Natriuretic Peptide Receptor Guanylyl Cyclase-A Protects Podocytes from Aldosterone-Induced Glomerular Injury

Abstract

Natriuretic peptides produced by the heart in response to cardiac overload exert cardioprotective and renoprotective effects by eliciting natriuresis, reducing BP, and inhibiting cell proliferation and fibrosis. These peptides also antagonize the renin-angiotensin-aldosterone system, but whether this mechanism contributes to their renoprotective effect is unknown. Here, we examined the kidneys of mice lacking the guanylyl cyclase-A (GC-A) receptor for natriuretic peptides under conditions of high aldosterone and high dietary salt. After 4 weeks of administering aldosterone and a high-salt diet, GC-A knockout mice, but not wild-type mice, exhibited accelerated hypertension with massive proteinuria. Aldosterone-infused GC-A knockout mice had marked mesangial expansion, segmental sclerosis, severe podocyte injury, and increased oxidative stress. Reducing the BP with hydralazine failed to lessen such changes; in contrast, blockade of the renin-angiotensin-aldosterone system markedly reduced albuminuria, ameliorated podocyte injury, and reduced oxidative stress. Furthermore, treatment with the antioxidant tempol significantly reduced albuminuria and abrogated the histologic changes. In cultured podocytes, natriuretic peptides inhibited aldosterone-induced mitogen-activated protein kinase phosphorylation. Taken together, these results suggest that renoprotective properties of the endogenous natriuretic peptide/GC-A system may result from the local inhibition of the renin-angiotensin-aldosterone system and oxidative stress in podocytes.

Aldosterone, beyond its effects on sodium reabsorption, has been shown to play an important role in causing cardiovascular complications including renal injury.^1,2 Proteinuria is not rare in patients with primary aldosteronism, suggesting that aldosterone can induce renal injury independent of circulating angiotensin II (AngII) levels.³ In experimental nephropathy models, aldosterone exerts direct effects on renal cells, leading to the progression of glomerulosclerosis and proteinuria.⁴ Several studies have revealed positive feedback by which aldosterone activates the renin-angiotensin-aldosterone system (RAAS), through inducing renin and angiotensin-converting enzyme gene expression,^5,6 or by activating AngII type 1 (AT1) receptor signaling.⁷ Furthermore, aldosterone upregulates reactive oxygen species (ROS) and activates mitogen-activated protein kinases (MAPKs) in the kidney and cardiovascular system.^7–9

Podocytes play a crucial role in barrier function as well as the pathogenesis of glomerular diseases, forming a branched interdigitating network with foot processes by the slit diaphragm.¹⁰ Genetic studies have proved roles of slit diaphragm–associated proteins, including nephrin and podocin, in various proteinuric disorders.¹⁰ Recently, podocytes have attracted greater attention as the target for aldosterone action. In fact, aldosterone causes foot process effacement and downregulation of nephrin and podocin, via the mineralocorticoid receptor.^7,11–13 In addition, the mineralocorticoid receptor in podocytes can be activated by Rac1 GTPase, which is sufficient to cause glomerular injury and proteinuria irrespective of aldosterone levels.¹³

The natriuretic peptide family consisting of atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP), and C-type natriuretic peptide possess potent diuretic, natriuretic, and vasodilating properties.¹⁴ ANP and BNP are secreted predominantly by the cardiac atrium and ventricle, respectively, upon cardiac overload.^14–16 They exert various biologic effects by acting on guanylyl cyclase-A (GC-A)/natriuretic peptide receptor-A (NPR-A), which is the major and possibly the only receptor for ANP and BNP, through the activation of cGMP/cGMP-dependent protein kinase.¹⁷ GC-A is abundantly expressed in blood vessels and the heart¹⁷; within the kidney, it is localized in glomeruli, thin limbs of Henle’s loop, cortical collecting ducts, and inner medullary collecting ducts.¹⁸ In glomeruli, GC-A is expressed in mesangial cells and podocytes.¹⁸ GC-A–deficient mice show chronic salt-independent elevation of BP by approximately 15–30 mmHg¹⁹ and cardiac hypertrophy,²⁰ with virtually no natriuretic or diuretic response to acute volume expansion.²¹ These studies clearly demonstrate essential roles for natriuretic peptide/GC-A signaling in BP regulation and acute volume handling through the kidney, but the role of GC-A in podocyte injury remains elusive.

