EGFR (Epidermal Growth Factor Receptor) is a protein that is ubiquitous in the human body. Aberrant activity of EGFR or its isoforms is implicated in a number of cancers, notably brain cancer. An isoform of EGFR, EGFRvIII (EGFR variant III), is particularly relevant to brain cancer since it is only naturally found in brain tumour tissue. However, the presence and activity of EGFRvIII is not well characterized. I hypothesize that the different levels of EGFRvIII expression and its expression relative to wild type EGFR in human brain tumour tissue can be used to diagnose the different stages and progression of disease in the glioblastoma multiforme (GM) type of brain cancer. The work presented in this thesis is an attempt to develop a method for the accurate and absolute quantitation of EGFRvIII from brain tumour tissue. Using iMALDI (immunoMALDI), which combines the high-specificity of MALDI mass spectrometry with antibody immunoaffinity enrichment, I have optimized and developed a highthroughput technique for sensitive, specific and quantitative detection and differentiation of EGFR and EGFRvIII. I have also demonstrated a proof-of-concept by applying this assay to the isolation and detection of these proteins from brain tumour tissue. / Graduate

Thesis advisor: Thomas N. Seyfried / It has long been posited that all cancer cells are dependent on glucose for energy, termed the "Warburg Effect". As a result of an irreversible injury to the mitochondria, cancer cells are less efficient in aerobic respiration. Therefore, calorie restriction was thought to be a natural way to attenuate tumor growth. Calorie restriction lowers blood glucose, while increasing the circulation of ketone bodies. Ketone bodies are metabolized via oxidative phosphorylation in the mitochondria. Only cells that are metabolically capable of aerobic respiration will thus be able to acquire energy from ketone bodies. To date, calorie restriction has been shown to greatly reduce tumor growth and angiogenesis in the murine CT2A, EPEN, and human U87 brain tumor models. Using the novel VM-M3 model for invasive brain cancer and systemic metastatic cancer, I found that though calorie restriction had some efficacy in reducing brain tumor invasion and primary tumor size, metastatic spread was unaffected. Using a bioluminescent-based ATP assay, I determined the viability of metastatic mouse VM-M3 tumor cells grown in vitro in serum free medium in the presence of glucose alone (25 mM), glutamine alone (4 mM), or in glucose + glutamine. The VM-M3 cells could not survive on glucose alone, but could survive in glutamine alone indicating an absolute requirement for glutamine in these metastatic tumor cells. Glutamine could also maintain viability in the absence of glucose and in the presence of the F1 ATPase inhibitor oligomycin. Glutamine could not maintain viability in the presence of the Krebs (TCA) cycle enzyme inhibitor, 3-nitropropionic acid. The data indicate that glutamine can provide ATP for viability in the metastatic VM-M3 cells through Krebs cycle substrate level phosphorylation in the absence of energy from either glycolysis or oxidative phosphorylation. I therefore developed a metabolic therapy that targeted both glucose and glutamine metabolism using calorie restriction and 6-diazo-5-oxo-L-norleucine (DON), a glutamine analog. Primary tumor growth was about 20-fold less in DON treated mice than in untreated control mice. I also found that DON treatment administered alone or in combination with CR inhibited metastasis to liver, lung, and kidney as detected by bioluminescence imaging and histology. Although DON treatment alone did not reduce the incidence of tumor metastasis to spleen compared to the controls, DON administered together with CR significantly reduced the incidence of metastasis to the spleen, indicating a diet/drug synergy. In addition, the phagocytic capabilities of the VM-M3 tumor cells were enhanced during times of energy stress. This allowed for the digestion of engulfed material to be used in energy production. My data provide proof of concept that metabolic therapies targeting both glucose and glutamine metabolism can manage systemic metastatic cancer. Additionally, due to the phagocytic properties of the VM-M3 cell line also seen in a number of human metastatic cancers, I suggest that a unique therapy targeting metabolism and phagocytosis will be required for effective management of metastatic cancer. / Thesis (PhD) — Boston College, 2010. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Biology.

Thesis (M. D.)--University of Hong Kong, 2007. / Title proper from title frame. Also available in printed format.

