Angiogenesis, the growth of new blood vessels, is vital to tumour growth. Prevailing dogma has been that tumours cannot grow without angiogenesis. Based on this premise, anti-angiogenic drugs are used clinically. However, the principle of angiogenesis as an absolute requirement for tumour growth has been challenged with reports that many tumours are entirely or partially non-angiogenic. This study describes and quantifies characteristics of non-angiogenic non-small cell lung tumours, demonstrates non-angiogenic growth in small-cell/neuroendocrine lung tumours and investigates the underlying pathogenetic processes by comparison with angiogenic lung tumours. Hypoxia is an important stimulus for angiogenesis. Differences in response to hypoxia may determine whether a tumour produces new vessels. In order to test this, levels of. necrosis, often considered a surrogate marker of hypoxic stress, were quantified but no difference in quantity of necrosis was found Moreover, immunohistochemical investigation of hypoxia and angiogenesis factors provided no unambiguous explanation for the differences in angiogenesis. Significant differences were seen, however, in fibrosis and inflammation, which were both greater in angiogenic tumours. Differences were greater for lymphocytes rather than cells of the ‘innate’ immune system. This provided an alternative hypothesis: angiogenesis occurs during wound healing and in the growth of granulation tissue, so it is possible that tumour angiogenesis is a response to factors produced by immune cells rather than the tumour itself. A tumour’s angiogenic status may, therefore, be determined by the response it provokes from the immune system. Further work to test this theory would compare levels of immunogenic factors such as Tumour Necrosis Factor and tumour cell surface antigens such as the HLA class I molecules. The study concludes with an investigation into the molecular basis of non-angiogenic growth using the technique of comparative genomic hybridisation (CGH) which allows amplifications and deletions of areas of DNA to be calculated. High-resolution array CGH was evaluated against conventional CGH, and the results compared with previous RNA studies from our laboratory. These revealed a set of genes with consistent changes in both RNA and DNA, several of which form part of known angiogenic and inflammatory pathways.

Understanding the perfusion of blood and drugs in tumours is fundamental to foreseeing the efficacy of treatment regimes and predicting tumour growth. In particular, the dependence of these processes on the tumour vascular structure is poorly established. The objective of this thesis is to derive effective equations describing blood, and drug perfusion in vascular tumours, and specifically to determine the dependence of these on the tumour vascular structure. This dependence occurs through the interaction between two different length scales - that which characterizes the structure of the vascular network, and that which characterizes the tumour as a whole. Our method throughout is to use homogenization as a tool to evaluate this interaction. In Chapter 1 we introduce the problem. In Chapter 2 we develop a theoretical model to describe fluid flow in solid tumours through both the vasculature and the interstitium, at a number of length scales. Ultimately we homogenize over a network of capillaries to form a coupled porous medium model in terms of a vascular density. Whereas in Chapter 2 it is necessary to specify the vascular structure to derive the effective equations, in Chapter 3 we employ asymptotic homogenization through multiple scales to derive the coupled equations for an arbitrary periodic vascular network. In Chapter 4, we extend this analysis to account for advective and diffusive transport of anticancer drugs delivered intravenously; we consider a range of reaction properties in the interstitium and boundary conditions on the vascular wall. The models derived in Chapters 2–4 could be applied to any drug type and treatment regime; to demonstrate their potential, we simulate the delivery of vinblastine in dorsal skinfold chambers in Chapter 5 and make quantitative predictions regarding the optimal treatment regime. In the final Chapter we summarize the main results and indicate directions for further work.

