The cognitive mechanisms which underlie the formation of confabulations remain a matter of debate. In the present programme of studies predictors of the presence and clinical course of confabulations, and affective biases in the content of confabulations, were examined. 24 confabulating brain injured amnesic patients, 11 non-confabulating brain injured amnesic controls and 6 healthy participants were assessed on confabulation, temporal context confusion (TCC), insight and mood measures. Confabulating participants were followed up for 9 months. In partial replication of previous findings (Schnider, 2008), TCC scores were raised in confabulating patients compared with healthy individuals. However, TCC was not good at discriminating between confabulating and other brain injured patients. Current results are consistent with the argument that TCC may be sensitive, but not specific to confabulation (Gilboa, 2010). A combination of poor insight and somewhat elated mood state predicted the presence of confabulations in the current sample sensitively and specifically. Initial elated mood score also predicted the clinical course of episodic confabulation. The present results indicated that elated mood and level of awareness into difficulties and well-being may influence ‘core’ mechanisms underlying confabulation (Gilboa, 2010). 4 True and false memories reported by patients were rated for affective content. Although many of them were evaluated as ‘neutral’, more confabulatory memories were labeled as either pleasant or unpleasant, than ‘true’ memories. Location of lesion in terms of whether focal ventro-medial frontal pathology was present or absent, had no effect on this affective bias. The affective state may contribute importantly to confabulation formation (Fotopoulou, 2010; Metcalf et al., 2010).

Biomarker discovery studies in plasma have revealed several proteins which may have clinical utility in the diagnosis and prognosis of Alzheimer’s disease. The qualification of these proteins as true biomarkers, sensitive and specific to the disease, requires rapid, high quality, quantitative assays. Selected reaction monitoring (SRM) has emerged as an accurate and precise method for targeted protein selection through signature peptide measurement in highly complex mixtures such as plasma. Quantitative SRM-based assays are an attractive alternative to immuno-based approaches where costs can be high and requisite antibodies unavailable. Furthermore, the highly multiplexed nature of SRM allows the quantitation of multiple peptides/proteins in one rapid analysis step. Herein, we combine SRM with isotopic versions of tandem mass tags (TMT) to provide an internal reference for quantitation, an approach termed TMT-SRM, for the validation of prognostic candidate biomarkers of Alzheimer’s disease. The panel of proteins included clusterin, complement C3, alpha-2- macroglobulin, gelsolin, fibrinogen gamma-chain, serum amyloid p-component and complement factor H. These proteins were previously found to be differentially expressed between plasma of Alzheimer’s disease subjects and non-demented controls. Additionally, apolipoprotein E4 was included in the panel as possession of the apolipoprotein e4 allele is the only unequivocal genetic risk factor for late-onset Alzheimer’s disease and the protein expression of this genetic association may translate as an Alzheimer’s disease biomarker. For technology validation, TMT-SRM was shown to strongly correlate with western blot measurements for one of the proteins, gelsolin, in a sample cohort. Furthermore, the performance characteristics were determined for each target analyte in the assay. For biomarker validation, a cohort of 90 subjects was selected as having Alzheimer’s disease (n = 60) or as non-demented controls (n = 30). To monitor disease progression, the Alzheimer’s disease group was sub-divided into two groups (n = 30 per group), rapid cognitive decliners and slow cognitive decliners, based on decline in Mini Mental State Examiniation score per year. Plasma alpha-2-macroglobulin expression was found to significantly increase in Alzheimer’s disease and the difference correlated with disease progression. No significant differences were observed for the remaining proteins in the panel. Western blot analysis was performed on the same cohort for alpha-2-macroglobulin and gelsolin. In accordance with the mass spectrometry results, gelsolin showed no significant differences between disease and controls groups, providing confidence that the mass spectrometry-based assay was reflecting the real expression of the protein in the cohort. However, in contrast to the mass spectrometry results, alpha-2- macroglobulin did not show a significant difference between disease and control groups, illustrating the fact that the TMT-SRM approach was able to detect more subtle changes in peptide/protein expression than western blot. Results were replicated on an analogous mass spectrometer with strong correlation, demonstrating the robustness and ease-of-transfer of the TMT-SRM assay. Here, we have demonstrated TMT-SRM to be a rapid, precise and selective strategy for the quantitation of candidate biomarkers of Alzheimer’s disease in plasma.

