Aldosterone and the mineralocorticoid receptor’s role in inflammation and vascular pathology in retinal disease

Purpose: Retinopathy of prematurity (ROP) is a significant cause of severe vision loss and blindness in children, and can be replicated in rodent models of the disease where it is known as oxygen induced retinopathy (OIR). A distinct feature of OIR is retinal neovascularization, which is also present in patients with proliferative diabetic retinopathy. Blockade of the renin angiotensin-aldosterone system (RAAS) via inhibitors of the angiotensin II receptor type 1 (AT1-R) or angiotensin-converting enzyme (ACE) in diabetic patients are associated with reduced incidence and progression of retinopathy. However, these interventions do not completely prevent retinal pathology. One reason for this may be incomplete suppression of the RAAS. For instance, aldosterone may still be present in plasma and tissues due to the aldosterone escape phenomenon. Indeed, in patients, blockade of aldosterone via the mineralocorticoid receptor (MR) combined with angiotensin II blockade, provides superior cardiorenal protection than monotherapy. Another consideration is an underlying critical mechanism for ROP, which is the upregulation of reactive oxygen species (ROS), with studies showing that global inhibitors of ROS such as apocynin reduced both retinal ROS levels and neovascularization. However, these findings do not seamlessly translate to humans, with strategies such as antioxidants such as vitamin E being associated with a significant risk of infection, which has precluded the widespread use of this therapy. The explanation for the failure of such antioxidant therapies is not entirely known, but it may be due to their lack of efficacy, and nonspecific scavenging of all ROS might lead to untoward effects. Thus, there has been substantial interest in identifying a more suitable ROS-related target for ROP. A likely candidate is NADPH oxidase (NOX) enzymes which solely function to produce ROS and are involved in various pathological processes. Of relevance to this thesis is that angiotensin II and aldosterone are stimulators of NOX isoforms. Lastly, a Western diet typified by high sodium levels has been viewed as a risk factor for a number of pathological conditions, such as hypertension, cardioavascular disease and diabetic retinopathy. The reasons for this are not completely understood, but studies suggest that it is likely due to oxidative stress, inflammation and disturbances of the RAAS. At present, the role for aldosterone, NOX isoforms and dietary salt in ROP is unclear. The current study was aimed to elucidate (1) the role of aldosterone in OIR using an aldosterone synthase inhibitor, FAD286, (2) the role of NOX isoforms using mice lacking NOX1, NOX2 and NOX4, a novel inhibitor of NOX1 and NOX4, GKT137831, and the association between NOX isoforms and the RAAS as well as (3) the role of dietary salt in OIR. Methods: To address Aim 1, the rodent model of OIR was induced in Sprague-Dawley (SD) rats by exposure to 80% oxygen from postnatal days (P) 0 to 11, followed by 7 days in room air. Treatment with either FAD286 (30 mg/kg/day) and the AT1-R blocker, valsartan (10mg/kg/day) was administered from P12-P18. Retinal neovascularization was evaluated by counting pre-retinal vessels on paraffin sections and by measuring the area of neovascular tufts on FITC-lectin stained retinal wholemounts. An enzyme-linked immunosorbent assay (ELISA) was used to quantitate retinal vascular endothelial growth factor (VEGF) levels. The density of microglia was evaluated by quantitating ionized calcium binding adaptor protein (Iba-1, a microglial marker) immunolabeling on retinal paraffin sections. Quantitative real-time PCR was employed to measure the expression of inflammatory factors, such as intercellular adhesion molecule-1 (ICAM-1), monocyte chemoattractant protein-1 (MCP-1), vascular adhesion molecular 1 (VCAM-1) and tumor necrosis factor-α (TNF-α) and RAAS components, such as aldosterone synthase. Finally, the cellular location of the aldosterone/system in the retina was determined by PCR using primary cultures of retinal cell populations, such as retinal ganglion cells (RGC), microglia, glial cells and vascular cells. To address Aim 2, the mouse model of OIR was induced in NOX1, NOX2, NOX4 knockout (KO) and their respective wild type (WT) control mice by exposure to 75% oxygen from P7 to P11, followed by 5 days in room air. Control animals were in room air from P0-P18. Retinal vasculopathy was evaluated by quantitating pre-retinal vessels on paraffin sections and neovascular tufts on retinal wholemounts. The extent of vaso-obliteration was quantitated by