The last major study of causes of vision loss in 600 eyes with uveitis was published over 10 years ago and there have been many advances in treatment over this time. In this thesis I undertook a study of 1594 patients (2593 eyes) with uveitis currently attending the clinic, 75% of whom were aged between 24 and 63 years. The type of uveitis, sight threatening complications that developed and treatment were followed from presentation to final follow up. At presentation, 16% of eyes had BCVA ≤ 6/18 (e.g. 6/18-6/36) and 14% of affected eyes had BCVA 6/60 or worse. At one year follow-up, we found 11% of eyes with vision loss to 6/18-6/36 and 8% of eyes with severe visual loss or blindness. In the group of eyes followed up for 10 years or more, 19% developed severe visual loss or blindness and 16% developed vision loss to 6/18-6/36. Chronic macular damage was the main cause of visual loss, accounting for both for visual impairment and for severe visual loss, accounting for 41% and 36% respectively. Cystoid macular oedema accounted for 29% in visual impairment and 19% in severe visual loss or blindness. When classified by uveitis types, CMO was the main cause of vision loss in intermediate uveitis (38%), glaucoma was the leading cause in anterior uveitis (32%), and chronic macular damage accounted for 46% in posterior/panuveitis. Additionally, I looked at the outcome and subsequent impact on vision of ocular surgery for cataract, glaucoma and vitreo-retinal procedures. Visual prognosis after cataract surgery was favourable in anterior and intermediate uveitis. Eyes which underwent glaucoma surgery had vision stabilised or slightly improved over time. The mean log MAR BCVA prior to glaucoma surgery was 0.53+/- 60, and 0.31+/- 49 at final follow-up visit. (P= 0.012). There was no statistically significant improvement in visual acuity in eyes which had undergone vitreo-retinal procedures. The mean logMAR BCVA were 1.1+/-0.82 and 0.87+/-0.80 respectively pre-operative and at last post- op visit. (P=0.28) The 3rd main results chapter looks at patients presenting with retinal vasculitis who had ischemia and the long term outcome for these eyes. Of the 106 eyes which developed ischemia, 24% had vision loss to 6/18-6/36 at presentation, 23% of these had BCVA 6/60 or worse. Chronic macular damage was the main cause of visual impairment and accounted for 36%, macular ischemia accounted for 67% of severe visual loss or blindness. I found that in most eyes with ischemia, visual loss developed early in the first 5 years and do not worsen with time.

Retinal degeneration is the leading cause of untreatable blindness in the developed world. Cell transplantation provides a novel therapeutic strategy to repair and restore the degenerate retina. Photoreceptor degeneration is possibly one of the most feasible disorders that could, potentially, be treated by cell transplantation, as the remainder of the retinal circuitry remains intact. Therefore, transplanted cells need only make a single efferent connection to the host’s sensory neurons. SO far, it has been shown that transplanted post mitotic photoreceptor precursors are able to functionally integrate into the adult mouse retina. Greater numbers of these functionally integrated cells would be required to restore visual responses in models of photoreceptor degeneration. The studies presented here aim to further our understanding of photoreceptor cell integration into the adult retina, in order to enhance the number of integrating cells and enable the improvement of cell transplantation into degenerate models. Initial investigation of integrated photoreceptor cell survival will examine the potential longevity of cell mediated retinal repair, and explore possible methods to improve upon this. Further characterisation of the host environment will investigate the importance of structural barriers, naturally present in the retina, and determine whether disruption of these barriers could lead to enhanced photoreceptor cell integration. Finally, the importance of extrinsic factors within the host environment will be examined, in relation to enhancing precursor cell survival and integration in the adult retina. These studies demonstrate that the immunological, structural and extrinsic characteristics of the host retina can effect photoreceptor precursor cell integration. By identifying and manipulating these factors it has been possible to significantly increase the number of integrating cells in the adult retina.

