To develop targeted brain tumour therapies that carry fewer side effects it is essential to understand brain tumour histogeneiss and identify the cell of origin. It has been suggested that neoplastic transformation of neural stem/progenitor cells in the adult brain can result in the development of intrinsic brain tumours. Using the Cre-LoxP system we have developed a model that allows us to inactivate tumour suppressor genes Rb, p53 and PTEN in neural stem/progenitor cells both in vivo and in vitro. In vivo recombination of these tumour suppressor genes in neural stem/progenitor cells resulted in the formation of intrinsic brain tumours, with the phenotype depending on the combination of recombined genes. Intracerebral engraftment of in vitro recombined neural stem/progenitor cells produced a similar tumour phenotype pattern. However, while in vivo recombination of Rb and p53 frequently caused brain tumours, in vitro recombined and engrafted Rb-/-; p53-/- neurospheres rarely generated tumours. The converse was true following in vivo recombination of Rb and PTEN. These results suggest that microenvironment influences tumour formation. An increased number of tumourigenic neural stem/progenitor cells expressed the cancer stem cell marker CD133 when compared to control cells. In vivo engraftment of purified CD133+ Rb-/-; p53-/-; PTEN-/- neural stem/progenitor cells resulted in the preferential development of gliomas. We also showed that astrocytes are not capable of forming tumours following either orthotopic recombination or in vitro recombination and engraftment. In conclusion, we found that (i) certain phenotypes of brain tumours develop from neural stem/progenitor cells in the subventricular zone, (ii) the combination of inactivated tumour suppressor genes is the most significant but not sole determinant of in vitro functional profile and behaviour in vivo, (iii) the cellular microenvironment of neural stem/progenitor cells influences their tumourigenic profile, and (iv) terminally differentiated astrocytes do not contribute to tumour development in this model.

Familial Alzheimer’s disease (FAD) is a valuable paradigm for the study of the more common sporadic AD (SAD). The two forms of the illness share many neuropathological, clinical and radiological characteristics but it is not yet possible to predict the onset of SAD or confirm its presence without histopathological analysis. Fully penetrant amyloid precursor protein (APP), presenilin 1 (PSEN1) or presenilin 2 (PSEN2) gene mutations permit both, and therefore lend themselves to clinical research with results which may be applied to the study of SAD. The identification of biomarkers of onset and progression are vital in the selection of research participants and in the rapid evaluation of new therapies. This thesis further characterizes FAD with a number of studies. These address the use of imaging biomarkers and help clarify the nature of the relationship between FAD and SAD. Two reports of novel pathogenic FAD mutations are included as well as other studies exploring the clinical and radiological (structural and molecular) phenotypes associated with FAD. Key findings are as follows: the novel PSEN1 p. L166del and S132A mutations are both associated with FAD; the APPV717G mutation can be associated with pure progressive amnesia reflected in an atypical structural imaging profile; APP locus duplication is a significant cause of early onset dementia in the UK and the recognised phenotype should be expanded to include early seizures and apparently sporadic disease; regional cortical thickness (CTh) decline accelerates after diagnosis in FAD mutation carriers (MC) and differences between MC and controls are detectable in presymptomatic mutation carriers more than 4 years prior to clinical diagnosis; APP and PSEN1 mutations may produce different temporal and topographical patterns of cortical change; increased 11C-PiB retention in a highly heterogeneous pattern may be detected in presymptomatic PSEN1 mutation carriers.

Human prion disease is associated with the accumulation in brain of an abnormal glycoprotein known as prion protein. Prions are far more resistant to physical and chemical inactivation than conventional pathogens. In 1996 a new disease, variant Creutzfeldt-Jakob disease was recognised as distinct from sporadic Creutzfeldt-Jakob Disease and initially showed a rapid rise in incidence. The potential for iatrogenic transmission of prion disease via surgical instruments, despite presumed adequate sterilisation, is well documented. The long asymptomatic incubation periods seen in human prion diseases therefore provide a significant risk of transmission. This project investigated the disinfection of prion-contaminated surfaces, with the goal of developing a process applicable to medical instruments on a large scale. It also aimed to investigate the susceptibility of different brain regions to infection and the spread of peripheral infection to brain, as well as the mechanisms of disease transmission from infected surfaces to tissues. Potential sterilisation procedures were tested on steel wires that had been incubated in infectious brain homogenate. These wires were then inserted into the brains of appropriate indicator mice, which were observed for the clinical signs of prion disease and subsequently investigated using immunohistochemistry. To investigate the susceptibility of different brain regions to infection, steel and plastic spheres were inserted stereotaxically and mice culled at different time points and subjected to immunohistochemical examination. Intraperitoneal and subcutaneous insertion of these spheres was used to investigate the travel of infectivity from peripheral tissues to the central nervous system. Attempts were made to develop a high sensitivity cell culture assay, to enable quantification of the amount of sterilisation produced by different methods, and to investigate their kinetics.