TRAP1 is a Heat Shock Protein (HSP) chaperone to retinoblastoma but also associated to the tumor necrosis factor receptor. HSPs are primarily up regulated in cancer. Work in our lab noted a down regulation of TRAP1 in some non-small cell lung cancers compared to normal lung. The first aim of this project was to evaluate the effect of the loss of TRAP1 on cell proliferation using a spheroid model. The presence of TRAP1 in spheroids promoted cell proliferation and a faster onset of hypoxia. This suggests an oncogenic role for TRAP1 since rapid hypoxia development equates to poor prognosis. Micro array analysis showed that TRAP1’s loss was associated with increased transcrpition of the Junctional Mediating and Regulatory protein (JMY). JMY possesses an oncogenic property due to its ability to facilitate cell motility. Additionally it has tumor suppressor activity in promoting p53 activation. The second aim of this project was to produce an anti-JMY antibody and use it to characterize JMY and additionally verify the association between TRAP1 and JMY. JMY was found to be widely expressed in normal tissues and in many types of tumors. In neoplastic tissues, comparing primary versus metastatic tumors, JMY was found to have significantly higher expression in the metastatic compared with the primary tumors. A pilot study showed that nuclear co-expression of JMY and P53 was associated with shorter overall survival suggesting that a possible tumorigenesis mechanism could be via a deregulation/mutation of JMY/p53 or both. Finally, using 3 dimensional constructions, I demonstrated the distinct morphological difference between an angiogenic tumor and a non-angiogenic tumor. Additionally, I showed a characteristic cytoplasmic p53 sequestration in the non-angiogenic phenotype that is absent in the angiogenic phenotype. This could be the mechanism that the non-angiogenic tumor uses to adapt to hypoxia. This would imply that there is a potential for cancers to escape therapy by switching between these 2 phenotypes.

ASPP2 has been identified as a haploinsufficient tumour suppressor in mice, and an activator of the apoptotic function of the p53 family. Yeast two-hybrid experiments have also shown that ASPP2 interacts with a large number of proteins involved in other major signalling pathways. The mechanism(s) of action of ASPP2 are therefore complex, and likely to involve more than just the stimulation of the apoptotic programme. Since a study previously conducted in our laboratory revealed that the deletion of ASPP2 in vivo leads to severe hydrocephalus in the J129/C57BL6 background (Vives et al., 2006), it can be hypothesised that ASPP2 safeguards the normal development of the mammalian central nervous system (CNS), in addition to its role as a tumour suppressor. Deletion of ASPP2 leads to the development of hydrocephalus, most probably by affecting tight junctions (TJs) in the choroid plexus, thereby impairing its blood-cerebrospinal fluid (CSF) barrier function. TJ defects are likely to be the underlying cause of the loss of cell polarity observed in the neuroepithelium of several areas of the CNS. As cell polarity plays a key role in multiple aspects of CNS development, ASPP2 appears to be required for the proper lamination of the cerebral cortex and retina.

Background Trastuzumab resistance imposes a major limitation to the successful treatment of HER2 positive breast cancer. The expression of HER4 and its prognostic value is controversial in breast cancer. Furthermore, its role in trastuzumab treatment and resistance in HER2 positive breast cancer has not been reported. Methods The effects of trastuzumab on HER4 cleavage and its localisation were studied in both parental and trastuzumab-resistant SKBR3 and BT474 cells using western blot, RT-PCR, nuclear fractionation and confocal microscopy. Tissue microarrays consisting of a cohort of HER2 positive breast cancer patients were stained for HER4 by immunohistochemistry and the results were correlated with patients’ outcome. This study also assessed HER4 expression in the tumor samples from a window study of trastuzumab alone or in combination with neoadjuvant chemotherapy in HER2 positive breast cancer patients. Results Trastuzumab treatment upregulated HER4 mRNA, and increased expression of both 80 and 180 kDa HER4 protein isoforms, and induced nuclear translocation of 80kDa HER4 protein isoforms, which the results similar to heregulin stimulation. This was also seen in trastuzumab resistant cells although HER4_80kDa and nuclear HER4 decreased upon overnight withdrawal of trastuzumab in resistant cell lines. In addition, knockdown of HER4 protein expression by specific siRNAs increased trastuzumab sensitivity and reversed trastuzumab resistance in SKBR3 and BT474 cells, confirming the importance of HER4 in trastuzumab response. This study also showed that trastuzumab-induced HER4 nuclear translocation is due to HER4 activation and cleavage since γ-secretase inhibitor (GSi) and neratinib prevented the process when combined with trastuzumab treatment, correlating with an increased apoptosis and decreased cell viability. There was also increased nuclear HER4 expression in tumors from both BT474 xenografts and from patients with breast cancer treated with trastuzumab monotherapy. Furthermore, nuclear HER4 predicted poor clinical response to trastuzumab monotherapy in patients undergoing a window study and was a poor prognostic factor in HER2 positive breast cancer. Conclusions This study suggests HER4 activation, cleavage and nuclear translocation play a key role in trastuzumab resistance in HER2 positive breast cancer. Nuclear HER4 could be a novel predictive and prognostic biomarker in HER2 positive breast cancer patients.