Aims: To describe the epidemiology, biometric characteristics and risk factors for primary angle closure disease (PAC & PACG) in European people. Methods: 1.Systematic reviews of PACG prevalence and acute angle closure (AAC) incidence studies; with prevalence modelling, and incidence trend analysis. 2.Analyses of EAGLE study data to investigate for differences in presenting characteristics by diagnosis , ethnicity and PACG severity. 3.Quantification of risk factors associated with PAC/ PACG by case-control analysis using EAGLE, EPIC-Norfolk and Liwan Eye Study data. 4.Genotype-phenotype correlation study of SNPs recently associated with PACG, and ocular biometry in participants of the EPIC-Norfolk Eye Study. 5.Investigation into the epidemiology and characteristics of EPIC-Norfolk participants with nanophthalmos: the “angle closure phenotype.” Results: 1.PAGC prevalence was estimated as 0.4% in those ≥40 years old. AAC incidence appears to be reducing, equivalent to halving over the past decade. 2.EAGLE participants with PACG were older, had longer axial length (AL) and thinner CCT than those with PAC. Chinese ethnicity was associated with a shallower anterior chamber depth (ACD), less hyperopia and thinner CCT. PACG severity was associated with IOP, ACD and Chinese ethnicity. 3.The odds of PAC/ PACG were 33 times and 15 times higher per 1mm shallower ACD for non-Chinese andChinese people respectively. For each 1mm less AL, the odds of PAC/ PACG were 2.7and 1.8 times higher for non-Chinese and Chinese respectively . 4.PACG risk locus, rs1015213, may exert at least part of its effect through an association with ACD in European populations. 5.There is no standardized definition for nanophthalmos. Small eyes were more common than expected and associated with visual impairment. Conclusions: PACG appears to have been historically under-recognised in populations of European descent. This series of work provides a basis for future angle closure disease risk models.

The Retinal Pigmented Epithelium (RPE) has many functions, one of which is the phagocytosis of shed photoreceptor outer segments (POS). This process is vital to the maintenance of both the RPE and photoreceptors. Outer segment shedding and internalisation are under circadian regulation, such that shedding is followed by a burst of phagocytosis at the onset of light. Annexin 2 is well placed to have a role in this process. Its direct involvement in actin dynamics and association to vesicle membranes during endocytosis may be significant in RPE outer segment phagocytosis, as this process requires extensive re-organisation of actin and re-distribution of membrane on the apical processes of the RPE. This thesis examines cell differentiation in two RPE cell lines and in primary porcine RPE cells, in order to evaluate the best system for conducting phagocytosis experiments. In vitro experiments provided evidence that annexin 2 localises to the phagocytic cup during POS internalisation but dissociates once internalisation is complete. Following knock down of annexin 2, phagocytosis was shown to decrease. Furthermore annexin 2 was shown to be phosphorylated during phagocytosis. We also found that c-Src is phosphorylated alongside annexin 2 and therefore may phosphorylate annexin 2, which contains a c-Src phosphorylation site. To investigate the circadian aspects of POS phagocytosis by the RPE, apical and basal phagosomes were quantified in the RPE from annexin 2 knock out and wild type eyes, harvested before and after light onset. Phagosomes from eyes harvested one hour after light onset were also mapped relative to Bruch’s membrane. The annexin 2 knock out animals lack the characteristic burst of phagocytosis one hour after light onset exhibited in wild type animals. Phagosomes were also retarded in the apical processes one hour after light onset, at the peak of phagocytosis, when they are normally internalised into the cell body for processing and degradation. Lysates from wild type eyes showed that annexin 2 is phosphorylated before light onset along with c-Src and FAK, key molecules in the RPE phagocytic machinery. Importantly, the absence of annexin 2 in knock out eyes delayed phosphorylation of c-Src and FAK. This delay in phosphorylation of two key RPE phagocytosis molecules may account for the delay in ingestion of outer segments into the cell body and the accumulation of phagosomes in the apical processes observed in the knock out animals. In conclusion, work in this thesis has demonstrated that annexin 2 is required for efficient RPE internalisation of rod outer segments both in vitro and in vivo. Annexin 2 is required for the timely phosphorylation of FAK and c-Src, which may account for the delay in POS internalisation observed in the annexin 2 knock out mice.

