Prostate cancer is one of the most common cancers in the Western world. The metastasis of prostate cancer cells is the major cause of prostate cancer-related death. Understanding which genes are altered in prostate cancer and their implications for cell signaling and function is crucial for designing therapy strategies. Our group has identified a high frequency of somatic missense mutations in the plexin-B1 gene in localized and metastatic prostate cancer. The mutations are correlated with elevated metastatic features such as increased cell invasion, adhesion and decreased cell collapse relative to wildtype. Given the fact that Sema4D/plexin-B1 regulates cell signaling through coupling with tyrosine kinases and small GTPases, I hypothesized that plexin-B1 mutations may contribute to prostate cancer cell progression through affecting these pathways. The aims of this study were to characterize the molecular features of the three mutated forms of plexin-B1, A5359G, A5653G, and T5714C and to determine how the mutations affect cell signaling. It was found, using co-immunoprecipitation assays, that none of the three mutations affected the binding of either ErbB-2 or Met to plexin-B1. Sema4D/plexin-B1 mediated cell motility depending on the stoichiometry of plexin-B1, ErbB-2 and Met in prostate cancer cells. In addition, positive evidence was found for the impact of wildtype and mutant plexin-B1 on PI3K/Akt and MAPK pathways. All the three mutants exhibited an increased binding ability to RhoD GTPase relative to wildtype in GST-pulldown assays. This increased binding to RhoD is also associated with increased cell motility as shown in transwell migration assays. My study suggests that the three mutants confer a gain of metastatic phenotype to the cells through inactivating inhibitory pathways and has lead to the construction of a new model describing the mechanism by which plexin-B1 mutations contribute to cancer progression.

Efforts continue to identify and validate novel risk factors / biomarkers for breast cancer and improve current risk prediction models in the general population due to ongoing issues with sensitivity and specificity. The overall goal of this PhD study is to add to this effort. Specific aims are to (1) examine which is the best source of getting notified for breast cancer diagnosis in the general population since accurate data is crucial for risk assessment studies (2) investigate the association of sex steroids, gonadotrophins and novel assays of sex steroid hormone receptor serum bioactivity (SB) in breast cancer (3) examine whether they can be combined to improve breast cancer risk assessment and (4) identify new DNA methylation markers that might add to such a strategy in the future. To achieve this, a nested case-control study was undertaken within UK Collaborative Trial of Ovarian Cancer Screening. 2629 trial participants were identified via cancer registry (CR) or self-reporting to have breast cancer. Diagnosis was confirmed by the treating clinician. The largest study was undertaken in England and Wales to examine completeness of breast cancer diagnosis within UKCTOCS. Analysis of complete data obtained in 1083 of these women showed CR to be more accurate than self-reporting but associated with time-delays. Serum samples from 200 eligible breast cancer cases identified through the process and 400 matched-controls were analysed for oestradiol, free-oestradiol, oestrone, androstenedione, testosterone, free-testosterone, progesterone, dehydroepiandrosterone sulphate (DHEAS), sex steroid hormone binding globulin (SHBG), luteinising hormone (LH), follicle stimulating hormone (FSH) and oestrogen receptor-α and -β and androgen receptor SB. Results showed that sex steroid receptor SB assays could add to breast cancer risk prediction. Additionally, the best oestrogen for breast cancer risk prediction is oestrone and the best androgen is testosterone. High testosterone and FSH levels up to 5 years prior to diagnosis predict breast cancer with high power and may have a synergistic effect. In a separate case control study of 189 paraffin-embedded breast tissue samples, 55 genes were investigated using MethyLight. DNA methylation alterations were found to be homogeneous in breast cancer with 13 genes being predictive of the disease, suggesting that such changes could be useful as future biomarkers. Further studies (already underway) involve using high-throughput technology to analyse serum DNA methylation changes and correlate these with the observed serum hormonal changes and build better breast cancer risk prediction models.

