The global aim of this thesis was to assess retinal vascular reactivity in glaucoma patients using a standardised hypercapnic stimulus. There is a suggestion of disturbance in the regulation of retinal and optic nerve head (ONH) hemodynamics in patients with Primary Open Angle Glaucoma (POAG), although much of the work to-date has either been equivocal or speculative. Previous studies have used non-standardised hypercapnic stimuli to assess vascular reactivity. To explain, hypercapnia induces hyperventilation which disturbs arterial oxygen concentration, an effect that varies between individuals resulting in the non-standardised provocation of vascular reactivity. Therefore, a normoxic hypercapnic provocation was developed to avoid additional and potentially uncontrolled vasoconstriction in what is thought to be a vasospastic disease. The development of a safe, sustained and stable normoxic hypercapnic stimulus was essential for the assessment of retinal arteriolar vascular reactivity so that repeated hemodynamic measurements could be obtained. Furthermore, most techniques used to measure vascular reactivity do not comprehensively assess retinal hemodynamics, in terms of the simultaneous measurement of vessel diameter and blood velocity in order to calculate flow. In this respect, this study utilized a technique that quantitatively assesses retinal blood flow and vascular reactivity of the major arterioles in close proximity to the ONH. The stimulus and vascular reactivity quantification technique was validated in healthy controls and then was clinically applied in patients with POAG. Newly diagnosed patients with untreated POAG (uPOAG) were recruited in order to avoid any confounding pharmacological effects and patients with progressive POAG (pPOAG) were also selected since they are thought to likely manifest vascular dysregulation. Finally, the results of the functional vascular reactivity assessment were compared to those of systemic biochemical markers of endothelial function in patients with untreated and progressive POAG and in healthy controls. Overall summary A safe, sustained, stable and repeatable normoxic hypercapnic stimulus was developed, evaluated and validated. In terms of the physiology of retinal vascular regulation, the percent magnitude of vascular reactivity of the arterioles and capillaries was found to be comparable in terms of flow. The new stimulus was successfully applied in POAG and in healthy controls to assess vascular reactivity and was also compared to plasma levels of ET-1 and cGMP. In terms of the patho-physiology of POAG, the study revealed a clear impairment of vascular reactivity in the uPOAG and pPOAG groups. There were reduced levels of plasma ET-1 in the uPOAG and ntPOAG groups. In addition, treatment with Dorzolamide improved vascular reactivity in the ntPOAG group in the absence of any change in the expression of plasma ET-1 or cGMP. Future work will address this apparent contradiction between the outcome of the functional vascular reactivity assessment and the biochemical markers of endothelial function in newly diagnosed POAG patients treated with Dorzolamide. Aims of chapters  Chapter 3: To determine the effect of hypercapnia on retinal capillary blood flow in the macula and ONH using scanning laser Doppler flowmetry (SLDF) in young healthy subjects.  Chapter 4: To describe a new manual methodology that permits the comprehensive assessment of retinal arteriolar vascular reactivity in response to a sustained and stable hypercapnic stimulus. The secondary aim was to determine the magnitude of the vascular reactivity response of the retinal arterioles to hypercapnic provocation in young healthy subjects.  Chapter 5: To compare the magnitude of vascular reactivity of the retinal arterioles in terms of percentage change of flow to that of the retinal capillaries using a novel automated standardized methodology to provoke normoxic, or isoxic, hypercapnia.  Chapter 6: To determine the magnitude of retinal arteriolar vascular reactivity to normoxic hypercapnia in patients with untreated POAG (uPOAG), progressive POAG (pPOAG) and controls. The secondary aim was to determine retinal vascular reactivity in newly treated POAG (ntPOAG, i.e. after treatment with 2% Dorzolamide, twice daily for 2 weeks).  Chapter 7: To compare plasma endothelin-1 (ET-1) and cyclic guanosine monophosphate (cGMP) between groups of patients with untreated primary open angle glaucoma (uPOAG), progressive POAG (pPOAG), newly treated POAG (ntPOAG) and controls. The effect of normoxic hypercapnia on plasma ET-1 and cGMP was also assessed. The functional measures of retinal blood flow and vascular reactivity were correlated with systemic biochemical markers of endothelial function. Methods Chapters 3, 4 and 5 were conducted on young healthy control subjects, where as Chapters 6 and 7 were conducted on patients with glaucoma and healthy controls.  Chapter 3: Subjects breathed unrestricted air for 15 minutes (baseline) via a sequential gas delivery circuit and then the fractional (percent) end-tidal concentration of CO2 (FETCO2) was manually raised for 15 minutes by adding a low flow of CO2 to the inspired air. For the last 15 minutes, FETCO2 was returned to baseline values to establish a recovery period. Heidelberg Retina Flowmeter (HRF) images centered on both the ONH and the macula were acquired during each phase.  Chapter 4: Subjects breathed air via a sequential gas delivery circuit for 15 minutes and the air flow was then manually decreased so that subjects inspired gases from the rebreathing reservoir until a stable 10-15% increase in FETCO2 concentration was achieved for 20 minutes. Air flow rate was then manually elevated so that subjects breathed primarily from the fresh gas reservoir to return FETCO2 back to baseline for the last 15 minutes. Retinal arteriolar hemodynamics was assessed using the Canon Laser Blood Flowmeter (CLBF) during all three breathing phases.  Chapter 5: Normoxic, or isoxic, hypercapnia was induced using an automated gas flow controller (RespirActTM, Thornhill Research Inc. Toronto, Canada). Subjects breathed air with PETCO2 normalized at 38 mmHg. An increase in PETCO2 of 15% above baseline, whilst maintaining normoxia, was then implemented for 20 minutes and then PETCO2 was returned to baseline conditions for 10 minutes. Retinal and ONH hemodynamic measurements were performed using the CLBF and HRF in random order across sessions.  