DNA replication initiation is co-ordinated with progression through the cell cycle via a two-step mechanism. The first step, termed origin licensing, involves the assembly of a pre-replicative complex (pre-RC) at origins of replication, in which the Mcm2-7 replicative helicase is loaded onto DNA in an inactive form. Origin unwinding and DNA synthesis is only initiated during the second step of this process, origin firing, which requires the recruitment of multiple ‘firing factors’, such as Sld3/7 and Cdc45, as well as the activity of two essential cell cycle regulated kinases, cyclin-dependent kinase (CDK) and Dbf4-dependent kinase (DDK). Whilst studies of the mechanism of origin licensing have been greatly furthered by the availability of an in vitro pre-RC assembly assay, the subsequent activation of the replicative helicase has not been well characterised, and how the firing factor proteins catalyse the complex topological changes required for Mcm2-7 helicase activation is unknown. In this study, I used an in vitro biochemical approach to investigate the function of DDK and Sld3/7 in origin firing. I reconstituted the recruitment of Sld3/7 to the pre- RC in vitro, which is thought to be the first step during Mcm2-7 helicase activation. I observed recruitment of Sld3/7 to be dependent on phosphorylation of the loaded Mcm2-7 complex by DDK, and was subsequently able to map the Mcm2-7 binding activity of Sld3/7 to a central domain of Sld3. By isolating a number of point mutants in Sld3 that were specifically defective in Mcm2-7 binding, I showed that the Sld3-Mcm2-7 interaction is essential for replicative helicase activation. Furthermore, I showed that the central portion of Sld3 contains a Cdc45 interacting site, which is also required for efficient replication initiation. Subsequently, I showed that Sld3 can interact with both Mcm6 and Mcm4 in a phosphorylation-dependent manner. Sld3 itself was shown to contain a novel phosphopeptide binding activity, and can interact with numerous phosphorylated residues throughout the N-terminal half of Mcm6. Elimination of these phosphorylation sites resulted in defects in both Sld3/7 recruitment to the pre-RC and replicative helicase activation. Thus, the novel DDK-dependent Sld3-Mcm2-7 interaction described in this study helps to explain the function of DDK during the early stages of origin firing. Additionally, the observation that Sld3 can interact directly with phosphorylated residues on Mcm2-7 indicates that Sld3 is a reader of DDK activity. As both an essential CDK substrate and a DDK reader, Sld3 thus functions as a point of intersection for the activities of CDK and DDK during Mcm2-7 helicase activation.

Ataxia telangiectasia-mutated (ATM), the protein kinase that is mutated in patients with ataxia telangiectasia (A-T), is a central player in the cellular response to DNA damage. ATM is activated following double-strand breaks by MRE11-RAD50-NBS1 complex (MRN), which acts as an initial sensor of DSB. ATM is also activated by stimuli that change chromatin structure such as chloroquine and hypotonic shock, which depends on the ATM-interactor protein (ATMIN) as a cofactor. In this study, I show that there exists competition between ATMIN and NBS1 for ATM binding and this can regulate ATM signalling. The absence of one cofactor can lead to enhanced signalling through the alternative pathway. Furthermore, I also characterise the role of an E3 ligase that regulate the interaction between ATMIN and ATM by mediating ATMIN mono-ubiquitination. Thus ATMIN ubiquitination could be an important step in ensuring robust ATM activation after IR. In addition, I also show that ATMIN is required for ATM signalling after replication stress to protect against genomic instability in the form of anaphase bridges and aberrant chromosome segregations. ATM is activated by replication stress and is required for the formation of 53BP1-containing nuclear bodies, which protect fragile sites. ATMIN interacts with WRNIP and RAD18 in a complex that is required to activate ATM at sites of stalled replication forks. Furthermore, using a genome-wide screening approach, I identified additional factors that regulate ATM signalling after replication stress. The 8-oxo-guanosine repair pathway could represent a link between ATM signalling and genomic instability. I show that the components of base excision repair pathway are required for the activation of ATM and formation of 53BP1 nuclear bodies that protect fragile sites in the maintenance of genome stability.

