The protection of chromosome ends, and in turn genome stability, is intimately linked with the ability of telomeres to replicate and segregate with high fidelity. It has been shown previously that Taz1, the fission yeast ortholog of mammalian TRF1 and TRF2, is critical for promoting passage of the replication fork through telomere sequences. Deletion of taz1+ results in stalled telomeric replication forks that fail to resume. These defective replication intermediates lead to telomere entanglements that fail to be resolved in mitosis at cold temperatures, leading taz1Δ cells to be cold sensitive. It is not yet known exactly what entanglements are, why they occur at cold temperatures, and how (or if) they are resolved. In this thesis telomere entanglements are further investigated. We found that in taz1∆ cells grown in the cold, entangled telomeres can be visualized as DNA masses persisting between separating sister chromatids during M-phase. We have observed sister chromatid masses which remain connected either by histone bound DNA resulting in aberrant chromosome segregation morphologies, or by ssDNA coated with Rad11RPA and Rad22Rad52. The former was observed to be highly dangerous, the latter can be processed and cells can continue to the next cell cycle. Rif1, a conserved replication/repair protein sparked our interest when we observed that rif1+ deletion suppresses taz1Δ cold sensitivity. However, the function of Rif1 was a mystery; revealing the role of Rif1 in cells lacking taz1+ formed the next part of this thesis. We have found that Rif1 acts in removal of telomeric entanglements rather than at the fork-stalling events that generate them. Although Rif1 is reported to bind telomeres in a Taz1-dependent manner, we observe Rif1 at the mid-zone during anaphase in wild type and taz1∆ cells, associated to thin DNA stretches. Artificially recruiting Rif1 to a taz1∆ telomere is lethal in the cold, indicating that the departure of Rif1 from the telomere is essential for viability and presumably for entanglement resolution. We propose a newly identified role for Rif1 in regulating the resolution of entangled DNA during M-phase.

Sprouting angiogenesis describes the formation of blood vessels from preexisting ones, a process guided by leading endothelial tip cells, followed by stalk cells. It has previously been established that the transient and dynamic specification of both phenotypes is mediated by DLL4/NOTCH signalling, which is primarily actuated by vascular endothelial growth factor (VEGF) and the corresponding vascular endothelial growth factor receptor 2 (VEGFR-2). The ability of a cell to outcompete its neighbour is determined by VEGFR signalling mediated production of DLL4 and subsequent activation of NOTCH in the adjacent cell. The stalk cell phenotype, induced by NOTCH activation, is reinforced by SMAD signalling through a crosstalk between both pathways. However, the downstream effectors of NOTCH and the molecular link to SMAD signalling in tip cell competition are unknown. During my PhD I have identified Neuropilin-1 (NRP1) as a critical NOTCHregulated determinant of tip/stalk specification. I have found that endothelial cells lacking one or both functional alleles of NRP1 are unable to form tip cells when competing with WT cells. Despite NRP1 having been previously described as a coreceptor of VEGFR-2, my data indicate that NRP1 functions independently of VEGFR-2 during tip/stalk selection. Furthermore, I have shown that inhibition of NOTCH is not sufficient to rescue the profound stalk cell phenotype of NRP1 deficient cells. Thus, I have identified NRP1 as the first bona fide downstream effector of NOTCH signalling in regulating angiogenesis. Furthermore, my data provides the missing link between NOTCH and SMAD signalling in stalk cell specification, as it shows NRP1 to be a negative regulator of the TGF- β/ALK5/SMAD2 pathway. Additionally, in contrast to NOTCH ablation, inhibition of ALK5 quantitatively restores the ability of NRP1 null cells to contribute to the tip position. I conclude that NRP1 is a key regulator of tip/stalk cell specification in sprouting angiogenesis; differential NRP1 levels act as key effector.