Relaxin (RLN2) promotes cell migration/invasion, cell growth, and neoangiogenesis through binding to the relaxin receptor RXFP1 in many types of cancers. However, there have been no studies to determine the role of this system in brain tumors, especially in Glioblastoma Multiforme (GB), the most lethal primary brain tumor in adults. GB is a systemic brain disease and aggressively invades brain tissue. In this study, we have identified RXFP1 receptor, but not RLN2, in GB cell lines and primary GB cells from patients. RLN2 treatment resulted in a significant increase in migration of GB cell line and primary GB cells. To determine molecular mechanisms that facilitate RXFP1-mediated migration in GB cells, we employed a pseudopodia assay and 2D LC-MS/MS to investigate the protein composition at cell protrusions (pseudopodia) during GB cell migration. We also observed the expression of known mediators promoting tissue invasion upon RLN2 treatment. We identified PGRMC1, a candidate protein from 2D LC-MS/MS as a novel relaxin target protein in RXFP1-expressing brain tumor cells. RLN2 treatment also caused an increase in cathepsin (cath)-B and -L and enhanced production of as the small Rho-GTPases Rac1 and Cdc42 in GB cells. Collectively, these findings indicate that RXFP1-induced cell migration is mediated by the upregulation and intracellular actions of Rac1, Cdc42 and by cath-B and cath–L who serve as matrix modulating factors to facilitate brain tumor cells migration. PGRMC1 also contributes to RXFP1-mediated cell migration through an as yet unknown mechanism. RLN2 is not present in the brain. We determined the role of a peptide ligand of RXFP1, the newly discovered C1q/TNF related peptide (CTRP)8-derived P74 peptide, in promoting migration in GB cells. Similar to relaxin, P74 was found to have pro-migratory effects on GB cells. The biological activity of this peptide was also similar to relaxin and caused the upregulation of cath-B, cath-D and cath-L in the primary GB cells, thus, indicating that P74 might serve as a novel RXFP1 activating peptide ligand. We conclude that RXFP1 receptor signaling plays a key role in brain tumors cell migration and invasion.

Relaxin (RLN2) promotes cell migration/invasion, cell growth, and neoangiogenesis through binding to the relaxin receptor RXFP1 in many types of cancers. However, there have been no studies to determine the role of this system in brain tumors, especially in Glioblastoma Multiforme (GB), the most lethal primary brain tumor in adults. GB is a systemic brain disease and aggressively invades brain tissue. In this study, we have identified RXFP1 receptor, but not RLN2, in GB cell lines and primary GB cells from patients. RLN2 treatment resulted in a significant increase in migration of GB cell line and primary GB cells. To determine molecular mechanisms that facilitate RXFP1-mediated migration in GB cells, we employed a pseudopodia assay and 2D LC-MS/MS to investigate the protein composition at cell protrusions (pseudopodia) during GB cell migration. We also observed the expression of known mediators promoting tissue invasion upon RLN2 treatment. We identified PGRMC1, a candidate protein from 2D LC-MS/MS as a novel relaxin target protein in RXFP1-expressing brain tumor cells. RLN2 treatment also caused an increase in cathepsin (cath)-B and -L and enhanced production of as the small Rho-GTPases Rac1 and Cdc42 in GB cells. Collectively, these findings indicate that RXFP1-induced cell migration is mediated by the upregulation and intracellular actions of Rac1, Cdc42 and by cath-B and cath–L who serve as matrix modulating factors to facilitate brain tumor cells migration. PGRMC1 also contributes to RXFP1-mediated cell migration through an as yet unknown mechanism. RLN2 is not present in the brain. We determined the role of a peptide ligand of RXFP1, the newly discovered C1q/TNF related peptide (CTRP)8-derived P74 peptide, in promoting migration in GB cells. Similar to relaxin, P74 was found to have pro-migratory effects on GB cells. The biological activity of this peptide was also similar to relaxin and caused the upregulation of cath-B, cath-D and cath-L in the primary GB cells, thus, indicating that P74 might serve as a novel RXFP1 activating peptide ligand. We conclude that RXFP1 receptor signaling plays a key role in brain tumors cell migration and invasion.