Cord blood (CB) has served as a source of haematopoietic stem cells (HSC) to treat haematological diseases for the last two decades. Low incidence and severity of graft-versus-host disease, and a robust graft-versus-leukaemia effect are some of CB advantages as a source for HSC. However, its main disadvantages are a limited number of HSC per unit and delayed immune reconstitution and infections after CB transplantation (CBT). In order to improve engraftment and immune reconstitution after CBT different approaches have been explored like HSC expansion, double CB transplantation, CBT plus third party donor HSC or intra-bone infusion. It seems as if the interaction of cord blood stem cells (CB SC) with other cells is required for CB SC to engraft. On this basis, the effect of accessory cells, particularly natural killer (NK) cells, on CB SC engraftment was analysed. It was observed that NK cell co-infusion with CB SC led to higher levels of engraftment in NSG mice. This in parallel with a higher level of expression of CXCR4 receptor, a higher migration index, a higher number of colony forming units (CFU) after long-term culture initiating cells assay (LTC-IC) and a trend to a higher number of CFU when CB SC were co-cultured with activated NK cells (aNK). The effect observed on CBSC clonogenic capacity was maintained even after treatment with pertussis toxin (PTX), which blocks G-protein coupled receptors signalling, such as CXCR4. The results presented in this work offer the basis for an alternative approach that could help to improve CBT.

X-linked lymphoproliferative (XLP) disease is a severe primary immunodeficiency. Immunodysregulatory phenomena are observed following EBV infection suggesting that defects exist in these effector populations. The gene defective in XLP is SAP (SLAM-associated protein), an intracellular adaptor protein that mediates signals through SLAM and other immunoglobulin superfamily receptors including 2B4. Cytotoxic T cells (CTLs) and natural killer (NK) cells play a major role in the normal immune response to Epstein-Barr virus (EBV) infection. EBV specific T cell lines (EBV-T cell lines) were generated from normal individuals and XLP patients and examined for CTL function in response to different stimuli. It has been shown that XLP patients can generate EBV-T cell lines that are phenotypically similar to those from unaffected individuals. XLP patient derived EBV-T cell lines showed a significant decrease in interferon-gamma (IFN-gamma) production in response to 2B4 and autologous EBV transformed lymphoblastoid cell line (LCL) stimulation but not in response to SLAM. Furthermore, XLP EBV-T cell lines demonstrated markedly decreased cytotoxic activity against autologous LCLs. By retroviral gene transfer of the SAP gene into XLP patient derived EBV-T cell lines, reconstitution of EFN-gamma production and cytotoxic activity has been shown, confirming the defects are SAP dependent. These studies demonstrate that in XLP the lack of SAP affects specific signalling pathways resulting in severe disruption of CTL function. In addition, SLAM and 2B4 expression on immune cell lineages has been investigated, the results suggest a wider range of 2B4 expression and deserve further investigation in relation to XLP molecular and cellular pathogenesis.

Homologous recombination (HR) is an essential DNA break repair mechanism that remains incompletely understood. HR is a complex multistep process initiated by the loading of RAD-51 recombinase as filaments onto single stranded DNA (ssDNA). This structure directly invades an intact homologous duplex, which serves as a template for repair DNA synthesis. Numerous positive regulators of HR have been described, including the Rad51 paralogs, but the mechanism of action of Rad51 paralogs in promoting HR is unknown. In this study, I have characterized the mechanism of action of a novel Rad51 paralog complex, RFS-1/RIP-1, from C. elegans. RFS-1 is a Rad51 paralog required for RAD-51 focus formation at stalled replication forks, indicating an early positive regulatory role in HR. I demonstrate that RFS-1 interacts with a nematode-specific orphan protein, RIP-1. I identify a cryptic Walker B ATPase-like motif within RIP-1, which is functionally important in establishing the RFS-1/RIP-1 interaction interface. rip-1 and rfs-1 mutant animals phenocopy for essentially all phenotypes analysed. Together these data suggest RFS-1/RIP-1 functions as a constitutive complex. I show recombinant RFS-1/RIP-1 can be purified and specifically binds ssDNA but lacks measurable ATPase activity. RFS-1/RIP-1 also strongly stimulates strand invasion activity by RAD-51, consistent with a pro-recombinogenic function in vivo. I define for the first time the mechanism of action underlying the intrinsic ability of Rad51 paralogs to stimulate HR. Using a combination of biochemical and biophysical approaches, notably electrophoretic mobility shift assays, stopped-flow reaction kinetics and nuclease protection assays, I show RFS-1/RIP-1 dramatically alters the properties of RAD-51-ssDNA filaments such that RAD-51 is more stably associated with ssDNA yet the ssDNA is more sensitive to nuclease degradation. RFS-1/RIP-1 exerts these effects primarily downstream of filament formation, ruling out a major role in RAD-51 loading. I propose RFS-1/RIP-1 remodels RAD-51-ssDNA filaments to a conformation poised for pairing with the template duplex and strand invasion.

