Proteomic analyses of centrosomes from diverse species are helping to identify conserved components of these organelles. Amongst these studies, a group of novel proteins has been identified and postulated as candidate core centriolar components, dubbed Poc proteins, for proteome of centriole. Meanwhile, studies on the Xenopus germ plasm protein, Xpat, led to the identification of an interacting protein, named Pix for protein that interacts with Xpat. Later sequence comparison revealed that Poc1 and Pix are the same protein and will here be referred to as Pix. Pix proteins localize to centrosomes and spindle poles in human cells and the basal bodies of Chlamydomonas and Tetrahymena. It has been suggested that Pix proteins are required for centriole duplication and length control, as well as potentially ciliogenesis. Pix proteins have also been localized to mitochondria in human cells where they might act as molecular adaptors to anchor microtubules to mitochondria. Human cells encode two Pix proteins and the aim of this thesis was to investigate the cell cycle-dependent regulation of Pix1 and Pix2 in human cells. For this purpose, we first generated polyclonal rabbit antibodies that were specific for either Pix1 or Pix2. We then showed for the first time that both Pix isoforms localize to centrosomes throughout the cell cycle and independently from one another. Using the antibodies, we also confirmed that Pix1, but not obviously Pix2, is not an intrinsic mitochondrial protein but localizes to the surface of mitochondria. Through a proteomic approach, we identified two molecular chaperons that potentially interact with Pix proteins, HSP90 and TCP1, and demonstrated that Pix proteins can form dimers. In addition, we produced the first experimental evidence that human Pix proteins are alternatively spliced, suggesting that additional, undiscovered Pix isoforms might be expressed in human cells. Finally, we found that Pix1 is phosphorylated in a mitosis-specific manner, and that this is potentially regulated by Cdk1. Thus, we propose that Pix proteins have a specific and previously unidentified role in human mitotic cell division.

Circulating cell-derived microparticles, released from cells during activation and apoptosis, are involved in inflammation, coagulation and endothelial dysfunction, all important processes in the development of cardiovascular disease. This project aimed to adapt and validate a flow cytometric assay to measure microparticles derived from various cell types, and to utilise this assay for the investigation of microparticles in healthy individuals and patients with cardiovascular-associated diseases. A lack of standardisation of pre-analytical variables has impeded the study of microparticles. Pre-analytical variables were analysed, and small changes in methodology were found to have a large impact on microparticle levels detected. Functional microparticle assays were also investigated, and results from these assays were found to be significantly associated with the quantitative results from the flow cytometry assay. Healthy individuals were recruited to establish a normal range. Interestingly, in healthy individuals, hypoxia induced by exposure to moderate altitude, was shown to cause a decrease in procoagulant, platelet-derived and red blood cell-derived microparticles. Obstructive sleep apnoea is a common syndrome, associated with an increased risk of cardiovascular disease. Microparticle levels were determined in two randomised controlled trials, investigating the impact of therapy in these patients. Initiation of treatment for six months in minimally symptomatic patients led to a decrease in procoagulant microparticles. Withdrawal of treatment for two weeks in moderate/severe patients led to an increase in endothelial-derived, granulocyte-derived and monocyte-derived microparticles. Finally circulating microparticles were investigated in patients with cardiovascular-associated conditions. Patients undergoing a dobutamine stress echocardiogram (DSE), for the identification of coronary artery disease, were studied. Patients with a negative DSE exhibited an increase in procoagulant, platelet-derived, endothelial-derived and red blood cell-derived microparticles during stress testing, which was not evident in patients with a positive DSE, suggesting that microparticles provide additional diagnostic value in this setting. In the rapidly developing field of microparticle analysis, the flow cytometric assay described in this thesis, is reproducible, flexible and correlates well with functional microparticle assays. It has also been shown to provide novel and potentially clinically relevant results in a variety of clinical conditions associated with cardiovascular disease.

Meiotic recombination occurs at relatively high levels at specific regions in the genome called hotspots. The transcription factor-dependent hotspots (α-hotspots) have been widely studied in yeast, and are beginning to be elucidated in mammals. The HIS4 hotspot activity in Saccharomyces cerevisiae requires binding of Bas1p, Bas2p, Rap1p and Gcn4p. Bas1p acts in conjunction with Bas2p to regulate basal level of transcription of their target genes, and can be stimulated under conditions of adenine starvation and accumulation of metabolites AICAR and SAICAR from the purine biosynthesis pathway. Gcn4p activates transcription of yeast genes in response to starvation for amino acids and purines. This study focused on the influence of nutritional starvation on HIS4 hotspot activity, and different mechanisms mediating this effect. Our data suggests that deletion of genes known to accumulate AICAR/SAICAR can stimulate recombination at HIS4 in a Bas1p-dependent manner. Furthermore, intracellular and extracellular starvation for adenine and amino acids also activates recombination at HIS4. In addition, moderate levels of starvation only affect recombination when chromatin is already hyperacetylated, by the inactivation of the Set2p methyltransferase. Bas1p plays an essential role in mediating the effect of starvation and the set2 mutation on recombination. We showed that Gcn4p is not required for HIS4 hotspot activity, but plays a modest role in the effect of starvation in an adenine auxotrophic strain. Additionally, the starvation effect is also mediated by an as yet unknown factor independent from Bas1p/Bas2p and Gcn4p. This work provides additional information regarding the regulation of a transcription factor-dependent hotspot activity, and factors influencing its activation. Furthermore, data in this study indicate that BAS1, and not BAS2 exhibit haploinsufficiency with respect to its function in activating meiotic recombination. This implies that Bas1p is rate-limiting for HIS4 hotspot activity.