measuring the area of vessel loss on retinal wholemounts. Vascular leakage was quantitated by measuring levels of albumin in the retina using ELISA. A rhodamine-conjugated concanavalin A (ConA) perfusion technique was used to stain leukocytes adhering to retinal vasculature, and the total number of leukocytes was counted. Superoxide levels were measured by dihydroethidium (DHE) labeling on retinal cryosections. The density of microglia was evaluated by Iba-1 immunolabeling on retinal paraffin sections. Quantitative real-time PCR was used to measure VEGF, erythropoietin (EPO), angiopoetin-2 and ICAM-1. To examine if RAAS blockade can attenuate NOX derived ROS, primary cultures of rat retinal microglia were exposed to hypoxia to mimic the in vivo environment of OIR. The levels of superoxide in microglial lysates were evaluated by DHE staining. The expression of NOX isoforms, angiogenic and inflammatory factors were quantitated by quantitative real-time PCR and a protein cytokine array. Finally, to investigate if specific NOX isoform inhibition is a potential treatment for ROP, a subset of SD rats was employed to study the effect of GKT137831 in a rat model of ROP. To address Aim 3, a low-salt (LS, 0.03% sodium) and normal-salt (NS, 0.3% sodium) diet were given to the mothers from gestational day (G) 20 to P18. Retinal neovascularization and vaso-obliteration were quantitated on paraffin sections and retinal wholemounts. ELISA was used to measure retinal VEGF levels and vascular leakage. Western blotting was used to quantitate EPO and phosphorylated extracellular signal-regulated kinase-1 and 2 (pERK1/2). Immunohistochemistry was used to evaluate the density of microglia (Iba-1), TNF-α and the expression of water and ion channels such as aquaporin (APQ) 4, the inwardly rectifying potassium channel Kir4.1 on retinal paraffin sections. Glial fibrillary acidic protein (GFAP) immonofluorescence labeling on paraffin sections was used to evaluate retinal gliosis. Quantitative real-time PCR was used to quantitate the mRNA levels of RAAS elements, such as angiotensinogen, renin, AT1-R, aldosterone synthase and MR, and water and ion transport channels, such as AQP1, AQP4, Kir4.1 and the epithelial sodium channel (ENaC)α. Results: Firstly, in the rat model of OIR, treatment with FAD286 significantly reduced retinal neovascularization, microglial density, angiogenic and inflammatory factors, such as VEGF, ICAM-1, VCAM-1, MCP-1 and TNF-α. FAD286 effectively reduced aldosterone synthase mRNA in the retina. MR and 11βHSD2 were expressed in RGC, microglia, glial, endothelial cells and pericytes, but aldosterone synthase was only detected in RGC, microglia and glial. Secondly, deficiency in NOX1 but not NOX2 and NOX4 is protective against retinopathy. In addition, NOX1 but not NOX2 and NOX4 had reduced retinal neovascularization, decreased retinal avascular vaso-obliteration and leukostasis. These effects in NOX1 KO mice with OIR were accompanied by a reduction in the density of microglia, retina vascular leakage, the expression of angiogenic and inflammatory markers as well as ROS levels, such as VEGF, EPO, angiopoeitin-2, ICAM-1, superoxide and nitrotyrosine levels. In cultured microglia, ROS levels and the expression of NOX isoforms as well as angiogenic and inflammatory factors were elevated in response to hypoxia, and these increases were reduced with RAAS blockade. NOX1 was expressed in RGC, glia, microglia and pericytes, but was absence in endothelial cells. NOX2 was expressed in all retinal types, but was not detected in glial cells. NOX4 was presence in all retinal cell types. GKT137831 showed similar retinoprotective effects in rats with OIR as NOX1 KO mice with OIR. In vitro, GKT137831 effectively reduced hypoxia induced ROS in microglia. Lastly, the LS diet exerted protective effects in the rat OIR model with the benefits of reduced neovascularization, vaso-obliteration, angiogenic factors (VEGF, EPO, pERK1/2), inflammatory factors (TNF-α), microglial density, retinal RAAS elements (renin, aldosterone synthase, AT1-R and MR), retinal gliosis, vascular leakage. In addition, OIR rats fed with NS showed an increase in the expression of water and ion transport channels, such as AQP4, Kir4.1 and ENaCα, which remained unchanged with a LS diet. Conclusions: The results of the thesis suggest that FAD286, GKT137831 and a LS diet are anti-angiogenic and anti-inflammatory in OIR and that NOX1 is a key player in OIR. Therefore, FAD286, GKT137831 and a low-salt diet are potential treatments for ROP and other retinal diseases characterized by neovascularization, such as diabetic retinopathy which is the leading cause of vision impairment in the adult population.