Common anticancer strategies often fail resulting in tumour recurrences and metastasis that cause the majority of cancer-related deaths. Only about 6% of colorectal cancer (CRC) metastatic patients live more than 5 years after diagnosis. Thus, elucidating the crucial mechanisms underlying drug resistance and metastasis is an urgent need in cancer biology and therapy. The tumour suppressor gene FBXW7 (also called hCDC4, FBW7, Sel10, Ago) is one of the most frequently mutated genes in CRC. FBXW7, as a receptor subunit of the SCF-E3 ligase complex, drives the ubiquitination and degradation of many oncogenic proteins influencing cell proliferation, differentiation, senescence and other crucial pathways. Our group and others recently demonstrated that murine intestinal Fbxw7 loss of function induced tumorigenic activity, however its exact position in the multistep model of colorectal carcinogenesis remains to be clarified. More recent data from Dr Nateri’s laboratory (Li et al., under review) suggested that ZEB2, transcriptional inducer of epithelial-mesenchymal transition (EMT), represents a specific target of FBXW7 and FBXW7 loss induces an EMT-like phenotype. We have therefore proposed that FBXW7 suppression may confer resistance to chemotherapy, stem-like cell properties and invasive and metastatic activity via EMT, following accumulation of ZEB2. To test our hypothesis, we investigated FBXW7 role in the occurrence of drug resistance in human CRC cells and three-dimensional murine organoids, after treatment with 5-fluorouracil (5-FU) and oxaliplatin, DNA damaging chemotherapeutic agents. We explored EMT activation due to loss of FBXW7 by examining epithelial and mesenchymal markers in response to drug treatment as well as stem-like cell properties by colonosphere formation assay. EMT-associated properties were also analysed following stable down-regulation of ZEB2 in CRC cells and organoids. Moreover, to explore a putative involvement of FBXW7 through accumulation of ZEB2 in invasive and metastatic potential, we performed in vivo xenograft analyses by intra-splenic (for liver and lymph node metastasis) and intravenous (migration/invasion in lungs) injection of human CRC cells deleted for FBXW7 and stably down-regulating ZEB2. Herein, we show that CRC cell lines and murine organoids carrying deletion of FBXW7 are significantly less sensitive to 5-FU and oxaliplatin treatments. FBXW7 loss of function by accumulation of ZEB2 confers EMT phenotype to CRC cells including reduction of E-cadherin, up-regulation of Vimentin, stem-like properties and induction of invasive and metastatic activity. This suggest that FBXW7-deleted CRC cells may represent cancer stem cells. Moreover, this study also aimed to employ and establish an innovative in vitro model to evaluate the cell type-specific responses of gut epithelium to chemotherapeutic agents and how these responses are influenced by FBXW7. Chemotherapeutic sensitivity of the different cell types in normal gut and tumour tissues is poorly understood, impeding the development of improved targeted therapies. We have therefore established the organoid culture which mimics the structure of both gut and colorectal tumours comprising all the cell types likewise the human in vivo situation. To further investigate the mechanism(s) of action of Fbxw7, we grew organoids from isolated small intestinal crypts of mice in which fbxw7 gene is conditionally knocked-out in the intestine (fbxw7∆G), in parallel with wildtype floxed-fbxw7 (fbxw7fl/fl) organoids. The fbxw7∆G organoids exhibited rapid budding events in the crypt region. To test organoids for drug response, we exposed organoids from fbxw7fl/fl and fbxw7∆G mice to various concentrations of 5-FU for 72 hours. 5-FU triggers phenotypic differences in organoids including changing shape, survival, resistance, and death. 5-FU, however, rescues the drug resistance phenotype of fbxw7∆G through the induction of terminal differentiation. Our results support the hypothesis that a differentiating therapy successfully may target FBXW7-mutated CRC cells.