Radiotherapy is the most widely used treatment for cancer, with 4 out of 10 cancer patients receiving radiotherapy as part of their treatment. The delineation of gross tumour volume (GTV) is crucial in the treatment of radiotherapy. An automatic contouring system would be beneficial in radiotherapy planning in order to generate objective, accurate and reproducible GTV contours. Image guided radiotherapy (IGRT) acquires patient images just before treatment delivery to allow any necessary positional correction. Consequently, real-time contouring system provides an opportunity to adopt radiotherapy on the treatment day. In this thesis, freely deformable models (FDM) and shape constrained deformable models (SCDMs) were used to automatically delineate the GTV for brain cancer and prostate cancer. Level set method (LSM) is a typical FDM which was used to contour glioma on brain MRI. A series of low level image segmentation methodologies are cascaded to form a case-wise fully automatic initialisation pipeline for the level set function. Dice similarity coefficients (DSCs) were used to evaluate the contours. Results shown a good agreement between clinical contours and LSM contours, in 93% of cases the DSCs was found to be between 60% and 80%. The second significant contribution is a novel development to the active shape model (ASM), a profile feature was selected from pre-computed texture features by minimising the Mahalanobis distance (MD) to obtain the most distinct feature for each landmark, instead of conventional image intensity. A new group-wise registration scheme was applied to solve the correspondence definition within the training data. This ASM model was used to delineated prostate GTV on CT. DSCs for this case was found between 0.75 and 0.91 with the mean DSC 0.81. The last contribution is a fully automatic active appearance model (AAM) which captures image appearance near the GTV boundary. The image appearance of inner GTV was discarded to spare the potential disruption caused by brachytherapy seeds or gold markers. This model outperforms conventional AAM at the prostate base and apex region by involving surround organs. The overall mean DSC for this case is 0.85.

Gliomas are the commonest type of primary malignant brain tumours of the central nervous system (CNS). The highly aggressive and infiltrative characteristics of gliomas render them one of the most lethal cancers. Amongst all, the most malignant form of glioma is glioblastoma multiforme (GBM), a World Health Organization (WHO) grade IV astrocytoma. Despite well-developed multimodal treatment including surgery, radiotherapy, and chemotherapy, the prognosis of GBM patients remains poor with median survival of just over one year. This high mortality rate is commonly the result of relentless tumour recurrence secondary to the tumour’s intrinsic resistance towards its standard chemotherapeutic agent temozolomide (TMZ). Prolyl 4-hydroxylase, beta subunit (P4HB) is an endoplasmic reticulum stress response (ERSR) chaperone protein that was previously found to be overexpressed in the chemoresistant glioma cell lines D54-MG and U87-MG. Differential expressions of numerous microRNAs (miRNAs) were also found between chemosensitive and chemoresistant glioma cell lines. As such, we surmised that the dysregulation of a P4HB-regulating miRNA may contribute to P4HB upregulation and therefore chemoresistance in glioma. MiR-210, a commonly dysregulated miRNA in various cancers, is one of the most highly downregulated miRNAs in chemoresistant glioma cells (compared to chemosensitive glioma cells), and, based on bioinformatics findings, may also regulate P4HB expression. MiR-210 was therefore selected for further investigations regarding its potential roles in glioma chemoresistance. The regulatory relationship between P4HB and miR-210 was subjected for verifications. With the use of quantitative real-time polymerase chain reaction (qPCR) and western blotting, the intrinsic expressions of P4HB and miR-210 were studied. The upregulation of P4HB in D54 and U87 chemoresistant glioma (compared to the parental) cell lines were found to correlate reciprocally with the downregulation of miR-210 in the same chemoresistant glioma cells. To delineate the potential regulatory role of miR-210, a gain of function approach was adopted. Transfection of a miR-210 mimic was performed into the D54 and U87 parental chemosensitive (D54-S and U87-S) and chemoresistant (D54-R and U87-R) cells, along with a negative control. The transfection efficiency of miR-210 as well as the subsequent P4HB expressions was verified. It was found that P4HB expression was downregulated as a result of miR-210 upregulation both at the mRNA and protein levels in glioma cells. Furthermore, the effects of miR-210 overexpression on chemoresistance in the glioma cells were tested by performing cell proliferation assay. Decrease in the half maximal inhibitory concentration (IC50) of TMZ were found in all cell lines overexpressing miR-210, suggesting that miR-210 upregulation may lead to P4HB inhibition, which would at least partially mediate an alleviation of glioma cells’ resistance towards its chemotherapeutic agent TMZ. In summary, miR-210 is downregulated in chemoresistant glioma cells in vitro. It plays a potential role in regulating P4HB expression, hence chemoresistance in GBM cells. Future investigations may focus on its mechanism of action and potentiality for therapeutic intervention. / published_or_final_version / Surgery / Master / Master of Medical Sciences