Neurofibrillary tangles consist of hyperphosphorylated tau, one of the pathological hallmarks of Alzheimer’s disease (AD). Alternative splicing of the tau pre-mRNA produces six tau isoforms with or without exon 10 (E10); tau4R or tau3R, microtubule binding repeats respectively. In normal human brains the ratio of 4R/3R tau is approximately one. Aberrant E10 splicing is also observed in some sporadic AD cases as well as Pick’s disease. The rationale for this study is that abnormal splicing may predispose to neurological diseases such as AD either through abnormal expression and/or activity of splicing factors that control tau alternative splicing. In particular, expression levels of splicing factors regulating tau pre-mRNA splicing may be altered in AD and contribute to pathogenesis. The AD model used was dependent on the presence or absence of TDP43 inclusions because approximately 30 % of AD cases have inclusions of the RNA binding protein, TDP-43 which may contribute to the clinical phenotypes observed in these cases. However TDP-43 is an RNA binding protein known to regulate alternative splicing. In addition, TDP-43 pathology has now been shown to co-occur with tau pathology in some tauopathies. Whether the TDP-43 inclusions in AD is incidental or whether it contributes to more severe clinical phenotype remains unresolved. This model was used to determine if TDP43 pathology in AD exacerbates the changes in splicing factor expression in AD cases with TDP43 pathology. Splicing factors including serine and arginine (SR) rich proteins that either repress or promote E10 inclusion and CELF proteins are of particular interest. The brain selective CELF3 promotes tau E10 inclusion in vitro by binding to an intronic element in tau pre-mRNA. The levels of expression of CELF proteins and SR proteins that modulate tau splicing were characterised in five brain regions (frontal cortex, temporal cortex, amygdala, hippocampus and cerebellum) using post-mortem brain tissue from AD patients. Tau E10 alternative splicing pattern was also analysed in these brain regions. We investigated the expression levels of CELF proteins (CELF1, CELF3 and CELF4) and SR proteins (SC35, SRp40, and SRp55) in AD cases with (+) and without (-) TDP-43 inclusions compared to aged-matched controls. Tau E10 alternative splicing pattern analysis revealed that tau4R was increased in the amygdala and hippocampus of AD brain. We found, by quantitative RT-PCR, that the expression level of CELF3 RNA is increased in the amygdala and frontal cortex of AD cases, regardless of TDP-43 inclusions. Interestingly, in the amygdala of ADTDP43- cases where an increase in tau4R is found, we also found an increased expression of CELF4, SRp40 and SRp55 RNA. Although CELF3, CELF4 and SRp40 promote tau E10 inclusion in vitro, SRp55 inhibits tau E10 inclusion in vitro; whether this occurs in vivo is unknown. The association of abnormal expression of SR proteins and CELF proteins as well as aberrant tau E10 splicing in brain regions affected during AD pathogenesis may be a contributing factor to disease. Independent of the role of SR proteins and CELF proteins in sporadic AD, these tau splicing factors may be therapeutic targets to correct aberrant tau splicing in tauopathies. Future work will determine whether these splicing factors (CELF3, CELF4, SRp40 and SRp55) show abnormal expression in other neurodegenerative diseases.

Stroke is a currently the major cause of locomotor disability and the 3rd most common cause of death worldwide. It occurs when a region of the brain is damaged due to compromised cerebral blood flow either from a clot or rupture of a blood vessel and as a result, the affected area of the brain is unable to perform its normal functions. Stroke often leads to arm disability and neglect of the opposite side of the body, for which there is an urgent need for effective, efficient treatment. This thesis utilises an endothelin-1 model of focal ischemic stroke to evaluate functional recovery in adult and aged rats following treatment with neurotrophin-3 (NTS). Endothelin-1 produces a reduction in local blood flow to levels that produce ischemic injury when applied to areas of the brain. Due to the reproducible nature and location of this injury, we can qualitatively and quantitatively monitor any increase in forelimb sensory or motor function. NT3 is essential for the development and function of locomotor networks and clinical trials for other conditions have shown that systemic, high doses of NT3 are well tolerated. Here we show in three independent, randomised, blinded studies that NT3 reverses sensory neglect and arm disability in rats, when treatment is initiated 24 hours after stroke (a clinically-feasible time-frame). This thesis demonstrates that NT3 induces spinal and supraspinal neuroplasticity: brain imaging during forearm stimulation shows reorganisation upon treatment. Recovery was associated with neuroanatomical plasticity in the spinal cord of spared corticospinal and serotonergic pathways, which express receptors for NTS. Nerve recording during CST stimulation of both hemispheres showed that NTS enhances motor output. We find no evidence that NTS causes pain by lack of sprouting of nociceptive related fibres or no change to cold allodynia responses. This thesis shows that intramuscular injection of AAV-NT3 improves function in elderly rats as well as in young rats, and that intramuscular infusion of recombinant NTS protein improves function in young adult rats, even when treatment is initiated 24 hours after stroke. Furthermore, it provides additional evidence for the beneficial use of NTS following CNS insult and provides novel evidence s for the use of this therapy following an ischemic stroke.

