Search

Non-Hodgkin’s lymphomas (NHL) are a spectrum of hematopoietic cancer accounting for 4% of new cancer diagnoses each year. Approximately 95% of all NHL are of B-cell origin; diffuse large B-cell lymphomas (DLBCL) and follicular lymphoma (FL) accounting for 30-40% and 20% of B-NHL respectively. Recent human mutation profiling and resequencing studies have shown that CREBBP and BCL2 are frequently mutated, early events in B-NHL that are often are concurrent. This thesis presents a study of their role in oncogenesis by generating a novel model of B-NHL overexpressing BCL2 in the haematopoetic compartment combined with the conditional heterozygous loss of Crebbp in the germinal centre B-cells. Loss of Crebbp significantly accelerated BCL2 driven lymphomagenesis. Characterisation of these lymphomas by flow cytometry demonstrated that they were phenotypically similar to mature human B-cell lymphomas of germinal centre origin. Additionally an insertional mutagenesis screen was performed in mice sensitised to B-NHL. Somatic MoMuLV caused mutations interact with the sensitising mutations, facilitating tumourigenesis and recapitulating the stepwise and accumulative progression of human disease. Provirus insertions were detected using a novel method of insertion site cloning called UMI-LM-PCR which amplifies the region flanking the provirus. Insertions were then mapped to the mouse genome flagging genes with a putative role in tumourigenesis. This method represents the highest throughput method to date and its applicability to Illumina sequencing permits the most comprehensive quantitative survey of subclonal mutations. This work confirms Crebbp is a tumour suppressor gene and that its loss cooperates with overexpression of BCL2 to accelerate lymphomagenesis. Through insertional mutagenesis screening of this model hundreds of putative cancer genes have been identified including Pou2f2 and Tfrc which have a propensity for mutation specifically in B-cell lymphomas deficient in Crebbp and overexpressing BCL2. These candidates now need to be validated to better characterise their role in B-cell lymphoma Open Access

Dopamine neurons in the substantia nigra pars compacta (SNc) and ventral tegmental area (VTA) regulate day-to-day activities such as reward-related learning, and motor control. Dysfunction of these neurons is implicated in disorders like Schizophrenia, addiction to drugs, and Parkinson’s disease. Dopamine neurons fire action potentials with variability in inter-spike intervals (ISI) interspersed with bursts of spikes, both firing patterns being a result of co-ordinated action of synaptic inputs and ion channels. Pharmacological inhibition of calcium-activated potassium (SK) channels increases the variability in their firing pattern, sometimes also increasing the number of spikes fired in bursts, indicating that SK channels play an important role in maintaining dopamine neuron firing regularity and burst firing. However, the exact mechanisms underlying these effects are still unclear. We hypothesized that the post-spike after hyperpolarisation (AHP) provided by the SK current controls and standardises the availability of other voltage-gated ion channels, thus timing successive action potentials at regular intervals. Here, I develop a biophysical model of a dopamine neuron incorporating ion channel stochasticity that enabled me to analyse availability of ion channels in multiple states during spiking. Decreased firing regularity is primarily due to a significant decrease in the AHP that in turn resulted in a reduction in the fraction of available voltage-gated sodium channels. Insufficient AHP causes a failure of sodium channels to recover from inactivation and, due to ion channel stochasticity, a variable number of sodium channels are available to fire a spike, in the absence of AHP, thus resulting in an irregular spike train. My model further predicts that inhibition of SK channels results in a depolarisation of action potential threshold along with an increase in its variability, thus suggesting that SK channels regulate spike threshold in midbrain dopamine neurons. Moreover, a quantitative description of action potential features during endogenous burst firing in midbrain dopamine neurons has not been reported, perhaps due to a lack of in vivo whole-cell recordings that can shed light on membrane dynamics. Because we had available in vivo whole-cell recordings from immuno-histochemically identified dopamine neurons in the VTA, I analysed these using mathematical techniques like Principal Components Analysis (PCA) and the stochastic model mentioned above, in order to provide quantitative descriptions of action potential features during the course of endogenous burst firing. I found that the first detectable change in membrane dynamics that perhaps leads to burst firing is a reduction in the AHP mediated by SK channels. This was accompanied by progressive depolarisation of action potential threshold with each spike within the burst due to reduced availability of sodium channels as a result of progressive decrease in recovery from inactivation. This also resulted in a decrease in rate of change of membrane voltage for each successive spike in a burst. Reduction in action potential height that is commonly observed during burst firing can also be explained by the reduced availability of sodium channels. My model predicts that a reduction in A-type potassium channel current during burst firing could facilitate the transient increase in instantaneous firing rate during bursts and the progressive reduction in sodium channel availability results in the cessation of firing, facilitating burst termination in these neurons. Open Access

Neuroinflammation is an early feature of a number of nervous system disorders. Inflammation in the brain is primarily mediated via microglial cells, which are active components of circuit development in the central nervous system. Schizophrenia is a psychiatric illness with deficits in perceptual, cognitive and emotional function. Prior to the onset of psychosis, there is a period of attenuated psychotic symptoms, where individuals experience sub threshold features of psychosis. This ‘ultra high risk’ period can provide unique opportunities to investigate the development of psychosis. It has been demonstrated through translocator protein (TSPO) positron emission tomographic (PET) imaging that microglial activity is elevated in chronic schizophrenia, however it is unknown whether this elevation is present prior to the onset of psychosis. It is also uncertain what effect antipsychotic medication has on microglia in vivo. This thesis is divided between clinical and animal investigation, the results can be split into four findings; Firstly we demonstrate that there is a higher binding of [11C]PBR28 (a novel TSPO PET ligand) in ultra high risk subjects and patients with schizophrenia compared to healthy controls. Symptoms in the ultra high risk subjects also correlate with the level of ligand binding. Secondly, brain volumes are not correlated with [11C]PBR28 binding or inflammatory cytokine levels in peripheral blood samples. The third finding of this thesis is that antipsychotic drug administration does not appear to alter cortical microglial cells in naïve and systemically inflamed animals. The final finding is that brain volume is reduced by antipsychotic medication. Together these findings demonstrate that inflammation is present in subjects experiencing subthreshold psychotic symptoms. The animal experiments suggest medicated patients would not be expected to have higher levels of microglial activity than their un-medicated counterparts. Further investigation is needed to determine the mechanism of cortical volume changes after medication and how this relates to TSPO. Open Access

The development of a whole organism from a single fertilised egg is a tightly orchestrated process at the molecular level. Signalling pathways stimulate the expression of transcription factors that in turn coordinate the activation, repression or priming of genes through multi-protein complex assembly and chromatin remodelling at regulatory regions during key developmental transitions. This project focuses on the role of the chromatin modifier Jmjd2c/Kdm4c in embryonic stem cell (ESC) pluripotency. Jmjd2c is a member of Jmjd2 H3K9-demethylase family, which is highly expressed in ESCs and during early mouse development, and was proposed to contribute to ESC identity. Here, Jmjd2c-depleted ESCs were generated and found to sustain an undifferentiated state. In contrast, these cells fail to execute multi-lineage differentiation, as evidenced by their inability to fully activate appropriate gene expression programs, but readily adopt an extra-embryonic endoderm-like phenotype under appropriate conditions. Moreover, Jmjd2c-knockout ESCs can transit to an early epiblast stage and be stably converted into epiblast stem cells (cEpiSCs) that harbour an immature state as highlighted by a lack of transcriptional gene priming for germ layer markers. In order to unfold the mechanistic basis for the observed differentiation defect, genome-wide distribution of Jmjd2c was analysed by ChIP-sequencing. This revealed prominent binding to the TSS of active and bivalently marked genes, concordant with recent reports. Remarkably, acquisition of multi-lineage priming in ESCs was accompanied by Jmjd2c recruitment at poised enhancers of lineage-affiliated genes. Surprisingly, these regions were co-occupied by the H3K9-methyltransferase G9a/Ehmt2, which physically interacts with Jmjd2c and Mediator, within chromatin-bound complexes. Critically, the absence of Jmjd2c in cEpiSCs was sufficient to destabilize the assembly of enhancer protein complexes, suggesting a scaffolding role independent of its H3K9-demethylase activity. Collectively, this study uncovered a novel role for Jmjd2c in regulating gene expression at exit of the pluripotent state. Open Access

Sepsis is the leading cause of direct maternal mortality in the UK, but also a leading concern worldwide. In the USA, severe maternal sepsis rates and the risk of sepsis-related death have doubled in the last decade. There appears to be an increased susceptibility to, and severity of, infections during pregnancy. However, the mechanisms behind the increased morbidity and mortality are not fully understood. Pregnancy is associated with acute physiological changes in most organ systems, in particular the cardiovascular and immune systems. In this thesis the haemodynamic response to infection and inflammation was investigated. It was hypothesised that there is an altered haemodynamic response to infection mediated by inflammatory pathways, and by targeting signalling pathways within the immune and cardiovascular systems maternal outcomes could be improved. In order to achieve to this, a murine model of endotoxaemia during pregnancy was employed to study the effects of inflammation on maternal haemodynamics, utilising a telemetric technique to monitor physiologic changes. Firstly, this model was used to demonstrate that inflammation in pregnancy can lead to marked hypotension and haemodynamic dysfunction that does not occur in non-pregnant controls. Supplementation with additional progesterone, a key hormone of pregnancy, did not exacerbate this hypotensive effect. Manipulation of the chemokine receptor CCR2 was able to reverse inflammation-induced hypotension in pregnancy, both by genetic knockout and pharmacological inhibition. Manipulation via genetic knockout of DDAH1, a component of the nitric oxide signalling pathway, whilst conferring an initial hypertension did not significantly impact the gestational haemodynamic profile. The marked fall in blood pressure in response to endotoxin may help to explain the greater susceptibility of pregnant women to sepsis. Maternal outcomes may be improved by modulating the immune response. Open Access

Germ cells represent the link between generations. At the time of fertilisation, sperm and egg come together to form a single cell, called the zygote, from which an entire organism is ultimately constructed. In mouse, two unique properties underlie the exceptional potential of gametes: 1) the ability to execute meiosis; and 2) the correct establishment of complementary epigenetic signatures (genomic imprints) in sperm and oocyte that enable post-implantation embryonic development in the next generation. At the molecular level, both the progression through meiosis and the establishment of sex-specific methylation imprints are critically dependent on a wave of global DNA demethylation that is initiated in primordial germ cells (PGCs), the embryonic precursors of the gametes, following their colonisation of the embryonic gonads. The work I present in this thesis focuses on the molecular mechanisms regulating this wave of DNA demethylation. Specifically, I conclusively show that DNA demethylation proceeds independent of the 5-methylcytosine oxygenase Tet1, previously proposed to act as the putative DNA demethylase. Rather, I show that, with respect to DNA demethylation, Tet1 appears to protect the hypo-methylated state in PGCs following the major loss of 5-methylcytosine. Most strikingly, however, with respect to germ cell development, Tet1 appears to play an important role in regulating the timing of the major developmental and transcriptional transitions that occurs in the mouse foetal germ line between embryonic day (E) 10.5 and E14.5, with PGCs derived from Tet1-knockout embryos showing increased stability of the demethylated pre-meiotic germ cell state. Beyond understanding the role of Tet1 in gonadal PGCs, I also present results focusing on the potential relationship between cell signalling, DNA demethylation and licensing for gametogenesis, with preliminary data suggesting a potential role for Akt-mTORC1 signalling in activating the expression of the key meiotic regulator Dazl. Open Access

