The olfactory bulb (OB) is the first processing structure in the olfactory system it receives direct sensory input from the olfactory sensory neurons and relays processed information to the olfactory cortex and other structures in the brain. The axons from sensory neurons expressing the same olfactory receptor molecule converge onto the same discrete structure on the surface of the OB, termed glomerulus. Each glomerulus forms a modular unit which confines the single apical tufts of about 25 mitral cells (MC). Understanding how intraglomerular cells communicate with one another will therefore improve our understanding of how the glomerular network processes the olfactory signal. Using whole cell recordings from mitral cells in acute mouse olfactory bulb slices I have examined various modes of MC-MC communication. MC dendrites are known to release glutamate from their apical and lateral dendrites which has been shown to depolarize their own presynaptic dendrites (self excitation SE). I first examined the anatomical locus of SE by investigating its properties in intact versus dendrotomized (those with amputated tufts) MCs. While dendrotomized cells lacked detectable self excitatory potentials I have found that all morphologically identified MCs with an intact apical tuft displayed robust SE. This form of SE was mediated by Ca2+ permeable AMPA receptors as it was completely blocked by their specific antagonist Naphthyl-acethyl-spermine (NAS). Electrical coupling between MCs was assessed by injecting large hyperpolarizing pulses. I have found that MCs that shared the same glomerulus were always bidirectionally coupled via gap junctions though the magnitude of coupling (CC) was variable across different pairs. Action potentials evoked in one MC also results in EPSP-like depolarizations in intraglomerular partners (lateral excitation LE). In contrast to the previously mentioned forms of communication LE was not reliably present in all intraglomerular pairs. I have found LE is mainly a unidirectional form of communication, although both bidirectional and complete lack of measurable LE can also occur. The magnitude of LE varied independently of that of other forms of communication (CC and SE amplitude) as well as with morphometric characteristics of the apical tufts (number of nodes, surface area). Despite being dependent on glutamate receptors LE contrasted with SE in that it was insensitive to NAS and therefore is a distinct form of chemical communication that relies on different AMPA receptor subtypes. I have also investigated whether in vivo-relevant patterns of theta-burst stimulation (TBS) could evoke significant long-term changes in LE efficacy similar to those observed at axo-dendritic connections. This form of communication showed a unique sensitivity to TBS in that the polarity and amplitude of change in efficacy was linearly and inversely correlated with the initial strength of LE. Small connections were enhanced while larger ones depressed, the change in efficacy was accompanied with a change in paired pulse ratio consistent with a presynaptic change in release probability after TBS. All together these data show that SE, electrical coupling and LE are independent forms of intraglomerular communication which might serve independent roles on glomerular processing. While SE and electrical coupling have been shown to facilitate spike synchrony across the network, LE provides the OB with a means of modulating intraglomerular excitation in response to in vivo -relevant patterns of activity.

Background: Ischaemic cardiovascular disease, including myocardial infarction and stroke, is the leading cause of morbidity and mortality worldwide. Atherosclerosis and coronary artery disease, which underpin ischaemic cardiovascular disorders, are characterised by chronic inflammation of the blood vessel wall and endothelial dysfunction. C-type natriuretic peptide (CNP) has recently been identified as an endothelium-derived hyperpolarising factor with anti-atherogenic properties. The studies described herein investigated the hypothesis that the vasoprotective profile of CNP includes opposing effects on endothelial and vascular smooth muscle cell proliferation and regulation of blood pressure. Methods: Cellular incorporation of bromodeoxyuridine was used to determine cell proliferation and immunoblotting was employed to assess expression/activity of intracellular signalling proteins in human umbilical vein endothelial cells (HUVEC) and rat aortic smooth muscle cells (RAoSMC). An endothelium specific CNP knockout (ecCNP KO) mouse model was developed and organ bath pharmacology utilised to assess vascular reactivity in vitro, and radiotelemetric monitoring used to determine blood pressure in vivo. Principal findings: CNP augmented HUVEC proliferation in a natriuretic peptide receptor (NPR)-Cdependent fashion by up-regulating the cell cycle promoter, cyclin D1. In contrast, CNP increased expression of the cell cycle inhibitors p21waf1/cip1/p27kip1 in RAoSMC and reduced cell growth; the pro- and anti-mitogenic effects of CNP were mediated in an extracellular signal-regulated kinase (ERK) 1/2-dependent manner. Vascular reactivity and endothelial function were disrupted in isolated aortae from female ecCNP KO mice compared to WT, whilst in males was unchanged. Female ecCNP KO mice were hypertensive. Conclusions: The anti-atherogenic properties of CNP are mediated in part by NPR-C and ERK 1/2 signalling, resulting in a differential regulation of cell cycle proteins that promotes endothelial cell proliferation and inhibits smooth muscle cell growth. Moreover, endothelium-derived CNP is key to blood pressure regulation in females. These data suggest that targeting CNP/NPR-C signalling may represent a novel approach for the treatment of cardiovascular disease.

