Experimental investigations at transcript, protein and enzyme activity levels revealed that l-C metabolism differs between the ovary and liver and also between the ovaries of different species. Variations considered primarily relate to the presence of MATII and the absence of BHMT within the ovary. Consequently, a mathematical model of hepatic methionine metabolism was modified to render it more representative for ovarian cells, particularly in the cow. Model predictions indicated that the ovary regulates methionine metabolism over a much narrower range of methionine input than the liver. Within this narrow range S-adenosylmethionine (SAM)-mediated transmethylation reactions are predicted to vary. Model predictions were tested in bovine granulosa cells cultured under various methionine concentrations (i.e. 0 to 500 J!M). Methionine adenosyltransferase 2A (MAT2A) transcript levels, de novo synthesis of SA M and the SAM:SAH (S-adenosylhomocysteine) ratio decreased with increasing methionine within the physiological range (i.e. 10 to 50 /!M). Homocysteine concentrations in spent culture media increased within the same range whereas progesterone synthesis and DNA methylation were unaffected by methionine dose. Results herein promote and support the importance of an optimal parental methionine status during the periconceptional period.

A universally conserved process called branching morphogenesis generates the final tree-like structure that is a characteristic of the salivary glands and other organs including the lungs, kidneys and mammary glands. Branching morphogenesis is the mechanism by which a single epithelial bud undergoes repetitive cycles of bifurcation in order to maximise its surface area within a confined space. During salivary gland development, alongside branching morphogenesis occurs lumen formation, which corresponds to the cavitation of the inside of the epithelial branches to form ducts. Branching and canalization take place in a controlled manner, so that branching is active at the distal ends of epithelial branches while lumen formation initiates at the proximal ends, and spreads distally. We looked at the function of Wnt signaling during embryonic development of the submandibular and sublingual salivary glands and found that FGF signaling acts upstream of Wnt signaling and that Wnt and FGF signals coordinate branching morphogenesis and lumen formation. To locate the sites of active Wnt/p-catenin signaling, we used Axin2LacZ mice and found that the pathway is active from early stages of salivary gland development in the mesenchyme and at later stages in the ductal epithelium corresponding to the time of branching morphogenesis and lumen formation, respectively. The use of an array of ex vivo gain and loss of function experiments revealed that Wnt/b-catenin signaling exerts an inhibitory effect on branching morphogenesis and that tight regulation of this pathway is critical for correct salivary gland development. We also present data showing critical interactions between the Wnt and FGF signaling pathways, whereby the endbuds remain devoid of Wnt/p-catenin signaling, a hallmark of ductal structures, through FGF mediated inhibition of this pathway.

I used the zebrafish embryo to investigate the morphogenesis of the developing somite and the behaviour of its different compartments in vivo via 3D confocal time-lapse microscopy. I demonstrated that the epaxial and hypaxial somite regions show distinct morphogenesis that lead to the chevron-shape of the somite. Evidence is presented that shows that the somite volume does not significantly change during its morphological transition, which suggests that growth does not contribute to somite shape morphogenesis. In vivo cell tracking revealed that cellular dynamics during somite morphogenesis constitute a potential force, capable of driving somite shape changes. Tracked anterior border cells (abc) were observed migrating to the lateral somite surface were they are though to form the dermomyotome. The observation of different abc migration behaviours led to the hypothesis that abc's could be a heterogeneous population of cells which segregate into different behaviour groups, depending on their later fate. Furthermore, slow muscle fibres were observed in vivo to migrate collectively as a chain of slow muscle cells that reaches from somite to somite along the anteroposterior body axis. The slow muscle cells maintain interdigitating contacts to adjacent slow muscle fibres from neighbouring somites and are polarized anteroposteriorly and mediolaterally and gravitate towards the anterior somite half during their migration. Somite borders are shown to be not necessary for slow muscle fibre migration but for slow muscle fibre polarisation and maintenance of the specific cell alignment during migration. Both, abc migration and slow muscle fibre migration are discussed as contributing factors of the observed shape change of the somite.

The mesencephalic trigeminal nucleus (MTN) forms part of the monosynaptic trigeminal circuit and is essential for eating and suckling in mammals. Little is known about how the MTN forms. For this thesis I aimed to elucidate the molecular and cellular basis of MTN development. I also aimed to investigate the role of the Fgf and Wnt signalling pathways in MTN development. The zebrafish was used as a model organism to investigate these aims. Putative MTN cells in zebrafish larvae were retrogradely labelled by applying Dil to the adductor mandibulae, a jaw closing muscle. Labelled axons projected from muscles via the trigeminal ganglion to cell bodies in the dorsal anterior mesencephalon, suggesting that the MTN does innervate jaw muscles in teleosts, contrary to previous studies. Molecular characterisation of the MTN in zebrafish revealed a similar expression profile as the mammalian MTN. To investigate whether MTN neurons are neural crest-derived, the neural crest was ablated, which resulted in an increase in MTN number. This suggests that the neural crest may play an inhibitory role in MTN development contradictory to previous studies in chick that suggested MTN neurons are derived from the neural crest. The role of the Fgf and Wnt signalling pathways was investigated by analysis of mutants and drug treatments where the pathways had been genetically or chemically manipulated. Down-regulation of Fgf signalling showed an increase in MTN neuron numbers, suggesting that Fgf signalling from the midbrain/hindbrain boundary inhibits development of the MTN in the midbrain. When the Wnt pathway was up-regulated there was also an increase in MTN neuron number. Based on the results from these experiments a model is proposed, in which Fgf signalling regulates the formation of MTN neurons in a spatial and temporal manner, and Wnt signals from the dorsal roof plate induce the proliferation of MTN precursor cells.

Mechanisms that result in directional tissue growth include cell proliferation, orientated cell division, an increase in cell size, a change in the cell density and cell death. These were investigated in the development of the mouse palatal shelves between stages E11.5 – E15.5. The aim was to develop and apply image processing tools to automatically analyse images of fixed specimens to reveal cellular behaviours that can be applied to understand directional growth and diagnosing the mechanism of its failure in mutant tissue in the palate and more generally. Techniques included immunohistochemistry after frozen or paraffin sectioning and imaging on the confocal microscope. Analysis of images was done using programs such as ImageJ, Volocity and R. This analysis was carried out on CD-1 wild-type mice to provide a thorough understanding of the mechanisms of cellular morphogenesis throughout normal palate development. It was then performed on mutant Tbx-1, Msx-1 and Wnt1-Cre;ERK2fl/fl mice all of which have a cleft palate phenotype to demonstrate how a better understanding of the links between genetic mutations, tissue phenotypes and cellular disruption can be gained.