The significance of DNA methylation beyond the gene promoter is little understood. High levels of intragenic methylation (IGM) correlate with active gene expression in lower organisms. Pyrosequencing assays across oestrogen receptor alpha (ESR1) in ESR1-positive and negative breast cancer cell lines showed higher IGM levels in ESR1-positive cells, with reduced ESR1 expression after demethylation with decitabine. However, different ESR1 IGM patterns were determined in breast cancer biopsies, probably because of cell heterogeneity. DNA methylation can reflect cellular environmental exposures. A methylation signature of smoking was determined in a European Prospective Investigation into Cancer and Nutrition epigenome-wide association study (EWAS, 450K array), in case-control pairs of breast (n=92) and colorectal cancer (n=95). Differentially methylated loci (p=1x107) were validated independently by pyrosequencing in peripheral mononuclear blood cells from 102 never, 45 former and 33 current smokers (n=180). Validated loci included intragenic loci in AHRR, involved in dioxin metabolism, and an intergenic region, 2q37, that was associated with breast cancer risk (p=0.003). Four loci comprised an index that outperformed the gold standard test for former smoking, cotinine. DNA methylation signatures and exposure levels vary between cell types, but few studies have exclusively investigated the target for breast carcinogenesis, the breast epithelial cell. Therefore, a proof-of-principle study into cells from breast milk samples (n=150 milk bank donors) characterised >97% as epithelial (EMA-positive) and >80% as ESR1-positive by flow cytometry. Extracted DNA and RNA underwent methylation, genotyping and expression analyses. ESR1 IGM patterns matched ESR1-positive tumour cell lines. Furthermore, 2q37 was hypomethylated in 19 former smokers compared to 69 non-smokers (p=0.026). To conclude, differentially methylated non-promoter loci may be functionally important, and useful as biomarkers, but tissue selection for prospective epigenetic studies should be carefully considered. Future large-scale breast milk sample collections will inform population-level studies into environmental exposures, breast cancer risk, and mammary biology. Open Access

Coronary angioplasty is a commonly used therapy to remove stenoses in patients with chronic stable angina. Although a successful procedure, there remain both the short-term (stent thrombosis and vessel injury) and long-term risks (late stent thrombosis and significant in-stent restenosis) such that indiscriminate use of stenting can be deleterious. A considerable body of evidence suggests clinical outcomes are improved when stenting is performed using invasive coronary physiology; if stenoses cause ischaemia revascularisation is preferential to deferment, while non-ischaemic stenoses may be treated with medical therapy and observation. Successfully treated vessels are less likely to have major cardiovascular events and patient events are improved. However, there remains no good way of readily and practically predicting whether placing a stent will be successful with normalisation of abnormal coronary physiology. Furthermore, there remains no good parameter to delineate the length of a stenoses and the amount of stenting required to achieve an optimal haemodynamic result. The behaviour of coronary physiology after coronary intervention remains poorly described, particularly in the resting state. Furthermore, human coronary disease is often complex with many coronary stenoses present. This confounds our physiological assessment of a given stenosis and contributes to the poor utilisation of physiological technologies in the patients. In this thesis, I apply phasic assessment throughout the cardiac cycle to describe the behaviour of coronary physiology in humans who undergo assessment and intervention in the catheter laboratory. The behaviour of coronary pressure, flow velocity and resistance will be assessed. I will assess how these parameters change after coronary intervention to develop models of predicting the change in coronary flow and pressure. I will then apply these models to a novel approach that allows assessment of physiological stenosis length, the identification of haemodynamic impact imposed by a single stenosis in diffusely diseased and tandem stenoses. The aim is to produce a new clinical tool that can be readily used in the clinical assessment of challenging coronary disease. Open Access

Asthma is a disease with increasing global prevalence and patients typically suffer from symptoms such as wheezing, breathlessness, chest tightness and cough. In allergic asthma, exposure to allergen can lead to episodes of bronchoconstriction known as the late asthmatic response (LAR).Despite this being a facet of asthma that patients find debilitating and the fact that this feature of asthma is often used in clinical research as an end point to study new therapeutic options, the mechanisms driving the LAR remain unclear. The aim of this thesis was to study the LAR. Initial studies utilised genetically modified (GM) mice missing key allergic effector cells. The data showed that CD4+ and CD8+T cells are central to the aetiology of the LAR. Further studies into the role of the B-cell – mast cell– IgE axis in the murine model indicated that an alternative model was required. In a rat model, it was found that a mast cell stabiliser Disodium Cromoglycate (DSCG), but not a mast cell degranulator, compound 48/80, suppressed the LAR. This suggested that DSCG may be inhibiting LAR independently of its activity on mast cells. As previous data has shown a role for airway sensory nerves and the TRPA1 ion channel in the LAR, the impact of DSCG on sensory nerve activity was investigated. The data indicated that DSCG can inhibit TRPA1 responses in airway sensory nerves and further investigations demonstrated that this was via its actions on the G-protein coupled receptor, GPR35. Overall, this body of work provides new insights into the driving mechanisms behind the LAR and may have uncovered the mystery of the mechanism of action of DSCG. From this, new therapeutic options for asthmatics may become available in the future. Open Access

Platinum resistance causes treatment failure in a variety of cancers including ovarian cancer. Epigenetic changes can contribute to platinum resistance, as the epigenome is highly divergent between platinum-resistant and -sensitive cells, and DNA demethylating agents partially resensitise resistant cells to platinum. The mismatch repair (MMR) gene MLH1 is epigenetically silenced in certain ovarian cisplatin-resistant cells and links between DNA damage and DNA methylation have been reported. This thesis aims to address the hypothesis that platinum-induced DNA damage causes aberrant DNA methylation, which can be selected for leading to drug resistance, and that the methylation response to platinum damage is influenced by MMR status. Two ovarian cancer cell lines, derived from A2780/CP70, containing a fragment of chromosome 3 with either actively expressed MLH1 (proficient) or an epigenetically silenced MLH1 (deficient) were used to derive individual clones after mock or 5 μM (IC95) cisplatin treatment. LINE-1 methylation was significantly increased only in MLH1-proficient clones after cisplatin treatment. Using the Infinium HumanMethylation450 BeadChip array, methylation was measured at ~480,000 CpGs genome-wide in mock- and cisplatin-treated MLH1-proficient and -deficient clones (n=11 per group). Cisplatin induced bi-directional methylation changes, biased toward hypermethylation, in MLH1-proficient and -deficient clones, with a higher number of DNA methylation changes in MLH1-proficient clones. No detectable genomic location or sequence specificity was observed. In peripheral blood mononuclear cell (PBMC) DNA of ovarian cancer patients treated with carboplatin as part of the SCOTROC1 clinical trial, significant methylation changes were detected using the Infinium methylation array when comparing matched samples collected pre-treatment and at relapse. These methylation changes were, again, biased towards hypermethylation post-platinum treatment and occurred preferentially in non-promoter regions of the genome. This thesis demonstrates for the first time that platinum induces bi-directional DNA methylation changes in cell lines and in PBMCs of ovarian cancer patients and there may be both MLH1-dependent and MLH1-independent mechanisms involved in this response. Open Access

Background Poverty drives TB rates but the current TB control approach is disproportionately biomedical. In 2015, the World Health Organisation’s End TB Strategy explicitly identified the need to address the social determinants of TB through socioeconomic interventions. However, evidence concerning poverty-reduction and costs-mitigation strategies is limited. My PhD research aimed to address this knowledge gap. Methods During this PhD, I aimed to develop as an independent researcher while addressing the social determinants of TB in impoverished shantytown communities of Callao, Peru, through integrated projects. The research was divided into two phases: 1) Final follow-up, data collection, analysis, and write-up of: a case-control study defining the TB-poverty association; an ecological study assessing poverty-related risk factors for TB infection and disease; and a cohort study identifying TB-related costs of TB-affected families and creating a clinically-relevant catastrophic costs threshold. 2) Conception, design, implementation, data collection, analysis, and write-up of a household-randomized controlled study of a socioeconomic intervention to improve TB cure and prevention. Results The first phase showed that TB remains a disease of people living in poverty, that “free” TB care was expensive for impoverished TB-affected families to afford, and that incurring catastrophic costs explained as many adverse outcomes as multi-drug resistant (MDR) TB. The second phase showed that, in households receiving the TB-specific socioeconomic intervention, TB-affected households were less likely to incur catastrophic costs, household contacts were more likely to start and adhere to TB preventive therapy, and TB patients were more likely to be cured. Conclusion In impoverished Peruvian shantytowns, poverty remains associated with TB and incurring catastrophic TB-related costs predicted adverse TB outcome. A novel TB-specific socioeconomic intervention reduced catastrophic costs and improved TB preventive therapy uptake and TB cure. The impact of the intervention on TB control will now be evaluated during the Community Randomized Evaluation of a Socioeconomic Intervention to Prevent TB (CRESIPT) study. Open Access

Early life is a period of increased vulnerability to infection, particularly of the respiratory tract. The neonatal adaptive immune system is immature, with a bias towards Th2/Th17 responses and against Th1. Very little is understood about the innate response in early life, especially at mucosal surfaces such as the lung. To address this, Toll-like receptor (TLR) ligands were administered intranasally to neonatal and adult mice. In neonates, pulmonary neutrophilic influx was barely detectable, and expression of inflammatory chemokines greatly attenuated. Administration of exogenous CXCL1 elicited a strong neutrophilic response, indicating that diminished chemokine production is a limiting factor for cellular inflammation in early life. An unbiased microarray approach revealed that whilst expression of immune-related genes was mostly suppressed in the naïve neonatal lung, antimicrobial peptides such as cathelicidin were over-expressed. These novel findings challenge the perception that the infant immune system is simply an ‘immature’ version of the adult system, with implications for development of vaccines and adjuvants for this vulnerable population. An important respiratory pathogen of infants is respiratory syncytial virus (RSV), which can cause bronchiolitis, and in clinical studies is linked with lung dysfunction such as asthma and wheezing later in life. However, the mechanisms by which early life RSV infection could cause these delayed sequelae are unknown, and the effect on the innate immune response has not been explored. The long-term consequences of early life RSV infection were investigated using a murine model. Following neonatal RSV infection, allergen challenge resulted in an exaggerated inflammatory response in the adult murine lung, indicating that pulmonary innate immunity was dysregulated. Further, expression of various immune response, apoptosis-related and circadian rhythm genes was found to be altered. These data provide new evidence that infantile RSV has a causative effect on later lung dysfunction and identify gene targets for further investigation. Open Access

Previous research has provided pre-clinical validation for inhibition of the enzyme N-myristoyltransferase (NMT) target as a novel treatment of fungal and parasitic infections. This thesis describes the discovery of novel NMT inhibitor series derived from high-throughput screens (HTS) of the Pfizer compound collection against NMTs from both Leishmania donovani and Plasmodium falciparum. The further development of two series using structure-based drug design will be discussed, culminating in a series of compounds with efficacy in a mouse malaria model. Following the initial HTS, the hits were triaged based on the orthogonal attributes of structural diversity of each series of analogues; synthetic ease; orthologue selectivity; and estimated ligand efficiency. Iterative screening resulted in a master set that was further triaged through dose-response assays, including selectivity screening against both human NMTs. Eight novel series of NMT inhibitors, half selective for Leishmania and half for Plasmodium NMTs were approved for disclosure by Pfizer. One Leishmania-selective series, based on a thienopyrimidine template, was selected for further development based on a physicochemical hypothesis for achieving cell-based activity. Synthesis of a close analogue of the HTS hit provided validation of the series, which was further developed through the iterative synthesis and testing. The key objective of greatly improved activity in the Leishmania cell-based assays was not achieved. Development of a Plasmodium-selective series of aminomethylindazoles was more successful. Co-crystallisation of a close analogue of the screening hit demonstrated an orthogonal binding mode to the previously published quinoline series. Use of a fragment-linking approach enabled the design and synthesis of novel, highly potent inhibitors with balanced activity against three Plasmodium spp. NMTs, while retaining selectivity over both human NMTs. Excellent translation into cell-based activity was achieved, resulting in several lead compounds with efficacy in a mouse malaria model following both intraperitoneal and oral dosing. Open Access

