Aldosterone and the mineralocorticoid receptor in renal injury: A potential therapeutic target in feline chronic kidney disease

Abstract There is a growing body of experimental and clinical evidence supporting mineralocorticoid receptor (MR) activation as a powerful mediator of renal damage in laboratory animals and humans. Multiple pathophysiological mechanisms are proposed, with the strongest evidence supporting aldosterone‐induced vasculopathy, exacerbation of oxidative stress and inflammation, and increased growth factor signalling promoting fibroblast proliferation and deranged extracellular matrix homeostasis. Further involvement of the MR is supported by extensive animal model experiments where MR antagonists (such as spironolactone and eplerenone) abrogate renal injury, including ischaemia‐induced damage. Additionally, clinical trials have shown MR antagonists to be beneficial in human chronic kidney disease (CKD) in terms of reducing proteinuria and cardiovascular events, though current studies have not evaluated primary end points which allow conclusions to made about whether MR antagonists reduce mortality or slow CKD progression. Although differences between human and feline CKD exist, feline CKD shares many characteristics with human disease including tubulointerstitial fibrosis. This review evaluates the evidence for the role of the MR in renal injury and summarizes the literature concerning aldosterone in feline CKD. MR antagonists may represent a promising therapeutic strategy in feline CKD.

been translated into therapeutic use more quickly and effectively than the use of MRAs. Whilst it is true that MR activation contributes to renal damage in the context of hypertension, a blood pressure-independent effect has been demonstrated in various models of kidney injury including subtotal nephrectomy (Ibrahim & Hostetter, 1998), ischaemia/reperfusion injury (Barrera-Chimal et al., 2015;Mejía-Villet et al., 2007;Ramírez et al., 2009), diabetic nephropathy (Bamberg et al., 2018), glomerulonephritis (Asai et al., 2005) and calcineurin inhibitor nephrotoxicity (Feria et al., 2003).
Chronic kidney disease is the most common cause of mortality in ageing cats (O'Neill et al., 2015) and may result in significant morbidity in affected individuals. Aetiology is usually unknown on an individual basis, but pathological characteristics, namely multifocal tubulointerstitial fibrosis and chronic mononuclear tubulointerstitial inflammation, are consistent (Chakrabarti, Syme, Brown, & Elliott, 2013;McLeland, Cianciolo, Duncan, & Quimby, 2015;Zini et al., 2014). Important differences exist between feline and human CKD, as cats exhibit a lower frequency of proteinuria and glomerulonephritis compared with humans, and different risk factors for disease development exist, with hypertension and diabetes mellitus being important in people (Jha et al., 2013), and frequent vaccination and dental disease identified in feline epidemiological studies (Finch, Syme, & Eliott, 2016;Greene et al., 2014). Tubulointerstitial fibrosis is the lesion best correlated with disease severity in both cats (Chakrabarti et al., 2013;Sawashima et al., 2000;Yabuki et al., 2010) and people (Hruby et al., 1998;Nath, 1992), however, and occurs early in feline CKD (McLeland et al., 2015). Although several clinicopathological findings, including proteinuria, anaemia and hyperphosphataemia, correlate with fibrosis severity and/or survival (Boyd, Langston, Thompson, Zivin, & Imanishi, 2008;Chakrabarti et al., 2013;Chakrabarti, Syme, & Elliott, 2012;Elliott, Rawlings, Markwell, & Barber, 2000;King, Tasker, Gunn-Moore, Gleadhill, & Strehlau, 2007;McLeland et al., 2015;Syme et al., 2006), causal and progression factors of feline CKD remain poorly understood. Recently, renal hypoxia/ ischaemia, perhaps episodic in nature, has been proposed to contribute to the initiation and progression of feline CKD (Cowgill et al., 2016;Jepson, 2016). This is supported by experimental models where renal ischaemia results in morphological changes akin to those observed in naturally occurring disease Schmiedt et al., 2012Schmiedt et al., , 2016. Aside from the feeding of a renal diet (Ross et al., 2006), currently no effective treatments exist which are proven to significantly slow feline CKD progression. One of the benefits of a renal diet is thought to be restriction of phosphate intake (Elliott et al., 2000;Ross, Finco, & Crowell, 1982). As such, it is important to understand factors which may be associated with disease advancement so that novel therapeutic interventions may be established.
This review provides an overview of the evidence supporting the deleterious role of aldosterone/MR activation in renal injury in laboratory animals and humans and discusses its potential relevance in the context of feline CKD.