The demand for unrelated haematopoietic stem cell (HSC) donors has risen threefold over the last decade, and is likely to continue to rise over the next 10 years. The time taken from diagnosis to transplant is recognised to adversely affect patient outcome, and provision of unrelated donors (UDs) has been identified as a key source of delay. Obstacles to provision of UD include: delays in referral to a transplant centre, awaiting sibling typing, lack of matched donors (particularly for those from ethnic minorities and/or with rare HLA phenotypes), low- or intermediate-resolution donor HLA typing, donor attrition from the registries, donor ineligibility on grounds of health and difficulties encountered transporting HSC across international borders. A prospective study of 401 patients explores timescales prior to transplantation, and the causes and duration of delays in the pre-transplant period. However, it also shows that the provision of transplantation to those of non-white Northern European descent is now similar to white Northern Europeans (WNE). However, donor attrition at the confirmatory typing stage is found to significantly impact on donor selection, and disproportionately affects WNE patients. A retrospective study of over 7000 donors identifies that donor ethnicity, gender and duration on the register are associated with increased donor attrition, and that a risk score based on these factors may help predict those donors most likely to defer. Donor medical deferrals form the basis of a project to improve the consistency and availability of medical guidance at Anthony Nolan, and this project was extended internationally to form the WMDA medical guidelines. Finally, a retrospective study of second donations made by Anthony Nolan donors shows those who donate by bone marrow are more likely to be called for a second donation.

The incidence of breast cancer has progressively increased, while survival rates have simultaneously improved. Young women with breast cancer are likely to suffer ovarian damage from chemotherapy, which can have a profound effect on their quality of life. At present, it is impossible to predict the functional lifespan of the chemotherapeutically damaged ovary as there is insufficient data. Ovarian reserve tests (ORT) have the potential to estimate the reproductive age of the ovaries, which would allow an accurate estimation of fertility status and the risk of premature ovarian failure. This project investigates the use of ORT in young women with breast cancer. To achieve this, biochemical and biophysical ORT were performed in a mixed longitudinal and cross-sectional study group comprising young women treated with chemotherapy for breast cancer and age-matched, regularly menstruating controls with proven fertility. Overall the results indicate potential for the use of inhibin B, anti-mullerian hormone, oestradiol and antral follicle count in the evaluation of these patients. An in vitro study was performed to supplement the clinical study, in which the effects of cytotoxic agents commonly used to treat breast cancer were examined by simultaneously applying equivalent doses of each to a breast cancer cell line and primary granulosa-luteal cell cultures respectively. The effects of these agents were measured in terms of cellular integrity (apoptosis) and functionality (hormones). Overall the results suggest variations in cytotoxicity (LD50) between breast and granulosa cells which have potential therapeutic implications.

Epithelial cells are connected to each other via intercellular junctions, which are established, remodelled and maintained by a complex molecular machinery. Aberrant regulation of cell-cell junctions leads to loss of tissue organisation and is a hallmark of cancer formation. Although many kinases that regulate the phosphorylation of junctional proteins have been described, much less is known about the reversal of phosphorylation by phosphatases. This thesis investigates the role of Drosophila ASPP, a scaffold protein that is localised at adherens junctions, as a regulatory subunit for PP1s. In vitro, ASPP can bind to PP1 using its RVXF motif and SH3 domain. In vivo, ASPP co-localises with the PP1a96A and PP1b9C isoforms at cell-cell junctions in the pupal retina and ASPP function is at least partially dependent on its ability to bind to PP1. Furthermore, ASPP can recruit two additional coiled coil containing scaffold proteins, RASSF8 and Ccdc85, to form trimeric complexes with PP1a96A. ccdc85 mutants that were generated in this work have a rough eye, similar to ASPP mutants, suggesting a similar function in vivo. Two potential substrates for the ASPP/PP1 complex were tested: (1) Yki a transcriptional co-activator that is part of the Hippo pathway and (2) Baz, a scaffold protein that is required for cell polarity. Although no evidence for dephosphorylation of Yki was found, Baz can be dephosphorylated in vitro. The well-described aPKC phosphorylation site (S980) and five additional serine/threonine residues are strongly dephosphorylated by ASPP/PP1. My work also identified three potential regulators/scaffolds of ASPP/PP1. The Hippo pathway kinase Wts can phosphorylate RASSF8, the E3 ubiquitin ligase Sina can ubiquitylate and degrade ASPP and the junctional protein Magi can associate with RASSF8 at adherens junctions. Finally, novel potential regulators, scaffolds or substrates of the ASPP/PP1 complex were identified through AP-MS experiments and tested for their ability to modulate the ASPP depletion phenotype in vivo.