Cellular environments are highly crowded due to the presence of nucleic acids, proteins and other cellular organelles that can occupy between 20%- 40% of the total cellular volume. Biomolecules, such as RNAs, require precise folding to the most compact native state in order to perform their specific function inside the cell. How crowding environments can influence the folding, stability and functions of RNA molecules is largely unknown. Generally, RNA folding and catalysis are studied in dilute in vitro conditions, which do not truly mimic cellular environments. The work of this thesis combines single-molecule fluorescence resonance energy transfer (smFRET) and bulk cleavage assays to determine the effects of molecular crowding agents in folding and catalysis of model small RNA enzymes, specifically the hairpin ribozyme and the large group II intron ribozyme. The single molecule data show that PEG favours the formation of the most compact (active) structure of RNA enzymes by increasing the docking rate constants. In addition, ribozyme folding and catalysis occurs at low concentrations of Mg2+ in the presence of crowding agents. Lastly, ensemble radiolabeled cleavage assays have shown that molecular crowding agents accelerate ribozymes’ catalysis significantly at lower concentration of Mg2+. This study show that crowding agents influence the stability and the function of ribozymes. These results are consistent with the idea that in vivo RNA enzymes have evolved to function optimally within the crowded environment of the cell. In addition, to understand RNA-protein interactions, the function of Dbp2, a DEAD-box helicase at single molecule resolution has been characterised. The results show that Dbp2 can bind and open double stranded RNA without ATP. In the presence of ATP, the opening and closing events increases, indicating Dbp2 recycling occurs faster. Finally, I have observed that the activity of Dbp2 reduces in the presence of the protein, Yra1. Open Access

Background/Aims: In order to mediate a diverse range of growth and metabolic functions, insulin receptor substrate (IRS) proteins recruit a complex network of intracellular signalling molecules including PI3-kinase and the Grb2 adaptor protein. These studies proposed to answer the question to what extent are the different phenotypes observed in Irs1 and Irs2 knockout mice due to the loss of the whole protein or due to the loss of specific signalling interactions between IRS proteins and either p85 (the regulatory subunit of PI3K) or Grb2? Determining which interactions mediate which biological responses is of interest to our understanding of insulin resistance and its associated pathologies. Methods: The strategy utilised for these studies was to mutagenise the IRS proteins in vivo with specific signalling defects with the aim of disrupting IRS1 or IRS2 signalling via two major insulin mediated pathways, the PI3K-Akt axis and the Grb2-Ras-Raf-MAPK cascade. The phenotypes of the mice were assessed in five main areas: growth, glucose homeostasis, energy homeostasis, cognitive behaviour and fertility. Results: The IRS1-PI3K mutants presented with a severe growth-defect and acute insulin resistance, though through compensatory β-cell expansion maintained normal glycaemia. They were also lean with low circulating leptin levels and displayed a severe defect in learning and memory. IRS2-PI3K mutants were glucose intolerant and insulin resistant with a reduction in β-cell area that resulted in hyperglycaemia and onset of diabetes around 3-months of age. Additionally, the mice had increased fat mass and high circulating leptin levels. In contrast, the Grb2 mutants displayed normal metabolic phenotypes. Conclusion: The data presented here revealed a primary role for IRS signalling via PI3K in regulating metabolic functions. Both the IRS-PI3K mutants largely phenocopy the corresponding Irs knockout mice. In contrast, the Grb2 mutants appeared metabolically normal, suggesting a relatively minor role for IRS-Grb2 interactions in metabolic and growth control. Open Access

Background Cardiac structure and function result from complex interactions between genetic and environmental factors. Population-based studies have relied on 2-dimensional cardiovascular magnetic resonance as the gold-standard for phenotyping. However, this technique provides limited global metrics and is insensitive to regional or asymmetric changes in left ventricular (LV) morphology. High-resolution 3-dimensional cardiac magnetic resonance (3D-CMR) with computational quantitative phenotyping, might improve on traditional CMR by enabling the creation of detailed 3D statistical models of the variation in cardiac phenotypes for use in studies of genetic and/or environmental effects on cardiac form or function. Purpose To determine whether 3D-CMR is applicable at scale, and provides methodological and statistical advantages over conventional imaging for large-scale population studies and to apply 3D-CMR to anthropometric and genetic studies of the heart. Methods 1530 volunteers (54.8% females, 74.7% Caucasian, mean age 41.3±13.0 years) without self-reported cardiovascular disease were recruited prospectively to the Digital Heart Project. Using a cardiac atlas-based software, these images were computationally processed and quantitatively analysed. Parameters such as myocardial shape, curvature, wall thickness, relative wall thickness, end-systolic wall stress, fractional wall thickening and ventricular volumes were extracted at over 46,000 points in the model. The relationships between these parameters and systolic blood pressure (SBP), fat mass, lean mass and genetic variationswere analysed using 3D regression models adjusted for body surface area, gender, race, age and multiple testing. Targeted resequencing of titin (TTN), the largest human gene and the commonest genetic cause of dilated cardiomyopathy, was performed in 928 subjects while common variants (~700.000) were genotyped in 1346 subjects. Results Automatically segmented 3D images were more accurate than 2D images at defining cardiac surfaces, resulting in fewer subjects being required to detect a statistically significant 1 mm difference in wall thickness. 3D-CMR enabled the detection of a strong and distinct regionality of the effects of SBP, body composition and genetic variation on the heart. It shows that the precursors of the hypertensive heart phenotype can be traced to healthy normotensives and that different ratios of body composition are associated with particular gender-specific patterns of cardiac remodelling. In 17 asymptomatic subjects with genetic variations associated with dilated cardiomyopathy, early stages of ventricular impairment and wall thinning were identified, which were not apparent by 2D imaging. Conclusions 3D-CMR combined with computational modelling provides high-resolution insight into the earliest stages of heart disease. These methods show promise for population-based studies of the anthropometric, environmental and genetic determinants of LV form and function in health and disease. Open Access

The precise division of DNA into two daughter nuclei is a critical event in every cell cycle. The pairing of sister chromatids in order to allow accurate segregation, is achieved by two primary mechanisms – the SMC complex cohesin binds as replication ends and holds sisters together with protein-mediated links; and the process of replication intertwines sister chromatids in a way that results in DNA-DNA mediated links called catenations. At the point of segregation, it is crucial that cohesin is destroyed and all catenations are resolved. Topoisomerase 2 (Top2) is the primary actor in S. cerevisiae able to resolve DNA inter- twines. This enzyme appears to maintain an equilibrium level of catenations between S phase and mitosis and it has been shown to be capable of both introducing and removing catenations by its strand passage mechanism. At the metaphase to anaphase transition, a shift to positive supercoiling may help to direct the action of topoisomerase enzymes towards full decatenation at the crucial moment of DNA segregation. Further, it has been shown that condensin, another SMC complex, has a pivotal role in this supercoiling transition. Here, work is presented to link this change in supercoiling to physical tension across sister chromatids aligned on the metaphase plate. The regulation of this transition requires Cdc5 kinase, Ipl1 kinase and acts independently of the Spindle Assembly Checkpoint. Furthermore, the supercoiling transition is linked to a global enrichment of Smc2 across the entire length of chromosomes as the spindles form and anaphase progresses. This enrichment is matched by Top2 binding and this Top2 enrichment depends directly on the action of Smc2. It is hypothesised that this recruitment of Top2 is dependent on catenations being revealed by the supercoiling action of Smc2. If Top2 is not active, then Smc2 continues to bind globally to DNA and further increase the level of positive supercoiling, suggesting that the complex is able to sense the presence of remaining catenations. Finally, a secondary last minute catenane resolution pathway may have been identified that occurs after the supercoiling transition and is regulated by the Anaphase Promoting Complex. Open Access

Type I interferons (IFNs) are the first line of defence against viral pathogens. They play a crucial role in activating antiviral responses to limit viral replication and viral spread. Human respiratory syncytial virus (RSV) can cause severe lung inflammation and this has been associated with polymorphisms in several innate immune response genes, including many that control the IFN system. In order to elucidate the importance of type I IFNs in regulating local inflammatory responses in the lung, their role in pulmonary immune responses to RSV infection was assessed in mice lacking the type I IFN receptor (Ifnar1-/-). Levels of proinflammatory cytokines and chemokines were markedly reduced during RSV infection. Similar results were obtained when in Ifnar1-/- mice were challenged with non-infectious innate stimuli such as Toll-like receptor agonists. It was further shown that the administration of recombinant IFN-α together with innate stimuli was sufficient to potentiate the proinflammatory cytokine production. Exploiting a reporter mouse strain expressing GFP under the control of the Ifna6 promoter, alveolar macrophages (AMs) were identified as the predominant source of type I IFNs during RSV infection. This type I IFN production solely relied on signalling through the cytosolic RIG-I-like receptors (RLRs), as AMs lacking MAVS, the adaptor downstream of RLRs, were unable to produce type I IFNs. Furthermore, Mavs-/- mice showed compromised production of proinflammatory cytokines and chemokines, and as a consequence had an increased viral burden and RSV-induced pathology. Interestingly, Mavs-/- mice had a marked deficiency in the production of monocyte chemoattractants and recruitment of inflammatory monocytes (infMo). The administration of recombinant CCL2 during RSV infection permitted monocyte recruitment into the lung of Mavs-/- mice and mediated an unexpected antiviral activity and reduced disease severity. This work emphasises a role for type I IFNs in antiviral immune responses by driving early proinflammatory cytokine responses and recruitment of antiviral monocytes, a yet underappreciated novel cell type responsible for dampening RSV-disease burden in vivo. Open Access

The classical view in systems neuroscience is that individual regions of the brain are specialised for the execution of specific tasks. This position is supported by a long history of lesion studies and more recently by a wealth of functional neuroimaging research, which has provided insights into how the anatomy of the brain relates to behaviour. However, in recent years this classical view has changed. Instead, new techniques have revealed intrinsic functional connectivity networks (ICNs) that reflect underlying patterns of structural connectivity, now considered to be the fundamental neural architecture supporting cognition and behaviour (Adams, 1982, Smith et al. 2009b, Honey et al. 2009). The activity of these ICNs is dependent on intrinsic fluctuations in their activity and the behavioural context, which dynamically reconfigure over time (Fox et al. 2005, Ances et al. 2009). Therefore, investigation of brain networks needs to consider not only the structural connections that constrain functional interactions, but also dynamic changes in functional interactions. An ongoing challenge, therefore, is to define a framework incorporating these neural dynamics that combines insights from structure, function and behaviour. Dynamical systems theory may provide a flexible framework that combines all of these levels. Current theory proposes that healthy neural dynamics operate in a multistable regime that promotes flexible information processing and behaviour (Kelso 2012, Tognoli and Kelso 2014, Shanahan 2012, 2010b, a, Friston 1997, Leech and Sharp 2014, Irner 2007). Large-scale multistable dynamics are likely constrained by underlying structural connections between brain regions. (Honey et al. 2009, Deco, Jirsa, McIntosh, Sporns, and Kötter 2009, Deco, Jirsa, and McIntosh 2011, Deco et al. 2008, Cabral et al. 2011, Cabral et al. 2012). However, it is unclear how such multistable dynamics may relate to behaviour, or how they are constrained by network structure. Moreover, whilst such dynamical accounts place considerable importance on the structural connectivity of the brain, it is unclear how the multistable dynamics are altered when pathological processes result in structural disconnections. In this thesis, I explore such dynamical accounts of the brain using a range of neuroimaging and computational approaches, and examine the implications of this level of description in one example of structural disconnection, namely the consequences of traumatic brain injury. Open Access