Neurons and glia (astrocytes and oligodendrocytes) are the two major cell types that make up the central nervous system (CNS). They are generated from precursor domains within the neuroepithelial germinal zone (ventricular zone, VZ) that surrounds the ventricles of the brain and the central canal of the spinal cord (the embryonic neural tube). In general, neurons are generated before glia. The intra-spinal circuits that control movement and locomotion are made up of different neuronal and glial elements that develop separately but come together to form interconnected functional units. To understand the logic of circuit development and ultimately circuit-driven behaviour, it is necessary to understand where and when each type of cell originates. To identify the products of the most ventral progenitor domain in the developing spinal cord, known as (Nkx2.2-expressing p3 domain), I made use of Cre-loxP technology. I generated a transgenic mouse line that expresses an inducible form of Cre recombinase (CreERT2) under Nkx2.2 transcriptional control and crossed this with a Cre-dependant reporter mouse to visualize p3-derived progeny. I confirmed that the p3 domain generates Sim1-expressing V3 interneurons, serotonergic interneurons as well as visceral motor neurons of the hindbrain. p3 progenitors also produce two spatially restricted subtypes of astrocytes, a few oligodendrocytes and ventrallypositioned ependymal cells. Unexpectedly, my studies also revealed that pre-ganglionic motor neurons of the sympathetic nervous system (SPNs, visceral motor neurons of the thoracic spinal cord), as well as a population of dorsally-located Sim1-expressing interneurons, are produced from Nkx2.2-expressing precursors. SPNs have been generally believed to originate from the same progenitor pool as HB9-positive somatic motor neurons (sMNs), defined by expression of Olig2 (pMN domain, immediately dorsal to p3). Supporting this idea, no spinal sMNs or SPNs are formed in Olig2-null mice. However, I found that Nkx2.2-expressing p3 precursors do not generate any HB9-positive sMNs, implying that sMNs and SPNs derive from distinct precursors - the latter from the most ventral part of the pMN domain that transiently co-expresses Nkx2.2 and Olig2. Thus, segregation of SPNs and sMNs occurs already in the neuroepithelium before their post-mitotic progenitors migrate away from the VZ into the ventral horns. This is how visceral and somatic MNs are known to develop in the brainstem, so my results provide a unifying theme to MN development at different levels of the neuraxis.

Epidermal growth factor (EGF) is a critical survival factor for mammary epithelial cells. Upon EGF withdrawal mammary epithelial cells undergo apoptosis, a process widely assumed to be a passive cellular response to a lack of growth factor induced survival signalling. Evidence presented in this thesis indicates that growth factor deprivation induced apoptosis is very much an active mechanism regulated by the ErbB receptor binding protein Mitogen inducible gene 6 (Mig6). The results show that Mig6 acts as an intracellular sensor of Epidermal Growth Factor receptor (EGFR) inactivation that in conditions of EGF deprivation induces mammary epithelial cell apoptosis by directly interacting with and activating cellular Abelson (c-Abl) tyrosine kinase. These findings implicate a role for Mig6 in the regulation of growth factor independent cell survival, a characteristic of many cancers.