Malignant haematopoiesis is usually associated with various genetic lesions such as chromosomal translocations or mutations in individual genes. This thesis investigates the genetic basis of two such syndromes, acute megakaryoblastic leukaemia (AMKL) and myelodysplastic syndrome (MDS). The most common constitutional aneuploidy with predisposition to leukaemia is trisomy 21, also known as Down Syndrome (DS). DS children have a 500-fold increased risk for AMKL. Somatic mutations acquired during foetal haematopoiesis in the GATA1 transcription factor are detected in megakaryoblasts from all the DS patients with AMKL. Here we show that the Gatal mutation co-operates with the chromosome 21 gene, ERGS, to immortalize foetal megakaryocyte progenitors with the phenotype of AMKL megakaryoblasts. We show that ERG-3 promotes megakaryopoiesis and acts as an oncogene and that progenitor cells harbouring a Gatal mutation plus ERG-3 or ERG-3 alone lead to rapid development of leukaemia in vivo. Our data support a model where trisomy 21 overexpressed genes, that promote foetal megakaryopoiesis, co-operate with mutations that arrest development and lead to DS-AMKL. This is also the first direct demonstration of the leukaemogenic activity of full length ERG protein, possibly explaining the poor prognosis of the acute myeloid leukaemias with high expression of ERG. DS patients are also predisposed to TMD (transient myeloproliferative disorder) and MDS (myelodysplastic syndrome) prior to AMKL development. Myelodysplastic syndromes are a group of clonal haematopoietic disorders, characterised by aberrant proliferation and differentiation of cells of the myeloid lineage resulting in severe cytopenia and dysplasia of myeloid, erythroid and megakaryocytes. The most common chromosomal translocation associated with MDS is the t(3 5)(q25q35) translocation. This rearrangement results in a fusion transcript comprised of nucleophosmin (NPM) gene and myeloid leukaemia factor 1 (MLF1) gene. To determine the importance of the NPM-MLF1 fusion protein in the development and progression of MDS to AML, its role in myelopoiesis and megakaryopoiesis was investigated. Our results show that NPM-MLF1 affects the differentiation of myeloid cells. Our preliminary data predicts that NPM may have a role in megakaryopoiesis and its interaction with NPM-MLF1 may affect the function of the fusion protein.

Microtubules are non-covalent polymers that form an essential part of the cytoskeleton in eukaryotic cells. Alternating phases of growth and shortening are essential for space exploration, force generation and facilitate rearrangements of the microtubule cytoskeleton in response to various stimuli. Microtubule associated proteins regulate filament dynamics and can transport cargoes. The mechanism of how microtubules grow, what triggers the transition from a growing to a shrinking microtubule, and the interplay between the various microtubule-associated proteins is only poorly understood. In vitro reconstitution approaches from purified components in combination with microfluidics techniques and simultaneous multi-colour total internal reflection florescence microscopy were employed to shed new light on the mechanism of microtubule dynamics and the interplay of proteins that bind specifically to growing microtubule ends. Tubulin undergoes conformational changes during incorporation into the polymer. Using a conformation-sensitive designed ankyrin repeat protein probe, it has been found here that these conformational changes occur at much later steps during incorporation into the polymer than previously appreciated. Growing microtubules switch to a rapid shortening phase unless their ends contain a stabilizing structure whose nature is not fully understood. The decay of this stabilizing structure was directly measured by rapid tubulin dilutions and predictions from several theoretical models have been tested. The density of a particular tubulin conformation recognized by microtubule End Binding proteins (EB1/Mal3) could be linked to filament stability. Microtubule end tracking proteins form a dynamic protein interaction network. Here, the molecular mechanism of several main players of these proteins that lead to growing microtubule end accumulation of the motor protein dynein has been elucidated by in vitro reconstitutions. The bottom up approach applied in this thesis yielded new information about fundamental properties of microtubule dynamics and gained new insight into the interplay of an important class of microtubule associated proteins.