Haematopoietic stem cell transplantation (HSCT) is an effective therapy for many malignant and non‐malignant diseases. However, adverse events such as infections, Graft‐versus‐Host Disease (GvHD) and relapse limit the wider use of HSCT. The identification of factors that predict the outcome of the transplant can be a vital tool to offer patients better options for treatment. Since there is a significant immunological contribution to the outcome of the transplant, it is of interest to ask to what extent the immune status of the patient prior to transplant might influence their subsequent recovery. After transplantation, the reconstitution of the T cell compartment in the patient relies on both expansion of existing T cells and the production of new ones via the thymus; in older patients this latter process is compromised by age‐dependent thymic involution. Improvements in thymus function leading to improvement in immune reconstitution after an HSCT may provide significant benefits, potentially reducing mortality from both infections and GvHD. In one study described in this thesis, an improvement in T cell reconstitution, in particular of the CD4 compartment, could be demonstrated after the administration of an LHRH agonist before transplantation from an allogeneic donor. Such an effect could not be demonstrated in an autologous setting. This may reflect a differential requirement for thymus function in the two transplant settings. In additional studies, various immune parameters including T cell subsets and cytokine profiles from the patient, that could affect the transplant outcome were analysed. The findings indicate that high levels of regulatory T cells (Tregs) as well as high levels of regulatory cytokines in patients pretransplant are factors that predict relapse after transplantation. It is likely that these act by suppressing anti‐tumour responses in the patient. These findings may provide a useful tool to stratify the patients into high and low risk categories prior to transplantation.

One of the key questions in developmental biology is how tissue growth is controlled to give rise to organs of specific sizes and shapes. Although some genes and pathways involved in the genetic and environmental control of tissue growth have been uncovered, the understanding of this process remains incomplete. In order to find new regulators of growth we carried out an in vivo RNAi screen in the Drosophila wing. I participated in the validation of candidate genes from the screen and identified the mask gene as an essential regulator of tissue growth acting in the Hippo signaling pathway. This pathway acts via the Yorkie (Yki)/Yes-associated protein (YAP) transcriptional co-activator to control tissue growth in both Drosophila and mammals. Yki/YAP translocates from the cytoplasm to the nucleus to activate target genes, a process that is negatively regulated by the Warts kinase, one of the core components of the Hippo pathway. I found that Mask is an essential positive regulator of Yki acting downstream of Warts. Mask is required for normal tissue growth, for the expression of Yki target genes and for the overgrowth phenotype caused by Yki overexpression. Mask binds to Yki and the two proteins translocate from the cytoplasm to the nucleus together in response to various stimuli. My results show that Mask acts in the nucleus to promote Yki target gene activation. Finally, Mask’s function appears to be conserved in humans, as two human homologues of Mask (hMask1 and hMask2) translocate with YAP to the cytoplasm upon cell contact inhibition, and we demonstrate that one of these homologues promotes YAP’s transactivation function.

A primary focus of the Molecular Neuropathobiology laboratory is the investigation of the long range trafficking of neurotrophins and neurotrophin receptors in motor neurons (MNs). The goal of my project was to deepen our understanding of the nature of the cellular machinery controlling long-range neurotrophin trafficking in MNs, by discovering new players involved in this process. In order to achieve this goal I performed an siRNA screen in MNs derived from mouse ES cells to monitor the cell surface binding and internalization of two fluorescently tagged reporters: the binding fragment of the Tetanus neurotoxin (HC), which enters an axonal transport compartment shared with neurotrophins and their receptors, and an antibody directed against the extracellular domain of the neurotrophin receptor p75NTR. A high-throughput, lipidbased siRNA transfection method was optimised for ES cell-derived MNs and used to screen a library of siRNAs directed against a pool of genes involved in endocytosis and membrane trafficking. The primary candidate genes were subsequently validated, and one gene in particular, Bicaudal D homolog 1 (BICD1), was selected for further analyses. BICD1 is a member of the Bicaudal D family, whose members function as molecular motor adaptors with pleiotropic roles in intracellular trafficking. Bicd1 expression at E12.5 and 13.5 was restricted to the nervous system, suggesting an important role for BICD1 in neurons. I used gene-trapped ES cells to derive MNs depleted of the BICD1 protein (Bicd1gt/gt MNs), which, when challenged with either HC or the p75NTR antibody, displayed an increased intracellular accumulation of both probes. Furthermore, I found that the level of neurotrophin receptors exposed at the plasma membrane was increased in these cells compared to their wild type counterparts, suggesting that BICD1 might be involved in the regulation of neurotrophin receptor dynamics in mammalian neurons. I also found that TrkB signalling upon stimulation with the brain-derived neurotrophic factor (BDNF) in Bicd1gt/gt MNs, was impaired. This suggests that the depletion of BICD1 not only affects the trafficking of neurotrophin receptors, but also their signalling capabilities. In conclusion, this thesis work has demonstrated that the concept of high throughput screening can be applied to cells notoriously difficult to handle and transfect, such as MNs. This approach was successful in unravelling a new role for BICD1 in neurons, where it appears to regulate the intracellular trafficking and signalling of neurotrophin receptors. Taken together with the in vivo expression data, these data suggest that BICD1 plays an important role in the development and function of the nervous system.