A renoprotective action of ANP has been shown early using animal models of ARF.²² A recent meta-analysis reveals that the administration of low-dose ANP may exert beneficial effects in clinical AKI.²³ We previously showed that chronic excess of BNP in mice prevents glomerular injury after subtotal nephrectomy,²⁴ and ameliorates proteinuria and histologic changes in immune-mediated renal injury²⁵ as well as in diabetic nephropathy.²⁶ In addition to exerting direct vasodilating and diuretic actions, natriuretic peptides act to antagonize the RAAS at multiple steps.^14,27,28 We postulate therefore that the beneficial effects of natriuretic peptides should be brought about, at least in part, by antagonizing local activation and/or action of the RAAS in the kidney.^24–26

To explore the role of natriuretic peptide/GC-A signaling in aldosterone-induced renal injury, we investigated renal findings of mice deficient in GC-A, along with the challenge of aldosterone and a high-salt diet.

Results

Aldosterone Infusion Causes Accelerated Hypertension in GC-A Knockout Mice

All mice were uninephrectomized and fed with 6% NaCl, with or without aldosterone infusion (0.2 μg/kg body weight per minute) for 4 weeks. Changes in systolic BP (SBP) in each group are shown in Figure 1 (time courses shown in Supplemental Figure 1). Aldosterone infusion in wild-type mice resulted in marginally higher SBP than vehicle (118.1±3.3 versus 116.6±4.5 mmHg at 4 weeks). GC-A knockout mice showed a significantly higher SBP compared with wild-type mice at baseline (126.2±2.7 versus 110.4±3.3 mmHg, P<0.05), and revealed marked hypertension after aldosterone infusion compared with vehicle (159.0±6.6 versus 123.0±4.5 mmHg at 4 weeks, P<0.01). Administration of hydralazine, spironolactone, or olmesartan in aldosterone-infused GC-A knockout mice resulted in reduced SBP to the same degree (135.7±3.1, 135.2±8.0, and 140.0±5.2 mmHg, respectively). In contrast, there was no significant SBP change with tempol (151.1±6.9 mmHg).

Figure 1: The comparison in SBP at 4 weeks after aldosterone or vehicle infusion. Data are from vehicle-infused wild-type mice (WT veh, n=5), aldosterone-infused wild-type mice (WT aldo, n=8), vehicle-infused GC-A knockout mice (GC-A KO veh, n=5), aldosterone-infused GC-A knockout mice (GC-A KO aldo, n=8), aldosterone-infused GC-A knockout mice treated with hydralazine (KO aldo+hyd, n=6), those treated with spironolactone (KO aldo+spi, n=5), those treated with olmesartan (KO aldo+olm, n=5), and those treated with tempol (KO aldo+temp, n=5). Aldosterone-infused GC-A knockout mice exhibited marked hypertension, and administration of hydralazine, spironolactone, or olmesartan in aldosterone-infused GC-A knockout mice resulted in reduced SBP to the same degree. In contrast, there was no significant SBP change with tempol. *P<0.05, **P<0.01, versus KO aldo, ^† P<0.05. WT, wild-type; veh, vehicle; aldo, aldosterone; KO, knockout; hyd, hydralazine; spi, spironolactone; olm, olmesartan; temp, tempol.

Body and Kidney Weights and Blood Parameters

Body weights and kidney weights are presented in Supplemental Table 1. GC-A knockout mice showed normal kidney weight at baseline but exhibited renal hypertrophy compared with wild-type mice as indicated by an increase in kidney weight per body weight at 4 weeks. Aldosterone infusion caused renal hypertrophy in both wild-type and GC-A knockout mice, and administration of spironolactone ameliorated such changes in GC-A knockout mice.

Serum aldosterone was markedly high (approximately 70-fold) and serum potassium was low in all aldosterone-infused groups (Supplemental Table 1). There was no significant difference in serum aldosterone, potassium, or creatinine levels among aldosterone-infused groups.

Aldosterone Causes Massive Proteinuria in GC-A Knockout Mice

At the basal level (−2 weeks), urinary albumin excretion was not different between wild-type and GC-A knockout mice (38.7±4.9 and 41.4±4.8 μg/mgCr, respectively) (Figure 2A). Aldosterone-infused wild-type mice showed a three-fold increase in urinary albumin excretion at 4 weeks compared with vehicle (129.2±25.4 versus 43.0±6.6 μg/mgCr, P<0.01). Surprisingly, aldosterone-infused GC-A knockout mice revealed 260 times higher urinary albumin excretion than vehicle-infused GC-A knockout mice at 4 weeks (17,559±6845 μg/mgCr versus 67.2±24.4 μg/mgCr, P<0.01). Administration of hydralazine in these mice significantly reduced BP, but failed to suppress albuminuria (Figure 2B). In contrast, spironolactone administration markedly reduced urinary albumin excretion by 95% in aldosterone-infused GC-A knockout mice (1069±411 μg/mgCr). Moreover, treatment with an angiotensin receptor blocker (ARB) olmesartan significantly reduced urinary albumin excretion by 70% (5450±1765 μg/mgCr). Treatment with tempol also reduced albuminuria by 60% (7125±2292 μg/mgCr). These results suggest that, in the absence of GC-A signaling, aldosterone stimulation can activate the renin-angiotensin system and oxidative stress in the kidney, leading to massive proteinuria.