Chapter 6: Retinal arteriolar vascular reactivity was assessed in patients with uPOAG, pPOAG (defined by the occurrence of optic disc hemorrhage within the past 24 months) and controls during normoxic hypercapnia. Using the automated gas flow controller, patients breathed air for 10 mins and PETCO2 was maintained at 38mmHg. Following this normoxic hypercapnia (a 15% increase in PETCO2 while PETO2 was maintained at resting levels) was induced for 15 mins and then for the last 10 mins PETCO2 was returned to baseline (post-hypercapnia) to establish recovery blood flow values. Retinal arteriolar diameter, blood velocity and blood flow was assessed using the CLBF in both patient groups and controls. A similar paradigm was repeated in the newly treated POAG group (ntPOAG, i.e. after treatment with 2% Dorzolamide, twice daily for 2 weeks).  Chapter 7: Blood samples were collected from the cubital vein of all participants (uPOAG, pPOAG, ntPOAG and controls) during baseline conditions (PETCO2=38mmHg) and then during normoxic hypercapnia (i.e. a 15% increase in PETCO2 relative to the baseline) using the paradigm described for Chapter 6. ET-1 and cGMP was assessed using immunoassay. Results  Chapter 3: The group mean nasal macula capillary blood flow increased from 127.17 a.u. (SD 32.59) at baseline to 151.22 a.u. (SD 36.67) during hypercapnia (p=0.028), while foveal blood flow increased from 92.71 a.u. (SD 28.07) to 107.39 a.u. (SD 34.43) (p=0.042). There was a concomitant and uncontrolled +13% increase in the group mean PETO2 during the hypercapnic provocation of +14% increase in PETCO2.  Chapter 4: Retinal arteriolar diameter, blood velocity and blood flow increased by 3.2% (p=0.0045), 26.4% (p<0.0001) and 34.9% (p<0.0001), respectively during hypercapnia. There was a stable ¬+12% increase in PETCO2 during hypercapnia and a concomitant -6% decrease in PETO2.  Chapter 5: Using an automated gas flow controller the co-efficient of repeatability (COR) was 5% of the average PETCO2 at baseline and during normoxic hypercapnia. The COR for PETO2 was 10% and 7% of the average PETO2 at baseline and during normoxic hypercapnia, respectively. Group mean PETCO2 increased by approximately +14.4% and there was only a +4.3% increase in PETO2 during hypercapnia across both study sessions. Retinal arteriolar hemodynamics increased during hypercapnia (p<0.001). Similarly, there was an increase in the capillary blood flow of the temporal rim of the ONH (p<0.001), nasal macula (p<0.001) and foveal areas (p<0.006) during hypercapnia. A non-significant trend for capillary blood flow to increase in the macula temporal area (+8.2%) was noted. In terms of percentage change of blood flow, retinal capillary vascular reactivity (i.e. all 4 analyzed areas = 22.4%) was similar to the magnitude of arteriolar (= 24.9%) vascular reactivity.  Chapter 6: Retinal arteriolar diameter, blood velocity and flow did not increase during normoxic hypercapnia in uPOAG compared to controls. Diameter and blood velocity did not change in pPOAG during normoxic hypercapnia but there was a significant increase in blood flow (+9.1%, p=0.030). After treatment with 2% Dorzolamide for 2 weeks there was a 3% (p=0.040), 19% (p<0.001) and 26% (p<0.001) increase in diameter, velocity and flow, respectively, in the ntPOAG group. Group mean PETCO2 increased by approximately +15% in all the groups and there was only a +3% increase in PETO2 during hypercapnia.  Chapter 7: Plasma ET-1 levels were significantly different across groups at baseline (one way ANOVA; p=0.0012) and this was repeated during normoxic hypercapnia (one way ANOVA; p=0.0014). ET-1 levels were lower in uPOAG compared to pPOAG and controls at baseline and during normoxic hypercapnia (Tukey’s honestly significant difference test). Similarly, ntPOAG group also showed lower ET-1 levels compared to the pPOAG and controls at baseline and during normoxic hypercapnia (Tukey’s honestly significant difference test). The cGMP at baseline and during normoxic hypercapnia across all groups was not different. In the control group, the change in ET-1 during normoxic hypercapnia was negatively correlated with change in retinal arteriolar blood flow (r = -0.52, p=0.04), that is, as the change in ET-1 reduced, the change in blood flow increased. A weak correlation was noted between change in cGMP during normoxic hypercapnia and the change in arteriolar blood flow (r = +0.45, p=0.08). Conclusions  Chapter 3: Hypercapnia resulted in a quantifiable capillary vascular reactivity response in 2 of the 3 assessed retinal locations (i.e., nasal macula and fovea). There was no vascular reactivity response of the ONH. It is critical to minimise the concomitant change in PETO2 during hypercapnia in order to obtain robust vascular reactivity responses.  Chapter 4: A technique to comprehensively assess vascular reactivity during stable and sustained hypercapnia was described. Retinal arteriolar diameter, blood velocity and blood flow increased in response to hypercapnia. The vascular reactivity results of this study served as a reference for future studies using the hypercapnic provocation and CLBF. Also, the concomitant change in PETO2 using the partial rebreathing technique was reduced compared to the manual addition of CO2 technique described in Chapter 3 but was still greater than optimal.  Chapter 5: A new automated gas flow controller was used to induce standardised normoxic, or isoxic, hypercapnia. The magnitude of vascular reactivity in both retinal arterioles and capillaries in response to the new hypercapnic stimulus was robust compared to the previous stimuli. There was a clear ONH vascular reactivity response in this study, unlike the result attained in Chapter 3. Although theoretically it is predictable that the percent magnitude of vascular reactivity of the arterioles and capillaries should be similar, this is the first study to show that they are indeed comparable. The magnitude of hypercapnia was repeatable and the concomitant change in PETO2 was minimal and physiologically insignificant.  Chapter 6: The normal response of the retinal arterioles and capillaries to normoxic hypercapnia is impaired in both uPOAG and pPOAG compared to controls. Short term treatment with 2% topical Dorzolamide for two weeks improved retinal vascular reactivity in ntPOAG. However, it is still unclear whether this improvement is a direct effect of Dorzolamide or as a secondary effect of the decrease in intraocular pressure (IOP).  Chapter 7: We found a reduction in the plasma ET-1 at baseline and during normoxic hypercapnia in the uPOAG and in the ntPOAG groups. This is the first study to show a lower plasma ET-1 level in uPOAG. The fact that this finding was repeated after 2 weeks treatment with Dorzolamide in the ntPOAG group further validates these results. It also suggests that Dorzolamide treatment does not impact ET-1 and cGMP measures, although it clearly results in an improvement of vascular reactivity. Correlation results suggest that as the change in ET-1 reduced during normoxic hypercapnia, the change in blood flow increased in the controls.