Long-term dopaminergic therapy in Parkinson’s disease (PD) can lead to motor complications including dyskinesia which can be treated with amantadine, an N-methyl-D-aspartate (NMDA) receptor antagonist. NMDA receptor activation is linked with nitric oxide (NO) production and changes in synaptic plasticity, suggesting a role in dyskinesia. This led to the hypothesis that NO, produced by neuronal nitric oxide synthase (nNOS), contributes to the occurrence and evolution of dyskinesia in PD. Therefore, these studies investigated the effects of nNOS inhibition, using the nNOS inhibitors ARR17477 and 7-nitroindazole (7-NI), on the dyskinesia expression and priming processes in rodent and primate models of PD following L-dopa and dopamine agonist treatment. To explore the role of nNOS inhibition on established dyskinesia, 6-OHDA-lesioned rats, primed to exhibit stable abnormal involuntary movements (AIMs), the rodent analogue of dyskinesia, were acutely challenged with nNOS inhibitors plus L-dopa or ropinirole. No reduction in AIMs was observed following nNOS inhibition. In order to investigate the potential for nNOS inhibitors to reduce the priming for AIMs, naïve 6-OHDA-lesioned rats were treated chronically with ARR17477 or 7-NI plus either L-dopa or ropinirole. Again, there was no beneficial effect of nNOS inhibition on the emergence of L-dopa- or ropinirole-induced AIMs. nNOS inhibition was also investigated in MPTP-treated primates, the gold standard behavioural model of PD. ARR17477 did not reduce the expression of established dyskinesia following L-dopa or ropinirole treatment. Similarly nNOS inhibition did not attenuate L-dopa-induced priming for dyskinesia in this model. In conclusion, inhibition of nNOS in 6-OHDA-lesioned rats and MPTP-treated primates did not reduce the expression or priming of L-dopa- or ropinirole-induced dyskinesia. These findings do not support a role for nitric oxide in processes underlying dyskinesia and suggest that nNOS inhibitors would not be beneficial in either preventing or attenuating motor complications in PD patients.

Over recent years the bacterial enzyme chondroitinase ABC (ChABC) has emerged as a promising experimental therapeutic for the treatment of spinal cord injury (SCI). In pre-clinical studies ChABC has repeatedly been shown to enhance functional recovery in a number of SCI models in both rodents and larger animals through its enzymatic degradation of inhibitory chondroitin sulphate proteoglycans (CSPGs). However, ChABC treatment has met with limited success in more traumatic, translational models of SCI, such as contusive or compressive injuries. As such injury models mimic the most common form of SCI in humans, it is important to show efficacy of experimental therapeutics in these models. The key aims of this thesis therefore, are to improve upon current methods of ChABC delivery and to assess the efficacy of optimised ChABC in a clinically relevant contusion injury model. The first series of experiments involved a detailed characterisation of the temporal pattern of functional and anatomical changes that occur following a moderate thoracic contusion injury. Changes in dorsal column sensory axon conduction were associated with early demyelination in the perilesional area and subsequent remyelination mediated primarily by Schwann cells. Further electrophysiological analysis revealed a population of viable dorsal column sensory fibres that remained unable to conduct at chronic post-injury time points, in which conduction could be restored following cooling of the lesion site. This established a reproducible and clinically relevant model, with multiple outcome measures and parameters with which to assess the efficacy of optimised ChABC. A gene delivery approach was applied to optimise the administration of ChABC. Sustained and widespread CSPG degradation was achieved using a lentiviral vector containing genetically engineered ChABC (LV-ChABC). This treatment resulted in significant improvements in injury pathology and functional recovery in the moderate thoracic confusion injury model. LV-ChABC treatment resulted in neuroprotection, improved behavioural function, increased spinal conduction through the contusion injury and enhanced plasticity below the level of the injury. Additionally, ChABC gene therapy was associated with modulation of the early post-injury immune response which may have contributed to its effects on neuroprotection, whilst the effect of long-term CSPG degradation were more likely to be responsible for the observed effects on plasticity and spinal conduction. Since recovery of upper limb function is a top priority for SCI patients, further assessments of LV-ChABC were carried out in a moderate contusion injury performed at the cervical level. This resulted in similar neuroprotective effects and functional recovery. In addition, electrophysiological assessments revealed some improved corticospinal tract function below the level of the injury. Finally, LV-ChABC also resulted in neuroprotection and improved spinal conduction in a more severe model of thoracic contusion injury, illustrating the robust effects of ChABC gene therapy in clinically relevant SCI models of varying severity and at different spinal levels. Thus, ChABC gene therapy achieves sustained and widespread degradation of growth inhibitory CSPG molecules following a single administration, resulting in significantly improved injury pathology and functional repair of the spinal cord in traumatic, clinically relevant models of spinal contusion injury. If safety issues associated with gene therapy can be addressed and efficacy can be demonstrated in experimental SCI models in larger animals, then ChABC gene therapy represents a promising candidate for clinical translation.