Cancer cells survive the harsh oxygen and nutrient deprivation of the tumour microenvironment through the selection of apoptosis-resistant and glycolytic clones (Cairns et al., 2011; Graeber et al., 1996). In particular, the integrated stress response (ISR) has been shown to be pivotal in cancer cell survival in vivo and the resistance of cancer cells to therapy (Harding et al., 2003). In recent years, it has become apparent that increased autophagy is one mechanism by which the ISR can confer resistance to stress (Kroemer et al., 2010). ATF4 is a major transcriptional effector of the integrated stress response in severe hypoxia (<0.01% O₂). ATF4 is a well-established regulator of genes involved in oxidative stress, amino acid synthesis and uptake, lipid metabolism, protein folding, metastasis, and angiogenesis. Recent work has demonstrated an important role of ATF4 in promoting resistance to severe hypoxia through the transcriptional upregulation of MAP1LC3B and ATG5, essential components of the autophagy machinery (Rouschop et al., 2009b; Rzyski et al., 2010). In this work, the author describes several novel ATF4 target genes, and examines their role in the regulation of autophagy and the resistance of cancer cells to severe hypoxia. In the first part of this thesis, the author shows that three BH3-only members of the BCL-2 family of proteins--HRK, PUMA, and NOXA--are upregulated in response to severe hypoxia in an ATF4-dependent manner. In particular, the author shows that the poorly described BH3-only protein HRK is a direct target of transcriptional activation by ATF4, and that HRK induces autophagy in severe hypoxia, thereby providing the first evidence that the integrated stress response can transcriptionally trigger the autophagy process. In contrast to the previously described role of HRK in apoptosis, this thesis demonstrates that HRK can play a pro-survival role in the context of breast cancer cells. In the latter part of this thesis, the author identifies the essential autophagy gene ULK1 as an ISR target. The author shows that ULK1 expression in severe hypoxia is transcriptionally upregulated through direct activation by ATF4. The author identifies ULK1 as a crucial regulator of autophagy and mitophagy in both normoxia and severe hypoxia and shows that ULK1 plays a pivotal role in cancer cell survival. Furthermore, it is shown that human breast cancer patients with high levels of ULK1 relapse earlier than those with low levels of ULK1, thereby identifying ULK1 as a potential target for cancer therapy.