Multiple sclerosis, a chronic immune-mediated disease of the central nervous system, is considered an autoimmune disease. Statins are safe cholesterol-lowering drugs and have immunomodulatory effects, ranging from reducing migration across the blood brain barrier to neuroprotective functions. To establish the immunological effects of 80mg daily of simvastatin, a randomised, placebo-controlled, double-blind clinical trial of 140 secondary progressive multiple sclerosis (SPMS) patients was set up, with screening visits at months 0, 6, 12 and 24. The effects of several parameters were investigated on peripheral blood mononuclear cells (PBMC), their function, the molecules involved in antigen presentation, lymphocyte regulation, and adhesion. Initially, the need for age-matched controls was demonstrated, due to age-dependent effects on various molecules associated with regulatory T cells, T cell proliferation, and CD49d expression on monocytes. Subsequently, PBMCs from SPMS patients compared with healthy controls showed decreases in CD4+CD25+, TGFβ and IL-10 expression, and the Th2 and Th17 T cell populations, while the regulatory CD4+FoxP3+ T cell population increased. Finally, 24 months of simvastatin treatment failed to demonstrate any major changes in the immune cells of SPMS patients, although the drug appeared to have a stabilizing effect on certain molecules. When comparing the simvastatin- to placebo-treated patients PBMCs these displayed increases in MAC-1 and HLA-DR expression on monocytes and CD40L and IL-4 expression on lymphocytes. Some of these changes were transient, whereas monocyte HLA-DR and T cell IL-4 expression increased at 24 months, suggesting simvastatin treatment can lead to long-term immunomodulatory changes. Additionally, an inhibitory effect of simvastatin in vitro was demonstrated on expression of CXCR3 and migration of mononuclear (THP-1) cells. Although statins induce significant effects in vitro, such effects were fewer with simvastatin therapy in SPMS patients.

Hearing-impaired people have difficulties in developing normal language skills because of their lack of vocal feedback. Visual feedback, as a substitute for vocal feedback, has been used for many years by speech therapists in rehabilitation schemes. However clients and therapists have to cope with problems such as negative reinforcement, frustration, lack of motivation, and the high cost associated with these visual feedback approaches. This thesis analyses these problems, and describes a novel approach to designing visual feedback for the rehabilitation of hearing impaired speech. This novel approach takes into account previous research on visual display design techniques, necessary for implementing user-friendly graphic interfaces, together with the experience and comments from speech therapists using both traditional methods and computer-aided systems, invaluable for understanding what is missing or wrong in present speech rehabilitation systems. The thesis details original experiments to investigate the best way to visually represent a specific speech feature. A novel experimental method is used where, instead of proposing different visual representations for the various speech features (such as loudness, pitch, vowel quality etc.) and assessing which work best, the various visual stimuli are shown to the subject without specifying the associated speech feature. In this way an intuitive connection between visual stimuli and speech features (the sound produced by the subject) can be characterised. The goal of the experiments is to identify the best associations for visual stimuli and speech features. The visual stimuli for each association is then used in a visual feedback scheme for that speech feature. The results of these studies are merged with real-time and system cost considerations in order to design and implement a set of modules for effective hearing-impaired speech rehabilitation. Trial results with a cohort of deaf subjects are presented for a system which includes a range of visual stimuli approaches.