Adeno-associated virus vector encoding codon optimised human factor IX (AAV-LP1-hFIXco) has demonstrated great promise for the treatment of patients with severe haemophilia B. However, in some patients treated with a high vector dose, hepatocellular toxicity was observed. To improve AAV vector potency, various scaffold/matrix attachment regions (S/MARs) were cloned at the 3’ end of a modified single-stranded (ss) AAV-LP1-hFIXco expression cassette. In a head to head comparison, a vector containing a S/MAR element from the human hypoxanthine-guanine phosphoribosyl-transferase gene in the forward orientation (ssAAV-LP1-hFIXco-HPRT-F) was found to mediate the highest levels of hFIX expression in mice. In comparison to animals transduced with a control vector containing no S/MAR, the ssAAV-LP1-hFIXco-HPRT-F transduced cohort expressed hFIX at 28-fold higher levels. This trend was reproducible in rhesus macaques where 10-fold higher FIX levels were observed following transduction with ssAAV-LP1-hFIXco-HPRT-F as compared to delivery of ssAAV-LP1-hFIXco-control vector. Through a deletion analysis, short regions from the IFNβ and HPRT S/MARs with potent enhancer activity were identified. This allowed for the in-silico elicitation of motifs with a potential role in S/MAR function and also minimised the space occupied by S/MARs within our AAV expression cassette. When cloned into a self-complementary (sc) AAV-LP1-hFIXco expression cassette, the 130bp region from the HPRT S/MAR (fragment 2b) was sufficient to enhance FIX levels in mice by 35-fold over that observed with a control self-complementary vector. Mechanistic studies showed that S/MAR elements enhanced AAV transgene expression by reducing heterochromatin marks (H3K9me2 and HP1α) in the promoter region, resulting in an increase in FIX mRNA levels by up to 20-fold. S/MARs therefore provide a novel inbuilt process for enhancing AAV mediated transgene expression by preventing epigenetic silencing of the provirus. As such, S/MARs offer the possibility to improve gene transfer to humans through using lower and potentially safer doses of AAV.

Chromosomal DNA must be faithfully duplicated only once per cell cycle in order to ensure genome stability. At the end of mitosis and during G1 phase of the cell cycle, eukaryotic cells assemble pre-replicative complexes (pre-RCs) on multiple replication origins distributed along their chromosomes in a process called origin licensing. During S phase, licensed origins are activated, DNA is unwound and the replicative machinery is recruited triggering origin firing. Cyclin-dependent kinases activity (CDKs), ubiquitin-dependent protein proteolysis and licensing factors inhibitors cooperate to promote origins firing and prevent replication origins from being reused within the same cell cycle. Mis-regulation of origin licensing contributes to genetic instability and is commonly observed in a variety of cancers. GEMC1 is a recently identified CDK target involved in the control of DNA replication in Xenopus laevis. GEMC1 is highly conserved in vertebrate organisms but nothing is currently known about its biological role in mammalian cells. In this study, the first characterisation of mammalian GEMC1 is presented. GEMC1 is a nuclear protein that is highly regulated during the cell cycle by a series of post- translational modifications, which include CDK-dependent phosphorylation and ubiquitination. Similar to its Xenopus homologue, GEMC1 interacts with TopBP1 and Cdc45 in S phase and its depletion strongly impairs origins firing, leading to the accumulation of DNA double-strand breaks (DSBs) in replicating mouse and human cells. These results confirm that mammalian GEMC1 has a positive role in chromosomal DNA replication. GEMC1 also forms a complex with the licensing inhibitor Geminin, both in vivo and in vitro, through its coiled-coil domain. Strikingly, GEMC1 overexpression in human U2OS cells induces DNA re-replication, chromosomal breakages that cause a G2/M arrest and centrosome amplification. All together, these results suggest that GEMC1 is a powerful DNA replication promoting factor that functions by stimulating origin firing and by antagonising Geminin dependent inhibition of origin licensing in vertebrate cells. These data also indicate that regulation of GEMC1 protein levels is critical for the maintenance of genome stability.