Microglia, the immune cells of the central nervous system (CNS), are important in the protection of the CNS, but may be implicated in the pathogenesis of neuroinflammatory disease. Upon activation, microglia produce reactive oxygen and nitrogen species; intracellular antioxidants are therefore likely to be important in their self-defence. Here, it was confirmed that cultured microglia contain high levels of glutathione, the predominant intracellular antioxidant in mammalian cells. The activation of microglia with lipopolysaccharide (LPS) or LPS + interferon-gamma was shown to affect their glutathione levels. GSH levels in primary microglia and those of the BV-2 cell line increased upon activation, whilst levels in N9 microglial cells decreased. Microglial glutathione synthesis is dependent upon cystine uptake via the xc- transporter, which exchanges cystine and glutamate. Glutamate is an excitatory neurotransmitter whose extracellular concentration is tightly regulated by excitatory amino acid transporters, as high levels cause toxicity to neurones and other CNS cell types through overstimulation of glutamate receptors or by causing reversal of xc- transporters. Following exposure to LPS, increased extracellular glutamate and increased levels of messenger ribonucleic acid (mRNA) for xCT, the specific subunit of xc-, were observed in BV-2 and primary microglial cells, suggesting upregulated GSH synthesis. An activation-induced decrease in N9 GSH levels suggests that this cell line is more susceptible to oxidative damage, and may be less able to upregulate GSH synthesis. Albumin, to which microglia may be exposed following blood-brain barrier damage, increased iNOS expression, glutamate release, xCT mRNA levels and intracellular levels of GSH and ATP in BV-2 and primary microglia. Primary and BV-2 microglial conditioned medium contained low levels of GSH, suggesting that microglia may release GSH. Modulation of microglial metabotropic glutamate receptors (mGluRs) may alter microglial activation and neurotoxicity. Here, stimulation of the neuroprotective mGluR5 and group III mGluRs caused a decline in GSH levels in BV-2 and N9 microglia, respectively. In contrast mGluR1 stimulation may increase BV-2 GSH levels. The work presented in this thesis therefore extends current knowledge regarding microglial GSH and its regulation, and contributes to the understanding of microglial neurotoxicity and neuroprotection.

Huntington's disease (HD) is a devastating inherited neurodegenerative disorder with an onset commonly in mid-adulthood. To date there are no disease-modifying treatments that have been shown to slow the progression or delay the onset of HD in humans; however, several compounds have shown promise in animal models of HD. To assess their efficacy in humans, robust and sensitive markers of disease progression are required. Biomarkers capable of detecting premanifest changes are critical for clinical trials of treatments to delay disease onset. Current clinical measures are limited by floor and ceiling effects and lack sensitivity to change with time especially in premanifest subjects. This thesis investigates the utility of volumetric magnetic resonance imaging (MRI) for tracking structural changes in premanifest and early-manifest HD. Several approaches are described including hypothesis-driven region-of-interest (ROI) investigations focused on the caudate nucleus, cingulate cortex and lateral ventricles, as well as hypothesis-free whole-brain analyses at the voxel-level. Where cross-sectional work has shown promise on the basis of group separation, investigations are extended to longitudinal analyses in order to track within-subject progression. This work included developing and validating a novel automated technique for quantifying change in caudate volume from registered serial MRI. A longitudinal voxel-based morphometry technique was adapted and applied to assess the regional progression of atrophy in HD over 27 months. Suitability of these measures for use as biomarkers of progression in large clinical trials is discussed in terms of sample-size requirements, automation of measurement and reliability.

