Rho-associated kinase 1 (ROCK1) is a serine/threonine kinase important for the regulation of the cellular cytoskeleton through the induction of actin stress fibres and acto-myosin contractility. The cleavage and subsequent activation of ROCK1 by caspase 3 during apoptosis is believed to cause many morphological phenomena associated with programmed cell death such as dynamic membrane blebbing. I now formally prove the necessity of ROCK1 cleavage for apoptotic blebbing by knocking-in a caspase cleavage resistant mutant of ROCK1 in a genetically modified model. In addition, animals homozygous for non-cleavable ROCK1 demonstrate a phenotype consistent with auto-immune disease suggesting that apoptotic blebbing is important to mediate rapid efferocytosis, which is a rapid phagocytic clearance of the cellular corpse, and thus maintain self-tolerance. Furthermore, apoptotic blebbing is important for the clearance of apoptotic cells and I demonstrate a novel mechanism for ROCK to mediate the release of factors participating in macrophage migration to dying cells. ROCK induced apoptotic blebs and bodies lose membrane integrity prior to secondary necrosis and leak intracellular material. Using quantitative mass spectrometry I identified numerous proteins that were previously unrecognized to be released during apoptosis. The release of protein was found to be impaired following ROCK antagonism with Y27632 which underscores the importance of ROCK activity in apoptotic protein release. One of these proteins, gelsolin, was released following caspase cleavage and encourages macrophage motility towards apoptotic cells. Finally, I now demonstrate that the three nonsynonymous somatic mutations in the ROCK1 gene identified in the Cancer Genome Project lead to elevated kinase activity and drive actin cytoskeleton rearrangements that promote increased motility and decreased adhesion, characteristics of cancer progression. Mapping of the kinase-interacting regions of the carboxy-terminus combined with structural modeling provides insight into how these mutations likely affect the regulation of ROCK1. Consistent with the frequency of ROCK1 mutations in human cancer, these results support the conclusion that there is selective pressure for the ROCK1 gene to acquire ‘driver’ mutations that result in kinase activation.

In this thesis high intensity focused ultrasound (HIFU) is utilized for cancer treatment (thermal mode) and treatment of ischaemic stroke (mechanical mode). These two applications were investigated in vitro and in vivo models. MRI was utilized to monitor the lesions created by HIFU either in thermal or cavitation mode in freshly excised lamb brain tissue in vitro, and in rabbit brain in vivo. Additionally, MRI was used to monitor lesions deep in tissue for both in vitro and in vivo exposures. All three MRI sequences used (T1-W FSE, T2-W FSE and FLAIR) were able to detect lesions. Both thermal and bubbly lesions were best monitored using T1-W FSE with excellent contrast, proving the potential of HIFU to treat reliably tumours in the brain. A HIFU system was also used to assist thrombolysis in cooperation with a thrombolytic drug such as recombinant tissue plasminogen activator (rt-PA) in vitro and in vivo. It was shown that higher intensity results to higher volume of dissolved clot, but there is a limit of the intensity to be used in order to avoid heating of the clot and the surrounding tissue. The goal in this study was to achieve temperature elevation not exceeding 1ºC (called safe temperature). It was found that the larger the beam area the larger the dissolved clot volume. Also, the lower the frequency, the larger the volume of the dissolved clot. The results reported herein point to the use of frequency around 0.5 MHz and pulsing to optimize thrombolysis and skull penetration and at the same time avoiding unwanted heating. Finally, an Acrylonitrile Butadiene Styrene (ABS) phantom skull model was developed in order to evaluate the propagation of ultrasound using a single element transducer. The skull model was appropriately designed so that it has the same attenuation as a human skull. It was demonstrated that using a frequency of 0.5 MHz versus 1 MHz, ultrasound propagation through the phantom skull was higher. Therefore, higher frequency has poor skull penetration and a small beam size at the focus, while low frequencies have better skull penetration but with the risk of reaching the unpredictable effect of cavitation. The developed system has proven to successfully create large lesions in the brain and at the same time, these lesions are successfully monitored with excellent contrast using MRI (T1-W FSE) enabling the accurate determination of the margins of these lesions. The results reported in this study point to the use of frequency around 0.5 MHz and pulsing to optimize thrombolysis and skull penetration and at the same time avoiding unwanted heating. For treating tumours located deep in the brain and for dissolving thrombus causing an acute ischaemic stroke, further extensive clinical studies will be needed before this technology is applied to humans.