Diabetic macular oedema, which can cause rapid visual deterioration, may not have early warning signs at times. Assessment of diabetic retinal complications is made chiefly by clinical examination combined with optical coherence tomography (OCT) and fundus fluoroscein angiography (FFA). However, assessment usually does not occur until the late stages of diabetic retinopathy (DR), and, as retinal neurologic changes precede clinical changes, as tested in this thesis, by the time clinical assessment is performed, much of the functional visual loss has already occurred. More robust diagnostic modalities are required to detect progression of retinopathy in the early stages, before irreversible damage has already happened, and advances in the treatment of diabetic macular oedema is imperative as the current standard treatment in the form of laser photocoagulation is ineffective in improving the vision as authenticated in the following chapters. In this thesis, both treatment and diagnostic strategies of diabetic macular oedema (DMO) are investigated. Although laser photocoagulation is effective in short term in treating diabetic macular oedema, its mechanism of action is unknown; is associated with considerable collateral damage; and long term visual prognosis is meagre at a mean change in visual acuity at 5 years of -5.23. The 3-year outcome was also inferior to the clinical trial results with more people gaining vision (≥ 15 letter gain) in the diabetic retinopathy clinical research network (DRCRN) group compared to this cohort (26% versus 9%). Furthermore, three times more patients lost vision (> 15 letter loss) in the real-life setting of this cohort compared to the clinical trial results of the DRCRN group (27% versus 8%, respectively). Therefore, improved preventative and treatment modalities are essential to prevent progression in the early stages and to improve functional vision in late stages. In an attempt to look for new treatment strategies, we hypothesized that retinal oxygenation by inhibition of dark adaptation in the rod photoreceptor, could possibly inhibit progression of diabetic maculopathy. Illuminated-mask treatment of individuals with early diabetic maculopathy revealed encouraging results that point to an inexpensive and non-invasive therapy. Whilst 19 out of 34 study eyes with cysts at the beginning of the trial improved, 11 out of 30 fellow eyes with no demonstrable cysts at the onset developed cystic macular changes towards the end of 6 month trial. In the final chapters the correlation of visual functions with anatomic appearance were examined. The results of functional assessments, including visual acuity, colour contrast sensitivity, and microperimetry, had variable relation to structural changes at the macula with OCT. Therefore, an urgent need remains for the development of reliable diagnostic and preventative tools for the early assessment and treatment of visual function defects related to diabetic macular oedema.

1. Introduction 1.1 Amblyopia 1.1.1 Definition Amblyopia or „lazy eye‟ has conventionally been defined as “a unilateral or bilateral decrease of visual acuity caused by deprivation of pattern vision or abnormal binocular interaction, for which no cause can be detected by physical examination of the eye and which in some cases can be reversed by therapeutic measures” (1). Clinically, amblyopia is defined as a reduction in best-corrected visual acuity to less than 6/9 monocularly in Snellen optotype or as a two-line difference or more in best-corrected visual acuity between the eyes in LogMAR optotype. This compares with findings in normal subjects, in which the interocular difference in best-corrected visual acuity has been found to be less than two lines (0.2 LogMAR optotype) in both infants and adults (2). However, clinical definitions are debated with different studies using different inclusion criteria for the amblyopic subjects participated. 1.1.2 Prevalence Amblyopia is a significant cause of unilateral visual deficit in childhood and is still considered as one of the most common causes of persistent unilateral visual impairment in adulthood, including populations in which advanced medical care is offered. The prevalence of amblyopia detected in children is estimated between 0.2-5.4% (3-30, 30-35) and in adults between 0.35-3.6% (36-41). It is also classified among the major causes of unilateral visual loss in visually impaired children (13, 42-47) and adults (48-60), in parallel with refractive error, retinal lesions, cataract, corneal opacities and age-related macular degeneration. However, prevalence estimates of amblyopia are affected by the criteria of visual loss used to define amblyopia, the socio-economic properties of the population, the efficacy of the applied screening programmes for amblyopia and amblyogenic risk factors and the effectiveness of the prescribed treatment regimens (1, 24, 61-72). 1.1.3 Aetiology Amblyopia is a form of cerebral visual impairment, in the absence of an organic cause (73-76). It is considered to derive from the degradation of the retinal image associated with abnormal visual experience during the developmental period of the visual system in infancy and early childhood (73-76). Children with anisometropia, strabismus or any other condition causing a reduction in the clarity of the image in one or both eyes, thereby disrupting equal binocular vision, are at risk of developing amblyopia (74-76). Amblyopia is therefore classified according to the type of pathology underlying the abnormal binocular interaction and/or form vision deprivation as (1, 77): (i) Anisometropic, in which a difference in the refractive error between the two eyes represents a risk for developing amblyopia due to creation of dissimilar images; (ii) Strabismic, in which the confusion and diplopia caused by the misalignment of the visual axes of the two eyes can lead to binocular rivalry and suppression of input from the deviating eye at the level of the visual cortex; (iii) Mixed, if anisometropic and strabismic amblyopia co-exist and, (iv) Stimulus deprivation, if there is some obstruction to vision during the sensitive period of visual development (opacities in the media e.g. cataract or severe ptosis). The results of the adult population study of Attebo et al (39) indicated that, the predominant cause of amblyopia was anisometropia in 50%, followed by strabismus in 19%, mixed in 27% and visual deprivation in 4%.