The phosphonoinosites play structural and functional roles in biological membranes; reversible phosphorylation of the inositol headgroup generates seven different headgroups which are involved in defining organelle identity and regulating signalling, trafficking and cytoskeletal integrity. The serine/ threonine kinase Akt lies at the heart of cell biology. It is activated by an interaction with the phosphoinositide PtdIns(3,4,5)P3 in response to growth factor stimulation. Akt controls cell growth, proliferation, survival and energy metabolism. Misregulated Akt activity is a hallmark of many different types of cancer and is a prominent feature of diabetes and obesity. A more detailed understanding of the mechanisms of Akt activation will be vital for the development of effective therapeutic interventions. Despite reports of the existence of phosphoinositides in intracellular compartments, it was thought that the only relevant site of Akt activation is PtdIns(3,4,5)P3 at the plasma membrane. In this thesis we took an interdisciplinary approach to assess the importance of intracellular membranes on Akt activation. We first optimised a drug inducible dimerisation system to acutely deplete PtdIns(3,4,5)P3 at the plasma membrane, leaving subcellular compartments untouched. We used HPLC coupled to tandem mass spectrometry to quantify cellular phospholipid composition, and developed a probe to detect PtdIns(3,4,5)P3 localisation by electron microscopy. We then used a FRET-based probe to monitor Akt localisation and conformation in plasma membrane-PtdIns(3,4,5)P3 depleted cells. Finally we used the inducible dimerisation system to recruit Akt to the early endosomes and nuclear envelope in order to determine whether localised Akt activation can take place away from the plasma membrane. We have shown for the first time that Akt interacts with PtdIns(3,4,5)P3 in subcellular compartments, resulting in localised activation and substrate phosphorylation; this suggests that intracellular Akt activation contributes to substrate specificity and signal duration, and may potentially be disrupted or upregulated in human disease.

Haematopoietic stem cell transplantation (HSCT) is used to treat malignant and non-malignant diseases. Apart from clinical and human leukocyte antigen (HLA)-related factors, non-HLA Immunogenetics is increasingly recognized to play a role in the outcome of HSCT. A gene that may be relevant for the outcome of HSCT is TGFB1, which encodes transforming growth factor β-1 (TGF-β1), a cytokine that is central in the regulation of numerous immune processes. In order to understand the role of TGFB1 polymorphism in HSCT, I studied the effect of this variation in the production of this cytokine by regulatory T cells (Treg) and also in the outcome of a cohort of HSCT patients and donors. This study has found evidence of differential surface expression of TGF-β1 by activated Treg bearing TGFB1 +29C as compared with those lacking this variant. Also, the analysis of a cohort of HSCT patients and donors revealed that patients carrying TGFB1 allele p001 had reduced overall survival and increased non-relapse mortality. Interesting additional discoveries include the lack of induction of TGFB1 messenger ribonucleic acid upon T cell receptor-mediated activation in Treg, the presence of preformed intracellular latent TGF-β1 both in Treg and effector cluster of differentiation (CD)4+CD25- cells, and the kinetics of transient surface latent TGF-β1 expression on Treg in the context of an in vitro suppression assay. The typing of TGFB1 alleles in more than 1000 unrelated samples gives the first large-scale glimpse of the diversity of TGFB1 polymorphism in the population. Likewise, this study discovered the presence of novel polymorphic positions and novel alleles of the human TGFB1 gene, adding to the current knowledge on the diversity associated with it. Additional contributions are also the design of a novel TGFB1 regulatory region and exon 1 sequencing-based typing strategy and the generation of an informatics tool for the easy assignment of allelic genotypes for this ~3000 base region based on the individual variant genotypes at 18 polymorphic positions.

Cyclin D1, a cell cycle regulator, plays an important role in cell cycle progression. As overexpression of cyclin D1 has been observed in various cancers, it has been considered that high level of cyclin D1 may cause uncontrolled cell proliferation and thus give rise to cancer. Interestingly, we showed that a POU transcription factor Brn-3b, which has been shown to contribute to uncontrolled cell growth and cause tumourigenesis when expressed at a high level, correlates to the level of cyclin D1 mRNA-when the level of Brn-3b mRNA increased, the levels of cyclin D1 mRNA increased when the level of Brn-3b mRNA decreased, the level of cyclin D1 mRNA also decreased. Thus, the strong correlation between Brn-3b and cyclin D1 indicates that Brn-3b, at least in part, contributes to uncontrolled proliferation by regulating the gene expression of cyclin D1.