A mutation in the charged multivesicular body protein 2B (CHMP2B) gene, identified in a kindred from the Jutland region of Denmark, segregates with affected family members with clinical presentations of frontotemporal dementia (FTD) and is absent in control populations (Gydesen et al., 1987; Gydesen et al., 2002; Skibinski et al., 2005). The mutation is a G>C transition in the splice acceptor site of exon 6 resulting in two novel splice variants CHMP2BInt5 and CHMP2BA10 leading to C-terminal truncation of the CHMP2B protein (Skibinski et al., 2005). Chmp2b knockout (Chmp2b-/-) mice and transgenic mice expressing either wild-type or C-terminally truncated mutant CHMP2B splice variants CHMP2BInt5 and CHMP2BA10 were generated with the aims of examining the normal function of Chmp2b and the effect of mutant CHMP2B species in vivo, as well as providing insight into a potential common FTD mechanism of disease. Quantification of Chmp2b protein in Chmp2b-/- mice demonstrates a significant (85%) depletion of endogenous Chmp2b in the mouse brain. No pathology is identified in the CNS or muscle tissue of these mice however, they do demonstrate significant motor and behavioural abnormalities. CHMP2BInt5 transgenic mice demonstrate neurodegenerative changes including progressive gliosis, accumulation of CHMP2B, p62 and ubiquitin inclusions which are negative for TDP-43 and FUS proteins, consistent with the inclusion pathology observed in patients with CHMP2B mutation. Furthermore, these mice have reduced survival and develop progressive axonopathy characterized by axonal swellings and accumulation of mitochondria and vesicles likely from the endosome¬lysosome and autophagy pathway, implicating altered axonal function in disease pathogenesis. This thesis describes the first mouse models of FTD-3 caused by CHMP2B mutation and presents evidence consistent with a gain-of-function effect unique to the CHMP2BInt5 isoform and provides new insights into the mechanisms of CHMP2B-induced neurodegeneration.

Oligodendrocytes are the principal target of immune attack in multiple sclerosis (MS), where the immune system attacks the myelin sheath, leading to oligodendrocyte demise, compromised axonal saltatory conduction and eventually neuronal degeneration. Microglia are also implicated in MS pathology and are present in lesions in an activated state. To study the effects of activated microglia on oligodendrocyte precursor cells (OPCs) or mature oligodendrocytes, primary cultures of microglia, oligodendrocytes, or co-cultures of the two cell populations were developed. Microglia activated with lipopolysaccharide (LPS), were toxic to OPCs and mature oligodendrocytes in co-culture. This toxicity was mediated by tumour necrosis factor α (TNF-α) released from microglia. Microglial activation with the blood-borne proteins (BBPs) fraction V albumin, fibrinogen and fibrin was also investigated. BBPs enter the brain after blood-brain barrier disruption, a typical event in MS, so their differential effect upon microglial activation and subsequently oligodendrocyte lineage cell maturation survival was studied. BBPs had distinct effects on microglial activation and oligodendrocyte cell toxicity. Fraction V albumin and fibrin were directly toxic to oligodendrocytes while microglial activation protected oligodendrocytes in co-culture. Fibrin also inhibited OPC maturation into myelinating oligodendrocytes. Fibrinogen activated microglia were toxic to OPCs and mature oligodendrocytes in co-culture. Attenuation of microglial induced oligodendrocyte death was attempted by blocking pathways of microglial activation such as the Rho-ROCK pathway and by modulating microglial activation by metabotropic glutamate receptor manipulation. Additionally, ROCK inhibition was able to attenuate LPS or BBP activated microglial expression of inducible nitric oxide synthase (iNOS). Modulation of microglial activation could prevent microglial induced oligodendrocyte toxicity and could lead to strategies to slow disease progression in MS patients by protecting mature oligodendrocytes from microglial induced death, or most importantly by enhancing survival and maturation of OPCs in MS lesions, where their recruitment to remyelinate neuronal axons is vital for disease remission.

Multiple sclerosis (MS) is an inflammatory demyelinating disease of the central nervous system that causes a range of neurological deficits expressed in a series of relapses separated by remissions. Surprisingly, our understanding of the mechanisms responsible for both relapses and remissions remains incomplete, even for the expression of major deficits. To explore the mechanisms, a model of MS (experimental autoimmune encephalomyelitis, EAE) has been studied neurologically, electrophysiologically and histologically at five time points along the course of the disease, i.e., 'preinduction', 'predisease, 'peak disease', 'remission', and 'relapse'. The animals were assessed daily to determine the magnitude of any neurological deficit, and electrophysiological techniques were developed to permit serial measures of axonal conduction, synaptic transmission, and the excitability of motor neurons. Also, to investigate whether inflammatory factors can modulate the expression of neurological deficits, a proinflammatory agent, lipopolysaccharide (LPS) was injected systemically in animals at the ‘remission’ phase to explore whether systemic inflammation exacerbates the disease. A selective inhibitor of inducible nitric oxide synthase (iNOS), 1400W, was also administered into animals with EAE at the onset of disease expression to examine whether inhibition of nitric oxide (NO) production ameliorates the disease. The findings have revealed that no electrophysiological deficits were detected before the onset of neurological deficits, but axonal conduction and synaptic transmission were significantly impaired at the ‘peak disease’. These measures remained reduced during the ‘remission’ and ‘relapse’, at which time motor neuronal excitability was also significantly decreased. It is intriguing that none of the electrophysiological measures have so far revealed changes that correlate closely with the expression of, and recovery from, neurological deficit. Present data also show that systemic inflammation induced by an injection of LPS in animals with EAE caused an acute relapse that peaked 3.5 hours post injection. However, no electrophysiological changes occurring at this peak were detected that could be responsible for the transient neurological deficits, and so the exact mechanisms underlying the deficits remain unknown. Also, based on the histological analysis revealing an exclusive expression of iNOS at ‘peak disease’ in animals with EAE, the consequences of selective inhibition of iNOS using 1400W were explored. This treatment significantly reduced the severity of the neurological deficits, implicating NO in their production. Importantly, electrophysiological and histological examination of 1400W-treated animals revealed that the inhibition of iNOS also provided significant protection from the loss of function, and degeneration, of axons, again implicating NO in a pathogenic role. In conclusion, these findings add further weight to the evidence that inflammation plays a key role in the pathogenesis of EAE, although the exact mechanisms underlying the production of the neurological deficits remain unknown. Potential explanations are discussed, including the possibility that the deficits may arise from spinal hypoxia. In addition, the inhibition of NO production may provide an effective protection from both the functional and pathological consequences of neuroinflammatory diseases such as MS.