Background: Lymphangioleiomyomatosis (LAM) is a rare progressive neoplastic cystic lung disease that primarily affects women of child bearing age leading to lung destruction, respiratory failure and death. Thought to be a consequence of dysregulated protease expression, cells of unknown origin accumulate in the lung, often forming clusters or nodules of cells with both melanocytic and smooth muscle properties. Some of these cells, known as LAM cells, have bi-allelic mutations in TSC2 resulting in constitutive mTOR activation. However LAM nodules are heterogeneous structures and genotyping analyses suggest that cells without LOH for TSC2 including wild-type fibroblasts are also common within LAM nodules. Hypotheses and aims: We hypothesise that LAM cells recruit wild-type fibroblasts and modify their properties to generate a permissive microenvironment, akin to a tumour stroma including the production and activation of matrix-degrading proteases which contribute to the destruction of the lung parenchyma. This study has therefore deigned in vitro co-culture models with an aim to study the expression patterns and activation of proteases in a LAM lung leading to matrix destruction. Another aim was also to characterise transcriptional differences normal human lung fibroblasts (NHLFs) and LAM-associated fibroblasts (LAFs) and to investigate changes in their gene expression when cultured together with a model LAM cell line, 621-101 angiomyolipoma cells which were derived from a LAM patient and have bi-allelic loss of TSC2. Methods: In vitro 2-dimensional (2D) and 3-dimensionalD (3D) co-culture models were designed and validated using fibroblasts characterised and isolated from 4 LAM lung donors, now termed LAFs, and 621-101 cells. The 3D extracellular matrix (ECM) incorporated the two cell types in a 10:1 ratio embedded in a basement membrane extract (BME) mimicking the lung matrix. An organotypic spheroid model was also developed incorporating both cell types thereby mimicking a LAM nodule. 6 LAM lung and 3 normal lung tissue donors were screened for candidate proteases in LAM pathology using qRT-PCR and identified upregulated proteases which may contribute to a role in LAM pathology. These findings were verified in the 2D and 3D in vitro models as well as ex vivo tissue using a variety of immunostaining techniques, activity assays and ELISA. Lastly, commercially bought NHLF (n=3) and LAF (n=3) were cultured in the presence or absence of 621-101 cells in the 2D Boyden chamber co-culture model. LAF and NHLF RNA was analysed using Affymetrix Human Genome U133 Plus 2.0 Arrays and Genomics Suite and Pathway (Partek). Findings were validated by multiplex assay and immunohistochemistry in 2D and 3D in vitro models and tissue respectively. Inflammatory cell migration and function was examined in co-culture model and LAM tissue. Results: The 3D BME model showed that TSC2-/- 621-101 cells and fibroblasts spontaneously form aggregates and clump together akin to a LAM nodule. The two cell types exhibited strong heterotypic cell-cell adhesive forces and resulted in strictly spherical spheroids. The 3D models designed all showed expression of markers of LAM nodules thereby representing LAM nodules in a dish. Of 30 proteases screened, cathepsin K gene expression was increased almost 15-fold in LAM lung compared to normal tissue and was also found to be elevated in 3D BME model. Cathepsin K in LAM tissue was expressed in the LAM nodules associated with cysts and was expressed exclusively by fibroblasts in the 3D spheroid model. As cathepsin K requires low pH for activity it was determined if LAFs and TSC2-/- cells can acidify the extracellular space. TSC2-/- cells but not LAFs decreased extracellular pH, over 24 hours and pH values < 7 were associated with increased cathepsin K activity in co-cultures. TSC2-/- cells expressed membrane transporters associated with extra-cellular acidification and inhibitors of the sodium bicarbonate co-transporters, carbonic anhydrases and mTOR reduced the pH gradient and decreased CTSK activity in co-cultures. Transcriptomic analysis using the 2D co-culture model showed 148 genes were significantly altered in both NHLF and LAF by 621-101 cells. Soluble factors from 621-101 cells induce pro-inflammatory transcriptional changes in both NHLFs and LAFs and pathway analysis showed enhanced chemokine signalling which highlighted stimulation of mainly the C-X-C motif chemokines and chemokine receptor signalling. The analysis identified 6 C-X-C motif chemokines all possessing a cognate receptor. The gene and protein expression of these chemokines was validated in the in vitro models and in ex vivo LAM lung tissue. Conclusions: The in vitro models are versatile and mimic the LAM lung nodule and environment. A potent matrix degrading protease possibly playing a role in LAM has been identified and using the in vitro models a possible mechanism of activation of CTSK resulting from a synergistic relationship between TSC2-/- cells and LAFs has been demonstrated. Also, soluble factors from the TSC2-/- LAM cell line elicit changes in gene expression in co-cultured fibroblasts. Chemokine signalling is associated with cell migration; elevated chemokine expression may be associated with the recruitment of inflammatory cells to the LAM nodule. The identification of these mechanisms and pathways opens up new avenues for therapeutic interventions in LAM.