| ALDOS TERONE AND THE MR
Aldosterone is a mineralocorticoid hormone produced primarily in the zona glomerulosa of the adrenal cortex whose major physiological function is to maintain sodium and potassium homeostasis and blood pressure control. Upon binding to the MR in the epithelial cells of the renal cortical collecting tubules and collecting ducts, aldosterone stimulates a cascade of events resulting in sodium reabsorption, and thus the maintenance of intravascular volume, and potassium secretion (Ponda & Hostetter, 2006). The major secretagogues of aldosterone are increased serum potassium concentration and angiotensin II (via the angiotensin type 1 receptor) (Beuschlein, 2013).
Components of the renin-angiotensin-aldosterone system (RAAS) are important on both a systemic and tissue-specific level (Nishiyama & Kobori, 2018;Siragy & Carey, 2010), and intrarenal aldosterone may act independently of circulating aldosterone levels. In fact, in humans and laboratory species, MR blockade has been shown to be beneficial in the absence of elevated plasma aldosterone levels (Du et al., 2009;Nagase, Matsui, Shibata, Gotoda, & Fujita, 2007;Nagase et al., 2006;Pitt, Remme, Zannad, & Neaton, 2003;Pitt et al., 1999) and renal MR expression is not correlated with serum aldosterone levels in people with CKD (Quinkler et al., 2005). CYP11B2, the gene which codes for aldosterone synthase, is expressed in the renal cortex of normal rats and is upregulated by angiotensin II (Xue & Siragy, 2005); other extra-adrenal sites of aldosterone synthesis include the brain, blood vessels and myocardium (MacKenzie et al., 2000;Takeda et al., 1995;White, 2003).

| ALDOS TERONE IN CK D
In the early stages of CKD, RAAS activation occurs as a compensatory response to maintain glomerular filtration rate (GFR); however, chronic activation is maladaptive and leads to progressive renal injury. Angiotensin II has historically been regarded as the major mediator of RAAS-induced renal injury, not only through its glomerular effects but also by activating proinflammatory and profibrotic pathways (Ames, Atkins, & Pitt, 2019;Eddy, 1996;Nishiyama & Kobori, 2018). Consequently, the current standard of care for CKD treatment in human medicine involves angiotensin II inhibition with angiotensin-converting enzyme inhibitors (ACEIs) and/or angiotensin type 1 receptor blockers (ARBs). Substantial and ever-increasing evidence in laboratory species and humans demonstrates that aldosterone also causes direct organ damage, particularly in the heart and kidneys. Aldosterone's pathophysiological actions are similar to and overlap with those of angiotensin II (Ames et al., 2019), and the interactions between the two are complex, meaning it can be difficult to discern their individual effects (Luther et al., 2012;Virdis et al., 2002).
Chronic kidney disease can be considered as a state of relative hyperaldosteronism. Increased plasma aldosterone levels are a risk factor for kidney injury in human clinical studies, and MRA treatment has been shown to be beneficial in numerous rodent models of renal disease and in human patients, for example by abrogating renal histopathological changes and reducing proteinuria and blood pressure.
Aldosterone's detrimental effects on the kidney predominantly occur via nonepithelial MRs, and importantly, can arise independently of aldosterone's effect on blood pressure (Fujisawa et al., 2004;Rafiq, Hitomi, Nakano, & Nishiyama, 2010). The proposed mechanisms underlying the detrimental effects of MR activation in the kidney are outlined in Figure 1. "Aldosterone breakthrough" is a phenomenon which further supports harmful effects of MR activation; this term applies to patients on ACEI/ARB therapy who experience plasma aldosterone concentrations that return to or exceed pretreatment levels following an initial reduction (Terata et al., 2012). Aldosterone breakthrough is associated with more severe proteinuria and a faster deterioration in renal function in people (Buglioni et al., 2015;Sato, Hayashi, Naruse, & Saruta, 2003;Schjoedt, Andersen, Rossing, & Tarnow, 2004). Aldosterone breakthrough is poorly characterized in veterinary species but has been documented in cats with hypertrophic cardiomyopathy treated with ACEIs (MacDonald & Kittleson, 2008), and preliminary studies have demonstrated that aldosterone breakthrough may occur in up to 33% of dogs with proteinuric renal diseases that are receiving ACEIs/ARBs (Ames, unpublished data Hess, 2017) and in healthy dogs receiving furosemide following ACEI or ARB treatment (Ames, Atkins, Lee, Lantis, & Zumbrunnen, 2015;Konta et al., 2018;.