Ischaemic injury of the central white matter (IICWM) has been relatively little studied despite its contribution to the pathology of several neurological diseases, including stroke and vascular dementia. Research has been hampered by a lack of suitable experimental models. Here, a new model of IICWM has been developed in the rat spinal cord in vivo, and the model has been used to explore the value of agents that block voltage-gated sodium channels, or the sodium/calcium exchanger (NCX). Endothelin-1 (ET-1, a potent vasoconstrictor) was injected into the spinal cord to induce local ischaemia, and hence injury to the spinal white matter as demonstrated by blockade of axonal conduction and change of axonal morphology. The sodium channel blocking agent phenytoin was found to improve axonal conduction during the first 4.5 hours post ET-1 injection. Importantly, the protection provided by phenytoin (fosphenytoin as prodrug) persisted for at least 3 days (the longest interval studied) in animals allowed to recover from anaesthesia. The selective reverse-mode blocker of the NCX, KB-R7943 was also found to improve axonal conduction during the first 4.5 hours post ET-1 injection. Another reverse-mode blocker of the NCX, SEA0400 was also neuroprotective. Furthermore, in the longer term, KB-R7943 protected axons at 3 days post ET-1 injection. None of the study agents (phenytoin, KB-R7943 and SEA0400) diminished the severity of ET-1-induced ischaemia, revealing the effect to be a true neuroprotection, rather than an unintended diminution of the ischaemic insult. The findings are consistent with an interpretation that intracellular sodium accumulation, probably via open sodium channels, may play a role in mediating IICWM in vivo, perhaps by promoting the lethal importation of calcium ions via the reverse-mode operation of the NCX. It is concluded that partial blockade of sodium channels, or the NCX, may be effective in protecting central axons from ischaemic injury.

This thesis presents experiments investigating the cognitive neuropsychological mechanisms underlying anosognosia for motor deficit, i. e. lack of awareness or underestimation of a deficit in motor functioning following brain injury. Two contemporary explanations of anosognosia were considered. First, an account based on the Somatic Marker Hypothesis (SMH; Damasio, 1994), which suggests that anosognosia might be explained by emotional disturbance leading to irrational reports regarding motor deficits, via an impairment in somatic feedback to the brain. The second account, proposed by Frith, Blakemore and Wolpert (2000a), provides a cognitive neuropsychological explanation of anosognosia. This -model proposes that motor awareness depends on predicting the expected sensory consequences of intended movements, and comparing these predictions with the actual sensory consequences of movement. Three research predictions regarding anosognosia arose from Frith et al. (2000a): (i) preservation of internal representations of movement involving the impaired limb, (ii) failure to register discrepancies between intended (predicted) and actual movement, and (iii) absent contrary information regarding the actual state of the motor system. Experiments investigating these contemporary accounts of anosognosia are presented. An initial, psychophysiological pilot study concerning the SMH failed to replicate effects alleged to occur in healthy individuals; therefore, subsequent inspection of the SMH in a patient cohort was not pursued. Subsequent cognitive neuropsychological investigations focussed on the three predictions arising from Frith et al. (2000a) in patients with anosognosia for motor deficit, control patients without anosognosia, and age-matched healthy individuals. The main results of these studies showed (i) distorted internal representations of movement, (ii) impaired monitoring of movements, and (iii) increased levels of unilateral neglect (i.e. ignoring one half of space) in anosognosia. Findings inform current accounts of anosognosia and are discussed in the context of a normal model of action control and awareness.