Aldosterone as a mediator of severity in retinal vascular disease: evidence and potential mechanisms.

Abstract

Diabetic retinopathy (DR) and retinal vein occlusion (RVO) are the two most common retinal vascular diseases and are major causes of vision loss and blindness worldwide. Recent and ongoing development of medical therapies including anti-vascular endothelial growth factor and corticosteroid drugs for treatment of these diseases have greatly improved the care of afflicted patients. However, severe manifestations of retinal vascular disease result in persistent macular edema, progressive retinal ischemia and incomplete visual recovery. Additionally, choroidal vascular diseases including neovascular age-related macular degeneration (NVAMD) and central serous chorioretinopathy (CSCR) cause vision loss for which current treatments are incompletely effective in some cases and highly burdensome in others. In recent years, aldosterone has gained attention as a contributor to the various deleterious effects of retinal and choroidal vascular diseases via a variety of mechanisms in several retinal cell types. The following is a review of the role of aldosterone in retinal and choroidal vascular diseases as well as our current understanding of the mechanisms by which aldosterone mediates these effects.

1. Introduction

1.1. Clinical unmet needs in retinal and choroidal vascular diseases

Diabetic retinopathy (DR) and retinal vein occlusion (RVO) are the two most common retinal vascular diseases and cause significant visual impairment in millions of people worldwide(Klein; Rogers et al., 2010; Yau et al., 2012). Vision loss due to DR and RVO are caused by two major complications: macular edema and progressive retinal ischemia leading to neovascularization. While current treatments including anti-vascular endothelial growth factor agents, corticosteroids and thermal laser have improved outcomes for both DR(Boyer et al., 2014; Campochiaro et al., 2011; Do et al., 2011; Elman et al., 2010; Korobelnik et al., 2014; Wells et al., 2016) and RVO(Boyer et al., 2012; Heier et al., 2014, 2012; Ramchandran et al., 2008; Retinal and Oc, 2009), there remain a sizable proportion of patients who experience incomplete response to current therapies. For example, incomplete resolution of macular edema is seen in 25–50% of DME and RVO patients despite continuous treatment in rigorous clinical trials. (Campochiaro et al., 2014; Scott et al., 2017; Wykoff et al., 2015) Given the complexity of retinal vascular diseases, no single therapy is universally effective and multimodal treatment is frequently needed(Allingham et al., 2017; Elman et al., 2010) Finally, progressive ischemia has been reported in spite of monthly anti-VEGF in both diabetes(Ip et al., 2012; Nguyen et al., 2012) and RVO(Wykoff et al., 2015) trials. Thus, there remains a significant clinical unmet need to better understand the pathobiology of these retinal vascular diseases and to identify novel therapies for patients afflicted by them.

Neovascular age-related macular degeneration (NVAMD) is a leading cause of blindness in the elderly. The advent of anti-VEGF therapies have revolutionized treatment of the disease(Patel et al., 2011). However, the majority of patients with NVAMD require life-long therapy in order to maintain their vision(Peden et al., 2015) and a significant proportion of patients experience persistent disease activity despite aggressive treatment with currently available drugs(Sharma et al., 2016). Additionally, treatment burden frequently leads to under-treatment in real world settings(Lad et al., 2014) which is associated with poor visual outcomes. Adjunctive therapies which address persistent disease activity despite anti-VEGF and/or which reduce treatment burden for patients are critical unmet needs.

Central serous chorioretinopathy (CSCR) commonly afflicts middle aged men but may be seen in any age or gender.(Bousquet et al., 2019; Haimovici et al., 2004) There is no FDA-approved treatment for CSCR, however systemic mineralocorticoid antagonists including spironolactone and eplerenone, verteporfin photodynamic therapy and macular thermal laser are all used off label with varying success(Bousquet et al., 2019). While many cases resolve spontaneously or respond well to treatment, the chronic form of CSCR causes persistent vision loss and is poorly responsive to treatment. An approved and highly effective treatment for CSCR, particularly the chronic form of the disease is needed for afflicted patients.