• Gastro-oesophageal reflux disease (GORD) is the commonest chronic disease in Western countries. Symptomatic GORD is the strongest risk factor for the development of oesophageal adenocarcinoma with obesity and male sex also linked to the development of neoplasia at this site. Recent decades have seen a significant increase in the incidence of this highly lethal cancer among Western populations with Scotland having the highest recorded incidence worldwide. • Human saliva has a high nitrite content derived from the entero-salivary recirculation of nitrate in our diet which has resulted from the increased use of nitrogenous fertilisers over the past 50-60 years. • The luminal chemistry produced at the gastro-oesophageal junction (GOJ) when swallowed salivary nitrite reacts with gastric acid, and most notably the production of nitric oxide (NO), may explain most of the physiological abnormalities that contribute to the pathogenesis of GORD. NO has been shown to reduce lower oesophageal sphincter (LOS) pressure, impair oesophageal clearance, delay gastric emptying and may be the final mediator of transient lower oesophageal sphincter relaxations (TLOSRs). Previous studies to investigate the role of this luminal chemistry in the pathogenesis of GORD show conflicting results. • In addition to the distal oesophageal acidification produced by traditional trans-sphincteric reflux, previous studies suggest ‘splaying open’ of the distal lower oesophageal sphincter following a meal may expose the gastric cardia and the most distal oesophageal squamous mucosa to the noxious effects of gastric acid. • Although the gastric cardia is an important site of pathology in the upper gastrointestinal tract, it is a complex and poorly understood area. It has been proposed, from autopsy studies, that cardia mucosa itself may be pathological and in fact an ‘acquired cardia’ due to metaplasia of the most distal oesophageal squamous mucosa. • A series of studies were designed to examine the effect of salivary nitrite on post-prandial GORD, gastro-oesophageal function and GOJ morphology in 20 healthy, asymptomatic adult volunteers using high-resolution pH manometry, an isotope gastric emptying breath testing and X-ray localisation of the squamo-columnar junction (SCJ). • Despite an excellent range of salivary nitrite concentrations extending over and above the normal physiological range no effect of salivary nitrite on gastro-oesophageal reflux, function or morphology was demonstrated. However, the studies did confirm, for the first time using high-resolution manometry, that distal opening of the LOS occurs after a meal. • The relationship of age and obesity to the SCJ position relative to the proximal border of the gastro-oesophageal high pressure zone (HPZ) was examined in 15 Helicobacter Pylori negative healthy volunteers. Strong negative correlations were seen between SCJ position relative to the proximal HPZ and increasing age, body mass index (BMI) and waist circumference (WC) respectively. These correlations were stronger in the male sub-group. • In 25 healthy volunteers, parietal cell density was measured from endoscopic biopsies taken from the macroscopic SCJ, 1cm distal to the SCJ, the gastric body and the gastric antrum. Again, a strong negative correlation was seen between increasing age and parietal cell density at the SCJ. This effect was localised to the SCJ and not seen at the other biopsy sites. • Our findings suggest that salivary nitrite does not alter gastro-oesophageal function, the integrity of the gastro-oesophageal barrier or gastro-oesophageal reflux in healthy volunteers. They confirm distal opening of the LOS after meals. The strong negative correlations between age and both SCJ position relative to the proximal HPZ and parietal cell density support the hypothesis of an ‘acquired’ cardia. The development of cardia mucosa may also be linked to obesity, visceral obesity and male sex. • Future work could examine the carcinogenic effect of salivary nitrite and its luminal chemistry but this would require large scale epidemiological research. Further, larger clinical studies are needed to investigate the role of distal opening of the LOS after meals and to improve our understanding of the gastric cardia. Such studies should focus on the role of obesity and posture.