A novel family of signalling regulators, Tribbles (Trb), have been recently identified and characterised. They possess a single kinase like domain but are catalytically inactive. Numerous interacting protein partners of Tribbles have been identified; they have been highlighted to play a role in many cellular and disease processes. Gab (Grb2 associated binder) proteins play a role in signalling events induced by receptor tyrosine kinases (RTKs). Gabl is a prototypic member of this family, which has been shown to have an important function in the mechanism ofPI3K/Akt activation. I demonstrated that Gab I and Trb2 interact, a protein-protein interaction that has not been reported previously. Based on the role of Gabl and Trb proteins in the regulation of P13K/Akt signalling, I examined the Gabl-Trb2 interaction in the wider context ofP13K and AktlPKB signalling. Subsequently, I demonstrated the Gabl-Trb2 protein interaction is regulated by PI3K1 Akt signalling. The cellular actin cytoskleton, functions in the control of cell motility and proliferation, morphology, and also cell-cell communication. Thus, it is inevitable that abnormalities in its regulation result in disease processes. A role for Gab 1 has been demonstrated previously in regulation of the actin cytoskeleton and membrane ruffles in response to RTK activation. The role of Gab! and Trb2, both individually and together in the regulation of membrane ruffling and actin cytoskeleton arrangement was investigated. Novel findings are presented that implicate a role for Trb2 in cellular morphology, and initial investigations were conducted to begin elucidating the potential signaling pathways involved. Overall this study demonstrates a role for Gab 1- Trb2 interaction in PI3K1 Akt signalling and cellular morphology. P13K signalling and actin cytoskeleton dynamics play a central role in many vascular disease processes. Thus, it is possible that investigations into Trb2's role in such processes may result in the identification of novel potential therapeutic strategies. xx

No description available.

The Cvt pathway is an unusual process which transports the two hydrolases aminopeptidase I (Apel) and a-mannosidase I (Amsl) from the cytoplasm directly to the vacuole. So far it has only been described in S. cerevisiae and closely related budding yeasts. Most of the components needed for the Cvt pathway, the Atg proteins. are also required for the autophagy pathway, a degradative system which delivers cytoplasmic constituents to the vacuole in a non-specific manner, induced under conditions such as nutrient starvation. Genes involved in these two pathways play a wide variety of roles in diseases such as cancer, cardiomyopathy, Huntington's and Parkinson's disease as well as the removal of intracellular pathogens and ageing.

Translin and its binding partner protein Translin-associated factor-X (TRAX) have been implicated in a diverse range of cellular processes. Translin has been demonstrated to bind to both DNA and RNA and appears to be involved in the recognition of specific sequences associate with breakpoint junctions of chromosomal translocations linked with the development of some human cancers. More recently, Translin and TRAX have been found to make a heterocomplex known as C3PO which has been implicated in passenger strand removal in the RNAi pathway and tRNA processing. Translin is conserved in the fission yeast and in this study we used this facile model system to further investigate the biological function of Translin. Initial work confirmed previous findings that demonstrated loss of Translin function alone has no measureable negative affect on fission yeast cells. However, we addressed the hypothesis that Translin functioned in a redundant pathway with the RNAi pathway component Dicer and found that when Dicer and Translin are disrupted there are enhanced levels of genome instability which are not due to failures in the DNA damage response. Further investigation demonstrated that this was due to an enhanced failure in the function of centromeric heterochromatin, which is regulated in part by the RNAi pathway. Here we present a model demonstrating that Translin is the key regulator of a previously inferred argonaute-dependent, Dicer-independent regulator of centromeric heterochromatin function and chromosome stability.

The interaction between a growth factor and the extracellular region of its receptor is the first step in triggering the intracellular signalling pathways that induce a change in cell phenotype. Several platelet-derived growth factors (PDGFs) each bind to and activate PDGF receptor (PDGFR) alpha, a transmembrane receptor tyrosine kinase that simulates proliferation and migration in cells of mesenchymal origin. PDGFR alpha itself activates several intracellular signalling pathways. Thus, the possibility exists that different growth factors interact with PDGFR alpha in distinct ways to elicit divergent cellular effects. This possibility was investigated by examining the interactions that occur between the extracellular region of PDGFR alpha and three growth factors: PDGF-AA, PDGF-BB and vascular endothelial growth factor-A165 (VEGF-A165). The extracellular region of PDGFR alpha, and fragments thereof, were recombinantly expressed by mammalian 293-EBNA cells. The affinities of the protein fragments for each of the three growth factors were assayed by solid phase binding analysis. Each growth factor had a different affinity both for the extracellular region of PDGFR alpha and the fragments of that region, indicating that they differentially interact with the receptor. These different growth factor/receptor interactions were further investigated by substituting charged amino acids for uncharged residues in the putative ligand binding domain of PDGFR alpha. The amino acids selected for substitution were analogous to residues that had been shown to form direct contacts with growth factors in growth factor/receptor systems that are evolutionarily related to PDGFR alpha. Mutants of the PDGFR alpha ectodomain were generated by site-directed mutagenesis, and their growth factor binding properties were subsequently assayed. Particular mutants had differential effects on growth factor binding to PDGFR alpha. For example, the K209A mutant had no effect on PDGF-AA binding to the receptor, compared to wild-type, but it had a lower affinity for PDGF-BB and a greater affinity for VEGF-A165, compared to wild-type. These data demonstrated that the examined growth factors form subtly different interactions with PDGFR alpha. Thus, it is likely that each growth factor induces a distinct conformational change in the receptor upon binding. In this way, different growth factors may elicit divergent cellular effects through the same receptor.