Introduction: Significant developments have been made in the treatment and prevention of cancer. Despite these developments, a third of the population of the UK will get cancer in their lifetime and a quarter of the population are likely to die from it, making it a leading cause of death in the UK. There is an increase in the numbers of cancers being largely driven by the ageing and expanding population. Conventional anti-cancer and anti-metastatic approaches such as chemotherapy and radiotherapy are effective because these approaches mostly inhibit cell division in proliferating cancer cells or make the tissue environment hostile to cancer cell growth and migration. However, the non-selectiveness of these approaches often causes serious side effects leading to the damage of healthy tissue. Significant gains have been made through the development of targeted therapies and early detection with the benefit of personalising medicine for a specific target thereby minimising harm. However, there is a problem of drug resistance in about 50% of patients with these forms of treatment. Progress in developing these treatments for disease remain slow, with traditional approaches inadequate and rational new therapies needed to tackle cancer. As a result, an alternative strategy for drug development such as using previously approved drugs for new medical indications is beginning to be explored. This strategy will potentially remove substantial risks, costs and time from the pathway of drug development. Recent evidence has shown the potential anti-neoplastic effects of some cheaper and safer medications. Among the best examples of this is the near 50% reduction in cancer specific mortality from colorectal cancer recently shown in those starting Aspirin after diagnosis, which has already led to a randomised controlled trial. In vitro and in vivo studies suggest a role for beta blockers and alpha blockers in inhibiting the proliferation and migration of cancer cells. A series of case-control and cohort studies was therefore conducted using the large population based Clinical Practice Research Datalink (CPRD) and associated datasets as the data source to explore the impact of beta and alpha blockers on cancer incidence and overall cancer mortality. Epidemiological studies on the effect of adrenergic blockers on cancer incidence have proved inconclusive with particularly limited evidence from large population based studies on cancer incidence. A case-control study was therefore conducted to assess the effect of adrenergic blockers upon incidence of cancers of the prostate, lung, bowel and breast. Additionally, epidemiological studies have examined the potential beneficial effects of adrenergic blocker on cancer survival, but these are still inconclusive with limited evidence from large population-based studies. Cohort studies were therefore carried out to examine the effect of beta and alpha blocker exposure post diagnosis on cancer specific and overall mortality. Furthermore, laboratory studies have demonstrated the effect of alpha blockers in reducing induced angiogenesis and suppress metastasis in mouse models. A study was therefore conducted to examine in detail the effect of alpha blockers on mortality outcomes in a cohort of prostate cancer patients and additionally considering their indication of use. Finally observational studies have investigated the anti-proliferative effects of beta blocker use on survival outcomes but not specifically in those without metastases who might be most likely to benefit. This study therefore investigated the effect of adrenergic blockers on mortality outcomes in a large population based UK cohort of non-metastatic colorectal cancer patients. The objectives of this thesis were: To test the hypothesis that: β-blocker use is associated with a reduced incidence of breast, lung, prostate and colorectal cancer. To test the hypothesis that: α-blocker use is associated with a reduced incidence of breast, lung and colorectal cancer. To test the hypothesis that: α-blocker use is associated with a reduction in cancer specific and overall mortality in prostate cancer patients. To test the hypothesis that: β-blocker & α-blocker use is associated with a reduction in cancer specific and overall mortality in patients with non-metastatic colorectal cancer. To test the hypothesis that: β-blocker use is associated with a reduction in cancer specific and overall mortality in those diagnosed with prostate, breast, lung and colorectal cancer. To test the hypothesis that: α-blocker use is associated with a reduction in cancer specific and overall mortality in those diagnosed with breast, lung and colorectal cancer. To test the hypothesis that: β-blocker & α-blocker use is associated with a reduction in cancer specific and overall mortality in those diagnosed with non-metastatic breast, lung and prostate cancer. Methods: A frequency matched case-control study was carried out using the Clinical Practice Research Datalink to assess the effect of adrenergic blockers upon incidence of prostate, lung, bowel and breast cancer. Amongst patients aged 18 years or older contributing at least 2 years of usable data between 01/01/1987 – 31/12/2012. Incident cases of relevant cancers and controls were selected and frequency matched 10:1 by age. Those with 2 or more prescriptions for alpha or beta blockers in the 2 years prior to cancer diagnosis were considered exposed and also assessed effect of the dose and duration of use. Logistic regression was used to adjust effect estimates for age, sex, smoking, alcohol use, and a number of potentially confounding co-morbidities and co-prescriptions. A cohort study of colorectal, lung, breast and prostate cancer patients was conducted and selected from linked UK Clinical Practice Research Datalink, Hospital Episode Statistics and National Cancer Intelligence data between 1998 and 2010. Beta blocker and alpha blocker exposure were assessed in the 6 months post cancer diagnosis and its effect on all cause and cancer specific mortality was assessed. Data were analysed using cox proportional hazards modelling. Confounding by age, sex, cancer stage, grade and important comorbidities and co-prescriptions was adjusted for. The indication of use, dose and pre-diagnosis exposure was also examined. Additionally, from