Introduction: People with Parkinson’s Disease (PD) experience a range of cognitive and neuropsychiatric symptoms (CNPS), including depression, fatigue and anxiety, hallucinations and dementia. Although common, CNPS are less well understood than motor symptoms. CNPS are rated by patients as among the most important and challenging features of their illness (Politis et al., 2010) and are closely associated with health-related quality of life (Gallagher, Lees & Schrag, 2010). Many spouses or partners of a person with PD become informal caregivers. Despite an increasing awareness of the implications of PD for carers, there is a paucity of information about the impact of CNPS on carers and on the patient-carer relationship. This study sought to investigate the nature, extent and impact of CNPS on the wellbeing of carers and patients as individuals and on their relationship satisfaction as a couple. Method: The study was exploratory and cross sectional focussing on a single time-point in the lives of 31 couples living with PD. Data collection was quantitative and involved using clinically valid screening tools, a semi-structured interview to assess the presence and intensity of CNPS and CNPS-related distress and self-report measures to assess participants’ mood, HADS distress, health-related quality of life, and relationship satisfaction. Results: The study showed that anxiety and depression was low for the majority of patients and carers, and most participants were satisfied with their relationships. Although patients and carers had similar levels of HADS distress and relationship satisfaction, health-related quality of life was poorer for patients than carers. The majority of participants rated patient physical disability as mild or non-existent. The most frequently reported CNPS among patients was decline in memory and attention followed by anxiety and apathy. Among carers, the most frequently reported CNPS was anxiety followed by depression and decline in memory and attention. Within couples, patients and carers agreed on levels of total CNPS intensity across the 14 CNPS assessed. Patients and carers within couples disagreed, however, in their views of the presence of specific symptoms. The highest rates of discordance were found for hallucinations followed by disinhibition, with over two-thirds of couples disagreeing on the presence of these symptoms. At least half of couples disagreed about the presence of irritability, agitation and aggression, apathy and delusions. Discordance was not associated with HADS distress or relationship satisfaction for either group. Patient HADS distress was positively predicted by patients’ and carers’ ratings of patient CNPS-related distress. Carer HADS distress was positively predicted by carers’ CNPS-related distress, caregiving-related distress and health-related quality of life. Carer relationship satisfaction was best predicted by caregiving-related distress. No predictors of patient relationship satisfaction were identified. Discussion: This study’s contribution to current understanding of couples’ experiences of CNPS and its clinical implications are discussed. Findings highlight that relying on the patient or proxy only to report CNPS is clinically invalid. The study demonstrates the importance of accurate assessment of patients’ and carers’ perceptions of CNPS and exploring any differences in these. It emphasises the need to pay careful attention to patient and carer distress. Given its clinical relevance, such information about patients’ and carers’ experiences of PD-related CNPS must be actively sought and used when formulating and planning interventions.