The central nervous system is a potential site of action for the Fusarium mycotoxin Fumonisin B1 (FB1), and is exemplified in horses by the disease equine leukoencephalomalacia. Structurally resembling sphingoid bases, FB1 inhibits ceramide synthase, an enzyme involved in sphingolipid metabolism, leading to accumulation of free sphinganine (Sa) and sphingosine (So). This investigation focused on FB1, Sa, So and the Fusarium mycotoxins fusaric acid (FA), moniliformin (MaN), zearalenone (ZEA), deoxynivalenol (DON), and T-2 toxin (T2). Effects of the Fusarium mycotoxins and sphingoid bases on the N2a neuroblastoma cell line were assessed using the methylthiazol tetrazolium (MIT) and ApoGlow™ assays. The MIT assay revealed significant differences between the viability of N28 control cells and the cytotoxic effects of FB1 (p=0.001), So (p=1.1 x10-6 ), Sa (p=1.9x10-6 ), MON (p=0.002), DON (p=0.04) and ZEA (p=0.003) on N28 cells between 5-250µM. The cytotoxic effects of FA did not differ significantly from controls (p=0.1). The ApoGlow™ assay revealed that in N28 cells, FB1 at 8µg.ml-1, FA at 128µg.ml-1, and (FBI+FA) combined induced growth arrest at 2 and 4µg·ml-1. Assessment of the effects of FBI and FA on the Jurkat leukaemic suspension cell line revealed that FB1 induced apoptosis at 1.56,12.5 and 50µ.ml-1, growth arrest at 100, 200 and 800µg.ml-1 and proliferation at 400µg.mg-1. Fusaric acid induced proliferation at 1. 56µg.ml-1, apoptosis at 3.15µg.mrl, growth arrest at 100 and 200µg.mrl, and necrosis at 800µg.ml-1. Combined, (FB1+FA) induced apoptosis at 1.56, 3.15,12.5 and 800µg.ml-1. Flow cytometry and fluorescence microscopy revealed that mycotoxins, Sa and So induced varying levels of apoptosis and necrosis in N28 cells. Acridine orange and ethidium bromide staining facilitated discrimination between viable, apoptotic and necrotic cells. Transition of the mitochondrial transmembrane potential was measured using Rhodamine 123 with propidium iodide, and the dual emission potential sensitive stain JC-1. Changes in mitochondrial membrane potential and plasma membrane integrity were expressed as increases or decreases in fluorescence intensity. An increase in mycotoxin concentration from 50 to 200µM was usually paralleled by a decrease in J-aggregate formation, suggesting a decrease in the ?¦¥m. Staining with Rh 123/PI indicated at specific concentrations whether N28 cells were either late apoptotic or necrotic reflected by the levels of PI uptake. No dose dependant mechanism of cell death was established using either method, as fluctuations were evident. Immunolocalisation of T2, ZEA and FB1 within cellular organelles that exhibited ultrastructural pathology provided correlation between mycotoxin exposure and effects. Multinucleate giant cells and retraction of cellular processes were observed. At the electron microscope (EM) level, FB1 was immunolocalised within microsegregated and peripherally condensed nucleoli, the nucleoplasm, distorted mitochondria and dilated endoplasmic reticulum (ER). The capacity of cells to incorporate mycotoxins and effect cytological changes represents a major factor in the potential for initiation of malignant transformation. Exposure of N2a cells to FB1 for 72 hours increased intracellular free Sa and depleted complex sphingolipids. Using High Performance Liquid chromatography (HPLC), acid hydrolysis revealed reduction in Sa from a level of O.6±0.12µM in control cells, to 02±0.lµM in cells exposed to 50µM and lOOµM FB1. Base hydrolyses revealed increase in free Sa: So ratios from 0.52±0.2 in control cells, to 1.14±0.2 and 1.4±0.3 in cells exposed to 50 and l00µM FB1 respectively. The Sa: So ratio in the complete culture media (CCM) increased from 1. 7±0. 3 for control cells to 2.0±0.2 and 2.50±0.4 for cells exposed to 50 and lOOµM FB1 respectively. Correlation coefficients between Sa: So ratios to FB1 exposure in CCM (R=0.75) and within cells (R=0.85), imply that the free Sa: So ratio within cells appears to be a better biomarker for FB1-induced disruption of sphingolipid metabolism in vitro, than the Sa: So ratio in CCM. Optimisation of HPLC analytical procedures improved recovery of FB I from spiked human sera to 95.8% (n=15) and detection limits to -5ng.ml-1 at a signal to noise ratio of 5:1. Optimisation of methods for recovery of Sa and So from spiked sera, led to recoveries of 77.9% and 85.0%, for So and Sa respectively at levels of spiking with lOng per 500µl of serum. Matched sera Sa:So ratios and FB1 levels in brain cancer and non-cancer subjects in KwaZulu-Natal were determined using these optimised methods. Fumonisin B1 was detected in sera of non-cancer (76.7±62.2nM) and brain cancer subjects (l07.38±116nM). Mean serum Sa:So ratios of 21 non-cancer subjects was 1.7±0.7. There was no correlation (R=0.26) between these variables in non-cancer subjects. The mean serum FB1 level in brain cancer subjects was 107.4±116nM (range 10.5-298nM) (n=50) and the mean Sa:So ratio (n=50) was 1.9±1.7 (range 0.40-8.16). No correlation was found between these variables in the brain cancer subjects either (R = -0.23). Fumonisin B1 was irnmunolocalised in 49 of 76 brain tumour tissue samples analysed using immunohistochemistry (IHC). Thirty-eight of the 76 specimens had matched serum FBI levels and Sa: So ratios, and 23 of these were positive for FB1 presence. Although not significantly different (p=0.ll), the FBI sera levels in the cancer group with FBI within the tumour tissue had higher levels of FB1 in sera than the IHC FB1 negative group. Fumonisin B1 was localised within irregular profiles of nuclei, elongated and swollen mitochondria and ER. Immunolocalisation of FB1 within organelles in the brain showing ultrastructural cellular pathology suggests FBI may be implicated in the aetiology of human brain carcinogenesis. / Thesis (Ph.D.)-University of KwaZulu-Natal, Durban, 2004.