the linked data sources above, a cohort of 3164 non-metastatic colorectal cancer patients was selected and conducted a cohort study using similar methods but additionally considering the cardio-selectivity of beta blocker drugs. Results: For the case-control study 18968 colorectal, 19082 lung, 21608 prostate and 29109 breast cancers were identified. There was no evidence of a protective effect of α or β blockade in lung and prostate cancer and found a slightly increased risk of colorectal and breast cancer in users. This was largely explained by the effects of confounding in a multivariate analyses with final OR estimates of lung, colorectal, breast and prostate cancer of 0.99, 95% CI [0.96-1.04]1.14, 95% CI [1.09 – 1.18]1.10, 95% CI [1.06 – 1.14]1.01, 95% CI [0.98-1.05] respectively for beta blocker exposure and 1.03, 95% CI [0.97 – 1.09]1.13, 95% CI [1.07 – 1.20]1.08, 95% CI [1.00 – 1.17] for alpha blocker exposure. Stratification by dose and duration did not reveal any statistically significant findings. For the cohort study of common solid cancers 15636 colorectal, 13646 lung, 23877 breast and 18654 prostate cancer patients were selected with a median follow up of 3.7, 0.6, 5.5 and 4.4 years respectively. There were no significant effects observed on all-cause mortality in any cancers and similarly no significant effects observed on cancer specific mortality in patients on betablockers compared to those who were not. For alpha blocker exposure, there were no significant effects observed on all-cause mortality in any cancers and similarly no significant effect on cancer specific mortality except in prostate (HR0.874, 95% CI [0.781 – 0.978]) and colorectal (HR1.878, 95% CI [1.108 – 3.182]) cancer patients. There were no clear significant effects observed by dose or prediagnosis exposure. Conclusion: In these large population-based case-control and cohort studies investigating the impact of beta and alphablocker use on cancer incidence and mortality, limited evidence was found to suggest that adrenergic blocker use prevents the incidence of common cancers. Indeed, a slight increased risk of colorectal and breast cancer was found which may reflect residual confounding and health seeking behaviours. Furthermore, beta or alpha blocker use post diagnosis was not associated with a decreased risk of cancer-specific or all-cause mortality in colorectal, lung or breast cancer patients or in those with non-metastatic colorectal cancer. However, our results do provide evidence that alphablockers are associated with a decreased risk of prostate cancer.

Nucleophosmin 1 (NPM1) is a multi-functional phosphoprotein, which shuttles between the nucleolus and the cytoplasm. It participates in many cellular processes including ribosome biogenesis and transport, centrosome duplication, and also contributes to the control of genomic stability. NPM1 interacts with many proteins including those that participate in DNA damage repair processes. C-terminal mutations in NPM1 occur in 35% of patients with acute myeloid leukaemia and are associated with a good prognosis. They are characterised by delocalisation of NPM1 into the cytoplasm. This may have a role in the sensitivity of mutant cells to chemotherapy via inactivation of the DNA damage response processes. Firstly, we aimed to identify potential differences in DNA repair in response to double-strand breaks (DSB) and alkylating inducers and then assess them through the comet assay in the wild-type and mutant NPM1 cell lines. The percentage and kinetics of DNA damage in both cell lines was identical, indicating that DNA lesions were repaired efficiently in both the wild-type and mutant cell lines. The γ-H2AX foci were also evaluated which increased to similar levels in mutant and wild-type cells lines after exposure to DNA damaging agents and decreased with similar rates when the cells were allowed 24 hours to repair the damage. Interestingly, following DNA damage, the amount of NPM1 increased significantly in NPMc+ cells, both in the nucleus and the cytoplasm, which was not seen in NPM1 wild-type cells. Next, we determined the subcellular localisation of APE1, which is a DNA repair enzyme in the base excision repair pathway and a transcriptional co-activator. APE1 has previously been shown to associate with NPM1. Using confocal microscopy, we found that the APE1 in both wild-type and NPMc+ mutant cell lines is predominately localised in the cytoplasm, while it is translocated into the nucleus after the cells were exposed to MMS, presumably to play a role in the DNA damage repair mechanism. As we could find no apparent difference in DNA repair between NPM1 mutated and wild type cells we went on to look at proteins involved in cell survival - BCL-2 and MCL-1. The NPMc+ mutant cell line expressed the highest level of mRNA BCL-2 and MCL-1 when compared to the NPM1 wild-type cell lines. The NPMc+ mutated cells have previously been shown to be sensitive to the effect of all-trans retinoic acid (ATRA). We looked to see whether ATRA has an effect on the anti-apoptotic proteins. The results demonstrate that the BCL-2 and MCL-1 levels were down-regulated to a greater level in the NPM1 mutant cell line than in the NPM1 wild-type cells. In patient samples, the BCL-2 and MCL-1 mRNA down-regulation was seen in 3/5 and 5/5 NPMc+ samples, respectively. Thus, these finding indicate that BCL-2 and MCL-1 mRNA expression is down-regulated following ATRA in NPMc+ mutant cells. Finally, siRNA was used to decrease levels of either total NPM1 or mutant NPM1 alone. Results showed that in the NPM1 mutant cells there was a down-regulation of BCL-2 mRNA, while in the NPM1 wild-type cells, no effect on the BCL-2 mRNA level was found. Furthermore, levels of P53 were up-regulated in the mutant NPM1 cells after knocking-down the total NPM1, whereas the wild-type cell line showed no change in the P53 level. These results provide evidence suggesting that down-regulation of NPM1 in the NPMc+ cell line increases total P53, possibly by interaction of ARF and HDM2, resulting in down-regulation of the BCL-2 in the mutated cells. In conclusion, perturbation of NPM1 subcellular localisation in NPMc+AML has no effect on the DNA repair mechanisms but we found differences in the anti-apoptotic proteins expression compared to NPM1 wild-type cells.