Poor memory for recent conversations is the commonest presenting symptom in patients attending a cognitive neurology clinic. They also frequently have greater difficulty following and remembering conversations in the presence of background noise and/or unattended speech. While the ability to participate in and recall conversations depends on several cognitive functions (language-processing, attention, episodic and working memory), without the ability to perform auditory scene analysis, and more specifically speech-stream segregation, recall of verbal information will be impaired as a consequence of poor initial registration, over and above impaired encoding and subsequent retrieval. This thesis investigated auditory attention and speech-stream segregation in healthy participants (‘controls’) and patients presenting with ‘poor memory’, particularly a complaint of difficulty remembering recent verbal information. Although this resulted in the recruitment of many patients with possible or probable Alzheimer’s disease, it also included patients with mild cognitive impairment (MCI) of uncertain aetiology and a few with depression. Functional MRI data revealed brain activity involved in attention, working memory and speech-stream segregation as participants attended to a speaker in the absence and presence of background speech. The study on controls demonstrated that the right anterior insula, adjacent frontal operculum, left planum temporale and precuneus were more active when the attended speaker was partially masked by unattended speech. Analyses also revealed a central role for a right hemisphere system for successful attentive listening, a system that was not modulated by administration of a central cholinesterase inhibitor. Therefore, this study identified non-auditory higher-order regions in speech-stream segregation, and the demands on a right hemisphere system during attentive listening. Administration of a central cholinesterase inhibitor did not identify any benefit in the present patient group. However, my research has identified systems that might be therapeutic targets when attempting to modulate auditory attention and speech-stream segregation in patients with neurodegenerative disease. Open Access

Human noroviruses (HuNoV) are a significant cause of viral gastroenteritis in man worldwide. Noroviruses are also associated with intestinal disease in multiple species, including dogs. Canine norovirus (CNV) was initially discovered in 2007 and the first aim of this thesis was to determine the prevalence of CNV in the UK dog population. qPCR screening of canine stool samples did not identify CNV RNA, but canine astroviruses (CaAstV) were serendipitously identified and subsequently characterized according to the second aim of this work. For serological screening, CNV virus-like particles (VLPs) to three CNV strains were produced. CNV circulation in the UK was confirmed by identification of CNV-specific antibodies in 60% of canine serum samples collected in 2012-2013. The third aim of this thesis was investigate to CNV interactions with host cells by identifying the cellular attachment factor for CNV. Synthetic carbohydrates and canine tissue samples were used to assess the binding specificity of CNV VLPs, and it was shown that antigens of the HBGA family were recognized. Phenotyping studies then demonstrated expression of HBGAs in dogs. As HuNoV also uses HBGAs to attach to cells, this raised concerns that dogs may be susceptible to HuNoV. Evaluating the zoonotic risk of enteric viruses in dogs was the final aim of this thesis. The susceptibility of dogs to HuNoV and hepatitis E virus (HEV) was determined by screening canine samples for the presence of HuNoV or HEV RNA and HuNoV or HEV-specific antibodies. Antibodies to both HuNoV and HEV were identified in dogs, and results confirmed HuNoV VLPs can bind to canine gastrointestinal samples. This data indicates that dogs are susceptible to HuNoV and HEV infections. In conclusion, this thesis has provided epidemiological and molecular characterization of CNV and CaAstV, in addition to highlighting the zoonotic potential for CNV, HuNoV and HEV in dogs. Open Access

The emergence of novel influenza A viral lineages to which there is limited or no immunity remains an on-going concern for human and animal public health. At least three of the pandemics reported in human hosts during the 20th and 21st century are believed to have emerged as a result of reassortment between human, avian and swine viruses circulating at the time. The emergence of novel lineages with pandemic potential is frequently believed to occur as a consequence of reassortment. The ability to predict when and how the next novel lineage with pandemic potential will arise remains challenging. Additionally, the dynamics of infection in systems where multiple subtypes are known to circulate remain poorly understood. Gaining insight into these biological processes is of vital importance if we are to understand how much co-infection is occurring, and hence the opportunity for reassortment. This thesis uses data analysis, novel algorithmic methods applied to genomic data and mathematical modelling to develop an understating of the dynamics of influenza A infection within non-human hosts. It seeks to analyse patterns of reported reassortment in influenza A across all hosts, geographic regions and time periods. Chapters 3 and Chapter 4 use data on reported reassortment events to identify additional, previously unreported suspected reassortant viral lineages, and develop a quantitative definition of significant reassortants. These chapters use publicly available sequence data to understand to what extent published data on reassortants reflects the possible extent of reassortment that can be detected with an iterative clustering algorithm. They develop methods that can be used to assign a rank score to suggest whether an individual isolate is a suspected reassortant. Methods presented in these chapters are applied to all publicly available influenza A sequence data from two different databases. The estimation of epidemiological parameters such as R0 and the time of virus introduction into a population (Ts) are of vital importance to infer in the event of an outbreak of a novel influenza virus n a poultry production facility. This thesis assesses the performance of different surveillance strategies to estimate R0 and Ts using simulated surveillance data from an outbreak of H7N9 in commercial poultry. Since the detection of novel H7N9, it has been isolated from numerous live bird markets. Live birds markets are known to harbour diverse subtypes of avian influenza within numerous different susceptible hosts. Co-infection with two different influenza viruses is a prerequisite for reassortment to occur therefore the final results chapter explores the conditions required to generate the multi-strain dynamics of different AI subtypes and the resulting co-infection dynamics that emerge. Taken together, through these topics addressed in this thesis we understand more broadly the possible true extent of reassortant detectable in publicly available sequence data. How surveillance data collection can be optimised to help estimate epidemiological parameters in the event of an outbreak, and the possible mechanisms that generate the multi-strain dynamics within a live bird market. Open Access

Nitric oxide (NO) is an important signalling molecule. One-way that NO’s production is regulated is through the asymmetric dimethylarginine (ADMA)-dimethylarginine-dimethylaminohydrolase (DDAH) pathway. One role of NO in the brain is as a negative feedback signal in ventral tegmental area (VTA) dopamine (DA) neurons. It has been shown through electrophysiology that in response to depolarisation NO diffuses retrogradely and potentiates GABA release, which in turn inhibits DA neuron activity. It has been hypothesised on the basis of electrophysiology and pharmacology experiments that NO is produced by VTA DA neurons. However although both DDAH1 and nitric oxide synthase (nNOS; which is the source of neuronal NO) are expressed in the VTA, it is not clear in which cell types. I, therefore, used immunohistochemistry to investigate nNOS and DDAH1 expression in the parabrachial pigmented area (PBP) of the VTA. The PBP is where most electrophysiological recordings from DA neurons in the VTA are carried out. I found that nNOS is expressed in the PBP of the VTA in close proximity to, but not colocalized, with DA neuron cell bodies and processes. This may indicate that NO production occurs in another cell type in response to VTA stimulation. Although DDAH1 expression was found in both DA and non-DA cells in the PBP, staining was present, although noticeably weaker, in DDAH1 global knockout mice. Thus no firm conclusion can be drawn on DDAH1 expression in the VTA. Open Access

Paracetamol is a popular antipyretic and analgesic medication. It is a known hepatotoxin in overdose and is the commonest cause of acute liver failure in the UK. There is significant variability in inter-individual susceptibility to the hepatotoxic effects of paracetamol, which is incompletely understood. This thesis describes work done in murine models of paracetamol-induced acute liver failure with the aim of identifying causes of variable susceptibility and understanding the immune response to liver injury. A quantitative trait locus (QTL) mapping approach was taken using a murine strain susceptible to paracetamol hepatotoxicity (C3H/HeH) crossed with a relatively resistant strain (C57BL/6). Novel QTLs on murine chromosome 17 and 18 were identified that associated with response to paracetamol. Within the loci a number of candidate genes were identified. A survey of 10 inbred mouse strains for their response to paracetamol was conducted and highlighted the large variability within each strain. It was hypothesised that this variability might be due to differences in intestinal microbiota. A study of the role of intestinal microbiota in paracetamol-induced liver failure was conducted, by comparing response in germ free (GF) mice and conventional (CV) controls. This demonstrated that the presence of intestinal microbiota influenced the sulphonation:glucuronidation ratio during paracetamol metabolism. Although the extent of liver injury as assessed by necrosis and liver enzyme elevation was the same in GF and CV mice, there was evidence of a protective effect of a sterile intestine. Finally a reverse genetics approach was taken to study the influence of the gene Slpi in the secondary immune response to liver injury. It was shown that absence of SLPI protected mice against peak liver injury. The work presented in this thesis highlights some potential sources of variability in response to paracetamol and a number of targets that with further research could have therapeutic potential. Open Access

During the course of development, specification of the three embryonic germ layers, ectoderm, mesoderm and definitive endoderm (DE), is a critical process by which pluripotent cells acquire the temporal and spatial information needed to form specialised tissues. Of these initial germ layers, the DE arises during gastrulation, which latterly gives rise to the liver, pancreas, lung and epithelial lining of the digestive tract. Elucidation of the molecular mechanisms that govern DE specification not only facilitates our understanding of developmental biology but also aids in the differentiation of human pluripotent stem cells to specific cell types for disease modelling and regenerative therapies. DE formation is largely driven by the cooperation of Activin/Nodal and Wnt/β-catenin signalling, however recent evidence has additionally implicated PI3K/Akt signalling in modulating this process. Although it has been previously reported that PI3K activation acts to antagonise the in vitro differentiation of DE, the molecular mechanisms responsible for this effect remains unclear. To address this issue, this study utilises pluripotent human embryonic stem cells (hESCs) as an in vitro model to interrogate the molecular underpinnings of DE formation through a fully defined differentiation protocol. Modulation of PI3K activity was found to reciprocally downregulate the activation of Smad2/3, which was mitigated in the presence of the PI3K inhibitor LY294002 (LY). Suppression of PI3K/Akt signalling prolongs the activation of Smad2/3 in response to Activin, promoting their nuclear accumulation and the enhancement of transcriptional activity, resulting in the upregulation of mesendoderm and DE gene expression. Activation of PI3K negatively impacts the activity of Smad2/3 via phosphorylation of the Smad2/3 linker T220/T179 residue, which is fully independent of Erk and CDK activity. Phosphorylation of this residue induces the recruitment of the E3 ubiquitin ligase Nedd4L to activated Smad2/3, which in turn promotes their ubiquitin-mediated degradation and attrition of activity. Inhibition of mTORC2 activity by both inhibitor supplementation and genetic manipulation, rather than modulation of Akt or mTORC1 activity, recapitulates the LY-mediated reduction of T220/T179 phosphorylation and increases the duration of Smad2/3 transcriptional activity, promoting a more robust mesendoderm and endoderm differentiation. These findings reveal a new and novel connection between the PI3K/mTOR and TGFβ/Activin pathways, which will greatly impact our understanding of both cell fate determination and the preservation of normal cellular functions. Notably, identification of mTORC2 as a key player in the regulation of this differentiation provides new avenues through which hESCs differentiation protocols can be improved for both regenerative and biomedical applications. Open Access