| Vascular effects of MR activation
The effects of MR activation on vascular function and structure is thought to be the major mechanism by which aldosterone causes renal injury (Duprez, 2007;Jaisser & Farman, 2016). MR activation in vascular endothelial cells and VSMCs results in endothelial dysfunction, increased oxidative stress (where the production of potentially damaging reactive oxygen species [ROS] exceeds endogenous antioxidant capacity) and ultimately vascular injury and remodelling, leading to reduced arterial compliance and vasoconstriction (Duprez, 2007;Gros et al., 2007;Jaffe & Mendelsohn, 2005;Nguyen Dahn Cat et al., 2010;Struthers, 2004).

| Effects on endothelial function
Endothelial dysfunction, characterized by impaired vasodilation, increased platelet and leucocyte adhesion, and decreased nitric oxide bioavailability, occurs secondary to MR activation in experimental rodent studies (Gromotowicz et al., 2011;Oberleithner et al., 2004).
Following MRA treatment, increased eNOS expression occurs and is associated with improved endothelial function and renal blood flow Sanz-Rosa et al., 2005).
Plasma concentrations of asymmetric dimethylarginine (ADMA), an endogenous eNOS inhibitor, are increased in cats with CKD, suggesting that endothelial dysfunction may also occur in this species (Jepson, Syme, Vallance, & Elliott, 2008) although no direct evidence for this has been established.
Chronic MR activation results in structural vascular changes.
Hypertrophic remodelling of renal small arteries occurs in aldosterone-infused rats, an effect inhibited not only by spironolactone but also by endothelin-1 type A (ET A ) receptor antagonism, indicating the likely underlying mechanism (Pu, Neves, Virdis, Touyz, & Schiffrin, 2003). MR blockade improves carotid intima-media remodelling in haemodialysis patients (Vukusich et al., 2010), decreases angiotensin II-mediated cardiac endothelial cell and VSMC hypertrophy (Hatakeyama et al., 1994), cerebral vascular remodelling in strokeprone rats (Rigsby, Pollock, & Dorrance, 2007) and arteriosclerosis in Dahl salt-sensitive rats . Vascular calcification is another feature of MR-induced vasculopathy (Jaffe & Mendelsohn, 2005;Voelkl, Alesutan, Leibrock, Kuro-o, & Lang, 2013); evidence suggests interplay between MR activation and the klotho fibroblast growth factor (FGF)-23 axis, which drives soft tissue and vascular mineralization in CKD-mineral and bone disorder (Voelkl et al., 2013;Zhang et al., 2016). Increased circulating FGF-23 concentrations were the strongest independent predictor of feline CKD progression and all-cause mortality in one study (Geddes, Elliott, & Syme, 2015). Although vascular calcification has not been demonstrated in cats with CKD, mineralization of other tissues occurs and serum calcification propensity (an in vitro assay which predicts vascular calcification in humans) increases with declining renal function (van den Broek, Chang, Elliott, & Jepson, 2018b). As MR activation is likely to contribute to CKD-mineral and bone disorder in cats as in other species, further rationale exists for the use of MRAs in the management of feline CKD. MRA treatment in people with end-stage renal disease is associated with a reduced risk of cerebroand cardiovascular events (Matsumoto et al., 2014) and a reduction in vascular mineralization and stiffness likely accounts for this.