More people are surviving cancer, and for longer. As a result, ‘cancer survivorship’ is a key policy and research issue. This exploratory study aimed to describe, and further understanding of, the cancer experience of individuals living five years or more post-treatment. A multiple-case study design was adopted. A narrative interview and semi-structured follow-up interview were held with thirteen participants and their ‘significant others’. ‘Restorying’ of narratives was followed by a cross-case analysis to explore similarities and differences across cases. The utility of liminality as a framework for understanding cancer survivorship was then explored. Little et al. (1998) argue individuals live in a state of ‘sustained’ liminality that persists until end-of-life. However, some individuals may ‘transcend’ the liminal phase. Individuals diagnosed with breast, gynaecological, prostate, testicular and colorectal cancer, five to sixteen years post-treatment, took part. A legacy of lingering uncertainty was evident across cases. The most common manifestation was fear of recurrence. A typology of the place of cancer was developed. Cancer was situated in the past, past-present or present-future. However, the place of cancer is not static; it oscillates between the past, present and future, and foreground and background of participants’ lives as a result of the lingering uncertainty and various ‘reality checks’ experienced. Most, but not all, participants live in a state of ‘sustained’ liminality. Perceiving the five-year survival marker as a ‘milestone’ is key to transitioning out of the liminal state. Some participants have put cancer (the disease) in the past, but consequences of treatment result in them living in an on-going state of physical liminality. Others perceive they are living with cancer within them and experience liminality existentially. However, whilst liminality is often construed negatively, it can be a catalyst for positive change to self. Implications for practice are positioned within the context of new models of ‘aftercare’ implemented in England. Holistic needs assessments at key transition points along the survivorship trajectory are crucial to providing tailored care within the context of individuals’ wider lives.

This thesis describes three different strategies employed with the aim of optimising targeted antibodies for the treatment of metastatic solid tumours. Whilst the search for improved predictors of response to anti-EGFR antibodies continues, paired primary and metastatic archived tissue from 32 patients with metastatic colorectal cancer was explored for the immunohistochemical expression of EGFR, pEGFR and pMAPK and activating mutations in KRAS, BRAF and PI3KCA. The resulting discordance between expression of pEGFR and pMAPK between primary and metastatic tissue CRC suggests they are unlikely to be useful biomarkers for response unless metastatic tissue is also analysed. Confirmation that mutations in KRAS, BRAF and PI3KCA are concordant in primary and metastatic tissue supports the analysis of archived primary tissue alone for mutation screening. PI3KCA mutations were shown to be present in patients with both wild-type and mutant KRAS, which provides both an additional method for resistance in wild type tumours and a mechanism for high resistance in those with mutant primary tumours, suggesting screening patients for all 3 mutations should be encouraged for future trials of anti-EGFR antibodies. The Phase I biodistribution study of Ley targeting immunoconjugate in advanced epithelial cancers, primarily explored the biodistribution and pharmacokinetics of the immunoconjugate CMD-193 (a humanised anti-Ley antibody conjugated with calicheamicin) in 9 patients with advanced Ley expressing solid tumours. Cycle one was trace labelled with 111In for biodistribution assessment, and subsequent cycles were administered every 3 weeks, to a maximum of 6 cycles, depending on toxicity and response. Tumour targeting was assessed using gamma camera imaging and single photon emission computerised tomography (SPECT), and PK analysis was based on gamma counting of 111In-CMD-193. There were 2 dose cohorts (1.0mg/m2 and 2.6mg/m2), and patients with Ley positive, measurable, advanced and treatment refractory malignancies, were eligible. Nine patients (6 in dose cohort 1, 3 in cohort 2) were enrolled (and received 1-6 cycles of treatment) before the study was terminated. Biodistribution imaging demonstrated initial blood pooling, followed by markedly increased hepatic uptake by day 2 (persisting to day 8), and fast blood clearance. This pattern was seen for all patients and dose levels. There was no significant uptake in tumour visualised in any patient. The overall T 1/2 β of 111In-CMD-193 was 102.88 ± 35.67 hours, with no statistically significant difference between the 2 dose levels. One patient had a partial metabolic response on 18F-FDG PET after 4 cycles, but no radiologic responses were observed. Myelosuppression and effects on liver function were the most significant toxicities, but no severe or unexpected toxicities were observed. The result of this trial highlight the importance of biodistribution and pharmacodynamic assessment in early phase studies of new biologics to assist in clinical development. The Phase I trial of oral capecitabine combined with 131I-huA33 in patients with metastatic colorectal cancer built on the previous development of the humanised antibody huA33 which targets the A33 antigen, known to be expressed in >95% of human colon cancers. This study used radiolabelled huA33 in combination with capecitabine chemotherapy to target chemoradiation to metastatic colorectal cancer, with safety and tolerability being the primary objective. Pharmacokinetics, biodistribution, immunogenicity, and tumour response were also assessed. Eligibility included measurable metastatic colorectal cancer, adequate hematological and biochemical function, and informed consent. An outpatient scout 131I-huA33 dose was followed by a single therapy infusion one week later, when capecitabine was commenced. Dose escalation occurred over 5 dose levels. Patients were evaluated weekly, with tumor response assessment at the end of the 12-week trial. Tumour targeting was assessed using gamma camera and single photon emission computerised tomography (SPECT) imaging. Nineteen patients were enrolled, and although the dose escalation protocol required an amendment following 2 dose-limiting toxicities in the second cohort, subsequent cohorts demonstrated good tolerability. Biodistribution analysis demonstrated excellent tumour targeting of the known tumour sites, expected transient bowel uptake, but no other normal tissue uptake. 131I-huA33 therefore achieves specific targeting of radiotherapy to sites of metastasis and can be safely combined with chemotherapy, providing a promising opportunity to deliver chemoradiation specifically to metastatic disease in colorectal cancer patients.