Rhinoviruses (RV) are the major cause of asthma exacerbations, with evidence supporting a synergistic relationship between viruses and allergen. Dendritic cells (DCs) mediate airway recognition of both viruses and allergens and instruct downstream immune responses. However, nothing is known about the role of DCs in viral exacerbations of asthma. We hypothesised that following viral infection, asthmatic patients have reduced airway recruitment of anti-viral type II mDCs and pDCs, in addition, these patients exhibited impaired DC anti-viral functionality. To address these hypotheses, moderate atopic asthmatic patients and healthy controls were experimentally infected with RV-16. Bronchoalveolar lavage (BAL) was collected at baseline, day 3 and day 8 post infection and DC populations were isolated and analysed using fluorescence activated cell sorting (FACS). At baseline, the number of type II mDCs, which cross prime CD8+ T cells and induce type I anti-viral responses, was significantly reduced in asthmatic patients. Following RV-16 infection, their recruitment to the airways was delayed compared to healthy controls. No difference in the recruitment of type I mDCs and pDCs was observed. Furthermore, we identified reduced anti-viral interferon reponses in purified peripheral pDCs from asthmatic patients in response to ex vivo RV-16 stimulation. IFN-α production was significantly reduced in asthma patients compared to healthy controls (p=0.028). IFN-λ production was lower in a sub group of asthmatics, although not significantly reduced compared to healthy controls. Deficient IFN-β production was identified in BAL cells from asthmatic patients following ex vivo RV-16 stimulation, compared to healthy controls (p=0.016), with a trend towards reduced IFN- λ. We aim to correlate these findings with clinical parameters, to determine whether deficient cellular recruitment and innate IFN responses have an impact on viral clearance, lung function and both upper and lower airway symptoms. This work will aid our understanding of the pathogenesis of asthma and the role of DCs in viral induced exacerbations. Open Access

Carbon nanotubes (CNTs) possess a number of unique properties and play a pivotal role in the nanotechnology industry with widespread applications in construction, medicine and electronics. Previous studies have demonstrated an association between CNT inhalation and adverse cardiovascular effects, however, it is currently unclear whether this is due to the release of pulmonary inflammatory mediators or particle translocation across the blood-gas barrier. A better understanding of CNT interactions at this barrier is therefore essential for developing safer nanomedicines and reducing occupational health risks associated with carbon nanotube manufacture. The working hypothesis of this study was that the physicochemical properties of multi-walled carbon nanotubes (MWNTs) would determine their toxicity and reactivity at the blood-gas barrier. To test this hypothesis, primary alveolar type-2 epithelial cells (AT2) and microvascular endothelial cells (HPMVEC) were isolated from resected human lung and employed alongside a unique immortalised human type-1 alveolar cell line (TT1). The cells were then exposed to functionalised MWNT to determine the cellular reactivity of the particles. Using this model, it was shown that MWNT exposure induced endothelial cell activation, characterised by a pro-inflammatory and pro-thrombotic response, and MWNT functionalisation was a key factor in determining the nature of this response. Acid-oxidised MWNTs induced the greatest endothelial activation, however as-received and thermochemically grafted MWNTs induced a lesser pro-inflammatory response that also coincided with an increase in ROS formation. Of the thermochemically grafted MWNTs, P(M4-VP) and P(PEGMA) MWNTs induced the greatest and smallest response respectively. At the epithelium, MWNTs were far less reactive. All MWNTs were non-toxic to AT2 cells, however a small cytotoxic and inflammatory response was observed in TT1s following chronic exposures (72h) and MWNTs were present in both vesicles and the cytoplasm of TT1s following 24h exposure. The thermochemically grafted P(M4-VP) MWNTs were shown to induce the greatest pro-inflammatory response during these chronic exposures. In vivo, disruption of cell-cell junction could have implications for particle translocation, inflammation and atherosclerotic lesion formation. Exposure to MWNTs was shown to increase the permeability of type 1, but not type 2 alveolar epithelial barriers. The greatest disruption was seen at the endothelial cell junctions, which was accompanied by depletion of adherens junction protein and rearrangement of the actin cytoskeleton. Finally, to assess whether epithelial-endothelial crosstalk could affect particle toxicity a co-culture model of the alveolar blood-gas barrier was constructed using a transwell system where alveolar type 1 and type 2 epithelial cells were separated from pulmonary microvascular cells by a porous membrane. There were disparities between the mono and co-culture findings and acid oxidised MWNTs applied to the epithelial surface of the co-culture induced a significant release of inflammatory mediator into both the apical and basolateral chambers of the co-culture system, indicative of inflammation in both the pulmonary and vascular systems. This highlighted the need to incorporate endothelial cells into NP toxicity assays, especially when considering the cardiovascular reactivity of a particle. This thesis has demonstrated that MWNT toxicity is both cell and MWNT surface functionalisation dependent, and that endothelial activation could be responsible for the adverse cardiovascular effects observed following CNT inhalation in vivo. These findings will aid the generation of safer nanomaterials and further the understanding of CNT induced cardiovascular reactivity. Open Access

Forced expiatory volume in one second (FEV1) is the most widely accepted marker in COPD. However, COPD is now well recognised as a multisystem and heterogeneous disease, and it is unlikely that a single marker could reflect the complex range of observed pathophysiologic changes. Exercise intolerance is a common symptom in COPD, and reflects both lung and skeletal muscle dysfunction. Logistical issues often mean physical performance is not routinely measured objectively in all clinical settings. Gait speed is an easy, quick to perform established measure of physical performance in the elderly, and has been recommended as a “vital sign” or “global marker of well-being” that reflects biological rather than chronological age. Functional decline is common to both ageing and COPD but gait speed has not been validated in COPD. In 586 patients with stable COPD, usual gait speed over 4 metres (4MGS) had excellent reliability and showed concurrent validity with well-established measures in COPD. Stratification according to slow 4MGS identified significant impairments in exercise capacity and poor health-related quality of life. Subsequent studies showed that 4MGS was responsive to intervention with pulmonary rehabilitation (n=301) and longitudinal decline at one year (n=162). Using both distribution- and anchor-based approaches the minimum clinically important difference was estimated to be 0.11ms-1. In a cohort of patients with stable COPD (n=402), slower 4MGS was associated with significantly increased risk of all-cause mortality at three years, and provided prognostic information independent of FEV1. In a prospective cohort of 213 patients admitted to hospital with an acute exacerbation of COPD, slower 4MGS was associated with significantly increased odds of all-cause readmission at 90 days. The 4MGS is a reliable measure of physical performance that has potential utility as a simple assessment tool to provide a more comprehensive assessment of patients with COPD. Open Access

Glioblastoma multiforme (GBM) are the most aggressive and devastating tumours of the brain and are essentially incurable. They are defined by diffuse invasion of the surrounding brain parenchyma along preexisting structures like the vasculature. Glioma stem cells (GSC) are thought to be largely responsible for tumour recurrence following treatment due to their high resistance to therapy, their ability to recapitulate tumours from single cells and their marked invasive potential. Here we show, that normal neural stem cell in the subventricular zone are compartmentalised by endothelial ephrinB2 and their proliferation limited through activation of p53 in an Eph signalling dependent manner. GSCs however evade both compartmentalisation and proliferation inhibition and are able to invade perivascularly. Intravital imaging, coupled with mechanistic studies in vitro revealed that upregulation of ephrinB2 in highly aggressive, mesenchymal GSCs enables escape from endothelial compartmentalisation through homotypic forward signalling. Surprisingly we also find that that ephrinB2 reverse signalling promotes tumourigenesis by mediating anchorage-independent cytokinesis through activation of RhoA. In preclinical models using human GSCs we show, that inhibition of ephrinB2 by RNA silencing or with ephrinB2-blocking antibodies strongly suppresses tumourigenesis of established glioblastoma by inducing cell-cycle arrest and blocking GBM/vascular interactions. Thus, ephrinB2 is an oncogene and represents an attractive candidate for anti-GBM therapies aimed at eradicating the GSC compartment by targeting both glioma invasion and proliferation. Open Access

Microwave and millimetre-wave technology has enabled many commercial applications to play a key role in the development of wireless communication. When dissipative attenuation is a critical factor, metal-pipe waveguides are essential in the development of microwave and millimetre-wave systems. However, their cost and weight may represent a limitation for their application. In the first part of this work two 3D printing technologies and electroless plating were employed to fabricate metal pipe rectangular waveguides in X and W-band. The performance for the fabricated waveguides was comparable to the one of commercially available equivalents, showing good impedance matching and low attenuation losses. Using these technologies, a high-performance inductive iris filter in W-band and a dielectric flap phase shifter in X-band were fabricated. Eventually the design and fabrication of a phased antenna array is reported. For microwave and millimetre-wave applications, system-on-substrate technology can be considered a very valuable alternative, where bulky coax and waveguide interconnects are replaced by low-loss transmission lines embedded into a multilayer substrate, which can include a wide range of components and subsystems. In the second part of this work the integration of RF MEMS with LTCC fabrication process is investigated. Three approaches to the manufacture of suspended structures were considered, based on laser micromachining, laser bending of aluminium foil and hybrid thick/thin film technology. Although the fabrication process posed many challenges, resulting in very poor yield, two of the solution investigated showed potential for the fabrication of low-cost RF MEMS fully integrated in LTCC technology. With the experience gained with laser machining, the rapid prototyping of high aspect ratio beams for silicon MEMS was also investigated. In the third part of this work, a statistical study based on the Taguchi design of experiment and analysis of variance was undertaken. The results show a performance comparable with standard cleanroom processing, but at a fraction of the processing costs and greater design flexibility, due to the lack of need for masks. Open Access

Amyloids are proteinaceous aggregates best known for their role in degenerative diseases involving protein misfolding. Research into amyloid has intensified in recent times due to its prominence in many debilitating human diseases and limited understanding of the causes. The discovery of functional amyloids in a broad range of species has enhanced our understanding of amyloid, of these the curli system of E. coli has been extensively studied, in this system CsgC was identified as a potent inhibitor of amyloid. An additional protein was discovered in some curli operons in other species termed CsgH and warrants further study. A morphologically similar but genetically distinct bacterial functional amyloid system was identified in Pseudomonas encoded by the fapABCDEF operon and termed amyloid-like fibres (Alf). The study of functional amyloid has the potential to provide insights into how amyloid can be controlled. The aims of this thesis were to investigate the novel functional amyloid system of Pseudomonas with a view to structural and functional characterisation of the individual components. The structure and function of the CsgH protein were also studied by nuclear magnetic resonance (NMR) and the ThioflavinT (ThT) amyloid fibrillation assay. Constructs were produced for all the Alf proteins and the more structured components, FapD and FapF, were optimised to produce constructs for structural study. The structure of CsgH was solved successfully using NMR and showed that the protein shared a similar tertiary structure to CsgC. The function of the CsgH was shown to be similar to CsgC inhibiting amyloid formation by CsgA at substoichiometric concentrations. Mutagenesis, ThT assay and NMR were used to show that CsgH and CsgA interact and that several charged residues have an important role in function. It was also interesting to note that CsgH was capable of inhibiting amyloid formation by the FapC amyloid protein of Pseudomonas. Open Access