Introduction Retinal vascular reactivity is impaired in patients with diabetes and is thought to be involved in the onset and progression of diabetic retinopathy (DR). Previous studies that have utilized hyperoxia to assess retinal vascular reactivity have been limited due to confounding factors associated with the administration of oxygen and have used a variety of different instruments to measure retinal blood flow. The influence of blood glucose at the time of blood flow assessment has also not been systemically investigated. The specific aims of each Chapter are as follows: Chapter 3: To compare three systems used to administer hyperoxia to human subjects. Chapter 4: To quantify the magnitude and timeline of change of retinal hemodynamic parameters induced by an isocapnic hyperoxic stimulus. Chapters 5, 6 and 7: To quantify the magnitude of change of retinal hemodynamic parameters induced by hyperoxia, hyperglycemia and combined hyperoxia / hyperglycemia, respectively, in groups of diabetic patients with no clinically visible, and mild-to-moderate, DR and in age-matched subjects without diabetes. Methods Chapter 3: Subjects breathed air followed by oxygen, or oxygen plus carbon dioxide using a non-rebreathing system, or air followed by oxygen using a sequential rebreathing system. The magnitude of change and variability of CO2 concentrations was compared between systems. Chapter 4: Baseline retinal blood flow data was acquired while the subjects breathed air using a sequential rebreathing system. An isocapnic hyperoxic stimulus was initiated and maintained for 20 minutes. Air was then re-administered for 10 minutes. Retinal blood flow measurements were acquired every minute over the course of the study. The magnitude of change of each hemodynamic parameter was determined by fitting individual data with a sigmoidal function. For Chapter 5, 6 and 7 diabetic patients with no clinically visible, and mild-to-moderate, DR were stratified into groups based upon their retinopathy status. Age-matched non-diabetic subjects were recruited as controls. Baseline retinal blood flow data was acquired while subjects breathed air. Retinal blood flow measurements were then acquired after exposure to (a) hyperoxia, (b) hyperglycemia and (c) combined hyperoxic / hyperglycemic stimuli. Change in hemodynamic parameters was compared between groups and correlated with objective measures of retinal edema. Results Chapter 3: The difference in group mean end-tidal CO2 levels between baseline and hyperoxia was significant for oxygen administration using a non-rebreathing system. The sequential rebreathing technique resulted in a significantly lower variability of individual CO2 levels than either of the other techniques. Chapter 4: An ~11% decrease of diameter, ~36% decrease of velocity and ~48% decrease of blood flow was observed in response to isocapnic hyperoxia in young, healthy subjects. A response time of 2.30±0.53 minutes and 2.62±0.54 minutes was observed for diameter and velocity, respectively. Chapter 5: Retinal blood velocity, flow, and WSR significantly decreased in response to isocapnic hyperoxia in all groups. The magnitude of the reduction of blood flow was significantly reduced with increasing severity of retinopathy. There was a significant relationship between baseline objective edema index values and retinal vascular reactivity. Chapter 6: A significant change in blood glucose level was observed for all groups. No significant change in any hemodynamic parameter was found in patients with diabetes and in age-matched subjects without diabetes. Chapter 7: Retinal blood velocity and flow significantly decreased in all groups in response to combined hyperoxic / hyperglycemic provocation. The vascular reactivity response was not significantly different across the groups. Conclusions Chapter 3: Control of CO2 is necessary to attain standardized, reproducible hyperoxic stimuli for the assessment of retinal vascular reactivity. Chapter 4: Arteriolar retinal vascular reactivity to isocapnic hyperoxic provocation occurs within a maximum of 4 minutes. Although there was a trend for diameter to respond before velocity, the response characteristics were not significantly different between diameter and velocity. Different response characteristics of the retinal vasculature to transmural pressure mediated autoregulation as opposed to metabolic mediated vascular reactivity are suggested. Chapter 5: The vascular reactivity response in terms of the reduction of blood flow relative to baseline was significant in all groups but the magnitude of the change in flow was significantly reduced with increasing severity of retinopathy. A loss of retinal vascular reactivity is indicated in patients with moderate DR without clinically evident diabetic macular edema (DME), and in patients with DME. Chapter 6: Unaltered retinal arteriolar blood flow was found 1 hour after glucose ingestion in patients with diabetes and in age-matched subjects without diabetes. These results do not support the theory that retinal blood flow is affected by an acute increase of blood glucose in diabetic patients and in subjects without diabetes. Chapter 7: The vascular reactivity response to a combined hyperoxic / hyperglycemic provocation produced a pronounced reduction in blood flow. Unlike the response to hyperoxia alone, the vascular reactivity response was not significantly different across the groups. This suggests that hyperglycemia may influence the retinal vascular reactivity response to hyperoxia.