*Please note: The abstract, acknowledgments, list of publications, table of contents and abbreviations in this PDF file appear in a different order to that of the print version of this thesis. Cancer patients, treated with systemic adjuvant chemotherapy, have described experiencing persistent deteriorations in cognition. The nature of these effects is unclear, and although a wide range of theories have been advanced, there is currently no treatment. This thesis uses an animal model to investigate the effects of a commonly prescribed chemotherapeutic agent, 5-fluorouracil (5-FU). The cognitive effects of 5-FU were examined using two behavioural tests, the object location recognition test (OLR) and the conditioned emotional response test (CER) both of which require input from the hippocampus, a brain region associated with memory. Memory consolidation by the hippocampus requires the continual production of new neurons (adult neurogenesis) from progenitor cells in the sub granular zone (SGZ) of the dentate gyrus. As an anti mitotic agent, 5-FU could be reducing the cell proliferation required for neurogenesis and this could be a cause of the cognitive deterioration. This hypothesis was tested by quantifying the numbers of proliferating cells (Ki67+) in the SGZ in sections together with the levels of doublecortin (DCX), a neurofilament expressed in developing neurons and brain- derived neurotrophic factor (BDNF), a factor required for new neuron survival and synaptic plasticity, by Western blotting. After developing the methodology (chapter 2); adult male Lister Hooded rats were given five i.v injections of 5-FU (25mg/kg) over a two week period and their behaviour and cellular aspects of the hippocampus compared with saline injected controls (chapter 3). 5-FU treated animals showed significant impairments in their performance of both the OLR and CER behavioural tests. Animals were sacrificed after the behavioural tests were performed and analysis showed they had significantly reduced numbers of dividing cells in the SGZ and non significant reductions in the levels of BDNF and DCX within the hippocampus. These results demonstrate that 5-FU treatment can produce cognitive impairments in this animal model which are similar in nature to those described by patients after chemotherapy. These behavioural changes are correlated with a reduction in the cell proliferation required for hippocampal neurogenesis providing support for the hypothesis that chemotherapy drugs are affecting this aspect of hippocampal function. In order to develop a treatment for the cognitive effects of chemotherapy the antidepressant fluoxetine was co-administered with 5-FU (chapter 4). This approach was based on recent evidence that fluoxetine can increase neurogenesis and protect neurons after damage. As with the experiment described above, performance in the CER test was impaired by five injections of 5-FU (25 mg/kg) as compared with saline treated controls. Similarly, animals treated with six injections of 5-FU (20mg/kg) were unable to discriminate between objects in novel and familiar locations in the OLR task. However co-administration of fluoxetine in drinking water (10mg/kg/day) for three weeks, starting a week before 5-FU treatment, prevented the impaired performance of this task found in the 5-FU only group. 5-FU chemotherapy caused a significant reduction in the number of proliferating cells in the SGZ compared to controls but this reduction was eliminated in the group co administered with fluoxetine. Fluoxetine on its own had no effect on proliferating cell number or behaviour. Moreover hippocampal BDNF or DCX protein levels in the co-treated group (5-FU+fluoxetine) were significantly increased compared to the 5-FU only treated group. These findings suggest that while 5-FU can negatively affect cell proliferation and hippocampal dependent memory, these deficits can be reversed by co- administration of fluoxetine. To understand the long term effects of chemotherapy, the cellular effects of 5-FU treatment were quantified one day, 2 and 6 weeks after the end of two weeks of 5-FU (20mg/kg) treatment (chapter 5). The results showed that 2 weeks of 5-FU treatment did not significantly reduce cell proliferation in the SGZ when quantified one day after the end of treatment. However proliferating cell numbers were significantly reduced compared to controls two and six weeks after the end of treatment. This suggests that 5-FU has a delayed effect on cell proliferation with its maximum effect two weeks after the end of treatment. Cell survival was quantified by BrdU labelling cells immediately prior to 5-FU treatment, and quantifying the numbers of BrdU positive cells at the different time points. BrdU+ cell numbers were significantly reduced at the end of treatment and continued to decline at 2 weeks but stabilised by 6 weeks. These results demonstrate that 5-FU has prolonged effects on neurogenesis after the end of chemotherapy treatment. The effects of 5-FU on cognition and neurogenesis are discussed and correlated with chemotherapy treated patient reports of continued cognitive impairment for months or years after completion of chemotherapy treatment.