1.2. Brief introduction to renin-angiotensin-aldosterone system in retina

Aldosterone is a steroid hormone which binds to the mineralocorticoid nuclear receptor (MR) where it regulates gene expression and downstream cellular responses to a variety of physiologic and pathologic stimuli. While aldosterone/MR signaling were initially characterized in renal physiology and the regulation of sodium and potassium homeostasis and blood pressure, it has become clear that both the production and action of aldosterone is far more widespread than initially believed.(Taves et al., 2011) Components of the renin-angiotensin-aldosterone-system (RAAS) have been identified in lymphoid tissues, the heart, vasculature, skin, central nervous system and the eye.(Taves et al., 2011) During the last two decades, evidence of an active renin-RAAS have been found in numerous ocular tissues.(Berka et al., 1995; Jennifer L Wilkinson-Berka et al., 2009; Zhao et al., 2010) The roles of upstream components of the RAAS including renin and angiotensin II in ocular pathobiology have been recently reviewed by Wilkinson-Berka.(Wilkinson-Berka et al., 2012) The purpose of this review will be to specifically examine the role and potential mechanisms of aldosterone in the pathobiology of retinal vascular disease.

1.3. Aldosterone and the mineralocorticoid receptor function in retina

Aldosterone classically functions via binding to the MR in the cytoplasm which allows MR to dimerize and translocate to the nucleus where it is known to interact with numerous other regulators of transcription to influence gene expression in target cells. Interestingly, the MR possesses equal affinity for both the mineralocorticoid hormone aldosterone and the glucocorticoid hormone cortisol, the latter of which is classically considered the ligand for the glucocorticoid receptor.(Arriza et al., 1987) Furthermore, cortisol is present in serum at a one hundred fold excess compared to aldosterone(Funder, 2017). This begs the questions of how aldosterone can specifically activate MR in its target tissues. While controversy remains regarding how MR achieves ligand specificity, it has been shown that aldosterone sensitive cells express 11-beta-hydroxysteroid dehydrogenase type 2 (HSD2) which inactivates cortisol leaving aldosterone free to bind MR.(Funder et al., 1988)

Since its initial description in renal physiology, signaling components of the RAAS have been found expressed in numerous organs including the eye, brain, lung and heart and is present in many cell types including the endothelium, pericytes, smooth muscle cells, glia, and neurons.(Taves et al., 2011) The rodent retina has been shown to express components of the RAAS in nearly all cell types including ganglion cells, microglia, Müller cells, endothelium, pericytes, bipolar, amacrine and horizontal neurons (Deliyanti et al., 2012b; White et al., 2015; Jennifer L. Wilkinson-Berka et al., 2009; Zhao et al., 2010). The presence of some components of the RAAS has been confirmed in post-mortem human retina but further studies are needed to confirm data obtained in rodent tissues.(White et al., 2015) Of particular interest is the fact that in the rat, Müller cells, retinal microvascular cells and retinal ganglion cells express aldosterone synthase, HSD2 and MR (Deliyanti et al., 2012b; Zhao et al., 2010), which would allow them to both synthesize and respond specifically to aldosterone. This finding raises several intriguing possibilities. First, the retina may be influenced by both systemic and locally produced aldosterone. Most currently published studies have examined the role of systemic aldosterone administration or suppression. These studies leave open the possibility that extraocular actions of aldosterone are partially responsible for the observed ocular effects. However, aldosterone has been shown to cause Müller cell swelling and redistribution of critical water and ion channels, aquaporin 4 (AQP4) and Kir4.1, respectively when administered intravitreally in rats.(Zhao et al., 2010) This suggests that at least some effects of aldosterone are mediated locally. There is still much to be learned about the potential role of locally produced aldosterone in retinal physiology and pathobiology. While the retina contains aldosterone synthetase, there is not yet proof that aldosterone is synthesized in the retina. Further, it is unknown whether aldosterone synthesis is altered during physiologic or pathologic conditions. Interestingly, local aldosterone synthesis has been demonstrated in the rat brain.(Gomez-Sanchez et al., 2005) This suggests that local aldosterone synthesis occurs within the CNS and provides tantalizing rationale for investigations into whether this holds true for the retina and other ocular tissues.