PURPOSE: Systemic hyperoxia has been induced using inspired gases in many studies to investigate vascular reactivity in the retinal vasculature. Technical limitations in the past resulted in inadequate control of systemic partial pressures of O2 and CO2, the latter of which tended to decrease secondary to induced hyperoxia. Recent development of a computerized gas delivery instrument has enabled the specific control of end-tidal CO2 (ETCO2) and fractional expired O2 (FeO2), independent of each other and of minute ventilation. The specific aims of each chapter are as follows: Chapter 3: To compare the magnitude and variability of the retinal vascular reactivity response to an isocapnic hyperoxic stimulus delivered using a manually-operated method to the newly developed computer-controlled gas sequencer. Chapter 4: To investigate the retinal hemodynamic response to incremental changes in hyperoxic stimuli during isocapnia. METHODS: Chapter 3: Ten young, healthy adults inhaled gases in a sequence of normoxic baseline, isocapnic hyperoxia, and normoxic recovery, using both gas delivery systems in random order. Chapter 4: Twelve healthy, young adults participated in a gas protocol consisting of 4 phases at varying fractional expired oxygen levels (FeO2): baseline (15%), hyperoxia I (40%), hyperoxia II (65%), and recovery (15%). End-tidal carbon dioxide (ETCO2) was maintained at an isocapnic level (~ 5%) throughout the experiment. In both Chapters 3 and 4, blood flow was derived from retinal arteriolar diameter and simultaneous blood velocity measurements of the superior temporal arteriole, acquired at 1-minute intervals during each of the phases of the gas protocol. RESULTS: Chapter 3: There was no interaction effect between the phases and gas delivery methods (p = 0.7718), but ETCO2 was significantly reduced during hyperoxia (p = 0.0002) for both methods. However, the magnitude of change in ETCO2 was physiologically insignificant i.e. <1%. The two systems differed in terms of FeO2 during hyperoxia, at a level of 85.27 ± 0.29% for the manual method, and 69.02 ± 2.84% for the computer method (p < 0.05). Despite this difference in oxygen concentrations, there was no difference in the vascular reactivity response for diameter (p = 0.7756), velocity (p = 0.1176), and flow (p = 0.1885) for equivalent gas phases between the two gas delivery systems. The inter-subject variability of retinal hemodynamic parameters was consistently lower using the computer-controlled gas sequencer. Chapter 4: Repeated measures ANOVA showed that there were significant influences of incremental changes in FeO2 on arteriolar diameter (p < 0.0001), blood velocity (p < 0.0001), and blood flow (p < 0.0001) in the retina. Paired t-tests of these retinal hemodynamic parameters during each phase in the gas sequence showed they were significantly different (p < 0.05) from each other, with the exception of baseline and recovery values. Incremental increases in FeO2 caused a linear decrease in group mean arteriolar diameter (R2 = 1, p = 0.002), group mean blood velocity (R2 = 0.9968, p = 0.04), and group mean blood flow (R2 = 0.9982, p= 0.03). CONCLUSIONS: Chapter 3: Inter-subject variability for virtually all retinal hemodynamic parameters was reduced using the computer-controlled method, presumably due to a higher degree of gas control. However, care needs to be exercised in the interprtetation of these results due to the relatively small sample size. A similar retinal hemodynamic response to isocapnic hyperoxia was induced using the two gas delivery systems, despite different levels of maximal FeO2. Chapter 4: Isocapnic hyperoxia elicits vasoconstriction and the reduction of retinal arteriolar blood flow in a dose-dependent manner over the range of FeO2 explored in this study.