Rationale: CMF (cyclophosphamide: CP; methotrexate: MTX; 5-fluorouracil: 5-FU) is a chemotherapy combination associated with the cognitive impairments which many cancer patients experience after treatment. A reduction in hippocampal neurogenesis is a known means by which cytotoxic drugs alter cognition and is the mechanism investigated in the present study. There is currently no way of treating or preventing the cognitive deficits produced by chemotherapy and a simple pharmacological approach to achieving this could potentially have significant benefits for patients. Objectives: The studies in the present thesis use an animal model to investigate the effects of the individual agents in the CMF combination on spatial working memory and the proliferation and survival of neural precursors involved in hippocampal neurogenesis. It was also investigated whether the cognitive impairment produced by chemotherapy could be reversed or prevented by the antidepressant fluoxetine. Methods: In 4 separate experiments, adult male Lister-hooded rats were chronically administered with CP (30mg/kg, 4 or 7 i.v. doses), MTX (75mg/kg, 2 i.v. doses) or 5-FU (25mg/kg, 5 i.p. doses). Some rats were co-administered with fluoxetine (10mg/kg/day, in drinking water) for different time periods. Spatial memory was tested using the novel location recognition (NLR) task and the spontaneous alternation in the T-maze memory tasks. Proliferation and survival of hippocampal cells was quantified using immunohistochemistry and the levels of doublecortin (DCX) and brain-derived neurotrophic factor (BDNF) were quantified in the hippocampus and frontal cortex using Western blotting. Neural stem cells (NSC) were also isolated from the adult mouse hippocampus, to examine the direct effects of 5-FU, fluoxetine and its active metabolite, norfluoxetine (0.01-100µM) in vitro. Results: Rats treated with 5-FU and MTX showed impairment in the NLR task but not the spontaneous alternation in the T-maze task. They also exhibited a reduction in cell proliferation (Ki67-positive cells) and survival (BrdU-positive cells) in the dentate gyrus, compared to saline treated controls, but no difference was seen in the levels of DCX of BDNF. The induced cognitive and cellular impairments were not seen when fluoxetine was co-administered with the chemotherapy. The impairments caused by 5-FU were counteracted when fluoxetine was co-administered before and during 5-FU treatment but not when it was only administered after treatment. CP did not impair performance in the NLR task or hippocampal cell proliferation; however it significantly reduced cell survival. 5-FU, fluoxetine and norfluoxetine all decreased cell viability in vitro. Conclusions: These results demonstrate that MTX and 5-FU have more pronounced effects on spatial memory and hippocampal cell proliferation than CP in the CMF combination. Furthermore these impairments can be reversed by fluoxetine in a mechanism of prevention but not recovery. Although further work is required, it would be beneficial to establish an in vitro model of chemotherapy-induced cognitive impairment to justify this conclusion and to investigate the potential benefits of fluoxetine in cancer patients.