Amplification of the tyrosine kinase receptor ErbB2 in breast cancer correlates with disease progression, poor prognosis and recurrence. The mechanisms of downstream ErbB2 signalling and their effects on tumour progression remain ambiguous and thus the elucidation of pathways involved in ErbB2-dependent transformation is essential to realise novel diagnostic/prognostic markers and therapeutic targets. Previous profiling studies of an ErbB2 over-expressing cell system and tumour tissues identified gene products potentially implicated in ErbB2-dependent breast cancer. Candidate proteins were selected from the previous profiling experiments with respect to their relationship with ErbB2 over-expression and/or growth factor-dependent modulation. Particular interest was placed on poorly characterised proteins that were previously unknown to function in ErbB2-dependent malignancy. The protein expression profiles of selected candidates were established in a panel of HMLEC and breast cancer cell lines. Previous findings from the gene expression profiling studies, including the role of the candidate IGFBP3 in growth factor receptor interplay were validated at the protein level in a breast cancer cell line and/or an ErbB2 over-expressing cell system. Validated targets were functionally characterised by evaluating the effect of siRNA-mediated silencing of candidate expression on cellular invasion, proliferation, anchorage-independent growth and adhesion in ErbB2 over-expressing breast cancer cell lines. Prospective candidate interaction partners were identified by determining the effect of siRNA-dependent gene silencing on global protein expression using an in vivo quantitative mass labelling approach (SILAC) combined with LC-MS/MS technology. Bioinformatics analysis of data was used to define the functional consequences of siRNA-mediated knockdown and to link changes with cellular phenotype. The effects of ErbB2 knockdown were also evaluated to validate its role in cellular transformation and malignancy. Finally, novel downstream ErbB2 signalling targets and putative sites of phosphorylation were determined using a combination of SILAC labelling, phosphopeptide enrichment, LC-MS/MS and bioinformatics analysis in an ErbB2 over-expressing model cell system.

Over the past ten years significant effort has been put in the identification of pharmacological targets that facilitate the selective targeting of cancer cells. The concept of synthetic lethality in combination with RNA interference (RNAi) technology provides an attractive platform for the identification of such drug targets. With this in mind I designed, set up, and executed a large-scale siRNA screen aiming at identifying genes that exhibit a synthetic lethal relationship with loss of the PI3K/AKT signalling cascade negative regulator PTEN. A PTEN-isogenic cell system derived from the breast epithelial cell line MCF10A was employed in this study, and screened with a siRNA library against the “Druggable genome”. Putative hits exhibiting a potential synthetic lethal relationship with loss of PTEN were identified by differential Z-score analysis, and validated in a panel of breast cancer cell lines segregated based on their PTEN status. This analysis identified several PTEN-loss synthetic lethal candidate genes whose further evaluation may reveal new insights in the biology of PTEN null cancer cells. The functional mechanism underlying one of the identified PTEN-loss synthetic lethal putative hits (CYTH1/PSCD1) was investigated in detail. Knockdown of CYTH1 selectively induced apoptosis in the PTEN-/- MCF10A cells, and biochemical and genetic evidence supported a potential synthetic lethal relationship with PI3K/AKT pathway activation due to loss of PTEN. Although definite confirmation that CYTH1 was the sole target mediating the identified PTEN-loss synthetic lethal interaction was not obtained, further investigation on the exact nature of the described PTEN-loss synthetic lethal relationship may have the potential to uncover previously unknown vulnerabilities of PTEN-deficient cancer cells that could be pharmacologically exploited.