The global aim of this thesis was to assess retinal vascular reactivity in glaucoma patients using a standardised hypercapnic stimulus. There is a suggestion of disturbance in the regulation of retinal and optic nerve head (ONH) hemodynamics in patients with Primary Open Angle Glaucoma (POAG), although much of the work to-date has either been equivocal or speculative. Previous studies have used non-standardised hypercapnic stimuli to assess vascular reactivity. To explain, hypercapnia induces hyperventilation which disturbs arterial oxygen concentration, an effect that varies between individuals resulting in the non-standardised provocation of vascular reactivity. Therefore, a normoxic hypercapnic provocation was developed to avoid additional and potentially uncontrolled vasoconstriction in what is thought to be a vasospastic disease. The development of a safe, sustained and stable normoxic hypercapnic stimulus was essential for the assessment of retinal arteriolar vascular reactivity so that repeated hemodynamic measurements could be obtained. Furthermore, most techniques used to measure vascular reactivity do not comprehensively assess retinal hemodynamics, in terms of the simultaneous measurement of vessel diameter and blood velocity in order to calculate flow. In this respect, this study utilized a technique that quantitatively assesses retinal blood flow and vascular reactivity of the major arterioles in close proximity to the ONH. The stimulus and vascular reactivity quantification technique was validated in healthy controls and then was clinically applied in patients with POAG. Newly diagnosed patients with untreated POAG (uPOAG) were recruited in order to avoid any confounding pharmacological effects and patients with progressive POAG (pPOAG) were also selected since they are thought to likely manifest vascular dysregulation. Finally, the results of the functional vascular reactivity assessment were compared to those of systemic biochemical markers of endothelial function in patients with untreated and progressive POAG and in healthy controls. Overall summary A safe, sustained, stable and repeatable normoxic hypercapnic stimulus was developed, evaluated and validated. In terms of the physiology of retinal vascular regulation, the percent magnitude of vascular reactivity of the arterioles and capillaries was found to be comparable in terms of flow. The new stimulus was successfully applied in POAG and in healthy controls to assess vascular reactivity and was also compared to plasma levels of ET-1 and cGMP. In terms of the patho-physiology of POAG, the study revealed a clear impairment of vascular reactivity in the uPOAG and pPOAG groups. There were reduced levels of plasma ET-1 in the uPOAG and ntPOAG groups. In addition, treatment with Dorzolamide improved vascular reactivity in the ntPOAG group in the absence of any change in the expression of plasma ET-1 or cGMP. Future work will address this apparent contradiction between the outcome of the functional vascular reactivity assessment and the biochemical markers of endothelial function in newly diagnosed POAG patients treated with Dorzolamide. Aims of chapters  Chapter 3: To determine the effect of hypercapnia on retinal capillary blood flow in the macula and ONH using scanning laser Doppler flowmetry (SLDF) in young healthy subjects.  Chapter 4: To describe a new manual methodology that permits the comprehensive assessment of retinal arteriolar vascular reactivity in response to a sustained and stable hypercapnic stimulus. The secondary aim was to determine the magnitude of the vascular reactivity response of the retinal arterioles to hypercapnic provocation in young healthy subjects.  Chapter 5: To compare the magnitude of vascular reactivity of the retinal arterioles in terms of percentage change of flow to that of the retinal capillaries using a novel automated standardized methodology to provoke normoxic, or isoxic, hypercapnia.  Chapter 6: To determine the magnitude of retinal arteriolar vascular reactivity to normoxic hypercapnia in patients with untreated POAG (uPOAG), progressive POAG (pPOAG) and controls. The secondary aim was to determine retinal vascular reactivity in newly treated POAG (ntPOAG, i.e. after treatment with 2% Dorzolamide, twice daily for 2 weeks).  Chapter 7: To compare plasma endothelin-1 (ET-1) and cyclic guanosine monophosphate (cGMP) between groups of patients with untreated primary open angle glaucoma (uPOAG), progressive POAG (pPOAG), newly treated POAG (ntPOAG) and controls. The effect of normoxic hypercapnia on plasma ET-1 and cGMP was also assessed. The functional measures of retinal blood flow and vascular reactivity were correlated with systemic biochemical markers of endothelial function. Methods Chapters 3, 4 and 5 were conducted on young healthy control subjects, where as Chapters 6 and 7 were conducted on patients with glaucoma and healthy controls.  Chapter 3: Subjects breathed unrestricted air for 15 minutes (baseline) via a sequential gas delivery circuit and then the fractional (percent) end-tidal concentration of CO2 (FETCO2) was manually raised for 15 minutes by adding a low flow of CO2 to the inspired air. For the last 15 minutes, FETCO2 was returned to baseline values to establish a recovery period. Heidelberg Retina Flowmeter (HRF) images centered on both the ONH and the macula were acquired during each phase.  Chapter 4: Subjects breathed air via a sequential gas delivery circuit for 15 minutes and the air flow was then manually decreased so that subjects inspired gases from the rebreathing reservoir until a stable 10-15% increase in FETCO2 concentration was achieved for 20 minutes. Air flow rate was then manually elevated so that subjects breathed primarily from the fresh gas reservoir to return FETCO2 back to baseline for the last 15 minutes. Retinal arteriolar hemodynamics was assessed using the Canon Laser Blood Flowmeter (CLBF) during all three breathing phases.  Chapter 5: Normoxic, or isoxic, hypercapnia was induced using an automated gas flow controller (RespirActTM, Thornhill Research Inc. Toronto, Canada). Subjects breathed air with PETCO2 normalized at 38 mmHg. An increase in PETCO2 of 15% above baseline, whilst maintaining normoxia, was then implemented for 20 minutes and then PETCO2 was returned to baseline conditions for 10 minutes. Retinal and ONH hemodynamic measurements were performed using the CLBF and HRF in random order across sessions.  