Background: Comprehension of informed consent poses greater challenges to clinical trial participants in The Gambia because of low literacy and absence of standardised formats for writing the local languages. This thesis reports the development and evaluation of a locally developed informed consent tool that addresses these challenges. Objectives: 1. Develop and validate an audio digitised tool for assessment of comprehension of informed consent. 2. Develop a multimedia consent tool for Gambian research participants. 3. Evaluate acceptability and ease of use of the multimedia tool. 4. Assess the effectiveness of the multimedia tool compared to ‘standard’ consent among participants in a clinical trial. Methods: A 34-item questionnaire was developed and audio-recorded in three major Gambian languages. This was digitised and validated among clinical trial participants in Gambian urban and rural areas. The informed consent document of a malaria drug trial was developed into a multimedia tool which integrated video, animations and audio narrations in three major Gambian languages. Acceptability and ease of use of the tool were assessed using quantitative and qualitative methods. Participants in the drug trial were randomised to either receive consent information through the multimedia tool or ‘standard’ procedure. Participant comprehension was assessed using the digitised questionnaire at baseline and follow-up visits. Results: The questionnaire was deemed to be valid and reliable (Cronbach’s alpha: 0.73- 0.79). Majority of the participants (70%) reported that the multimedia tool was clear and easy to understand. Participants in the intervention arm had significantly higher comprehension scores than those in the control arm at baseline and follow-up visits. Higher comprehension scores were associated with being a male participant (p=0.03), resident in a peri-urban area (p=0.02) and having basic formal education (p=0.005). Male participants (OR = 0.29, 95% CI: 0.12-0.70, p=0.006) and living in a peri-urban area (OR= 0.33, 95% CI: 0.13-0.82, p=0.017) were independent predictors of comprehension. Survival analysis showed that participants in the intervention arm took longer time to drop to 50% of the baseline comprehension scores than those in the control arm (hazard ratio=0.22, 95% CI: 0.16-0.31). Conclusions: A customised multimedia tool was more effective in delivering consent information and sustaining participant comprehension than ‘standard’ consent procedure. Further research is needed to compare the tool with conventional consent method in other sub-Saharan Africa settings.

Vascular cell senescence has been observed in vascular healing, atheroma and advanced age. Other groups have demonstrated that senescence is a process characterized by metabolic dysfunction. We therefore investigate whether it may occur as a consequence of metabolic stress. Controversy existed as to whether nitric oxide was able to prevent senescence in vascular cells. As nitric oxide is able to inhibit mitochondrial respiration, we investigated its effects both as an oxidant and metabolic stressor. We observed that cells exposed to nitric oxide donors senesce over a concentration range of DETA-NO 0.5-0.75mM or GSNO 1mM, showing an exponential increase in senescence until succumbing to cell death as concentrations of DETA-NO increase. This was confirmed by cell morphology, senescence-associated β-galactosidase activity, total β-galactosidase activity, cell proliferation assays and expression of p16INK4a and p21WAF. We investigated whether this effect could be reproduced by cells induced to express iNOS in vitro. Using these tools, we investigated the mechanism of NO-induced senescence looking at soluble guanylate cyclase activation, increased generation of ROS, protein transnitrosation and glutathione depletion. We investigated ways in which nitric oxide-induced senescence could be circumvented by using antioxidants (NAC, selenomethionine, uric acid), activating AMP kinase with metformin and AICAR and reducing protein transnitrosation by exposing the cells to cold light. We used pharmacological means to mimic lysosomal dysfunction and to stimulate autophagy. My findings show that nitric oxide can cause senescence in vascular cells by a mechanism that involves protein transnitrosation. The effect of nitric oxide on senescence is independent of ROS generation, AMP kinase activation, soluble guanylate cyclase activation and glutathione depletion. My findings may have important implications in vascular disease, particularly in cases of short periods of high nitric oxide production, such as sepsis, but also in the case of low-grade chronic inflammation such as that seen in atherosclerosis. However, their physiological relevance needs to be confirmed. The pathway by which protein transnitrosation induces senescence has yet to be fully elucidated.