Amytrophic lateral sclerosis (ALS) is a fatal neurodegenrative disease characterised by selective degeneration and death of motor neurons. The architecture of neurons makes them dependent upon the proper transport of protein and organelle cargoes, especially through axons (axonal transport). Indeed, disruption to axonal transport is a very early, pathological event in ALS. A proline to serine substitution at position 56 in the vesicle-associated membrane protein-associated protein B (VAPB; VAPBP56S) causes some dominantly inherited familial forms of motor neurone disease including ALS type-8. How VAPBP56S causes ALS is not properly understood. In this thesis, I investigated the effect of VAPBP56S on axonal transport of mitochondria in primary rat cortical neurons and primarybmouse motor neurons. Using time-lapse microscopy, I showed that expression of VAPBP56S but not wild-type VAPB in neurons selectively disrupts anterograde axonal transport of mitochondria. Anterograde axonal transport of mitochondria is mediated by the microtubule-based molecular motor kinesin-1. Attachment of kinesin-1 to mitochondria involves the outer mitchondrial membrane protein Rho GTPase-1 (Miro1) which acts as a sensor for cytosolic calcium levels ([Ca2]c); elevated [Ca2+]c disrupts mitochondrial transport via an effect on Miro1. To gain insight into the mechanism underlying the VAPBP56S effect on mitochondrial transport, I monitored [Ca2+]c levels in VAPBP56S expressing primary rat cortical neurons. Expression of VAPBP56S but not VAPB increased resting [Ca2+]c in these cells. Moreover, the amounts of tubulin but not kinesin-1 that were associated with Miro1 were reduced in VAPBP56S compared to VAPB transfected HEK293 cells. Also, expression of a Ca2+ insensitive mutant of Miro1 rescued defective mitochondrial axonal transport and restored the amounts of tubulin associated with the Miro1/kinesin-1 complex to normal in VAPBP56S expressing HEK293 cells. Thus VAPBP56S may perturb axonal tranport of mitochondria by disrupting Ca2+ homeostasis and affecting the interaction of Miro1/kinesin-1 with tubulin.

Prospective Memory (PM), or the ability to act upon delayed intentions, is cognitively complex as it requires a combination of mnemonic, attentional and executive abilities. PM tasks can be particularly challenging for people with cognitive impairment, and it is important to identify effective means of rehabilitation. Errorless Learning (EL) is an encoding technique that results in superior recall and recognition memory performance compared with ‘errorful’ learning in people with memory impairment. This so-called ‘Errorless Learning advantage’ (ELA) has been attributed to implicit memory processes (Page et al., 2006), and there is a basis for predicting a similar beneficial effect on PM performance. However, PM tasks vary in their retrieval demands, some involving environmentally-cued retrieval of a cue-action association (referred to as Event-based PM tasks), and some requiring self-cued retrieval of the action to be performed (referred to as Time-based PM tasks). Event-based PM performance may, therefore, be seen to rely more upon mnemonic processes, and Time-based PM performance on more executive processes. Given there is no evidence suggesting an ELA for executive tasks, differential effects of EL on Time- and Event-based PM tasks were predicted. This study investigated these predictions. Fourteen participants with neurological memory impairment completed four computer-based PM tasks in a within-subjects 2x2 factorial experiment, with each factor having two levels: encoding method (Errorless, Errorful), and PM task type (Timebased, Event-based). A significant ELA was observed for Event-based PM (d=.63), but not for Time-based PM (d=-.01), and the interaction between encoding condition and task type approached significance (d=.41). Errorless Learning also resulted in reduced accuracy in participants’ retrospective estimates of how many opportunities there had been to perform the PM tasks, suggesting that encoding manipulations can affect metacognitive awareness of PM performance. These findings extend the existing evidence for the benefits of Errorless Learning within cognitive rehabilitation, by showing for the first time that EL can benefit future action in addition to performance on purely retrospective learning and retrieval tasks. There are also clear clinical implications of these results; day-to-day Event-based PM tasks (e.g. take your medication with breakfast, check you’ve got your keys before you go out the front door), if learned with Errorless methods, are more likely to be acted upon than tasks where errors have been made during learning.

Background: Numbers of older people are increasing rapidly in most low and middle income countries and there is a pressing need for adequate information on dementia and cognitive disorders in these regions. Mild cognitive impairment is increasingly recognized as an important ‘transition’ prior to dementia onset, but is poorly understood outside Western settings, as are key constructs underlying this concept: namely, subjective memory complaints, informant-reported memory deficits and the relationship between cognition and disability. Methods: Data were analysed in relation to these questions from a series of catchment area surveys of older people carried out following identical methodologies in Cuba, Mexico, Dominican Republic, Peru, Venezuela, India and China, involving over 15,000 participating residents aged 65 years and over. Measurements had been rigorously assessed for cross-cultural applicability and were identically administered. Results: Normative data for cognitive function are described and compared, followed by the prevalence of amnestic mild cognitive impairment. Substantial variations were found between sites in the prevalence of subjective memory complaints and informant-reported memory deficits, and in their associations with dementia, and with cognitive function in participants without dementia. Variation was also found in the association between cognitive function and informant-reported disability in participants. For example, subjective memory complaints in China were relatively rare but much more strongly associated with dementia and/or cognitive function than in other sites. Conclusions: The high level of between-site variability in the associations in question suggests that mild cognitive impairment as a construct is strongly influenced by cultural factors which need to be taken into account when interpreting it or applying it in healthcare.