Chapter 6: Retinal arteriolar vascular reactivity was assessed in patients with uPOAG, pPOAG (defined by the occurrence of optic disc hemorrhage within the past 24 months) and controls during normoxic hypercapnia. Using the automated gas flow controller, patients breathed air for 10 mins and PETCO2 was maintained at 38mmHg. Following this normoxic hypercapnia (a 15% increase in PETCO2 while PETO2 was maintained at resting levels) was induced for 15 mins and then for the last 10 mins PETCO2 was returned to baseline (post-hypercapnia) to establish recovery blood flow values. Retinal arteriolar diameter, blood velocity and blood flow was assessed using the CLBF in both patient groups and controls. A similar paradigm was repeated in the newly treated POAG group (ntPOAG, i.e. after treatment with 2% Dorzolamide, twice daily for 2 weeks).  Chapter 7: Blood samples were collected from the cubital vein of all participants (uPOAG, pPOAG, ntPOAG and controls) during baseline conditions (PETCO2=38mmHg) and then during normoxic hypercapnia (i.e. a 15% increase in PETCO2 relative to the baseline) using the paradigm described for Chapter 6. ET-1 and cGMP was assessed using immunoassay. Results  Chapter 3: The group mean nasal macula capillary blood flow increased from 127.17 a.u. (SD 32.59) at baseline to 151.22 a.u. (SD 36.67) during hypercapnia (p=0.028), while foveal blood flow increased from 92.71 a.u. (SD 28.07) to 107.39 a.u. (SD 34.43) (p=0.042). There was a concomitant and uncontrolled +13% increase in the group mean PETO2 during the hypercapnic provocation of +14% increase in PETCO2.  Chapter 4: Retinal arteriolar diameter, blood velocity and blood flow increased by 3.2% (p=0.0045), 26.4% (p<0.0001) and 34.9% (p<0.0001), respectively during hypercapnia. There was a stable ¬+12% increase in PETCO2 during hypercapnia and a concomitant -6% decrease in PETO2.  Chapter 5: Using an automated gas flow controller the co-efficient of repeatability (COR) was 5% of the average PETCO2 at baseline and during normoxic hypercapnia. The COR for PETO2 was 10% and 7% of the average PETO2 at baseline and during normoxic hypercapnia, respectively. Group mean PETCO2 increased by approximately +14.4% and there was only a +4.3% increase in PETO2 during hypercapnia across both study sessions. Retinal arteriolar hemodynamics increased during hypercapnia (p<0.001). Similarly, there was an increase in the capillary blood flow of the temporal rim of the ONH (p<0.001), nasal macula (p<0.001) and foveal areas (p<0.006) during hypercapnia. A non-significant trend for capillary blood flow to increase in the macula temporal area (+8.2%) was noted. In terms of percentage change of blood flow, retinal capillary vascular reactivity (i.e. all 4 analyzed areas = 22.4%) was similar to the magnitude of arteriolar (= 24.9%) vascular reactivity.  Chapter 6: Retinal arteriolar diameter, blood velocity and flow did not increase during normoxic hypercapnia in uPOAG compared to controls. Diameter and blood velocity did not change in pPOAG during normoxic hypercapnia but there was a significant increase in blood flow (+9.1%, p=0.030). After treatment with 2% Dorzolamide for 2 weeks there was a 3% (p=0.040), 19% (p<0.001) and 26% (p<0.001) increase in diameter, velocity and flow, respectively, in the ntPOAG group. Group mean PETCO2 increased by approximately +15% in all the groups and there was only a +3% increase in PETO2 during hypercapnia.  Chapter 7: Plasma ET-1 levels were significantly different across groups at baseline (one way ANOVA; p=0.0012) and this was repeated during normoxic hypercapnia (one way ANOVA; p=0.0014). ET-1 levels were lower in uPOAG compared to pPOAG and controls at baseline and during normoxic hypercapnia (Tukey’s honestly significant difference test). Similarly, ntPOAG group also showed lower ET-1 levels compared to the pPOAG and controls at baseline and during normoxic hypercapnia (Tukey’s honestly significant difference test). The cGMP at baseline and during normoxic hypercapnia across all groups was not different. In the control group, the change in ET-1 during normoxic hypercapnia was negatively correlated with change in retinal arteriolar blood flow (r = -0.52, p=0.04), that is, as the change in ET-1 reduced, the change in blood flow increased. A weak correlation was noted between change in cGMP during normoxic hypercapnia and the change in arteriolar blood flow (r = +0.45, p=0.08). Conclusions  Chapter 3: Hypercapnia resulted in a quantifiable capillary vascular reactivity response in 2 of the 3 assessed retinal locations (i.e., nasal macula and fovea). There was no vascular reactivity response of the ONH. It is critical to minimise the concomitant change in PETO2 during hypercapnia in order to obtain robust vascular reactivity responses.  Chapter 4: A technique to comprehensively assess vascular reactivity during stable and sustained hypercapnia was described. Retinal arteriolar diameter, blood velocity and blood flow increased in response to hypercapnia. The vascular reactivity results of this study served as a reference for future studies using the hypercapnic provocation and CLBF. Also, the concomitant change in PETO2 using the partial rebreathing technique was reduced compared to the manual addition of CO2 technique described in Chapter 3 but was still greater than optimal.  Chapter 5: A new automated gas flow controller was used to induce standardised normoxic, or isoxic, hypercapnia. The magnitude of vascular reactivity in both retinal arterioles and capillaries in response to the new hypercapnic stimulus was robust compared to the previous stimuli. There was a clear ONH vascular reactivity response in this study, unlike the result attained in Chapter 3. Although theoretically it is predictable that the percent magnitude of vascular reactivity of the arterioles and capillaries should be similar, this is the first study to show that they are indeed comparable. The magnitude of hypercapnia was repeatable and the concomitant change in PETO2 was minimal and physiologically insignificant.  Chapter 6: The normal response of the retinal arterioles and capillaries to normoxic hypercapnia is impaired in both uPOAG and pPOAG compared to controls. Short term treatment with 2% topical Dorzolamide for two weeks improved retinal vascular reactivity in ntPOAG. However, it is still unclear whether this improvement is a direct effect of Dorzolamide or as a secondary effect of the decrease in intraocular pressure (IOP).  Chapter 7: We found a reduction in the plasma ET-1 at baseline and during normoxic hypercapnia in the uPOAG and in the ntPOAG groups. This is the first study to show a lower plasma ET-1 level in uPOAG. The fact that this finding was repeated after 2 weeks treatment with Dorzolamide in the ntPOAG group further validates these results. It also suggests that Dorzolamide treatment does not impact ET-1 and cGMP measures, although it clearly results in an improvement of vascular reactivity. Correlation results suggest that as the change in ET-1 reduced during normoxic hypercapnia, the change in blood flow increased in the controls.