| Effects on blood pressure
Traditionally, aldosterone was believed to increase systemic blood pressure solely by sodium and volume retention. However, it is now known to act directly on the vasculature, as discussed above, and also on the central nervous system (Duprez, 2007;Shavit et al., 2012). Aldosterone potentiates vasopressor-induced vasoconstriction in vitro (Michea et al., 2005; Nguyen Dahn Cat et al., 2010) but has little or no effect on blood pressure or systemic vascular resistance in healthy people (Farquharson & Struthers, 2002;Wehling et al., 1998); it is proposed that counteractive vasodilatory nitric oxide-dependent pathways lost in the presence of endothelial damage attenuate aldosterone's effect on vascular tone (Arima et al., 2004;Uhrenholt et al., 2004).
Sodium and volume retention caused by MR activation contributes to renal damage (including vascular and glomerular sclerosis, tubular damage and inflammation) in rodent experimental models of hypertension (Blasi et al., 2003;Nishiyama et al., 2004;Sun et al., 2006), and protection conferred by MRA blockade can occur partly due to decreases in systolic blood pressure (Du et al., 2009;Martín-Fernández et al., 2016;Zhou et al., 2011). Hypertension is observed in 19%-65% of cats with CKD (Acierno et al., 2018).
Although hypertension has not been independently associated with CKD progression or survival (Chakrabarti et al., 2012;Jepson, Brodbelt, Vallance, Syme, & Elliott, 2009;Syme et al., 2006), it is likely that untreated hypertension results in more severe renal injury and disease progression, as in people (Jamerson & Townsend, 2011). The strong association between hypertension and proteinuria also tends to "mask" significant associations between blood pressure and CKD progression in multivariate models. MRAs are effective in reducing blood pressure in people with CKD and end-stage renal disease (Bianchi et al., 2006;Bolignano, Palmer, Navaneethan, & Strippoli, 2014;Pisoni et al., 2012;Shavit et al., 2012), although some studies have shown no effect, likely due to differences in treatment duration and patient inclusion criteria (Chrysostomou, Pedagogos, MacGregor, & Becker, 2006;Rachmani et al., 2004;Sato et al., 2003Sato et al., , 2005. Hypertensive human CKD patients have more severe renal injury, lower creatinine clearance and higher serum aldosterone concentrations than their normotensive counterparts but interestingly no difference in renal MR or Sgk-1 expression (Quinkler et al., 2005). Plasma aldosterone levels are also increased in hypertensive CKD cats when compared to normotensive cats (Jensen, Henik, & Brownfield, 1997;Jepson, Syme, & Elliott, 2014;Mishina et al., 1998). The first-line treatment for feline hypertension is the calcium channel blocker amlodipine; although amlodipine can cause RAAS activation and aldosterone breakthrough in dogs (Ames, Atkins, Lantis, & Zum Brunnen, 2016), its effect on RAAS in cats is less clear with one study showing increased plasma renin activity but not plasma aldosterone in hypertensive cats postamlodipine treatment compared with pretreatment (Jepson et al., 2014). MRAs may have additional benefits with regard to reducing proteinuria in this population, however, as in people . Hypomagnesaemia is associated with systemic hypertension in cats with CKD (van den Broek, Chang, Elliott, & Jepson, 2018a) and MR activation may provide the link between these factors, as urinary magnesium excretion is stimulated by aldosterone (Barr et al., 1995) and aldosterone secretion is inhibited by increased circulating magnesium levels (Atarashi, Matsuoka, Takagi, & Sugimoto, 1989).