Combinations of common germline low-moderate susceptibility alleles may be responsible for some of the 90% of ovarian cancer (OC) cases not explained by known risk genes. These alleles may also affect survival of OC patients. The effects of 34 tagging single nucleotide polymorphisms (tSNPs) from candidate oncogenes (BRAF, ERBB2, KRAS, NMI and PIK3CA) and 63 tSNPs from “functionally” relevant genes (AIFM2, AKTIP, AXIN2, CASP5, FILIP1L, RBBP8, RGC32, RUVBL1 and STAG3) on the risk and survival of OC sufferers were evaluated with ~1,800 cases and 3,045 controls. Associations were found between disease risk and NMI rs11683487 (P-dominant=0.004) and RUVBL1 rs13063604 (P-trend=0.0192). These associations were not independently validated with additional samples, however, they remained significant when the results from both stages of genotyping were combined (P<0.05). Global tests of association with OC risk were significant for BRAF, ERBB2, CASP5 and RUVBL1 (P-global<0.05). However, there was no evidence of an excess of significant associations from 340 SNPs investigated with the admixture maximum likelihood test (P-trend=0.068). BRAF, FILIP1L, KRAS, RBBP8 and RUVBL1 were also associated with the survival of all OC cases (P<0.05). When analysis was restricted to the 4 main histological subtypes of OC, additional associations were identified. Although these results are of particular interest, they were based on relatively small numbers of samples and have not been corrected for multiple testing, therefore they should be treated with caution. The results from the secondary objective of the project, to evaluate whole genome amplification (WGA) of DNA and SNP multiplex platforms, are also described. To conclude, associations were identified between candidate oncogenes and functionally relevant genes on the survival and susceptibility of ovarian cancer. The performance of WGA DNA on SNP multiplex genotyping platforms highlighted the importance of comparing WGA DNA with corresponding gDNA in order to ascertain quality of genotyping on the platform.

The reduced engraftment potential of cytokine cultured haemopoietic stem/progenitor cells from adult mobilised peripheral blood has been associated with their defective homing to bone marrow niches. In this work, using established in vivo systems and a novel ex vivo model, an additional cytokine-induced attachment defect is described that reduces the retention of these cells in the bone marrow, post-transplantation. This defect was found to be related to specific niche ligands and was not caused by downregulation of their respective receptors on the expanded cells. CD26 is a protease that cleaves SDF-1 abrogating its chemotactic effect. CD26 inhibition on the transplanted cells was not sufficient to reverse the engraftment defect, although infusion of the inhibitor in immunodeficient animals, together with ex vivo treated cells, significantly increased engraftment. Finally, mobilised peripheral blood stem/progenitor cells were found to express neuroreceptors and their expression was altered after exposure to cytokines. Epinephrine pre-treatment of these cells rescued their adhesion to specific niche ligands, increased their short-term homing and improved their long-term engraftment in immunodeficient animals.

Elaborate DNA repair mechanisms have evolved, allowing cells to repair damages in their genomes. Nucleotide excision repair (NER) removes a variety of helix-distorting lesions, including those caused by ultraviolet (UV) irradiation. NER operates via two subpathways. Transcription-coupled repair (TC-NER) rapidly removes transcription-blocking lesions in the transcribed strand (TS) of active genes, and in the yeast Saccharomyces cerevisiae depends on the factors Rad26 and Rpb9. Lesions in untranscribed DNA, including the non-transcribed strand (NTS) of active genes are removed slower by global genome repair (GG-NER). Besides activating specific DNA repair systems, DNA damage also leads to a global cellular response, known as the DNA damage checkpoint (DDC). Cell-cycle progression is temporarily stopped after DNA damage to allow sufficient time for repair and prevent replication or segregation of damaged chromosomes. The DDC is a complex signal transduction cascade involving a number of protein kinases, the central players in budding yeast being Mec1 and Tel1, the homologues of human ATR and ATM, respectively. Besides inhibiting cell-cycle progression, accumulating evidence suggests that DNA repair systems are also influenced by the checkpoint. I have investigated the rates of repair of UV lesions in checkpoint deficient strains of Saccharomyces cerevisiae and found that NER is significantly inhibited on both strands of an active gene in the absence of Mec1. The effect on NTS repair seems to be caused by deficient de novo synthesis of repair factors, whereas TC-NER is influenced mainly by post-translational modification of one or more pre-existing proteins. I have characterised a checkpoint-dependent phosphorylation of Rad26, and have shown using point mutants that this phosphorylation increases the TC-NER capacity of cells, establishing a new link between NER and the checkpoint. In addition to these results about the interplay between the DDC and NER pathways, preliminary data from two unrelated projects will be presented. One was an attempt to establish a system for analysis of NER factor recruitment to an artificial, highly UV-damage-prone DNA sequence. The other focussed on the regulation of UV-induced degradation of Rpb1, the largest RNA Polymerase II (RNAPII) subunit, by the DDC.