2. Pathobiology of aldosterone in retinal and choroidal vascular diseases

2.1. Pathobiology of RVO and potential role of aldosterone

Retinal vein occlusion is the second most common retinal vascular disease and is a major cause of vision loss in adults.(Rogers et al., 2010) RVO is thought to be caused by a combination of anatomic and local or systemic risk factors which include age, hypertension, diabetes, obesity, glaucoma and others.(Hayreh et al., 2001; “Risk factors for branch retinal vein occlusion. The Eye Disease Case-control Study Group.,” 1993, “Risk factors for central retinal vein occlusion. The Eye Disease Case-Control Study Group.,” 1996) Central retinal vein occlusion in caused by thrombus formation within the central retinal vein which is likely multifactorial.(Green et al., 1981; Hayreh et al., 2001) In the case of branch vein occlusion, thickening of the arterial wall at a site of arteriovenous crossing results in turbulent blood flow which, in combination with the systemic factors above leads to thrombotic occlusion of the vein(Frangieh et al., 1982; Hayreh et al., 2001). Following loss of retinal venous blood flow, a combination of hydrostatic forces and cellular signaling secondary to partial tissue hypoperfusion result in vascular leakage and hemorrhage in the areas drained by the affected vein.

As noted above, hypertension is a major risk factor for RVO. Notably, prevalence of primary hyperaldosteronism in general hypertensive patients is reported to range from approximately 4% in untreated hypertensives to 8% in treated hypertensives(Hannemann et al., 2012) and as high as 21% in patients with refractory hypertension.(Clark et al., 2012) Elevated serum aldosterone has also been reported in patients with metabolic syndrome(Briet and Schiffrin, 2011), obesity(Goodfriend et al., n.d.) and diabetes(Joseph et al., 2018) which have also been associated with RVO. That high levels of aldosterone are specifically associated with RVO has not been established but it is notable that ischemic RVO has been reported as the presenting compliant in a patient with primary hyperaldosteronism.(Atik et al., 2017)

Clinically, RVO is separated into ischemic and nonischemic types with ischemic RVO displaying worse hemorrhage, more severe macular edema and carrying worse prognosis for vision.(Hayreh, 2014; Panretinal and Occlusion, 1995) To date, the pathobiology driving ischemic RVO has not been elucidated. Interestingly, Wycoff and colleagues reported high incidence of progressive retinal ischemia even in treated RVO(Wykoff et al., 2015) however, the mechanisms driving this remain unknown. Additionally, a significant proportion of eyes treated with first-line anti-VEGF therapy fail to completely resolve macular edema and have persistent visual dysfunction.(Campochiaro et al., 2014; Scott et al., 2017) We have recently examined the effects of low dose, systemic aldosterone administration in a mouse model of RVO.(Allingham et al., 2018) This study was designed to evaluate the role of aldosterone in retinal vascular disease independent of effects on blood pressure. Accordingly, mice were kept on normal salt diet which we and others have shown to minimize confounding effects due to increased blood pressure ((Gu et al., 2017), Allingham et al. unpublished data). We found that, compared to controls, aldosterone exposed mice demonstrated more severe retinal edema and intraretinal hemorrhage (Figure 1). Further, aldosterone exposure caused extensive retinal ischemia. These clinical findings were associated with more severe Müller cell injury and fluid pump dysfunction as well as increased retinal infiltration by mononuclear phagocytes. Taken together, the findings in aldosterone exposed mice with RVO are reminiscent of ischemic RVO in humans, suggesting that aldosterone is involved in the biology of ischemic RVO and that Müller cell injury and mononuclear phagocyte infiltration could mediate aldosterone effects. Thus aldosterone could be a target for treating ischemic RVO.

Figure 1:

Aldosterone causes more severe intraretinal hemorrhage and retinal edema in a mouse model of vein occlusion.