Numerous large clinical trials of cardiovascular risk lowering agents have been conducted in the hope of reducing the excess cardiovascular risk found in patients with diabetes mellitus. However, the relationship between glucose and cardiovascular disease remains complex and various areas require further study. Even in patients with diabetes, an individual’s cardiovascular risk is highly variable depending on other clinical characteristics, the assumption that glucose is a continuous risk factor has often been based on weak evidence from relatively short studies, the effect of commonly used cardiovascular risk lowering agents often has unexpected effects on new-onset diabetes and statins have not yet been studied in detail, and whether glucose-lowering therapies actually reduce cardiovascular risk has remained a contentious issue despite the conduct of large clinical trials. Furthermore, the realisation that the combination of diabetes and chronic heart failure, a common complication of coronary disease, carries a particularly poor prognosis suggests that prediction of diabetes in this population may be clinically valuable. Aims: I aimed to address the following different, though related, questions regarding glucose and cardiovascular disease: 1. Are anticipated cardiovascular event rates in diabetes endpoint trials actually achieved? Is it possible to easily identify patients with diabetes that are at particular risk of events (information that is crucial to investigators who wish to design clinical trials)? 2. Is fasting glucose concentration independently and convincingly associated with increased risk of cardiovascular events in those without diabetes? 3. Do statins, the most commonly prescribed medications worldwide, have any influence on the risk of developing diabetes? 4. If statins do indeed affect new-onset diabetes, is there any evidence of a dose-dependent effect? 5. How effectively can clinicians predict the development of diabetes in chronic heart failure using commonly recorded clinical information? 6. Does intensive glucose-lowering therapy reduce the risk of cardiovascular events in patients with diabetes? Methods: To address these questions three approaches were used, namely (i) systematic review of previously published data from large cardiovascular endpoint trials conducted in patients with diabetes; (ii) analyses of existing datasets from two large clinical trials; (iii) meta-analyses of published and unpublished data from large clinical trials. Results and interpretation: 1. In a systematic review of 29 trials with 116,790 patients with diabetes, it was apparent that the majority of large cardiovascular endpoint trials conducted in patients with diabetes vastly overestimated the likely cardiovascular event rates in initial power calculations. Introduction of (i) previous history of cardiovascular disease and/or (ii) presence of proteinuria, as binary trial inclusion criteria, provides a simple and effective way to identify patients at high risk, something that is sought after for appropriate clinical trial power calculations. 2. In a population of 6,447 men without diabetes at baseline, impaired fasting glycaemia was not associated with increased risk of cardiovascular events over 15 years. Similarly, when baseline fasting glucose values <7.0mmol/L were split into quintiles, patients in the highest quintile were at similar risk of all vascular endpoints to those in the lowest. By contrast, impaired fasting glycaemia was a powerful risk factor for developing diabetes. 3. A meta-analysis of published and unpublished data from most large placebo- and standard care-controlled statin trials, which included data for 91,140 trial participants without diabetes at baseline, revealed that statin therapy is associated with a 9% higher risk for developing diabetes. 4. A subsequent meta-analysis of unpublished data from five large trials comparing intensive statin therapy with moderate dose therapy found that intensive statin therapy increases the risk of developing diabetes by 12% compared to moderate dosing, in keeping with a dose-dependent effect. While statin therapy remains effective at reducing cardiovascular risk it appears that patients on statin therapy, especially those on intensive regimens, should be considered for diabetes screening. 5. In an analysis of data for 1,620 patients with chronic heart failure and no diabetes at baseline studied for 2.8 years, the strongest predictors of new-onset diabetes were similar to those in the general population. In particular, the combination of HbA1c and body mass index provided a c-statistic of 0.79. 6. In a meta-analysis of published data for 33,040 patients with diabetes who participated in clinical trials comparing intensive glucose-lowering therapy with standard therapy, non-fatal myocardial infarctions were reduced by 17% on intensive therapy but no other cardiovascular endpoints were reduced. Death rates were similar in both groups. Conclusion: While diabetes is associated with excess cardiovascular risk, risk varies considerably depending on other risk factors. Glucose is, at best, a weak risk factor in those without diabetes, and glucose-lowering in patients with diabetes has only yielded a modest reduction in non-fatal myocardial infarctions but not other events; by contrast, measures of glycaemia are powerful predictors of new-onset diabetes in patients with and without chronic heart failure. Finally, the relationship between glucose and vascular disease is further complicated by the fact that numerous medications designed to reduce cardiovascular risk appear to have surprising effects on the risk of developing diabetes.

International Telemetering Conference Proceedings / October 17-20, 1994 / Town & Country Hotel and Conference Center, San Diego, California / It is known that a PCM telemetry system features with good accuracy, great dynamic range, and low noise. And when more than fourteen data channels are required, the PCM is generally the best choice. The paper describes the general ideas involved in developing a PCM telemetry system with fibre transmission.