Intensive Care Unit Acquired Paresis (ICUAP) is a major health concern facing today’s critical care patients. This syndrome of severe muscle wasting and weakness is associated with significant morbidity and mortality, resulting in a substantial clinical and financial burden. However as yet, other than early mobilisation and rehabilitation, there are no available treatments. The aim of this work was to identify molecular mechanisms in the aetiology of ICUAP with a view to finding novel therapeutic targets. In a prospective study of high-risk elective cardiothoracic patients, prolonged elevation of circulating GDF-15, a stress-induced cytokine and TGF-β family member, was associated with acute muscle wasting. Following on from this work, the hypothesis that GDF-15 might be a driver of muscle wasting in ICUAP was tested in an observational study of 20 ICUAP patients and 7 elective cardiothoracic controls. Both circulating GDF-15 and muscle GDF-15 mRNA expression were elevated in patients compared to controls. In addition microRNAs essential for muscle homeostasis and protection against TGF-β signalling were found to be down-regulated. In vitro experiments demonstrated that GDF-15 could increase expression of muscle atrophy-related genes and supress these microRNAs. Increased sensitivity to TGF-β signalling secondary to down-regulation of muscle microRNAs driven by GDF-15 hypothesised as an aetiological mechanism of muscle wasting in critical illness. Supporting this hypothesis, transfection of miR-181a into C2C12 cells reduced the response to TGF-β1 and increased TGF-β signalling was found in the muscle of ICUAP patients. 9 An interventional study of neuromuscular electrical stimulation (NMES) in ICU demonstrated that although increased mobility in critical care patients can improve outcomes there is a risk that NMES may cause muscle damage. Finally preliminary data, suggested that miR-181a is a potential biomarker for ICUAP. GDF-15 and microRNAs are potential therapeutic targets and this work supports the need for further investigation into these pathways in ICUAP. Open Access

Infections with parasitic trypanosomatids cause serious neglected tropical diseases, which include cutaneous leishmaniasis (CL), causing skin ulceration, and human African trypanosomiasis (HAT), which causes flu-like symptoms followed by neurological features and is fatal if left untreated. Current disease management in both cases is inappropriate owing to suboptimal diagnosis and limited drugs with variable efficacy and severe adverse effects. Further research into their pathogeneses is critical for the discovery of novel diagnostic tools and therapeutic targets. Experimental Leishmania (L.) major infection in two murine CL models with differing pathologies, and HAT caused by Trypanosoma brucei (T. b.) rhodesiense infection, were investigated from a systemic metabolic perspective, using 1H nuclear magnetic resonance (NMR) spectroscopy. This was combined with cytokines and (for CL models) faecal microbial diversity measures, as additional markers of host response. In addition to the discovery of a range of Leishmania infection-associated markers, including putative parasitic end-products alanine and succinate, significant divergences in measures were observed between the self-healing and non-healing CL mouse models. Such differences included variations in glucose and lipid metabolism, in inflammatory status, and in faecal Clostridia and Gammaproteobacteria abundance. Striking differences were also observed in the degree of statistical links between these measures, using newly developed correlation network analyses, implying that a rapid and inter-connected systems response was required for effective parasitic elimination. This project also represents the first study to provide a comprehensive description of host metabolic changes in HAT. A total of 16 and 31 significant discriminatory metabolites were determined in plasma via 1H NMR spectroscopy and ultra-performance liquid chromatography/mass spectrometry (UPLC-MS), respectively, including membrane phospholipids and amino acids, linked with inflammation. These findings provide a unique insight into underlying metabolic interactions that may contribute to pathology, and reveal a set of infection-related markers with possible diagnostic potential for CL and HAT. Open Access

Small area studies are commonly used in epidemiology to assess the impact of risk factors on health outcomes when data are available at the aggregated level. However the estimates are often biased due to unmeasured confounders which cannot be taken into account. Integrating individual-level information into area-level data in ecological studies may help reduce bias. To investigate this, I develop an area/ecological level propensity score (PS) to integrate individual-level data and then synthesise the area-level PS with routinely available area-level datasets, such as hospital episode statistics (HES) and census data. This framework comprises three steps: 1. Individual level survey data is used to obtain information on the potential confounders, which are not measured at the area-level. Using a Bayesian hierarchical framework I synthesise these variables and calculate PS at the ecological level, taking into the account the correlation among the potential confounders. 2. The calculated PS is included as a scalar quantity in the regression model linking environmental exposure/risk factors and health outcome. As PS has no epidemiological interpretation, I introduce a number of flexible functions to allow for nonlinear effects, such as fixed-knot splines, reversible jump MCMC (RJ) and random walk (RW). 3. As real surveys are typically characterized by a limited coverage compared to small area studies, I impute the ecological PS in the areas with no survey coverage. I propose two new imputation models: random walk and cluster imputation (including a) regression tree and b) profile regression ) to relax the assumption of linearity, and through simulations, both imputation models are proven to produce better results than the traditional linear imputation model. I conclude that integrating individual-level data via PS is a promising method to reduce the bias intrinsic in ecological studies due to unmeasured confounders and I introduce a real application on small area studies for evaluating the effect of air pollution on CVD hospital admissions in England. Open Access

Rationale: Exposure to indoor NO2 has been implicated as a cause of respiratory symptoms suggestive of asthma but evidence has been inconsistent. Objective: To review the existing evidence and examine the effects of indoor NO2 in adult asthma-related symptoms using data from an adult multi-centre cohort (ECRHS) followed up for 20 years. Methods: Studies on respiratory health and indoor NO2 were systematically reviewed and meta-analyses performed. Cross-sectional analyses within a sub-set of ECRHS participants with indoor NO2 measurements were conducted to assess the associations of asthma severity and wheeze prevalence with NO2. A regression model was developed to predict indoor NO2 for a larger ECRHS sample without indoor NO2 measurements. GEE analyses were conducted to examine the long-term effects of gas cooking and modelled indoor NO2 on wheeze and asthma score. To investigate the effect of gas-generating NO2 peaks on asthma exacerbation a panel study was also piloted using a new-to-market portable NO2 sensor. Main results: The systematic review identified 50 studies, mainly in children. Results of meta-analyses suggested a significant association between 12-month period prevalence of wheeze and indoor NO2. Within ECRHS prevalence of wheeze but not asthma severity was associated with measured indoor NO2. Long-term associations of asthma-related symptoms with predicted indoor NO2 exposure but not gas cooking were significant. Interpreting this is difficult as the latter analyses (gas cooking) included a larger number of centres and some heterogeneity across centres was observed in the analysis on asthma score. Gas appliances, outdoor NO2, monthly temperature and country were the main predictors of indoor NO2. Evaluation of the pilot study recommends better recruitment strategies and independent calibration of NO2 sensor. Conclusions: There is some evidence for a link between indoor NO2 and asthma-related symptoms. Health risks may be small but are applied to a substantial proportion of the population. Open Access

Active whisking is an important model sensorimotor behavior, but the function of the cerebellum in the rodent whisker system is unknown. I have made patch clamp recordings from Purkinje cells, granule cells and interneurons in vivo to identify whether and how cerebellar activity encodes kinematic features of whisking including the phase and set point. I show that Purkinje cell spiking activity changes strongly during whisking bouts. On average, the changes in simple spike rate coincide with or slightly precede movement, indicating that the synaptic drive responsible for these changes is predominantly of efferent (motor) rather than re-afferent (sensory) origin. Remarkably, on-going changes in simple spike rate provide an accurate linear read-out of whisker set point. Thus, despite receiving several hundred thousand discrete synaptic inputs across a non-linear dendritic tree, Purkinje cells integrate parallel fiber and interneuron input to generate precise information about whisking kinematics via linear changes in firing rate. In addition, I demonstrate that evidence is consistent in both granule cells and interneurons activity within the circuit, providing a linear encoding regime of whisker set point. Open Access

The development of colorectal cancer (CRC) occurs sequentially through the accumulation of genetic changes, or mutations, resulting in unrestrained cellular proliferation and survival. The genetic changes leading to CRC are well defined, with aberrant activation of the Wnt signalling pathway and loss of the tumour suppressor p53 playing key roles. This project aimed to investigate the role of the orphan nuclear receptor liver receptor homologue-1 (LRH-1) in the development of CRC. Previous work in this laboratory identified LRH-1 as an important mediator of the estrogen response in breast cancer cells. LRH-1 has also been implicated in the development of CRC, where it promotes cell cycle progression and synergises with β-catenin. Here I provide evidence for novel mutations in the DNA binding domain of LRH-1 in CRC. I found that CRC-associated LRH-1 mutations affect its ability to bind LRH-1 response elements. To better define the mechanisms of LRH-1 action in CRC, gene expression microarray analyses were performed in two CRC cell lines following siRNA-mediated LRH-1 knockdown. Several previously undescribed candidate LRH-1-regulated genes were identified and validated. While I found no evidence for crosstalk between LRH-1 and the Wnt pathway in the CRC cell lines examined, pathway analyses coupled to gene expression profiling suggested a role for LRH-1 in p53 signalling. LRH-1 silencing was shown to be associated with growth inhibition and up-regulation of the cell cycle inhibitor p21 in CRC cells. This occurs in a p53-dependent manner and was not observed in cell lines where p53 is mutated or deleted. I further demonstrated LRH-1-mediated modulation of p53 activity at the p21 promoter. Collectively, this work demonstrates a novel role for LRH-1 in the suppression of p53 signalling in CRC tumours that retain wild-type p53, and identifies LRH-1 as an important prospective target for treatment of these tumours. Open Access

Mitochondria are important to the pharmaceutical industry. Mitochondrial perturbation is central to the aetiology of many poorly treated human diseases (e.g. neurodegenerative diseases, diabetes and cancer) and hence mitochondria are increasingly becoming therapeutic targets. However, drug-induced mitochondrial toxicity (DIMT) has also gained recognition as a significant cause of safety related attrition (including multiple post-market withdrawals). The present work ultimately aimed to discern the mechanisms downstream of DIMT in vitro and in vivo with two tool compounds which act via inhibition of the mitochondrial electron transport chain (ETC) (GSK932121A) or via uncoupling of mitochondrial oxidative phosphorylation (GSK2617971A). The first objective was to compare mitochondrial functional responses between in vitro (treatment of mitochondria isolated from control animals or in vitro cell lines) and in vivo scenarios using oxygen consumption rate (OCR) and calcium retention capacity (CRC) assays. The direction of change in these assays was often different between the in vitro and in vivo situation. Whilst seemingly contradictory, these differences reflect an adapted state in vivo which results from the initial insult in combination with compensatory changes made by the tissue to maintain energy production. In contrast, only the initial, unconfounded, response to treatment is observable in vitro. Integration of both is key to the future risk assessment of mitochondrially-active drug candidates. The second objective was to use a systems approach to build candidate adverse outcome pathways (AOP)/modes of action (MOA) for both tool compounds. AOP/MOAs were constructed based on key events observed in vivo. With both compounds, effect on body temperature, increased lactate/pyruvate ratio, gene expression changes indicative of a fasted state (induction of gluconeogenic genes) and mitochondrial ultrastructure effects were key events observed in the adverse outcome. Future work using additional compounds with mitochondrial activities would help validate the proposed AOP/MOAs. The third objective was to highlight markers of underlying mitochondrial perturbation with utility for future pre-clinical monitoring in the absence of overt DIMT. A number of mitochondrial functional changes (e.g. effect on CRC) were observed prior to the onset of overt toxicity in vivo, with both compounds and these endpoints have potential use in tailored future studies with mitochondrially-active drug candidates. These findings add to the knowledgebase on DIMT, particularly with regard to the candidate AOP/MOAs which have been postulated. This will facilitate the selection of more sensitive and functional endpoints for use in future toxicology studies with drugs where mitochondrial impairment is suspected. Open Access