2.2. Pathobiology of diabetic retinopathy and potential role of aldosterone

There is considerable evidence that upstream elements of the RAAS play a role in diabetic retinopathy. Specifically, RAAS inhibition with angiotensin converting enzyme inhibitors or angiotensin receptor blockers have shown positive effects on diabetic retinopathy even when compared to other modalities of blood pressure management.(Wang et al., 2015) While these effects could be mediated at they systemic level via blood pressure reduction, there is clear evidence that the retina expresses the necessary machinery to respond to upstream elements of RAAS(Jennifer L Wilkinson-Berka et al., 2009), that components of the RAAS are upregulated in response to experimental diabetes,(Okada et al., n.d.; Satofuka et al., 2009) and that inhibition of RAAS ameliorates the findings in animal models of diabetic retinopathy.(“Retinopathy and Nephropathy in Patients with Type 1 Diabetes Four Years after a Trial of Intensive Therapy,” 2000, “Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. UK Prospective Diabetes Study Group.,” 1998; Sjolie et al., 2008; Wilkinson-Berka, 2006) In spite of these observations and the fact that aldosterone represents a key downstream component of the RAAS, the specific role of aldosterone in development of diabetic retinopathy remains unknown. However there are several suggestive observations from studies of other models which share similarities with diabetic retinopathy and macular edema. Specifically, aldosterone mediates more severe proliferative retinopathy in the oxygen induced retinopathy (OIR) model and aldosterone inhibition ameliorates this phenotype(Deliyanti et al., 2012a; Jennifer L Wilkinson-Berka et al., 2009). Additionally, aldosterone causes severe ischemia, hemorrhage and retinal edema in an RVO model.(Allingham et al., 2018) The OIR model shares many similarities with proliferative diabetic retinopathy and RVO recapitulates features of diabetic macular edema. Taken together, proliferative retinopathy and macular edema are responsible for the vast majority of vision loss in diabetics. Thus, the validity of aldosterone as a therapeutic target in diabetic retinopathy and diabetic macular edema deserves further exploration. One challenge faced by researchers in the field is the lack of a diabetic animal model that recapitulates severe nonproliferative retinopathy and macular edema; current models recreate only the earliest and latest manifestations of the disease(Robinson et al., 2012; Su et al., 2000)

2.3. Pathobiology of choroidal vascular diseases and the role of aldosterone

Central serous chorioretinopathy (CSCR) is a choroidal vascular disease that commonly afflicts middle aged men but can be seen across a wide spectrum of ages and in both genders. The role of aldosterone and the mineralocorticoid receptor in the pathobiology of CSCR has been recently reviewed (Bousquet et al., 2019) and will be covered briefly here. CSCR is classically seen in patients who have been exposed to systemic glucocorticoid medications or who are under stress leading to increased endogenous glucocorticoid signaling (Haimovici et al., 2004). As the mineralocorticoid receptor binds both aldosterone and glucocorticoids, it is hypothesized that inappropriate activation of MR leads to CSCR. This hypothesis is supported by numerous clinical studies which have demonstrated the efficacy of MR antagonists spironolactone and eplerenone in treating CSCR ((Bousquet et al., 2019, 2013; Herold et al., 2014)). Zhao and colleagues were the first to demonstrate that intravitreal injection of aldosterone in rat led to choroidal vasodilation and accumulation of subretinal fluid, the hallmarks of CSCR in humans ((Zhao et al., 2012)). As local activation of MR by aldosterone can replicate CSCR in animals, it is tempting to speculate that local inhibition of MR could effectively treat the disease without exposing patients to the systemic effects of MR antagonists. Further study to evaluate this possibility is warranted.

Recently, a novel role for aldosterone/MR in choroidal neovascularization (CNV) has been reported. CNV is a choroidal vascular disease in which choroidal neovessels invade Bruch’s membrane, the sub-retinal pigment epithelial space and subretinal space leading to severe central vision loss due to exudation and hemorrhage involving the fovea. Neovascular age-related macular degeneration (NVAMD) is the most commonly seen disease manifesting CNV and is the leading cause of blindness in older people. CNV may also be present in ocular inflammatory conditions and myopic degeneration.