| Ischaemic kidney injury
Renin-angiotensin-aldosterone system activation is both a potential cause and effect of renal hypoxia/ischaemia. RAAS-driven glomerulosclerosis, haemodynamic adaptive alterations and arteriolosclerosis reduce renal capillary oxygen delivery (Hollenberg, 2004;Nangaku, 2006). Uninephrectomy plus ischaemia in rats leads to greater plasma aldosterone levels, hypertension, proteinuria and glomerulosclerosis compared with equivalent surgical reduction alone (Ibrahim & Hostetter, 1998). Sgk-1 expression, indicating MR activation, is upregulated in vitro in human embryonic kidney cells and in vivo in mice exposed to hypoxia (Rusai et al., 2009).
Mineralocorticoid receptor activation has been investigated experimentally in renal ischaemia/reperfusion injury in rodents and the potential therapeutic use of MRAs in this setting is relevant to the hypothesis that renal ischaemia contributes to feline CKD initiation and progression Cowgill et al., 2016;Jepson, 2016). Table 1 Ramírez et al., 2009). Activation of the Rho/Rho-kinase pathway, resulting in calcium-sensitization and smooth muscle contraction, also plays a role in aldosterone's vasoconstrictive and profibrotic effects following renal ischaemia Ramírez et al., 2009;Sun et al., 2006). MRAs likewise provide protection against ischaemic injury in other tissues (Fujita et al., 2005;Oyamada et al., 2008;Ozacmak, Ozacmak, Barut, Arasli, & Ucan, 2014).

Reference Species Model/population Results
Pérez-Rojas et al.

| Oxidative stress
Data suggest that oxidative stress is a central mechanism by which aldosterone/MR activation causes renal damage (

| Renal inflammation and fibrosis
Renal injury induced by aldosterone/MR activation is characterized by heightened inflammation and fibrosis, and MRAs abrogate these changes in both preclinical and clinical studies. Whether aldosterone directly contributes to inflammation and fibrosis or whether these TA B L E 2 Studies investigating the effects of MR antagonism on proteinuria and glomerular damage

Mineralocorticoid antagonist investigated Results
Preclinical studies Aldigier et al. (2005) Rats 5/6 nephrectomy Spironolactone 84% increase in GS index (compared with 157% in controls), GS regression in some rats BP increased despite spironolactone; effects on GS were enhanced when BP was controlled by antihypertensives Asai et al. (2005) Rat model of glomerulonephritis Spironolactone, also looked at the effect of cilazapril (ACEI) Reduced proteinuria (to the same degree as cilazapril) Bamberg et al. Reduced podocyte injury (evidenced by foot process effacement, induction of desmin and attenuation of nephrin) Delayed progression of proteinuria and GS, as did tempol Nishiyama et al. (2004) Rats, aldosterone/salt treatment Eplerenone, also looked at effect of tempol (antioxidant) Reduced proteinuria, as did tempol Nishiyama et al. (2005) Cultured rat mesangial cells Eplerenone Attenuated aldosterone-induced ERK1/2 phosphorylation Prevented the cellular proliferative and deforming effects of aldosterone Nishiyama et al. (2010) Diabetic rats Eplerenone, also looked at the effect of telmisartan (ARB) Decreased proteinuria, GS and podocyte injury Synergistic effect with telmisartan Rocha et al. (2000) AngII and L-NAME treated (nitric oxide synthase inhibitor) and salt-loaded rats occur predominantly secondary to vascular injury is somewhat uncertain, although some experimental data suggest the latter (Rocha et al., 2000). plasminogen activator inhibitor-1 (Brown, Nakamura, et al., 2000), epidermal growth factor and its receptor (Krug et al., 2003;Sheng et al., 2016), matrix metalloproteinase-2 (Martín-Fernández et al., 2016) and TGF-β1 (Fujisawa et al., 2004;Kadoya et al., 2015;Lai et al., 2006;Sun et al., 2006) (2015) Meta-analysis of patients with DN Spironolactone (in addition to ACEI or ARB) Reduced 24-hr urinary albumin/protein excretion and UACR Significantly reduced BP was also reported, therefore proteinuria reduction may have been partly due to BP-lowering effects Pitt et al. (2013) RCT, open-label; heart failure patients with mild or moderate CKD