ER positive (ER+) and HER2 negative (HER2-) breast cancers are routinely treated based on estrogen dependence. CDK4/6 inhibitors in combination with endocrine therapy have significantly improved the progression-free survival of patients with ER+/HER2- metastatic breast cancer. Gene expression profiling in ER+/HER2- models was used to define the basis for the efficacy of CDK4/6 inhibitors and develop a gene expression signature of CDK4/6 inhibition. CDK4/6 inhibition robustly suppressed cell cycle progression of ER+/HER2- models and complements the activity of limiting estrogen. Chronic treatment with CDK4/6 inhibitors results in the consistent suppression of genes involved in cell cycle, while eliciting the induction of a comparable number of genes involved in multiple processes. The CDK4/6 inhibitor treatment shifted ER+/HER2- models from a high risk (luminal B) to a low risk (luminal A) molecular-phenotype using established gene expression panels. Consonantly, genes repressed by CDK4/6 inhibition are strongly associated with clinical prognosis in ER+/HER2- cases. This gene repression program was conserved in an aggressive triple negative breast cancer xenograft, indicating that this is a common feature of CDK4/6 inhibition. Interestingly, the genes upregulated as a consequence of CDK4/6 inhibition were more variable, but associated with improved outcome in ER+/HER2- clinical cases, indicating dual and heretofore unknown consequence of CDK4/6 inhibition. Interestingly, CDK4/6 inhibition was also associated with the induction of a collection of genes associated with cell growth; but unlike suppression of cell cycle genes this signaling was antagonized by endocrine therapy. Consistent with the stimulation of a mitogenic pathway, cell size and metabolism were induced with CDK4/6 inhibition but ameliorated with endocrine therapy. Together, the data herein support the basis for profound interaction between CDK4/6 inhibitors and endocrine therapy by cooperating for the suppression of cell cycle progression and limiting compensatory pro-growth processes that could contribute to therapeutic failure.

RNA Polymerase III (Poll III) contributes to about 10% of nuclear transcription and is essential for the synthesis of short untranslated transcripts, including tRNA and 5S rRNA. Pol III deregulation has been implicated in driving cellular proliferation and transformation, along with increased expression of a number of Pol III specific transcription factors and transcripts. The significance of Pol III in clinical pathology, including that of human malignancies, remains to be formally tested. Using prostate cancer as a model, two key components of the Pol III complex, namely Brf1 and tRNAiMet, were investigated. The expression patterns of Brf1 and tRNAiMet were studied in this thesis using immunohistochemistry (IHC) and in situ hybridisation (ISH) respectively. Brf1, a subunit of transcription factor IIIB (TFIIIB), was detected predominantly in the nucleus with heterogenous staining intensity, its presence ranging from weak to strong. Examination across a wide range of tissue types and organs, both normal and tumour tissues revealed high levels of Brf1 expression in the epithelium. In addition, Brf1 expression could also be detected in connective tissues and, to a lesser extent, in muscular and nervous tissues. A number of tumour types exhibited elevated expression of Brf1 protein relative to their normal control tissues. These included prostate adenocarcinoma, B-cell lymphoma and, interestingly, tumours arising from connective tissues (sarcoma, fibrosarcoma and chondrosarcoma). As expected, tRNAiMet expression, as revealed by ISH analysis, was mostly observed in the cytoplasm, although some nuclear staining was also present. Similar to Brf1, tRNAiMet expression was detected predominantly in epithelial tissues such as the skin epidermis, prostate gland and epithelial lining of the cervix. A number of tumours were found to overexpress tRNAiMet. These included breast ductal carcinoma, oesophageal carcinoma and melanoma. The clinical impact of Brf1 and tRNAiMet overexpression was examined using tissue microarrays (TMA) containing tissue samples obtained from patients with prostate cancer (PCa) and benign prostate hyperplasia (BPH). Collectively, data from two independent patient cohorts, Glasgow TMA: BPH (n=21), PCa (n=151) and Newcastle TMA: BPH (n=113), PCa (n=365), showed that Brf1 expression was upregulated in prostate cancer relative to BPH (p=0.0034). Brf1 expression was not found to be associated with the following clinical and biologic parameters: Gleason sum score (indicative of tumour differentiation and morphology, p=0.653); prostate specific antigen (PSA level, indicative of tumour bulk or volume, p=0.381) and Ki-67 expression (signifying cellular proliferation, p=0.034). However, and importantly, within the prostate cancer patient cohorts studied, high Brf1 expression was associated with a significantly less favourable survival outcome (Kaplan Meier analysis, p<0.001). Together, the data presented in this study support the relevance of Brf1 and tRNAiMet overexpression as part of Pol III deregulation in tumours, especially of epithelial origin. This study also suggests the potential application of Brf1 as a prognostic marker in cancer, however, this warrants a further study.