To exploit pluripotent cells for regenerative medicine applications it will be necessary to understand the molecular mechanisms that govern pluripotency and lineage commitment within these cells. The mechanism by which LIF sustains ‘naïve’ pluripotency in mouse embryonic stem (mES) cell has recently been delineated; LIF signals to the core circuitry of pluripotency (Oct4, Sox2 and Nanog) via Jak/STAT3 and PI3K/Akt-mediated expression of Klf4 and Tbx3 respectively. E-cadherin has been shown to be required for LIF-dependent mES cell pluripotency since cell lines exhibiting low/no E-cadherin expression maintain pluripotency via Activin/Nodal. However, these cells maintain expression of the core circuitry of pluripotency, thus the role of E-cadherin in pluripotency remains elusive. To investigate this, we have characterised an E-cadherin negative proliferating stem (ENPS) cell line, generated by seeding wt mES cells at low density in the absence of LIF. These cells exhibit Activin/Nodal-dependent pluripotency marker expression but fail pluripotency tests such as EB differentiation and chimera generation and microarray analysis shows they lack naïve transcripts and express early lineage markers. We have also exploited two additional E-cadherin negative cell lines (Ecad-/- and EcadRNAi mES cells) to delineate the molecular mechanisms connecting E-cadherin to the core circuitry of pluripotency. These cells exhibit decreased expression of pluripotency markers Klf4 and Nanog, the latter a direct consequence of a lack of E-cadherin-mediated STAT3 activation. Interestingly, both ENPS and Ecad-/- mES cells can be ‘rescued’ to a naïve pluripotent state upon LIF stimulation. In ENPS cells, LIF supplementation induces restoration of E-cadherin expression, LIF-dependent pluripotency and EB and chimera generation abilities. In Ecad-/- mES cells, LIF supplemention restores LIF-dependent pluripotency via N-cadherin, thus demonstrating a novel role for N-cadherin in mES cell pluripotency. At high passage, ENPS cells (like some cancer cells) exhibit methylation of the E-cadherin promoter and PI3K-dependent increased survival compared to wt mES cells. Our findings provide a potential mechanism for the role of E-cadherin in induced pluripotent stem (iPS) cell generation, since STAT3 phosphorylation has recently been shown to be a limiting factor in this process. In addition, our data suggest E-cadherin can be manipulated to direct differentiation for regenerative medicine applications since ENPS cells exhibit a lineage bias towards neuroectoderm at the expense of endoderm specification.

This thesis describes experimental and theoretical investigations into new particle handling and separation methods and techniques. It makes a major contribution to the rapidly expanding field of cell separation technology. A novel dielectrophoretic cell separation system has been developed, which is capable of processing large sample volumes (~50mL) in a flow through system. Previously reported dielectrophoretic cell separator systems typically process sample volumes in the 100mL range. The electrode configuration developed for this work allows the isolation and concentration of single particle types from large sample volumes; a method which could be further developed into a new rare-cell separation technology. In addition, a new technique of particle fractionation was developed termed ‘Dielectrophoretic Chromatography’. A cell separation chip was designed and built using standard micro-fabrication techniques. Experimental work was undertaken to demonstrate the function and limitations of the device. Numerical modelling of the particle motion in the device is presented and compared with experimental work for a number of different particle types, applied voltages and fluid flow rates. The dielectrophoretic separation system comprises a microfluidic channel, of cross-section 100mm x 10mm and length 50mm, with two sets of interdigitated microelectrode arrays. The first set of arrays, with characteristic electrode size 40mm, called a focussing device, has electrodes patterned onto the top and bottom surfaces of the flow channel. The second electrode array, which is part of the same device, has an electrode array patterned only on the bottom of the channel. Two sizes of secondary electrode array were used 20mm and 40mm. AC voltages (from 1V to 10V peak) are applied to the microelectrode, with a frequency between 10kHz to 180MHz. A dielectrophoretic force is exerted on the particles as they flow along the channel. The first electrode array uses negative dielectrophoresis to focus the stream of particles entering the device into a narrow sheet (one particle diameter thick) midway between the upper and lower channel surfaces. The second electrode array, down stream from the first is separately controllable.