Finerenone vs. spironolactone
Finerenone was equivalent to spironolactone in decreasing albuminuria Finerenone was associated with a lower incidence of hyperkalaemia and worsening renal function Rachmani et al. (2004) Patients with DN and hypertension Spironolactone, cilazapril or their combination Spironolactone was superior to cilazapril in reducing UACR Co-therapy more effective than either drug alone BP-independent effects Sato et al. (2003) Patients with DN Spironolactone (in addition to ACEI) Reduced urinary albumin excretion by 40% Effect higher in patients with aldosterone breakthrough BP independent Sato et al. (2005) CKD (DN and non-DN, BP controlled) Spironolactone (in addition to ACEI) Reduced urinary albumin excretion, effect greater in diabetic vs. nondiabetic patients (46% vs. 29%) Reduced urinary collagen type IV Tylicki et al.
In people with CKD, renal MR and Sgk-1 expression are positively correlated with TGF-β1 and MCP-1 expression, and serum aldosterone levels with renal fibrosis (Quinkler et al., 2005).
Spironolactone reduces urinary TGF-β1 levels and markers of fibrosis and tubular injury in renal biopsies in this population (Guney et al., 2009;Tylicki et al., 2008) and also urinary type IV collagen in patients with diabetic (Sato et al., 2005) and nondiabetic nephropathy (Furumatsu et al., 2008). A tendency for reduced tubulointerstitial fibrosis was also demonstrated in a small study of paediatric patients with chronic allograft nephropathy receiving eplerenone (Medeiros et al., 2017). Given that the dominant histopathological features of feline CKD are tubulointerstitial fibrosis and inflammation (Chakrabarti et al., 2013), it is proposed that MR blockade in this species would be beneficial in reducing these lesions and resultant disease progression.

| FURTHER COMMENTS ON MR A S IN H UMAN CK D AND END -S TAG E RENAL DISE A SE
It is important to note that although numerous studies have investigated the effect of MRAs in human CKD patients, most have focused on the reduction in proteinuria and hypertension. To date, no studies have evaluated primary end points which allow conclusions to be made about whether MRAs reduce mortality or slow CKD progression. Two small studies have suggested the latter, however, based on a slower decline in estimated GFR (eGFR) compared with control groups (Bianchi et al., 2006;Tylicki et al., 2008). Enrolment is ongoing for a trial designed to evaluate the effect of finerenone on disease progression in patients with diabetic nephropathy (NCT02540993).
Additionally, studies investigating MRA treatment in severe CKD are still limited, although a meta-analysis of dialysis patients found a reduction in mortality with the addition of MRA treatment (Quach et al., 2016). This is proposed to be due to improved cardiac function and reduced cardiovascular events. Table 3 summaries the studies investigating MRAs in the context of cardiovascular outcomes in renal disease.

| Possible adverse effects of MRAs
MRAs have the potential to reduce renal blood flow and GFR. Small decreases in eGFR are not infrequently reported in people receiving MRAs, likely reflecting reversal of hyperfiltration (Bolignano et al., 2014;Pisoni et al., 2012;Schjoedt et al., 2004). Although "worsening renal function" (based on eGFR) was described in large cardiovascular trials, mortality rates remained improved (Pitt et al., 1999(Pitt et al., , 2003Zannad et al., 2011).
Hyperkalaemia is a concern with MRA treatment in human medicine, preventing their prescription in many instances (Maggioni et al., 2013). Individual studies report various effects on plasma potassium concentrations following MR blockade, including no difference between placebo and treatment (Epstein et al., 2006;Gonzales Monte et al., 2010;Sato et al., 2003) and increased hyperkalaemia incidence (Ando et al., 2014;Bianchi et al., 2006;Chrysostomou et al., 2006;Pisoni et al., 2012;Quach et al., 2016;Rachmani et al., 2004). Even though meta-analyses conclude that MRAs (in addition to ACEIs and/or ARBs) increase the risk of hyperkalaemia, the mean increase in potassium levels with treatment is very small compared with placebo (0.26 mM) (Bolignano et al., 2014) and compared with baseline (0.19 mM) (Currie et al., 2016).
Many trials excluded patients with high-normal baseline circulating potassium concentrations, however. Even when statistically significant, increases in serum potassium are deemed "clinically modest," and generally, the benefits of MR blockade are deemed greater than the risk of clinically relevant hyperkalaemia (Pisoni et al., 2012;Pitt et al., 1999Pitt et al., , 2003. Other adverse effects of spironolactone are related to its anti-androgenic and progestogenic properties and include gynaecomastia, impotence, menstrual irregularities and mastalgia (Kolkhof & Borden, 2012;Matsumoto et al., 2014;Pitt et al., 1999;Ponda & Hostetter, 2006). These effects are not reported with eplerenone due to its increased MR selectivity (Ando et al., 2014;Pitt et al., 2003;Zannad et al., 2011) and would not be an issue in treating a cat population which are predominantly neutered.