In developed countries, breast cancer is the commonest malignancy among women. Understanding the mechanisms involved in breast cancer progression and the influence of the microenvironment on cancer cell proliferation, results in better treatments. This study aimed to optimise breast cancer xenograft rates using a novel chamber developed for tissue engineering purposes. The established tumours were subjected to enzyme digestion, creating a single cell suspension, which was then injected into immunocompromised mice at primary, metastatic and intra-cardiac sites. The resulting tumours in the mammary fat pad (MFP) and bone were compared using species-specific reverse-transcription polymerase chain reaction (RT-PCR) and cDNA microarray, to examine the influence of the microenvironment on gene expression. The achieved xenograft graft rates of 25% were similar to those previously reported. The matrix metalloproteinase family of enzymes (MMPs) degrade extracellular matrix, influencing invasion and migration of malignant cells. RT-PCR results showed that the majority of the MMPs expressed in the cancers were stromal rather than tumour in origin. MT1-MMP, MMP-2 and MMP-11 had significantly higher expression levels in the MFP than in the bone, but MMP-9 was expressed more in the bone than MFP. There was also an up-regulation of stromal production of MT1-MMP and MMP-13 in the MFP in the presence of tumour. This may have significance when considering which MMPs are the most appropriate targets for inhibition during cancer treatment. The most significant of the differentially expressed genes on microarray analysis were trefoil factor 1 (TFF1) and insulin growth-factor binding protein 3 (IGFBP-3), both expressed significantly more in tumours from the MFP than the bone. The thesis presented demonstrates some of the complexities of tumour-stromal interactions and supports Paget’s seed-soil theory, confirming in several ways the variation in gene expression in breast cancer between primary and metastatic sites.

The aim of the study was to identify molecular markers that can determine progression of low grade glioma. This was done using various approaches such as IDH1 and IDH2 mutation analysis, MGMT methylation analysis, copy number analysis using array comparative genomic hybridisation and identification of differentially expressed miRNAs using miRNA microarray analysis. IDH1 mutation was present at a frequency of 71% in low grade glioma and was identified as an independent marker for improved OS in a multivariate analysis, which confirms the previous findings in low grade glioma studies. IDH1 mutation was associated with MGMT promoter methylation when partially methylated tumours were grouped with methylated tumours. Grade II and grade III tumour comparison analysis revealed 14 novel significant miRNAs with differential expression. A miRNA signature was shown for histological subtypes, oligoastrocytoma and anaplastic oligoastrocytoma, following the miRNA expression analysis in grade II and grade III tumors based on histology. Oligoastrocytoma presented a more similar profile to oligodendroglioma, but anaplastic oligoastrocytoma was more similar to anaplastic astrocytoma. Five novel miRNAs were identified in grade III tumours, when comparing IDH1 mutant and IDH1 wild type tumours. Analysis of paired samples of primary/recurrent tumours revealed that additional genomic changes may promote tumour progression. For each of the pair, the two samples were genomically different and in each case, the reccurent tumours had more copy number aberrations than the corresponding primary tumours. Cell cultures derived from the tumour biopsies were not representative of the low grade glioma in vivo, which was evident from the differences identified in the miRNA expression and copy number changes in the paired samples. IDH1 mutation present in tumour biopsies was not maintained in their respective cell cultures. These findings give an insight into the molecular mechanisms involved in the tumourigenesis of low grade glioma and also tumour progression.

Cancer is a generic term for a large group of diseases where cells of the body lose their normal mechanisms for growth so that they grow in an uncontrolled way. One of the most common treatments of cancer is chemotherapy that aims to kill abnormal proliferating cells; however normal cells and other organs of the patients are also adversely affected. In practice, it's often difficult to maintain optimum chemotherapy doses that can maximise the abnormal cell killing as well as reducing side effects. The most chemotherapy drugs used in cancer treatment are toxic agents and usually have narrow therapeutic indices, dose levels in which these drugs significantly kill the cancerous cells are close to the levels which sometime cause harmful toxic side effects. To make the chemotherapeutic treatment effective, optimum drug scheduling is required to balance between the beneficial and toxic side effects of the cancer drugs. Conventional clinical methods very often fail to find drug doses that balance between these two due to their inherent conflicting nature. In this investigation, mathematical models for cancer chemotherapy are used to predict the number of tumour cells and control the tumour growth during treatment. A feedback control method is used so as to maintain certain level of drug concentrations at the tumour sites. Multi-objective Genetic Algorithm (MOGA) is then employed to find suitable solutions where drug resistances and drug concentrations are incorporated with cancer cell killing and toxic effects as design objectives. Several constraints and specific goal values were set for different design objectives in the optimisation process and a wide range of acceptable solutions were obtained trading off among different conflicting objectives. Abstract v In order to develop a multi-objective optimal control model, this study used proportional, integral and derivative (PID) and I-PD (modified PID with Integrator used as series) controllers based on Martin's growth model for optimum drug concentration to treat cancer. To the best of our knowledge, this is the first PID/I-PD based optimal chemotherapy control model used to investigate the cancer treatment. It has been observed that some solutions can reduce the cancer cells up to nearly 100% with much lower side effects and drug resistance during the whole period of treatment. The proposed strategy has been extended for more drugs and more design constraints and objectives.