Cachexia is common in patients with incurable cancer, particularly of the lung and upper-gastrointestinal tract, and impacts adversely on treatment options, morbidity, quality of life and survival. Current management of cancer cachexia is inadequate and progress is required. This thesis explores the use of exercise as a proactive supportive therapy with a focus on maintaining physical function. The first piece of work was a systematic review of the use of therapeutic exercise in patients with or cured of cancer. Across 65 exercise studies, the median [IQR] rates of uptake, adherence and completion were 63 [33–80]%, 84 [72–93]% and 87 [80–96]% respectively, with no characteristic influencing the proportion of patients taking up or completing a programme. The main reasons reported for refusal were lack of interest or the impracticality of the programme and for withdrawal were medical complication or deterioration. Overall, only about half of patients offered an exercise programme completed one. This review highlighted a need to modify existing programmes or explore novel alternatives if exercise is to be acceptable to the majority of patients. The second study explored exercise preferences in patients with incurable cancer. A questionnaire was used to determine patients’ perceived capability and preparedness to undertake six different exercise programmes, each illustrated by video clips and accompanying text, and preferences for the delivery of the most preferred programme. All 200 patients considered themselves physically capable of undertaking an exercise programme and two-thirds were prepared to undertake one at that moment in time. The most preferred type of exercise was neuromuscular electrical stimulation (NMES) 36 [35−44]%, followed by walking 22 [16−30]% and resistance training 19 [13−26]% and the majority preferred to undertake exercise at home, alone and unsupervised. This survey suggested that it is realistic to offer therapeutic exercise programmes to patients with incurable cancer and provided rationale to explore NMES in this group. The third study was a randomised controlled pilot study of NMES in patients with non-small cell lung cancer. Sixteen patients were randomised to a control group, which received usual care, or NMES group, which received daily stimulation to the quadriceps for up to 30min (frequency 50Hz, on phase 11−25%) for four weeks. All patients found the NMES device acceptable and median (range) adherence to the recommended programme was 80% (69-100). In the NMES group, quadriceps muscle strength and free-living physical activity improved by a mean of 7.4 Newton metres (22%) and 136 steps (11%) respectively, whilst exercise endurance deteriorated by a mean of 20 metres (4%). This compared favourably with the control group however none of the differences were statistically significant. These findings suggested NMES was an acceptable type of exercise and that further study is warranted in patients with lung cancer. The final piece of work was a feasibility study into the use of a lightweight ActivPAL™ monitor to measure physical activity level. The aims were to determine if this form of assessment is acceptable to patients, the optimal period of monitoring and to explore the added value of the monitor's energy expenditure (EE) estimate over a simple step count. Sixty patients with lung or upper-gastrointestinal cancer wore a monitor for one week. All but one found the monitor acceptable and mean [95% CI] adherence was 98 [94−100] %. Mean daily step count and EE values measured over 2 and 4 days were significantly higher than those from 6 days (p<0.01). Step count was strongly related to stepping EE and non-stepping EE. The ActivPAL™ monitor was shown to be an acceptable method of assessing physical activity level. A mean daily step count obtained over 6 days was recommended for use in future cachexia studies. Collectively, this work supports the use of therapeutic exercise and highlights a particular role for novel approaches, e.g. NMES, which may be more acceptable to patients. Findings can be used to guide future research which ultimately will determine if therapeutic exercise can help patients with cancer to maintain their level of physical activity and independence for as long as possible.

Modulation of DNA base excision repair (BER) has the potential to enhance response to chemotherapy and improve outcomes in tumours such as melanoma, glioma and pancreatic cancer. APE1, a critical protein in BER that processes potentially cytotoxic abasic sites (AP sites), is a promising new target in cancer. In the current study, my aim was to develop small molecule inhibitors of APE1 for cancer therapy. An industry-standard high throughput virtual screening strategy was adopted. The SYBYL8.0 (Tripos, St Louis, MO, USA) molecular modelling software suite was used to build inhibitor templates. Similarity searching strategies were then applied using ROCS 2.3 (Open Eye Scientific, Santa Fe, NM, USA) to extract pharmacophorically related subsets of compounds from a chemically diverse database of 2.6 million compounds. The compounds in these subsets were subjected to docking against the active site of the APE1 model, using the genetic algorithm-based programme GOLD2.7 (CCDC, Cambridge, UK). Predicted ligand poses were ranked on the basis of several scoring functions. The top virtual hits with promising pharmaceutical properties underwent detailed in vitro analyses using fluorescence-based APE1 AP-site cleavage assays and counter screened using endonuclease IV AP-site cleavage assays, fluorescence quenching assays and Whole cell extract AP-site cleavage assays. Biochemical APE1 inhibitors were then subjected to detailed cytotoxicity analyses. Several specific APE1 inhibitors were isolated by this approach. The IC50 for APE1 inhibition ranged between 50 nM and 25 µM. I also demonstrated that APE1 inhibitors lead to accumulation of AP sites in genomic DNA and potentiated the cytotoxicity of alkylating agents in melanoma, glioma and pancreatic cancer cell lines. I have also shown that APE1 inhibitors induce delay in cell cycle progression and caused delay in cancer cell growth. I also demonstrated that APE1 knockdown by shRNA results in decrease cancer cell growth and enhanced cell killing by alkylating agent in Glioma cell line. This PhD project provides evidence that APE1 is an emerging drug target and could have therapeutic application in patients with melanoma, glioma and pancreatic cancer.