| ALDOS TERONE/MR AC TIVATI ON IN FELINE CK D
Understanding aldosterone's ability to promote renal injury in laboratory animals and humans provides a convincing basis for its potential role in feline CKD. There is limited information available regarding aldosterone/MR activation in this species. Although reference ranges for plasma aldosterone concentrations have been determined, the pulsatile nature of aldosterone release and effect of diet (sodium and potassium intake) may contribute to large intra-and interindividual variation (Buranakarl, Mathur, & Brown, 2004;Syme et al., 2007;Yu & Morris, 1998). Primary hyperaldosteronism, either due to adrenal gland neoplasia (Ash, Harvey, & Tasker, 2005) or due to hyperplasia (Javadi et al., 2005), is recognized in cats and, as in people, is associated with progressive renal disease and histopathological changes encompassing hyaline arteriosclerosis, glomerulosclerosis and tubulointerstitial fibrosis (Javadi et al., 2005).
As in laboratory species and human patients, RAAS activation is an important factor in the pathogenesis of feline CKD (Ames et al., 2019). Plasma renin, aldosterone, angiotensin I and angiotensin II are increased in cats with experimentally induced CKD following renal ischaemia/reperfusion injury (Watanabe & Mishina, 2007).
RAAS activation is further exacerbated by low sodium intake in this model (Buranakarl et al., 2004) and also occurs in cats with naturally occurring CKD which are transitioned onto (relatively sodium-restricted) renal diets (Syme, 2003). Although experimental data support RAAS activation in feline CKD, it may not directly translate to naturally occurring disease, as plasma renin activity and aldosterone concentrations do not differ between normotensive azotaemic CKD cats and nonazotaemic age-matched controls (Jepson et al., 2014). Mishina et al. (1998) reported increased circulating renin, angiotensin II and aldosterone levels along with increased blood pressure in cats with CKD, although it is unclear whether the groups were age-matched. As in people and rodents, local (intrarenal) RAAS is likely of importance; three studies to date have investigated this using immunohistochemistry in naturally occurring feline CKD (Mitani, Yabuki, Sawa, Chang, & Yamato, 2013;Mitani, Yabuki, Taniguchi, & Yamato, 2013;Taugner, Baatz, & Nobiling, 1996). Renin expression was not associated with azotaemia severity or histopathological lesions (Taugner et al., 1996).
Tubular and interstitial angiotensin II, but not ACE or ACE2 expression, was correlated with glomerulosclerosis and tubulointerstitial inflammation (Mitani, Yabuki, Sawa, et al., 2013;Mitani, Yabuki, Taniguchi, et al., 2013). Intrarenal aldosterone has not been examined, although assessment of renal 11β-HSD activity has been attempted by urinary cortisol-cortisone ratio measurement; cats with CKD had a lower ratio, not supportive of the hypothesis that decreased excretion of active glucocorticoid may potentially reflect excessive MR stimulation in this population (Walker, Elliott, & Syme, 2009).
Aldosterone appears to be associated with feline systemic hypertension, a common finding in cats with CKD. Plasma aldosterone levels are higher in hypertensive azotaemic cats than nonhypertensive cats with and without renal disease (Jensen et al., 1997;Jepson et al., 2014). Lower plasma potassium tends to be a risk factor for feline hypertension in epidemiological studies, providing support for MR activation (Jepson et al., 2009;Sansom, Rogers, & Wood, 2004;Syme, Barber, Markwell, & Elliott, 2002), although blood pressure is not directly associated with plasma or urinary aldosterone concentrations (Syme, Barber, et al., 2002;Syme et al., 2007;Williams et al., 2013). Increased plasma aldosterone concentration is not seemingly driven by plasma renin activity, as cats with concurrent CKD and hypertension have variable or decreased renin compared with controls, resulting in increased aldosterone-to-renin ratios (Jensen et al., 1997;Jepson et al., 2014;. Given that increased circulating aldosterone in cats with concurrent CKD and hypertension does not appear to be secondary to increased renin or hyperkalaemia, alternative explanatory mechanisms include primary adrenal-dependent pathology, local MR activation, altered sensitivity to stimuli which dictate aldosterone release or reduced aldosterone degradation (Buranakarl et al., 2004).