Calcineurin (CaN), a member of ser/thr protein phosphatase, was cloned from bovine retina. The peptide sequence of CaN A subunit is consisted of 511 amino acid residues. A 10 amino acid (A-T-V-E-A-I-E-A-D-E-A) deletion before the autoinhibitory domain was observed in bovine retina CaN A compared to bovine brain CaN A. The study on CaN activity and regulation demonstrated that different metal ions have different effects on its phosphatase activity. Ni2+ was found to be the strongest stimulator while Zn2+ was found to inhibit CaN phosphatase activity. Mn2+ was a relatively less effective stimulator compared to Ni2+. Fe2+ was also able to stimulate CaN phosphatase activity; in contrast, a previous study found Fe2+ slightly inhibited bovine brain CaN activity. The residues at 97-201 were found to be essential for bovine retina CaN A phosphatase activity. The residues at 407-456 also had an inhibitory effect on CaN A phosphatase activity in addition to the previously known auto inhibitory domain at 457-480. These observations suggest that bovine retina CaN A might possess some distinct structural characteristics compared to bovine brain CaN A.

Purpose: To investigate whether retinal thickness varies with refractive error. Also secondary to see if there is any difference in retinal thickness between the right and left eye. Methods: The inclusion criteria for the study was subjects without any pathologies, age between 18-45 and refractive error of maximum +0.75 SER and the myopia had no limit, as well as no astigmatism higher then -1.00D. Subjects, which fitted the inclusion criteria for the study, was shown to the OCT room were retinal thickness measurements were acquired first on the right and then left eye. To get the same reading area, the same setup was used and the fixation point was always centered for each patient. After all subjects had undergone the same method the results were analyzed using t-test and regression analysis. Results: The analysis showed a difference between emmetropic eyes and myopic eyes in the peripheral retinal thickness, having the myopes being significantly thinner. The inter myopic analysis showed no difference in retinal thickness in any of the points. This could however be due to the smaller sample size. The comparison between right and left eye showed a good symmetry between the two eyes both in the emmetropic and the myopic group. Conclusions: From this study we can conclude that the myopic group has a thinner peripheral retinal thickness than the emmetropic group. Central retinal thickness is not significantly different but could be due to the smaller sample size. There is no difference in retinal thickness between right and left eye.

Vertebrate photoreceptors are highly specialized sensory neurons that utilize a modified cilium known as the outer segment to detect light. Proper trafficking of proteins to the outer segment is essential for photoreceptor function and survival and defects in this process lead to retinal disease. In this dissertation I focus on two aspects of protein trafficking, intracellular vesicular trafficking in photoreceptors and retinal pigmented epithelial (RPE) cells and how it relates to the human disease choroideremia (CHM), and the trafficking of proteins through the photoreceptor cilium. The human retinal degenerative disease choroideremia (CHM) is caused by mutation of the Rab escort protein-1 (REP1) gene, which is required for proper intracellular vesicular trafficking. However, it was unclear whether photoreceptor degeneration in this disease is cell-autonomous, due to defective opsin transport within the photoreceptor, or is noncell-autonomous and a secondary consequence of defective RPE. Utilizing the technique of blastomere transplantation and a zebrafish line with a mutation in the rep1 gene, I show that photoreceptor degeneration in CHM is noncell-autonomous and is caused by defective RPE. The molecular machinery responsible for protein trafficking through the photoreceptor cilium remained unclear for a long time. Recent studies found Intraflagellar Transport (IFT) is the process that mediates cilia formation and transport of proteins through a cilium, and further analyses showed IFT is important for trafficking proteins to the outer segment. However, many details about how IFT works in photoreceptors remained unclear. By analyzing zebrafish harboring a null mutation in the ift57 gene, I show that Ift57 is only required for efficient IFT, and that the Ift57 protein plays a role in the ATP-dependent dissociation of kinesin II from the IFT particle. Lastly, I investigate the role of retrograde IFT in photoreceptors, a process that had yet to be investigated. By utilizing antisense morpholino oligonucleotides to inhibit expression of cytoplasmic dynein-2 (the molecular motor that mediates retrograde IFT) , I show that retrograde IFT is required for outer segment extension and the recycling of IFT proteins.