Figure 2: Urinary albumin excretion in wild-type or GC-A knockout mice with or without aldosterone. (A) Aldosterone infusion resulted in a marked increase of albuminuria in GC-A knockout mice. Data from wild-type mice with vehicle (WT veh, open triangles), wild-type mice with aldosterone (WT aldo, filled triangles), GC-A knockout mice with vehicle (KO veh, open squares), and GC-A knockout mice with aldosterone (KO aldo, filled squares) are shown. (B) Administration of spironolactone (spi, open circles), olmesartan (olm, filled circles), or tempol (temp, filled diamonds) significantly reduced albuminuria in aldosterone-infused GC-A knockout mice. Hydralazine (hyd, open diamonds) did not reduce albuminuria. WT veh, n=5; WT aldo, n=8; KO veh, n=5; KO aldo, n=8; KO aldo+hyd, n=6; KO aldo+spi, n=5; KO aldo+olm, n=5; and KO aldo+temp, n=5. **P<0.01, KO veh versus KO aldo, ^† P<0.05, ^†† P<0.01, versus KO aldo.

Renal Histologic Changes in Aldosterone-Infused GC-A Knockout Mice

We examined renal histology at 4 weeks after aldosterone administration. In superficial glomeruli, aldosterone-infused wild-type mice exhibited marginal mesangial expansion with mild glomerular hypertrophy (Figure 3A). Renal histology of GC-A knockout mice showed virtually no significant difference from that of wild-type mice at baseline; after aldosterone infusion, GC-A knockout mice exhibited marked mesangial expansion with glomerular hypertrophy (Figure 3, A and B and Supplemental Figure 2A). Although hydralazine administration failed to ameliorate such changes, treatment with spironolactone, olmesartan, or tempol resulted in inhibited mesangial expansion in these mice.

Figure 3: Histologic examination of glomeruli in aldosterone-infused GC-A knockout mice and the localization of GC-A. Light microscopic analyses were performed at 4 weeks after aldosterone administration, stained with periodic acid-Schiff. (A) Representative views of superficial glomeruli. Aldosterone-infused wild-type mice (WT aldo) showed marginal mesangial expansion with mild glomerular hypertrophy. Aldosterone-infused GC-A knockout mice (GC-A KO aldo) exhibited marked mesangial expansion with glomerular hypertrophy. Administration of hydralazine in GC-A knockout mice (GC-A KO aldo+hyd) did not affect glomerular changes. Administration of spironolactone (GC-A KO aldo+spi), olmesartan (GC-A KO aldo+olm), or tempol (GC-A KO aldo+temp) in GC-A knockout mice improved histologic changes. (B) Mesangial area in superficial glomeruli at 4 weeks. (C) Representative views of juxtamedullary glomeruli. Aldosterone-infused GC-A knockout mice showed severe segmental sclerosis and marked glomerular hypertrophy. Administration of spironolactone, olmesartan or tempol ameliorated glomerular changes. (D) Mesangial area in juxtamedullary glomeruli at 4 weeks. (E) Degree of glomerulosclerosis in juxtamedullary glomeruli. (F) Immunohistochemical study for GC-A. GC-A was positive at podocytes (arrows) and distal tubules (arrowheads). WT veh, n=5; WT aldo, n=8; KO veh, n=5; KO aldo, n=8; KO aldo+hyd, n=6; KO aldo+spi, n=5; KO aldo+olm, n=5; and KO aldo+temp, n=5. **P<0.01, KO veh versus KO aldo, ^† P<0.05, ^†† P<0.01, versus KO aldo. WT, wild-type; veh, vehicle; aldo, aldosterone; KO, knockout; hyd, hydralazine; spi, spironolactone; olm, olmesartan; temp, tempol.

In juxtamedullary glomeruli, there were minor glomerular abnormalities in aldosterone-infused wild-type mice; in contrast, aldosterone-infused GC-A knockout mice exhibited severe segmental sclerosis and marked glomerular hypertrophy (Figure 3C). Again, these changes were unaltered with hydralazine but were alleviated with spironolactone, olmesartan, or tempol (Figure 3D and Supplemental Figure 2B). The number of sclerotic glomeruli was also significantly reduced with spironolactone, olmesartan, or tempol (Figure 3E). Localization of GC-A was examined by immunohistochemistry. Wild-type mice revealed the presence of GC-A presumably at podocytes, distal tubules, and collecting ducts, which was apparently unaltered with aldosterone infusion (Figure 3F).

We next examined renal fibrotic changes in these mice (Figure 4). Whereas wild-type mice with aldosterone showed slight fibrotic changes with mild tubular atrophy, aldosterone-infused GC-A knockout mice exhibited tubular dilation with marked protein cast deposition and significant tubulointerstitial fibrosis. These changes were not changed with hydralazine but ameliorated with spironolactone, olmesartan, and less effectively with tempol (Figure 4, A and B).