Grid cells fire action potentials at regular intervals in space, giving rise to a spectacularly regular and stable hexagonal arrangement of firing fields (Hafting et al., 2005). For this reason they have been proposed to represent a neural code for path integration (McNaughton et al., 2006). Grid cells have primarily been found in layer II of the medial entorhinal cortex (MEC) (Hafting et al., 2005). In this thesis I explore the dendritic properties of putative grid cells in MEC layer II and how they may contribute to generating the grid cell firing pattern. To assess the spatial and temporal dynamics of dendritic integration I have used patterned two-photon glutamate uncaging in vitro in combination with somatic whole cell recordings. My findings suggest that the principal neurons of MEC are highly excitable, exhibiting supralinear summation of near-simultaneous inputs and fast and slow dendritic spikes. Supralinear summation is timing-dependent and inputs are summated in a linear manner if separated by 8 ms time intervals. In order to understand the biophysical mechanisms of supralinear summation I blocked NMDA receptors and voltage-gated sodium channels (VGSCs) with D-AP5 and TTX respectively. Both supralinearity and dendritic spikes were abolished in the presence of both blockers, while TTX alone reduced supralinearity and abolished fast but not slow dendritic spikes. This suggests that fast dendritic spikes are largely mediated by VGSCs and slow dendritic spikes by NMDA receptors. Furthermore, I have assessed dendritic integration in physiologically relevant conditions by injecting current waveform to produce in vivo-like membrane potential dynamics, recorded when an animal was crossing a firing field of a MEC II principal neuron in a virtual environment (Schmidt-Hieber & Häusser, 2013). In vivo-like membrane potential dynamics increased supralinearity of the integral of EPSPs and probability of dendritic spikes. These findings have been integrated in a continuous attractor network model of grid cell firing by Christoph Schmidt-Hieber, to assess their relevance for the grid cell rate and temporal code, that revealed that supralinear dendritic integration increases grid cell rate code robustness and fast dendritic sodium spikes increase the precision of the temporal code (phase precession) of grid cells. To conclude, in this thesis I demonstrated that dendrites of principal neurons of MEC layer II integrate synaptic inputs in a highly supralinear manner, mediated by the VGSCs and NMDARs and boosted by putative dendritic spikes. Both supralinearity and proportion of dendritic spikes are increased under in vivo-like membrane potential dynamics. These findings suggest the hypothesis for the intracellular mechanisms that mediate the robustness of grid cell firing.

This thesis examines the role played by small RNAs, in particular micro RNAs (miRNAs), in peripheral sensory neurons and their effects on regulating the neural transcriptome. The principal approach to investigating small RNAs was to ablate the RNAse III family endoribonuclease dicer. Dicer is a critical enzyme responsible for the maturation and functional activity of many species of small RNAs, such as miRNAs, and short interfering RNAs (Kim, Han & Siomi, 2009). In an earlier study using the Cre-lox P system in mice, dicer was conditionally ablated from the small diameter Nav1.8 expressing sensory neurons. This generated mice with otherwise normal developmental characteristics but a phenotype of reduced sensitivity to cold stimuli. Furthermore, following inflammatory insult, these animals displayed significant analgesia in comparison to their Cre negative littermate controls (Zhao et al., 2010). In this thesis, the role of dicer in all sensory neurons was investigated. Initially this was carried out using a Cre under the control of the Advil gene (Minett et al. 2012). This Cre is expressed in all sensory neurons and ablation of dicer using this Cre proved fatal. Dicer is therefore essential in early neural development and survival since dicer loss at the embryonic stage is fatal. In order to circumvent the fatal developmental loss of dicer, a newly developed Cre under the control of the Advil gene but conjugated to a mutated human oestrogen receptor ERT2 was employed (Lau et al., 2011). This allowed for normal adult development of the mouse and Cre mediated gene ablation only following the administration of the ERT2 agonist tamoxifen. Pan sensory neuron loss of dicer in adult mice had no detrimental effect on overall survival, or survival of the dorsal root ganglion (DRG) neurons, and was not associated with abnormal responses to acute thermal and mechanical stimuli. Following inflammatory and neuropathic insults, no behavioural differences were detected between dicer deficient or wildtype mice. In a study of the DRG transcriptome following pan sensory neuron dicer ablation using microarray and quantitive real-time reverse-transcription PCR (qRT-PCR), many genes associated with sensory neurons were down-regulated. The loss of neuron specific transcripts associated with the ablation of dicer is paradoxical since dicer products are canonically considered to negatively regulate gene expression. Further mining of the microarray data showed that a suppression factor responsible for silencing neuronal genes in non-neuronal cells termed the neuron restrictive silencing factor (NRSF) was significantly up-regulated in the dicer ablated DRG sensory neurons. Associated with this was a significant increase in its essential co-factors, small C-terminal domain phosphatase (Scp1). It was proposed that Dicer ablation in sensory neurons lead to an increase in NRSF and Scp1 production, which ultimately silences the expression of neuronal genes. In order to test this, neuronal genes that are understood to be regulated by NRSF such as Nav1.8 and Edg7 were investigated following miRNA transfection into dicer ablated DRG neurons that specifically target NRSF and Scp1. Furthermore, NRSF sequestration in dicerablated neurons suggested an enhancement of Nav1.8 expression. However, these experiments met with some technical challenges. This is the first study investigating the loss of dicer in the adult sensory neuron. The behavioural differences between the Nav1.8-Cre conditional embryonic dicer knockout and the Advillin-CreERT2 conditional adult dicer knockout suggest that dicer plays a critical role in the developing neuron but not the adult neuron. This therefore suggests that pharmacological modulation of dicer activity has little therapeutic value as a means to treat pain.