While the role of aldosterone in CNV is just starting to be explored, the group of Behar-Cohen recently reported that systemic and local treatment with spironolactone reduced CNV in the rat laser induced CNV model. Interestingly, they further showed that adjunctive treatment with oral spironolactone resulted in reduced disease activity in people whose NVAMD was not controlled by standard of care anti-VEGF drugs.(Zhao et al., 2019) These findings suggest that MR inhibition could be a therapeutic target in severe forms of CNV which fail to respond to current therapies, which is a significant clinical unmet need.

3. Potential mechanisms of aldosterone in retinal vascular diseases

What mechanisms could drive aldosterone-mediated severity in retinal vascular disease? While the role of aldosterone as a mediator of retinal vascular disease is increasingly accepted, the specific mechanisms by which aldosterone mediates retinal vascular disease are just starting to be uncovered. Given the pleiotropic effects of aldosterone and the fact that MR is expressed in many retinal cell types, it is likely that the effects of aldosterone are multifactorial and are mediated by several cell types (Figure 2).

Figure 2:

Potential pathobiologic mechanisms of aldosterone in retinal vascular disease.

3.1. Impact of aldosterone on retinal microvascular cells

The most obvious target cells are those of the vasculature itself. The effects of aldosterone on vascular cells including endothelial cells (EC), vascular smooth muscle cells and pericytes have been extensively studied by vascular biologists wishing to understand the pathogenic mechanisms of aldosterone in hypertension and cardiovascular diseases.(Biwer et al., 2019; Moss and Jaffe, 2015) Current proposed mechanisms include upregulated reactive oxygen species (ROS), enhanced vascular inflammation, increased thrombosis, and fibrosis.(Biwer et al., 2019) Of note, ROS are thought to play a role in both RVO and diabetes. Aldosterone has been shown to upregulate transcription and plasma membrane translocation of several components of the nicotinamide adenine diphosphate (NADPH) oxidase complex including Nox2, Nox4(Caprio et al., 2008; Jia et al., 2016), p22phox(Jia et al., 2016; Schäfer et al., 2013) and p47phox(Nagata et al., 2006) in endothelium which results in enhanced production of ROS. Whether aldosterone also upregulates ROS production in other retinal cell types remains unknown. Enhanced thrombosis and fibrosis are as yet, not well studied in retinal vascular disease but are interesting potential mechanisms for future study. In support of some of these mechanisms being relevant to the ocular microvasculature, Berka-Wilkinson initially reported the presence of a functional RAAS in retinal microvascular cells and characterized the effects of aldosterone on both cultured ECs and in the OIR model. She showed dose dependent increases in EC tubulogenesis in response to aldosterone and demonstrated that neovascularization in OIR was increased by aldosterone and inhibited by MR antagonists or aldosterone synthase antagonists.(Deliyanti et al., 2012b; Jennifer L Wilkinson-Berka et al., 2009) There are significant similarities between the oxygen-induced retinopathy model and other ischemic retinopathies such as proliferative diabetic retinopathy and ischemic retinal vein occlusion which suggests that these findings may be generalizable to multiple retinal vascular diseases. In addition, Zhao and colleagues showed that endothelial deletion of MR resulted in reduction in severity of CNV (Zhao et al., 2019) providing further evidence in support for the hypothesis that actions at the level of the microvasculature are important in MR-mediated vascular disease. These observations support aldosterone as a potential mediator of retinal vascular cellular dysfunction in diabetes, RVO and other vascular diseases such as CNV.