Oculocerebrorenal syndrome of Lowe (OCRL) is caused by mutations in a phosphatidylinositol 5-phosphatase, OCRL1, and is believed to lead to an elevation of its preferred substrate, PI(4,5)P2. To date, much of the work on OCRL1 has centred on its role at Golgi and endosomal membranes. However, there is also evidence of plasma membrane activity for OCRL1, where its PI(4,5)P2 substrate is known to be highly abundant. PI(4,5)P2 regulates a wide array of downstream cellular functions such as cytoskeletal dynamics, membrane trafficking and signalling. The tight regulation of PI(4,5)P2 levels and localisation, like other phosphoinositides, provides a framework upon which many of these cellular processes work. In this thesis, effects of OCRL1 loss have been tested through siRNA depletion of OCRL1, focussing where possible on multiple PI(4,5)P2-dependent mechanisms, and also focussing on cells forming polarised epithelia. Firstly, we have visualised the localisation of PI(4,5)P2 in living HeLa cells lacking OCRL1 through immunostaining for Annexin A2, which showed a marked translocation to the plasma membrane. This change in distribution of Annexin A2 suggested that OCRL1 depletion may have an effect on intracellular calcium dynamics as well as PI(4,5)P2 localisation. We also used a GFP-chimera of the well characterised PI(4,5)P2-binding pleckstrin homology domain of PLCδ1. This showed no difference in localisation upon OCRL1 depletion. As OCRL1 is highly enriched at the TGN, we fused the pleckstrin homology domain of PLCδ1 to a mutated pleckstrin homology domain of OSBP known to bind ARF1 at the TGN, to act as a coincidence detector for PI(4,5)P2 at the TGN. This construct also showed no reproducible effect of OCRL1 depletion. Secondly we tested the effect of loss of OCRL1 on cytosolic calcium levels. Using two phospholipase C (PLC) agonists, and a SERCA pump inhibitor, we found no consistent differences in calcium handling upon depletion of OCRL1. Thirdly, we have assessed the potential specialised role that OCRL1 has in polarised epithelial cells, which might relate to the clinical picture in Lowe Syndrome. We found that OCRL1 targets the tight junctions of immortalised lines and primary cells. Through co-immunoprecipitation, we found OCRL1 in complexes with the tight junction scaffold protein ZO-1. Most significantly, we found that depletion of OCRL1 in human polarised epithelial cell lines interfered with epithelial differentiation, reducing cell number and altering morphology, to produce large flat cells. We attribute this phenotype, stronger than any other so far described experimentally, to a defect in tight junction maturation.

The initial encoding of direction by mammals occurs in striate cortex by neurons with small receptive fields that are tuned to narrow bands of the spatiotemporal frequency spectrum. Individual neurons are unable to signal the global direction of 2D motion and are instead sensitive to the 1D component of motion perpendicular to a moving edge. To compute 2D velocity, it is necessary to integrate over a range of 1D velocity sensors. In this work I probe the ability of the visual system to compute 2D velocity from a range of stimulus classes, including naturally contoured scenes, natural scenes and a global-Gabor array. My research shows that the motion stream is highly sensitive to the distribution of local orientations present in a moving image, but is largely insensitive to their spatial second-order statistics. I present a computational model of two-dimensional motion processing that is able to derive precise estimates of 2D motion directly from complex natural scenes. The model produces errors when confronted with stimuli composed of anisotropic orientation configurations and is able to capture many of the biases and errors experienced by human observers. Finally, I argue that observers’ misperceptions of 2D motion does not reflect a sub-optimal 2D motion strategy, but reflects a compromise between the competing requirements of defining motions in a spatially discrete manner across space, and the ability to accurately estimate 1D motions, on which the computation of 2D velocity must rely.