It is well known that existing and prior inflammatory events in the lung can adversely affect subsequent inflammatory events even though the inciting antigen might be different. For example, viral infections can enhance secondary bacterial infections and also the symptoms of asthma causing increased morbidity and mortality. Understanding changes in the lung micro-environment that lead to altered lung responsiveness is important given the increasing incidence in allergic diseases and the emergence of pandemic influenza strains. Elucidating the molecular mechanisms of these prolonged alterations could lead to new therapeutic strategies. Susceptibility to bacterial super-infections has been attributed to reduced toll-like receptor (TLR) responses in the inflammation-experienced lung. The triggering receptor expressed on myeloid cells (TREM) family of innate immune receptors fine-tune TLR responses, but have yet to be studied in the context of influenza and secondary bacterial infection. Targeting specific immune signalling receptors or signalling kinases involved in the inflammatory process in the lung by use of inhaled drugs could provide specific treatment of inflammatory lung diseases minimising effects in other organs or tissues. The aim of this thesis was to investigate the role of TREMs during influenza and secondary bacterial infection and to determine the effects of inhaled signalling kinase inhibitors in models of viral exacerbations of allergic inflammation.TREM-2 expression was shown to be upregulated during the peak of influenza-induced inflammation. Furthermore, soluble (s)TREM-1 in bronchoalveolar lavage (BAL), lung and serum coincide with heightened susceptibility to bacterial infection post-influenza virus. These novel findings suggest that targeted manipulation of TREM-1 (or prevention of its cleavage into the soluble form) and TREM-2 has the potential to restore the inflammatory tone of the airspaces to prevent subsequent infectious complications. Narrow-spectrum signalling kinase inhibitor TC2 reduced BAL eosinophils and total and allergen-specific serum IgE production in house dust mite (HDM)-induced allergic inflammation. Furthermore, a new model of viral infection during HDM-induced allergic inflammation was developed. Surprisingly, exposure to HDM prior to respiratory syncytial virus (RSV) infection led to reduced RSV-induced weight loss, damage and proinflammatory mediator release, which correlated with increased TREM-2 expression in the early stages of infection. Treatment with the kinase inhibitor TC2 during viral infection in allergen-exposed mice, however, still reduced airway hyperesponsiveness. Thus, whether severe asthma is characterised by additive inflammatory conditions or exacerbation of an underlying problem by infection, kinase inhibitors still hold promise in the clinic. Open Access

Chronic exposure to arsenic causes negative health outcomes, particularly malignant neoplasms of the skin, lung and bladder. Although epidemiological data has associated arsenic exposure to cancer, a clear molecular mechanism has remained elusive. This thesis studied the impact of arsenic trioxide (ATO) exposure on histone acetylation and microRNA expression at both tolerated and toxic levels in vitro to determine an epigenetic-based mechanism of carcinogenesis. This thesis outlines a framework for identifying tolerated and toxic ATO exposures, as a prerequisite to epigenetic characterisation. Tolerated ATO exposure increased cellular survival, anchorage-independent colony formation, cell-cycle progression and proliferation in HEK293T cells. HEK293T and UROtsa cells treated with tolerated ATO exhibited global H3K9 hyperacetylation at 3 hours and global H3K9 hypoacetylation at 72 hours. This was mediated by an imbalance in the intracellular HDAC2 to PCAF mRNA expression ratio. Global H3K9 hypoacetylation occurred for both tolerated and toxic exposures, giving poor mechanistic differentiation between these separated cellular outcomes. Chromatin immunoprecipitation identified PCAF recruitment, E2F1 binding and H3K9 acetylation at the FOS proto-oncogenic promoter leading to an elevation in FOS mRNA levels at the tolerated concentration only. This thesis also reports ATO-induced chromatin relaxation in HEK293T cells followed by a return to nominal levels for the tolerated concentration. This is in contrast to the toxic exposure, which leads to clear chromatin condensation and apoptosis. This thesis postulates that arsenic-induced global H3K9 hypoacetylation is caused by a miR-372 -mediated attenuation mechanism targeting PCAF mRNA, as predicted through bioinformatic analysis. In summary, tolerated ATO exposure resulted in measurable perturbations in both global and promoter-specific histone acetylation in addition to the aberrant expression of microRNAs, which led to cellular transformation over toxicity. Open Access

Colorectal cancer (CRC) is the third most common cancer worldwide with over a million new cases each year. Eighty percent of CRC cases have no genetic predisposition and result from a series of epigenetic and genetic alterations, referred to as ‘sporadic’ CRC. Diets rich in red or processed meats (containing dietary carcinogens) as well as bowel inflammation are a major risk factors for disease, however the underlying mechanisms are not yet well understood. Upregulation of pro-inflammatory cytokines and cytochrome P450 (CYP450) enzymes that activate dietary carcinogens as well as microRNA (miRNA) dysregulation have been observed in CRC tissue. Interleukin-6 (IL6) is a multifunctional cytokine thought to play numerous roles in tumour progression including activation of tumour-promoting signalling pathways, epigenetic gene regulation and alteration of DNA repair mechanisms; thus presence of IL6 may be key to promoting CRC tumour progression. I hypothesised that elevated levels of IL6 in the tumour microenvironment (TME) could alter miRNA expression and induce local activation of dietary carcinogens, thereby stimulating epigenetic and genetic changes that promote CRC. The current project investigated this hypothesis using a mechanistic approach with in vitro cell culture techniques. Phenotypic changes in response to IL6 treatment were observed and the underlying causes were determined by performing genetic and epigenetic studies. The findings from this project identified three potential mechanisms of IL6-mediated CRC promotion, which involve miRNAs, STAT3 signalling and DNA methylation. First, IL6 promoted dietary carcinogen-mediated DNA damage by inducing CYP450 expression. Second, IL6 stimulated CRC cell proliferation, migration and invasion through gene and miRNA expression changes. Third, IL6- mediated CRC-immune cell crosstalk in the TME resulting in the maintenance of IL6 secretion. Taken together, these data suggest that IL6 plays multiple roles in promoting CRC. Understanding these molecular events could lead to better prevention and therapeutic strategies. Open Access

Severe asthma and chronic obstructive pulmonary disease (COPD) are chronic inflammatory diseases of the lung both characterised by airway limitations but with differing pathophysiology. Asthma is normally well controlled with the use of corticosteroids (CS). However in 5 – 10% of all patients with asthma - severe asthmatics - and in most patients with COPD, the response to CS is poor and these patients are relatively CS-insensitive. p38 mitogen-activated protein kinase (MAPK) is a kinase cascade whose heightened activity has been reported in severe asthma and COPD. In this thesis I hypothesised that increased p38 MAPK activity contributes to CS-insensitivity in severe asthma and COPD Induced p38 MAPK activity was increased in peripheral blood mononuclear cells (PBMC) from patients with severe asthma compared with non-severe asthma and in patients with COPD compared with smoking subjects. The ability of dexamethasone to suppress induced pro-inflammatory cytokine release was impaired in PBMC from patients with severe asthma compared with those with non-severe asthma, and in patients with COPD compared with smoking subjects, demonstrating CS-insensitivity in these patients. The inhibition of p38 MAPK activity with GW856553 improved the ability of dexamethasone to suppress induced cytokine release in PBMC from patients with severe asthma or COPD. To investigate the molecular mechanisms by which p38 MAPK activity may be involved in CS-insensitivity, the effect of p38 MAPK inhibition on the phosphorylation of glucocorticoid receptor (GR) at serine 211 residue was determined and was revealed to be p38 MAPK dependent in PBMC from patients with COPD. MAPK phosphatase (MKP)-1 is an anti-inflammatory mediator that is induced by CS and can regulate p38 MAPK activity through dephosphorylation of its serine/threonine residues. Baseline and induced expression of MKP-1 was reduced in PBMC and monocytes from patients with severe asthma compared with non-severe asthma. Impairment of MKP-1 induction by CS may therefore be a mechanism through which CS-insensitivity manifests itself in severe asthma, however, siRNA knockdown of MKP-1 in monocytes from normal subjects resulted in an increase of induced pro-inflammatory cytokines, but it did not reach significance. CS and long-acting β2 agonists (LABA) are used in combination to treat patients with COPD or severe asthma. The effect of fluticasone propionate (FP) or the novel CS, fluticasone furoate (FF), alone or in combination, with a novel ‘ultra-LABA’ vilanterol trifenatate, on suppression of induced cytokine release was examined and compared between severe asthmatics and non-severe asthmatics, and between patients with COPD and healthy smokers. FF, compared with FP, was superior in its ability to suppress induced cytokine release in PBMC from all groups. Moreover, suppressibility of FF in combination with vilanterol, on induced cytokine release, was significantly enhanced, compared with FP in combination with vilanterol, or vilanterol alone, in PBMC from severe asthmatics or COPD. Open Access

The increasing worldwide incidence of obesity presents a significant economic burden on healthcare resources. The recent obesity epidemic may be attributed to the increasing availability of high fat and high sugar foods in our modern environment. These foods are energy dense, and are thus processed by the mesocorticolimbic system as rewarding substrates. The increased hedonic experience encoded by the activation of the mesocorticolimbic system during consumption of these energy dense foods may result in overconsumption. This may result in non-homeostatic energy intake, precipitating obesity. Melanocortins are neuropeptides which signal anorectically to reduce food intake and increase energy expenditure. Preliminary data within our group found that intra-ventral tegmental area (VTA) melanocortin administration reduced acute food intake in ad libitum fed rats. The work described in this thesis has determined that this effect is mediated via the melanocortin 4 receptor (MC4R), and is dependent on both nutritional status and diurnal phase. Furthermore, melanocortins were found to reduce acute food seeking behaviour in mice. Intra-VTA administration of melanocortins resulted in the increased activation of neurons in both the VTA and downstream mesocorticolimbic areas, including the nucleus accumbens shell, amygdala, lateral septal nuclei, ventral pallidum (VP) and the prefrontal cortex. Finally, the restoration of MC4R signalling in specific regions of the mesocorticolimbic system of MC4R null mice was used to investigate the role of MC4R in reward and energy homeostasis. Restoring MC4R signalling in the VP decreased chronic food intake but had no effect on body weight. General activity was decreased in these mice, suggesting melanocortin signalling in the VP plays a role in activity and energy expenditure. These findings suggest that melanocortin signalling in the mesocorticolimbic system may serve to maintain energy homeostasis and thus prevent the development of obesity. Further work is required to characterise the specific neuronal circuits responsible. Open Access

Globally there are around 10 million workers who routinely weld and are exposed to a metal fume aerosol of small particles of largely metal oxides and gases. Their medium- and long-term cardiorespiratory risks are incompletely understood. This thesis describes a programme of novel research carried out in a large Middle Eastern shipyard. Four studies were completed. First, a systematic review of published cross-sectional surveys (N: 38) in welders were identified; a meta-analyses of 34 publications showed a clear excess of asthma (OR 1.65; 95% CI 1.14, 2.37) and chronic bronchitis (OR 1.92; 1.50, 2.45) that could not be explained by smoking. In contrast, there was remarkably, little summary evidence of deficits in lung function in welders. Second, a cross-sectional survey of shipyard workers (N: 529, response rate 93%) showed no evidence of any decrements in lung or cardiac function in welders but suggested that symptoms of respiratory infections in winter were more common in this group (OR 2.65; 1.22, 5.78) than in matched referent workers. Support for this novel finding came from the third study, of clinical records held by the yard’s medical centre over an eleven-year period (N: 15,954). Consultations for respiratory infections were 72% higher in welders compared to office workers; there was a clear exposure-response relationship across workers with different intensities of welding. The fourth study, of a panel of 50 shipyard workers (response rate 100%) suggested a small cross-shift decline in lung function and increase in arterial stiffness. These findings suggest, for the first time, that welders are susceptible to a wide range of respiratory infections. High quality, collaborative research in this setting is feasible; future studies could usefully examine in more detail – including microbiological - the relationship between welding fume exposure and respiratory infections. The potential health-benefits to the global community of welders are substantial. Open Access