3.2. Aldosterone-mediated retinal inflammation

A second likely mechanism is aldosterone-mediated endothelial expression of ICAM1, increased expression of MCP1 and other inflammatory mediators leading to recruitment of macrophages into the retina in retinal vascular diseases. Macrophage infiltration into the retina has been reported in a variety of animal models of retinal disease including RVO(Allingham et al., 2018; Ebneter et al., 2017) AMD(Caicedo et al., 2005b), retinopathy of prematurity(Jennifer L. Wilkinson-Berka et al., 2009), and diabetes.(Rangasamy et al., 2014) Macrophage infiltration has also been demonstrated in pathology specimens taken from human eyes with dry and neovascular macular degeneration(Lad et al., 2015) and diabetes.(Zeng et al., 2008) Aldosterone exposure has been shown to increase infiltration of macrophages into retinal tissues following RVO and in OIR. This may be explained by the observation that aldosterone causes increased expression of chemotactic factors such as MCP-1 as well as upregulated expression of endothelial leukocyte adhesion proteins such as ICAM-1 which together could mediate macrophage recruitment to tissues.(Caprio et al., 2008; Jennifer L Wilkinson-Berka et al., 2009) Once in the retina, the function of macrophages is less clear. In models of neovascular AMD, macrophages have been shown to have deleterious effects including Müller cell injury, synaptic dysfunction.(Caicedo et al., 2005a) However, the roles of infiltrating macrophages and resident microglia remain incompletely understood in most vascular diseases affecting the retina. This is an area deserving of further study.

3.3. Effects of aldosterone on retinal Müller glial cells

Finally, Müller glial cells are a target of aldosterone which is unique to the retina. Müller cells are known to express MR at high levels and several groups have characterized them as aldosterone responsive both in vitro (Zhao et al., 2010) and in vivo.(Allingham et al., 2018; Zhao et al., 2010) Müller cells serve a variety of functions which are critical for maintenance of normal vascular function, water homeostasis and the extracellular milieu.(Reichenbach and Bringmann, 2013) Aldosterone could impact Müller cells via a variety of mechanisms including altered gene expression of proteins required for critical physiologic functions or aldosterone could also cause indirect injury to Müller cells via upregulated inflammation, reactive oxygen species (ROS) or other means. Along these lines, we have shown aldosterone to cause increased glial fibrillary acidic protein (GFAP), a marker of Müller cell injury and activation(Allingham et al., 2018). We and others have also shown that aldosterone alters expression and localization of both aquaporin 4 (AQP4) and Kir4.1(Allingham et al., 2018; Zhao et al., 2010), water and potassium channels which are specifically localized to perivascular membranes and to the foot plate and which together are responsible for Müller cell fluid pump function.(Reichenbach et al., 2007). This colocalization of AQP4, which facilitates transmembrane water transport with Kir4.1 which specifies the site of potassium efflux is critical for Müller cell mediated fluid transport (Figure 3). The release of potassium into the extracellular space by Kir4.1 creates a localized osmotic gradient which drives water out of Müller cells via AQP4. These channels are specifically colocalized by binding to members of the dystrophin associated protein complex which is found specifically at the foot plate and perivascular subdomains of Müller cells ((Reichenbach et al., 2007)). Water exported at these sites is thought to leave the retina via diffusion into the vitreous or via transport into capillaries. We have found that both AQP4 and Kir4.1 lose their specific subcellular localization within Müller cells when mice are exposed to systemic aldosterone. This loss of appropriate AQP4 and Kir4.1 localization is hypothesized to inhibit Müller cell fluid pump function which leads to severe intraretinal fluid accumulation and macular edema in the context of retinal vascular leakage. This hypothesis is supported by the finding that retinal edema is more severe following RVO in mice exposed to aldosterone(Allingham et al., 2018). In support of these observations, Zhao and colleagues reported that intravitreally administered aldosterone caused altered expression and localization of both AQP4 and Kir4.1 as well as retinal edema in rats which did not have disease-mediated vascular leakage.(Zhao et al., 2010) This suggests that Müller cells fluid pump function is required for basic water homeostasis as well as response to vascular injuries in which retinal vessels become leaky. Clearly, Müller cells have multiple roles in the maintenance of retinal homeostasis, many of which are dysregulated in retinal vascular diseases. Future studies should explore whether aldosterone antagonism could improve Müller cell function in models of retinal vascular disease including RVO.

Figure 3:

Disruption of retinal water homeostasis results from an imbalance in fluid influx due to vascular leakage and fluid export by Müller cell pump function. Aldosterone causes redistribution of AQP4 and Kir4.1 leading to loss of Müller cell pump function. This is one mechanism by which aldosterone may cause severe macular edema.