-- / Ultrasound imaging is one of the most important ways of imaging the human body. The mainly application of ultrasound is for the soft parts of the body, due to the properties of tissues in the propagation of the ultrasounds. The brain with its normal shape is an appropriate organ for ultrasound imaging. But intraoperative ultrasound imaging may be effective in tumor localization and it can be used throughout surgery to monitor the extent of tumor resection. The ability to locate precisely a deeply situated intracranial lesion intraoperatively can reduce the risk of damage to normal tissue, assist in determining the extent of tumor resection and reduce the time surgery. Intraoperative ultrasound holds great promise, but if it is to be used to its fullest extent, further modification of transducers must be developed. For this reason in this project we have done different simulations via the program Field II in order to evaluate the most appropriate transducers for the best depiction that can be achieved. We have to mention that those different experiments are based to the same code but with different specific parameters in every experiment.

Three-dimensional (3D) visualization in cancer control has seen recent progress due to the benefits it offers to the treatment, education, and understanding of the disease. This work identifies the need for an application that directly processes two-dimensional (2D) DICOM images for the segmentation of a brain tumor and the generation of an interactive 3D model suitable for enabling multisensory learning and visualization. A new software application (M-3Ds) was developed to meet these objectives with three modes of segmentation (manual, automatic, and hybrid) for evaluation. M-3Ds software was designed to mitigate the cognitive load and empower health care professionals in their decision making for improved patient outcomes and safety. Comparison of mode accuracy was evaluated. Industrial standard software programs were employed to verify and validate the results of M-3Ds using quantitative volumetric comparison. The study determined that M-3Ds‘ hybrid mode was the highest accuracy with least user intervention for brain tumor segmentation and suitable for the clinical workflow. This paper presents a novel approach to improve medical education, diagnosis, treatment for either surgical planning or radiotherapy of brain cancer. / Thesis / Master of Science (MSc)