| MRA use in cats
The optimal way to inhibit RAAS activation in feline CKD has yet to be determined, and to date, treatment has consisted of ACEI and/or ARB therapy. In many countries, the ACEI benazepril is licensed for treating proteinuria associated with CKD in cats and the ARB, telmisartan, is licensed for feline hypertension and proteinuria treatment (Coleman et al., 2019;Glaus, Elliott, Herberich, Zimmering, & Albrecht, 2019). Benazepril ameliorates glomerular capillary hypertension, increases GFR and reduces proteinuria in a partial renal ablation model (Brown et al., 2001), and reduces proteinuria in naturally occurring CKD (King, Gunn-Moore, Tasker, Gleadhill, & Strehlau, 2006;Watanabe & Mishina, 2007).
Telmisartan is as efficacious as benazepril in reducing urine protein/creatinine ratio in clinical cases (Sent, Gössl, Elliott, Syme, & Zimmering, 2015). Although ACEIs and ARBs successfully reduce proteinuria, a factor associated with reduced survival (King et al., 2007;Kuwahara et al., 2006;Syme et al., 2006), the present studies investigating these drugs in feline CKD have important limitations (e.g. are underpowered or not designed to test longterm outcomes) which prevent definitive conclusions from being made about their effect on CKD progression and prognosis in cats (King et al., 2006;Sent et al., 2015;Watanabe & Mishina, 2007).
Aldosterone breakthrough has not been studied in cats with CKD receiving long-term ACEIs or ARBs.
Two studies have investigated spironolactone in feline cardiac disease. Relevant to CKD pathology, feline hypertrophic cardiomyopathy is characterized by significant interstitial fibrosis and arteriosclerosis (Fox, 2003). In a small study of hypertrophic cardiomyopathy in Maine Coons, four of 13 treated cats developed severe ulcerative facial dermatitis approximately 2.5 months into treatment which the authors attributed to spironolactone (MacDonald & Kittleson, 2008). The dosage used in this study (2 mg/kg twice daily) was twice the recommended dosage in dogs (Guyonnet, Elliott, & Kaltsatos, 2010), and feline herpesvirus was not sufficiently ruled out as a possible cause. One cat also developed myelodysplasia. Cutaneous drug reactions are sporadically reported in people receiving spironolactone (Gupta, Knowles, & Shear, 1994) and spironolactone-induced agranulocytosis, and aplastic anaemia is also recognized (Ibáñez, Vidal, Ballarín, & Laporte, 2005). A second study reported no dermatological adverse effects of spironolactone (1.7-3.3 mg/kg once daily) over a 15-month treatment period, and the prevalence of adverse events was similar between the treatment and placebo groups (James et al., 2018). The risk of hyperkalaemia with ACEI and spironolactone co-therapy is emphasized in veterinary medicine, although combination therapy appears well-tolerated in cats and dogs with heart failure (James et al., 2018; Lefebvre et al., 2013).

| CON CLUS IONS
Given the expanding evidence base from in vitro and in vivo experimental studies and from human medicine, it seems likely that aldosterone and MR activation is an important player in the pathogenesis of feline CKD. It must be noted, however, that experimental models may not be directly translatable to the clinical situation and that differences in CKD pathogenesis exist between humans and cats. Furthermore  None of the authors has any other financial or personal relationships that could inappropriately influence or bias the content of the paper. CEVA Animal Health, who market spironolactone for the treatment of canine congestive heart failure caused by valvular regurgitation, played no role in the preparation of this manuscript.

AUTH O R CO NTR I B UTI O N
SS, CWJ and JE were responsible for the writing of this manuscript and have read and approved the final manuscript.