CXCR4 and CXCL12 are known to work in concert to mediate cell proliferation, survival and migration. These functions have a bearing on physiological cellular events such as haematopoietic cell recruitment to, and survival within, the bone marrow (BM) compartment. This axis can mediate analogous functions within a malignant setting, helping to organize tumour niches and promoting tumour cell survival and disease progression. Adoptive T cell therapy is well established in the treatment of haematopoietic cancers, and is under investigation for solid malignancies. Barriers to these therapies include off target toxicities such as GVHD and loss of T cell function or persistence. Approaches to target transferred therapeutic T cells to the sites of disease and increase their function and engraftment are desirable. I have constructed a retroviral vector to enable transduction of murine T cells to produce constitutive over-expression of CXCR4. This construct was used alongside control vectors, to investigate the therapeutic potential of CXCR4 over-expressing T (TCXCR4) cells. In vitro, murine TCXCR4 cells exhibited stable expression of CXCR4 and increased chemotaxis to CXCL12. Further in vivo studies observed preferential accumulation of T cells within the BM from 24hrs after transfer. By 2 weeks, TCXCR4 cells outcompete control counterparts in BM accumulation 15-fold. To further track TCXCR4 cells within the BM, intra-vital imaging experiments in osteoblast (Collagen-1α-GFP) mice identified preferential homing to the endosteal niche. In murine tumour models, allogeneic TCXCR4 cells mediated greater anti-tumour responses against sub-cutaneous CXCR4+ tumour. Of note, greater graft-versus-lymphoma (GVL) effect was achieved without the induction of greater graft-versus-host disease. This approach has also demonstrated preferential control of splenic metastasis of BM disease in a further murine model of CXCR4+ malignancy. In memory recall experiments, performed using OT-I CD8+ T as donor cells, TCXCR4 cells once transferred in vivo maintained higher proportions of CD44Hi CD62LHi CD127+ phenotype cells, underwent higher levels of proliferation and produced higher levels of IFN-γ, early post vaccination with relevant antigen. In conclusion, over-expression of CXCR4 in T cells increases their engraftment in the BM and the induction of GVL. This initial data suggests that the strategy of inducing CXCR4 over-expression in therapeutic T cells is an attractive approach to enhance malignant disease targeting and favour long-term engraftment of transferred cells.

During embryonic development multipotent cells are specified to give rise to the different tissues of the body. This process depends on a tightly controlled network of signalling pathways. Importantly, tissues, which require differential activity of these pathways, can be induced in close proximity, thus suggesting an intricate spatial and temporal control of pathway activation. One of the pathways crucial for tissue specification is the bone morphogenetic protein (BMP) signalling pathway. Its role in the patterning of the ectoderm is well understood during gastrulation but unclear for later stages of development. Using zebrafish and Xenopus as model organisms, I investigated the spatial and temporal control of BMP activity after gastrulation at the border of the neural plate as progenitor cells emerge that give rise to cell types such as melanocytes and smooth muscle cells, as well as the olfactory epithelium and the lens. I identified new players that regulate the formation of distinct domains of BMP signalling and thereby enable the specification of adjacent tissues with different requirements for BMP activity. Previously, Snw1 was identified as a crucial factor for neural crest specification during development. Overexpression and depletion of Snw1 in Xenopus and zebrafish embryos leads to a loss of the neural crest cell population. Snw1 was identified as a regulator of BMP activity at the neural plate border, but not as a core component of the pathway downstream of the receptor. Snw1 is involved in a step between transcription and expression of the BMP ligands and since depletion of Snw1 in zebrafish increases bmp2b ligand transcription but prevents expression of the protein. I have further dissected the function of Snw1 and shown that Snw1 is in a complex with components of splicing machinery, as well as several chromatin remodeling and transcriptional elongation factors. I have used RNAseq to identify additional Snw1 targets.

During mitosis, spindle microtubules nucleate from the spindle pole body (SPB; centrosome in humans) and capture the chromosomes at the kinetochores. Many players, including the Ndc80/Hec1 outer kinetochore complex, the TACC-TOG microtubule-associated proteins, and various motor molecules, have been implicated in this process. Mis-regulation (e.g. overexpression) of these proteins leads to chromosome segregation errors and aneuploidy, a hallmark of human cancer. Despite this knowledge, the underlying molecular mechanisms by which these proteins function and interact remain largely elusive. Targeting of Alp7/TACC to the SPB is essential for mitotic spindle assembly. Using Schizosaccharomyces pombe as a model system, I have identified the SPB- targeting sequence within Alp7/TACC and the conserved pericentrin-like protein Pcp1 as one of the loading platforms for Alp7/TACC at the SPB. Upon successful spindle assembly, the spindle microtubules elongate and capture the kinetochores, which is a crucial step for faultless chromosome segregation. Having isolated and characterised a novel ndc80 mutant (ndc80-NH12), I have uncovered the roles of the Ndc80 internal loop in binding to the Alp7/TACC-Alp14/TOG complex. This binding ensures proper kinetochore-microtubule attachment and subsequently recruits the kinesin-8-protein phosphatase I (Klp5-Klp6-PP1) complex to the kinetochores. This recruitment in turn silences the spindle assembly checkpoint (SAC) and promotes timely and rapid chromosome movement during anaphase A. These results unveil the importance of the Ndc80 internal loop and its binding partners in regulating kinetochore-microtubule attachment and controlling ordered mitotic progression. Intriguingly, by overexpressing different Ndc80 constructs, I have found a possible causal link between the Ndc80 internal loop and tumourigenesis. I propose that the Ndc80 loop, when overexpressed, would sequester away its binding partners, thus disrupting their normal functions and leading to aneuploid formation. My results provide new insights into the molecular organisation at the kinetochore-microtubule interface, and have significant implications for future therapeutics and drug development.