4. Conclusions and future directions:

The field is currently limited by several gaps in our knowledge. One is the open question as to which effects of aldosterone are mediated locally and which are attributable to systemic effects on other target tissues or cell types. For example, aldosterone is known to alter the effector function of peripheral blood monocytes(Martín-Fernández et al., 2016) which are known to invade the retina in RVO, diabetes and AMD. Under certain conditions, aldosterone causes elevated blood pressure which is a known contributor to vascular disease in the retina. On the other hand, there is evidence to support both local activity of aldosterone in retina and efficacy of local aldosterone antagonism in retinal disease models(Zhao et al., 2019, 2010). This is an important area deserving of further study because it will dictate whether therapies directed against aldosterone must be systemic or if local treatments such as intravitreal, periocular of topical routes are viable. The study of locally administered aldosterone and aldosterone antagonists such as spironolactone coupled with use of cell-specific MR deletion will allow us to gain further specificity in our understanding of the pathomechanisms of aldosterone in retinal and choroidal vascular disease.

A second significant gap in knowledge is comprehensive identification of the specific gene products which mediate the deleterious effects of aldosterone in ocular disease. Further, because of the myriad cell types which are capable of responding to aldosterone, it is not yet known which specific cell types are the primary mediators of disease and whether all effects of aldosterone are undesirable. This is important because specific understanding of specific cellular and genetic targets of aldosterone/MR could identify drugable pathways for treatment of retinal diseases. Efforts should be made to confirm and further characterize the cellular expression patterns of the RAAS seen in rodents using human ocular tissues. In addition, single cell RNA-seq as well as proteomic approaches could be utilized to gain more granular understanding of specific cellular targets and genetic responses to aldosterone in health and disease.

Finally, there is increasing evidence that aldosterone may play a roll in eye diseases which are not traditionally considered vascular in nature. For example, two recent studies have independently found that systemic aldosterone treatment in rats causes a dose-dependent loss of retinal ganglion cells ((Nitta et al., 2013; Takasago et al., 2018). Further, Nitta and colleagues found that treatment with spironolactone improved RGC survival independent of effects on intraocular pressure. Clearly novel and unexpected roles for aldosterone are still waiting to be discovered.

In conclusion, despite the numerous recent advances in treatment for retinal vascular diseases, persistent macular edema, progressive ischemia, and incomplete visual recovery remain clinical unmet needs. Aldosterone has gained increasing attention as a mediator of disease severity in models of ischemic retinopathy and macular edema and we are just beginning to understand the molecular mechanisms behind these observations. Further studies such as those outlined above focused on aldosterone-mediated effects on the retinal microvasculature, Müller cells, macrophage infiltration, as well as other ocular cell types are warranted and may yield novel mechanistic insights as well as potential targets for therapy.

Highlights:

• Retinal vascular diseases such as diabetic retinopathy and retinal vein occlusion manifest severe forms for which our current treatments are inadequate

• There is compelling clinical rationale for targeting aldosterone in retinal vascular disease

• Numerous retinal cell types including microvascular cells and Müller glial cells are responsive to aldosterone

• Aldosterone has been shown to increase the severity of several animal models of retinal vascular disease

• Mechanisms by which aldosterone increases severity of retinal vascular disease include microvascular cellular dysfunction, increased inflammation, Müller glial dysfunction and possibly others

Abbreviations:

AMD

age-related macular degeneration

AQP4

aquaporin 4

DME

diabetic macular edema

DR

diabetic retinopathy

EC

endothelial cell

HSD2

11β-hydroxysteroid dehydrogenase type 2

ICAM

intercellular adhesion molecule

MCP-1

monocyte chemoattractant protein 1

MR

mineralocorticoid receptor

NADPH

nicotinamide adenine diphosphate

OIR

oxygen induced retinopathy

RAAS

renin-angiotensin aldosterone system

ROS

reactive oxygen species

RVO

retinal vein occlusion

VEGF

vascular endothelial growth factor