　　Proliferative vitreoretinopathy (PVR) remains the major cause of failure of retinal detachment surgery. Retinal pigment epithelial (RPE) cells have been suggested to play a major role in the pathogenesis of PVR. Numerous studies have employed pharmacological means to modulate cellular activities in attempts to inhibit the process. Recent attempts using adjunctive therapy during PVR surgery that consisted of 5-fluorouracil (5-FU) and low molecular weight heparin showed some promise in preventing PVR but the concern is that prolonged 5-FU treatment may have a toxic effect. On the other hand, lectin from the edible mushroom Agaricus bisporus (ABL) was found to inhibit growth of RPE cells in a potent manner without apparent cytotoxicity. This lectin could be a candidate to modulate anomalous proliferation of RPE cells while the mechanism for the observed inhibition is unknown. 　　In our study, we investigated whether RPE cells treated with 5-FU or ABL would attenuate cellular proliferation, cell migration, cell adhesion and cell-mediated contraction rates. Further, we investigated if complementary inhibition for the above cellular activities could be obtained when RPE cells were treated with ABL after the short treatment using 5-FU. We also explored the possible mechanisms through which ABL inhibited RPE cell proliferation. 　　ARPE-19 and primary human RPE cells were treated with 5-FU or vehicle for 10 minutes. Cells were then maintained in culture medium supplemented with or without ABL. The rate of cellular proliferation was measured by a tetrazolium salt assay. Effects on cell adhesion were investigated through loading RPE cells onto the strips coated with collagen I or fibronectin. Cell migration was investigated using a scratch wound model. The effect on cell-mediated contraction was assessed using a free floating collagen I matrix. Cytotoxicity of 5-FU and ABL was determined by the live/dead assay. 　　To elucidate the mechanism through which ABL inhibited RPE cell proliferation, we investigated cell cycle distribution patterns using flow cytometry. Phosphorylation statuses of Erk, Jnk, p38, Akt as well as p53 and Cyclin D expression level were investigated by Western blotting. 　　Both 5-FU and ABL inhibited RPE cell proliferation. Only ABL promoted cell adhesion towards collagen I in hRPE3 cells. ABL was found to attenuate the rate of cell migration. Cell-mediated collagen gel contraction was attenuated by 5-FU only. Complementary inhibition in cellular proliferation and cell-mediated collagen gel contraction was observed when both 5-FU and ABL were applied. No significant cell death was observed after treatment with 5-FU, ABL or both. 　　ABL was found to reduce the amount of cells present at S phase. Akt and Erk were found to be hypo-phosphorylated and hyper-phosphorylated respectively after ABL treatment. The expression levels of phosphorylated-Jnk, phosphorylated-p38, p53, and Cyclin D1 were not altered when compared with the control. 　　These results showed that 5-FU and ABL complement with each other on inhibiting the wound healing activities of RPE cells in vitro without apparent cytotoxicity. They suggested a possible new treatment modality for PVR. ABL hypo-phosphorylated Akt and this observation is in line with the fact that ABL could attenuate cell proliferation. / published_or_final_version / Anatomy / Doctoral / Doctor of Philosophy

Diabetic retinopathy (DR) was characterized by numerous hyperglycemia-dependent cellular and pathological changes in the retina, including retinal ischemia/reperfusion (I/R) injury. To determine the role of the 2nd enzyme of polyol pathway in relation with the pathogenesis in ischemic retinopathy, SDH deficient mice, C57BL/LiA, that lacked SDH activity, was used to study the pathogenesis of diabetic retinopathy, which also included I/R injury. Wild type and SDH-deficient mice were subjected to I/R injury by transiently occluding middle cerebral artery for two hours and twenty-two hour of reperfusion. The rationale of this study was to investigate the effect by blocking the conversion of sorbitol to fructose by SDH null mutation (SDH -/-), leading to accumulation of sorbitol level and reduction of oxidative stress, as demonstrated by the polyol pathway. Results: After induction with transient MCAO, there was increase in the thickness of OLM to ILM ipsilateral SDH+/+ compared with contralateral SDH+/+ (from 84 +/- 1 to 96 +/- 2 μm) while that of ipsilateral SDH-/- compared with contralateral SDH -/- (from 77 +/- 2 to 90 +/- 2 μm) suggested that there was edema after ischemic reperfusion injury. The result showed that there was increased cellular edema in ipsilateral retina of both SDH +/+ and SDH -/- retina after transient MCAO. The level of immunoreactivity against Aquaporin-4 and nitrotyrosine in studying the presence of oxidative stress; glutamine synthetase and glutamate in studying the toxicity of astrocyte glutamate; sarco-endoplasmic reticulum Ca2+-ATPase (SERCA) in studying the regulation Ca2+ homeostasis was determined using immunohistochemistry. For all the antibodies, there was similar immunoreactivity level between the contralateral side of both SDH+/+ and SDH -/- mice. For the SDH+/+ group, there was increase in signal in the ipsilateral retina in comparison with the contralateral one. On the other hand, for the SDH-/- group, similar result was observed. There was increase in signal and it was found more in the ipsilateral retina in comparison with the contralateral retina. Finally, in the ipsilateral retina of both SDH +/+ and SDH -/- mice, increased immunoreactivity was found in both but their difference was not statistically significant. This concluded that SDH deletion and subsequent accumulation of sorbitol metabolites did not contribute significantly in the role of pathogenesis of ischemic retinopathy especially in mice after I/R injury. / published_or_final_version / Anatomy / Master / Master of Medical Sciences

The process of cell fate determination, which establishes the vastly diverse set of neural cell types found in the central nervous system, remains poorly understood. During retinal development, multipotent retinal progenitor cells generate seven major cell types, including photoreceptors, interneurons, and glia, in an ordered temporal sequence. The behavior of these progenitor cells is influenced by the Notch pathway, a widely utilized signal during embryogenesis which can regulate proliferation and cell fate decisions. To examine the underlying genetic changes that occur when Notch1 is removed from individual retinal cells, microarray analysis of single cells from wild type or Notch1 conditional knockout retinas was performed. Notch1 deficient cells downregulated progenitor and cell cycle marker genes, while robustly upregulating genes associated with rod genesis. Single wild type cells expressed markers of both rod photoreceptors and interneurons, suggesting that these cells were in a transitional state. In order to examine the role of Notch signaling in cell fate specification separate from its role in proliferation, Notch1 was genetically removed specifically from newly postmitotic cells. Notch1 deficient cells preferentially became cone photoreceptors at embryonic stages, and rod photoreceptors at postnatal stages. In both cases, this cell fate change occurred at the expense of the other cell types normally produced at that time. In addition, single cell profiling revealed that Inhibitor of differentiation 1 and 3 genes were robustly downregulated in Notch1 deficient cells. Ectopic expression of these genes during postnatal development in wild type retinas was sufficient to drive production of progenitor/Müller glial cells. Moreover, Id1 and 3 partially rescued the production of Müller glial cells and bipolar cells in the absence of Notch1, even in newly postmitotic cells. We propose that after cell cycle exit, retinal precursor cells transition through a period in which they express marker genes of several different cell types as they commit to a fate, likely endowed by their progenitor cell. Specifically, cells that will become bipolars or Müller glia depend on Id-mediated Notch signaling during this transitional state to take on their respective fates.