Penile cancer is a rare urogenital malignancy associated with an unusual geographical variability. Unifying hypotheses to explain this diversity include variations in human papilloma virus (HPV) infection and distinct socioeconomic influences. However, our understanding of the molecular drivers for penile carcinogenesis remains deficient. Furthermore, current prognostic and predictive tools for the clinical management of men with penile cancer have limited discriminatory roles. Radical surgery remains the cornerstone of treatment and effective adjuvant treatments are lacking. Despite improvements in surgical techniques and diagnostic methods, clinical outcomes appear to have plateaued in recent times. The DNA replication licensing factors are important cell cycle regulatory proteins which are showing promise as novel biomarkers and therapeutic targets in a broad range of tumour types. It may be possible to exploit the unique properties of the replication licensing system; as a relay station for upstream, signalling pathways, in order to develop utilitarian biomarkers and therapeutic targets for managing men with PeScc. In this thesis, I show that there is aberrant regulation of the DNA replication licensing system in PeScc. Importantly, I demonstrate that the dysregulation of these molecules is associated with aggressive, genomically unstable tumour phenotypes, which in turn translates into important prognostic information with regards to patient survival. Furthermore, I have shown that multiparameter expression analysis of these proteins allows assessment of cell cycle kinetics in individual pathological specimens, parameters linked to the biological behaviour of these tumours. This novel form of cell cycle biomarker analysis may also be of value as a predictive test for assessing the therapeutic effect of cell cycle phase specific anti-cancer agents or mechanistic drugs targeting the cell cycle machinery. Collectively, these studies highlight the prognostic, predictive and therapeutic utility of the DNA replication licensing system in PeScc.

The occurrence of survival, or time-to-event, data is commonplace in medical research, where interest lies in the time it takes from a given baseline, for an event of interest to occur, and the factors that are associated with it. For example, this could be the effect of a treatment on the time to death since diagnosis of cardiovascular disease. The primary aim of this thesis is to develop parametric methods for the analysis of complex survival data, including the extension to joint models of longitudinal and survival data, to provide a number of advantages over the commonly used semi-parametric Cox model. New and current methodology is often assessed using simulation studies; however, often in the field of survival analysis they are simplistic and fail to reflect biologically plausible scenarios. In this thesis a general algorithm for simulating complex survival data, from any given hazard function, is proposed and assessed. A general framework for the parametric analysis of survival data is then developed, utilising numerical quadrature, illustrated in detail using the special case of restricted cubic splines to model the baseline hazard and time-dependent effects. Extensions to the framework including cluster robust standard errors and excess mortality models are also considered. Finally, the joint longitudinal-survival modelling framework is extended to incorporate the Royston- Parmar survival model, and a mixture of two parametric distributions, both evaluated through simulation, utilising the proposed simulation algorithm, showing advantages over more simple parametric approaches. The estimation of joint models, using Gaussian quadrature, is also evaluated through an extensive simulation study. Throughout the thesis, user friendly software is developed to implement the methodological components, allowing statisticians and non-statisticians alike, to apply the methods directly. A variety of clinical datasets in the areas of cancer, cardiovascular disease and liver cirrhosis are used to exemplify the proposals.