Ao longo dos últimos anos, tornou-se evidente que o início e a progressão do câncer dependem de vários componentes do microambiente tumoral, incluindo células imunes e inflamatórias, fibroblastos, células endoteliais, adipócitos e matriz extracelular. Estes componentes e as células neoplásicas interagem entre si e trocam sinais pró e antitumor. O presente estudo teve como objetivo analisar o secretoma de células neoplásicas sob estresse e seu efeito no microambiente tumoral. Para este fim, duas linhagens celulares de carcinoma epidermóide de cabeça e pescoço foram cultivadas em duas condições de estresse: hipóxia e radiação. Os meios condicionados por estas células (secretoma 1) e o seu controle foram utilizados para cultivar células neoplásicas e fibroblastos humanos normais da cavidade oral. Os resultados sugerem que os sinais derivados das células neoplásicas em resposta a estresse dirigem a expressão gênica e proteica, bem como o comportamento celular das células vizinhas. Foram identificadas 38 proteínas celulares e nove proteínas secretadas com expressão aumentada e 61 proteínas celulares e 70 secretadas com expressão reduzida em células neoplásicas sob estresse hipóxico. Também foram identificadas 59 proteínas celulares e 29 proteínas secretadas com expressão aumentada e 59 proteínas celulares e 19 secretadas com expressão reduzida em células neoplásicas e fibroblastos humanos normais tratados com o meio condicionado por células sob estresse hipóxico. O secretome de células sob estresse não foi capaz de induzir proliferação de células neoplásicas e fibroblastos humanos normais, mas promoveu migração e invasão. Os resultados podem contribuir para o melhor entendimento do efeito dos fatores parácrinos liberados pelas células neoplásicas sobre a expressão gênica, bem como sobre o comportamento das células tumorais e estromais / Over the past years, it has become evident that cancer initiation and progression depends on several components of the tumor microenvironment, including inflammatory and immune cells, fibroblasts, endothelial cells, adipocytes, and extracellular matrix. These components and the neoplastic cells interact with each other providing pro and antitumor signals. The present study aimed to analyze the secretome of cancer cells under stress and their effect on the tumor microenvironment. For this purpose, two cell lines from head and neck carcinomas were cultured in two stress conditions - hypoxia and radiation. The medium conditioned by these cells (secretome 1) and their control were used to grow untreated neoplastic cells and normal human fibroblasts from oral cavity. Our results showed that signals derived from cancer cells in response to stress drive gene and protein expression and cell behavior. Thirty-eight overexpressed cellular and 9 secreted proteins, and 61 underexpressed cellular and 70 secreted proteins were identified in neoplastic cells under hypoxic stress. Fifty-nine overexpressed cellular and 29 secreted proteins, and 59 underexpressed cellular and 19 secreted proteins were identified in neoplastic cells and normal human fibroblasts treated with the medium conditioned by cells under hypoxic stress. The secretome of cells under stress was not able to induce proliferation of cancer cells and normal human fibroblasts, but promoted migration and invasion. The results may contribute to understand the effect of paracrine factors released by neoplastic cell on gene expression as well as on stromal and tumor cells behavior