Student Number : 9105022E - PhD thesis - School of Molecular and Cell Biology - Faculty of Science / SNAMA is the Drosophila melanogaster homologue of a group of proteins that are known to bind p53 and the retinoblastoma protein (Rb). This multi domain protein consists of a conserved N-terminal domain called Domain With No Name (DWNN), a zinc finger, a cysteine rich RING finger-like domain, a probable p53 binding region, and a glutamic acid-rich and lysine-rich region. These associated domains indicate that SNAMA plays an important regulatory role in the cell and may function in RNA processing and in apoptosis. The DWNN domain was first identified in Cytotoxic T-cell resistant Chinese hamster ovary (CHO) cells using promoter trap mutagenesis to screen for genes involved in apoptosis. Subsequently, this domain was identified in other eukaryotic organisms including animals and plants. The SNAMA transcriptional unit consists of 9 exons and 8 introns that code for a 1231 amino acid protein with the 76 residue N-terminal DWNN domain. The DWNN domain has a 23.5% sequence identity to the ubiquitin protein and a predicted folded structure similar to ubiquitin. Western blots identified multiple bands indicative of ubiquitin tagged proteins. Taken together this suggests a role in the ubiquitin pathway either as an ubiquitin domain protein or the DWNN domain of SNAMA tagging other proteins. The cysteine rich RING finger-like domain has a histidine to serine substitution at the fourth position of the putative RING finger and represents a distinct class of RING finger-like proteins that could have ubiquitin ligase activity. Northern blot analysis identified a single 4.6 kbp transcript expressed abundantly throughout development early in embryogenesis but reduced in older embryos and in adult male and females. SNAMA probably interacts with Dmp53 as a suppressor of apoptosis or a negative regulator of an activator of apoptosis. It is a vital gene required for development, as the mutant P-element insertion line in which the Pelement is inserted in the first intron of SNAMA is lethal when homozygous. Acridine orange staining of these mutant flies showed a direct correlation between the presence of SNAMA and apoptosis. An increase in the levels of apoptosis occurred in embryos with relatively low levels of SNAMA expression. The mode of this action is either direct, or via other proteins that are involved in the apoptotic pathway.

Collagen is the ligand for the discoidin domain receptor-1 (DDR1), a receptor tyrosine kinase that is over-expressed in Hodgkin lymphoma. However, the role of DDR1 in diffuse large B cell lymphoma (DLBCL) is not known. I showed that DDR1 is over-expressed in a subset of DLBCL where it positively correlates with expression of its collagen ligands, and negatively correlates with expression of mitotic spindle genes. DDR1 correlated genes also overlapped with three aneuploidy signatures and DDR1 expression correlated significantly with autosomal aneuploidy index. RNAseq analysis revealed that over-expression of DDR1 in primary germinal centre B cells down-regulated expression of CENPE, an essential component of the mitotic spindle checkpoint that when inactivated leads to chromosome mis-segregation and aneuploidy. CENPE expression was also significantly reduced in primary DLBCL. Moreover, I showed that the constitutive activation of DDR1 in an in vitro lymphoma model led to aneuploidy. Finally, I showed that DDR1 can be inhibited by three small molecules and established the basis for in vivo model to test these inhibitors in DLBCL xenograft. My data provide evidence that DDR1 can induce aneuploidy in B cells, and as such identify a mechanism to potentially explain the link between chronic inflammation and lymphomagenesis.

Hepatic stem cells are found in the liver at all stages of development, including adult, and have the potential to give rise to daughter cells of biliary, hepatocytic and pancreatic lineages. Hepatic stem cells contribute significantly to liver regeneration, but may be associated with development of primary liver cancer, and are being investigated as a cellular therapy for liver failure. This thesis describes the development of methods for the immunohistochemical quantification of hepatic stem cell activation, allowing assessment of the association between hepatic stem cell activation in needle biopsy tissue and subsequent development of HCC in a retrospectively identified cohort of cirrhotic patients. A murine dietary model of NASH and HCC was developed and characterised in detail demonstrating progressive hepatic stem cell activation with increasing injury severity. We then went on to describe and prospectively isolate a resident population of stromal stem/progenitor cells in adult, uninjured mouse liver with the potential to give rise to both myofibroblasts, and under selective conditions, epithelial stem/progenitor cells. Finally, we demonstrated the isolation of hepatic stem cell from explanted cirrhotic liver and normal common bile duct and assessed their utility as a source of hepatic stem cells for cellular or regenerative therapy.