Introduction: Non-Hodgkin Lymphoma (NHL) is rising in incidence. Treatment of this genetically heterogeneous disease has toxic side effects and significant numbers of relapsers / non-responders. BCL2, an anti-apoptotic protein, is commonly overexpressed in NHL as a result of the t(14;18) translocation. A number of BCL2 inhibitors have shown success in clinical trials but variable efficacy has meant that none have been licenced for use. Methods: Retroviral insertional mutagenesis (RIM), using Moloney Murine Leukaemia Virus (MoMuLV) in transgenic mice overexpressing BCL2, was used to identify putative target genes deregulated alongside BCL2 in lymphomagenesis. This project aimed to update MoMuLV integration site identification and sequencing, allowing quantification of integration site clonal abundance. Cohorts of mice were sacrificed at time points prior to disease onset in order to interrogate integration site kinetics. To test the oncogenic potential of candidate genes, C57BL/6 Vav-BCL2 p53+/- mouse B cells were retrovirally transduced with genes of interest and transplanted into mice to study the speed of lymphoma onset. Results & Conclusions: A novel, high throughput, quantitative library preparation and sequencing protocol compatible with an Illumina platform was validated. RIM screening in BCL2 transgenic and wild-type mice identified different insertion sites profiles, detecting known oncogenes and tumour suppressor genes as well as novel candidate genes involved in pathways of lymphoid organ development, B-cell activation and differentiation. Study of insertion kinetics over time showed three patterns of clonal abundance and also allowed the study of specific gene deregulation prior to disease onset. Overexpression of Cd86 slowed disease onset whilst Ildr1 expedited disease onset suggesting the former is a tumour suppressor gene and the latter an oncogene. Discovering genes mutated with BCL2 in lymphoma may help to explain the lack of efficacy of BCL2 inhibitors and also identify novel therapeutic targets. Open Access

Gene expression is a process that is tightly regulated by many factors. Different genes are transcribed not only in a cell specific manner but are also differentially expressed at different stages of the cell cycle. Cnot3 is part of the CCR4-NOT deadenylation complex, which is involved in the turnover of mRNAs in the cytoplasm and has also been shown to have roles in regulating transcription and cell proliferation and in maintaining ES cell pluripotency. Previous work demonstrated that Cnot3 interacts directly with Aurora B kinase and is phosphorylated by Aurora B in an in vitro assay. Aurora B and Cnot3 co-localise at active gene promoters in resting B cells. Since Aurora B is a cell cycle kinase, I have developed a cell synchronization system to analyse the role of the Cnot3-Aurora B interaction at different stages of the cell cycle in primary B cells. Using this system, I have demonstrated that the interaction between Cnot3 and Aurora B varies during cell cycle progression. In vitro analysis showed that the interaction occurs through the NOT box domain of Cnot3. Mass spectrometry analysis of Cnot3 interactors, performed on nuclear extracts from B cells in the early G1 and G2 phases of the cell cycle, identified interactions with many factors that are known to have roles in transcription regulation and RNA processing. Interaction of Cnot3 with Histone H1 was confirmed using a peptide binding assay, suggesting a potential role in chromatin organization. Cnot3 was also shown to interact with Xrn2, a 5’-3’ exoribonuclease that is involved in RNA turnover and termination of transcription. ChIP analysis demonstrated promoter binding of Cnot3 at a number of cell cycle stages. Cnot3 shows cell cycle dependent binding to promoters of a wide range of active genes, including promoters that are not directly involved in cell cycle regulation. Genome wide analysis using ChIP sequencing revealed changes in the binding profiles of Cnot3 at promoters and enhancers during cell cycle progression. A Cnot3 conditional knock out mouse has been generated, which will be used to test the functional importance of these observations. Open Access

Subpial demyelination in cerebral cortical grey matter is associated with clinical progression in multiple sclerosis and is suggested to result from diffusion of pro-inflammatory cytokines from areas of meningeal inflammation into the cortex. In order to test this hypothesis we have developed an animal model of subpial demyelination driven by meningeal inflammation, involving delivery of cytokines into the subarachnoid space (SAS). Dark Agouti rats were immunised with a subclinical dose (10µg) of recombinant mouse myelin oligodendrocyte glycoprotein followed 21-24 days later by injection of TNF (1.25-5µg) and IFN-γ (75-300ng) into the SAS of the sagittal sulcus. The presence of the cytokines in the SAS resulted in acute demyelination and inflammation followed by resolution of pathology. This supports the hypothesis that cytotoxic/pro-inflammatory molecules diffuse from areas of meningeal inflammation into the underlying cortex resulting in microglial activation and subpial demyelination. Increasing the doses of TNF and IFN-γ resulted in increased extent, but not duration, of pathology due to the acute presence of the cytokines. We conclude that a chronic inflammatory milieu in the CSF/meningeal compartment is required to achieve chronic microglial activation and subpial demyelination and neuronal loss. In order to achieve the chronic presence of the cytokines in the SAS a high titre VSV-G-pseudotyped lentiviral vector carrying the enhanced green fluorescent protein (eGFP) gene under control of the CMV promoter was tested. It induced extensive and long-term, up to 12 weeks, eGFP expression in the sagittal sulcus in the absence of long-term microglial activation in naïve animals. Expression was localised to astrocytes, leptomeningeal cells and a small number of pyramidal neurons. The vector did not induce non-specific demyelination and inflammation in animals immunised with a subclinical dose of rmMOG. In order to achieve more localised expression, the vector was injected with collagen hydrogel. The hydrogel delayed eGFP expression but increased its spread along the anteroposterior axis. The distribution and duration of expression appeared optimal for achieving the chronic presence of TNF and IFN-γ in the CSF/meningeal compartment required to develop this novel model, if expression at the injection site could be increased using the hydrogel, which requires optimisation. We propose that the chronic presence of the cytokines will result in chronic meningeal inflammation and cortical grey matter pathology, allowing evaluation of the role of cytotoxic/pro-inflammatory molecules. The identity of several of the cytotoxic/pro-inflammatory molecules suggested to diffuse from areas of meningeal inflammation were also identified in post-mortem MS meninges using PCR arrays. Expression of CXCL13, IL5RA, IFNG and CXCL9 were increased, and that of CXCL1 decreased, in MS patients, consistent with an inflammatory milieu in the CSF/meningeal compartment and suggesting that these molecules may represent novel therapeutic targets for modulating meningeal inflammation and cortical pathology. Open Access

Schizophrenia is a potentially devastating mental illness with a complex aetiology, in which the odds ratios for environmental risk factors for the disorder are greater than the odds ratios of any single gene hitherto identified. Within schizophrenia, striatal dopamine dysfunction has been proposed to underlie the development of psychosis. The Aberrant Salience hypothesis provides an explanatory model based on empirical findings to explain how psychotic symptoms may arise from striatal hyperdopaminergia, whereby multiple risk factors converge to elevate striatal dopamine synthesis capacity as the Final Common Pathway to psychosis. Two important epidemiological risk factors for the disorder are chronic cannabis use and longterm psychosocial stress, both of which have evidence supporting effects on the dopamine system. Environmental risk factors are by their very nature modifiable, and so this thesis examined whether these environmental risk factors were associated with the same dopaminergic abnormalities that have been observed in schizophrenia with 3,4-dihydroxy-6- [18F]-fluoro-l-phenylalanine Positron Emission Tomography. This thesis also examined whether cannabis users exhibit aberrant salience processing using a behavioural task, the Salience Attribution Task. This thesis found that long-term cannabis use was associated with reduced dopamine synthesis capacity and no relationship was found between striatal dopamine synthesis capacity and cannabis-induced psychotic-like symptoms. Whilst cannabis use was not associated with increased aberrant salience processing, there was a relationship between cannabis-induced psychotic-like symptoms and aberrant salience processing. This thesis found that long-term psychosocial stress is associated with reduced dopamine synthesis capacity, although this finding may be due confounding factors. However, a positive relationship was observed between childhood and recent adult stressors and dopamine synthesis capacity. These findings call into question the hypothesis that cannabis increases the risk of psychosis by inducing the same changes observed in schizophrenia, although there some evidence to support the hypothesis that psychosocial stressors do increase risk via this mechanism. Open Access

Asymmetric dimethylarginine (ADMA) is an endogenous inhibitor of nitric oxide synthases. ADMA is metabolised by dimethylarginine dimethylaminohydrolase (DDAH) of which there are two isoforms with the same catalytic activity but different tissue expression patterns. Clinically ADMA is relevant as increased levels are seen in multiple pathologies, most notably cardiovascular disease and renal disease. Recent studies have shown increased plasma levels of ADMA in patients with type 2 diabetes and obesity, although these studies did not distinguish between an associative or causal relationship. In human genetic studies, polymorphisms in both DDAH1 and DDAH2 have been associated with type 2 diabetes. Furthermore, mice overexpressing ddah1 have increased insulin sensitivity and overexpression of DDAH2 has been reported to increase insulin secretion from pancreatic β-cell lines. This thesis aims to investigate the relationship between ADMA and the development of insulin resistance and in particular establish whether there is a causal link between elevation of ADMA and the onset of insulin resistance. My work has utilised DDAH deficient mouse strains to mimic loss of DDAH activity and consequent elevation of ADMA. Thorough analysis of glucose homeostasis in global Ddah1 knockout mice by hyperinsulinemic-euglycemic insulin clamp showed no significant dysfunction in insulin sensitivity in these animals. However on a high fat diet, increased ADMA resulted in enhanced adiposity and enhanced insulin resistance compared to fat fed controls. These differences were linked to increased expression of mTOR and SREBP1c in liver tissue and white adipose tissue. Conversely study of Ddah2 knockout animals showed no obvious change in insulin secretion on a normal diet but a marked difference in response to a high fat diet. Ddah2 knockout mice have higher metabolic rates and reduced fat mass compared to fat fed controls. This increased metabolic rate was associated with striking changes in metabolic markers in the brown adipose tissue of Ddah2 knockouts indicating an increase in mitochondrial number and activity. This work suggests that although ADMA does not alone cause insulin resistance it does alter the body’s response to a high fat diet that increases the pathology of the condition. This work shows, for the first time, clear and independent roles for DDAH1 and DDAH2, with respect to the control of metabolism and energy expenditure. Open Access

The research reported in this thesis investigated the neural systems involved in auditory attention, and in particular how these may differ from those recruited for visual attention. One current leading theory of attentional control postulates that a single frontoparietal network (the ‘dorsal attention network’ or DAN) subserves top-down attention to all sensory modalities. However, there is an abundance of published evidence which contradicts this claim (which is discussed herein). This thesis reports the results of three studies. In the first study, I investigated auditory attention whilst controlling for crossmodal and executive factors which may have confounded the interpretation of previous studies. In the second study, I investigated whether another crossmodal factor, the control of eye movements, may also have contributed to the controversy regarding auditory attention. Lastly, I investigated whether some regions of the brain contain multiple overlapping signals, a finding which could explain how a cortical region might display ‘amodal’ properties, and participate in multiple cognitive functions simultaneously. As a whole, this thesis provides evidence that the DAN is a predominantly visuospatial attention network whose recruitment during auditory attention reflects indirect crossmodal mechanisms rather than the direct modulation of auditory information. In addition, this thesis provides evidence that a candidate frontotemporal network, which links executive regions of the prefrontal cortex with temporal auditory association areas, subserves top-down attention to non-spatial auditory features. Open Access