Figure 4: Histologic examination of tubulointerstitial fibrosis by Masson’s trichrome-stained sections at 4 weeks. (A) Representative views of tubules and interstitial fibrosis. Aldosterone-infused wild-type mice (WT aldo) showed mild tubulointerstitial fibrosis with tubular atrophy. Aldosterone-infused GC-A knockout mice (GC-A KO aldo) exhibited tubular dilation with protein casts and tubulointerstitial fibrosis. These changes were not improved with hydralazine (GC-A KO aldo+hyd), but ameliorated with spironolactone (GC-A KO aldo+spi), olmesartan (GC-A KO aldo+olm), or tempol (GC-A KO aldo+temp). (B) Fibrotic area at 4 weeks. WT veh, n=5; WT aldo, n=8; KO veh, n=5; KO aldo, n=8; KO aldo+hyd, n=6; KO aldo+spi, n=5; KO aldo+olm, n=5; and KO aldo+temp, n=5. ^# P<0.05, WT veh versus WT aldo, **P<0.01, KO veh versus KO aldo, ^† P<0.05, ^†† P<0.01, versus KO aldo. WT, wild-type; veh, vehicle; aldo, aldosterone; KO, knockout; hyd, hydralazine; spi, spironolactone; olm, olmesartan; temp, tempol.

Podocyte Injury in Aldosterone-Infused GC-A Knockout Mice

We then evaluated podocyte injury in these mice. Electron microscopic analyses revealed that wild-type mice with aldosterone exhibited slightly widened podocyte foot processes without thickening of the glomerular basement membrane (GBM) (Figure 5). Vehicle-infused GC-A knockout mice showed thickened GBM without widening of foot processes. GC-A knockout mice with aldosterone showed foot process effacement with irregular thickening of GBM, and theses changes were not improved with hydralazine, but reversed with spironolactone and also lessened with olmesartan (Figure 5, A and B). Tempol administration improved width of foot processes.

Figure 5: Electron microscopic analyses of glomeruli in aldosterone-infused GC-A knockout mice at 4 weeks. (A) Aldosterone-infused wild-type mice (WT aldo) showed almost normal foot process structure. Vehicle-infused GC-A knockout mice (GC-A KO veh) had normal podocyte foot processes. Aldosterone-infused GC-A knockout mice (GC-A KO aldo) showed foot process effacement with irregular thickening of glomerular basement membrane. These changes were not improved with administration of hydralazine (GC-A KO aldo+hyd), but ameliorated with spironolactone (GC-A KO aldo+spi), olmesartan (GC-A KO aldo+olm), or tempol (GC-A KO aldo+temp). (B) GBM thickness and foot process width (n=3, each). ^# P<0.05, WT veh versus WT aldo, **P<0.01, KO veh versus KO aldo, ^† P<0.05, ^†† P<0.01, versus KO aldo, ^¶ P<0.05, KO veh versus WT veh. WT, wild-type; veh, vehicle; aldo, aldosterone; KO, knockout; hyd, hydralazine; spi, spironolactone; olm, olmesartan; temp, tempol.

The expression of podocyte markers nephrin and podocin was decreased in aldosterone-infused GC-A knockout mice at 4 weeks with or without hydralazine, and such changes were ameliorated with spironolactone, olmesartan, or tempol (Figure 6, A and B).

Figure 6: Immunostaining for nephrin and podocin. (A) Aldosterone-infused wild-type mice (WT aldo) showed normal linear staining for nephrin and podocin. Aldosterone-infused GC-A knockout mice (GC-A KO aldo) exhibited decreased expression of nephrin and podocin at 4 weeks. The expression of both nephrin and podocin was recovered by administration with spironolactone (GC-A KO aldo+spi), olmesartan (GC-A KO aldo+olm), or tempol (GC-A KO aldo+temp), but not with hydralazine (GC-A KO aldo+hyd). (B) Quantification of the mean intensity and the area of fluorescent staining for nephrin and podocin, and the products of multiplying the intensity and the area are shown. WT veh, n=5; WT aldo, n=8; KO veh, n=5; KO aldo, n=8; KO aldo+hyd, n=6; KO aldo+spi, n=5; KO aldo+olm, n=5; and KO aldo+temp, n=5. #P<0.05, WT veh versus WT aldo, **P<0.01, KO veh versus KO aldo, ^† P<0.05, ^†† P<0.01, versus KO aldo, ^¶ P<0.05, KO veh versus WT veh. WT, wild-type; veh, vehicle; aldo, aldosterone; KO, knockout; hyd, hydralazine; spi, spironolactone; olm, olmesartan; temp, tempol.

Gene Expression and ROS in Glomeruli of Aldosterone-Infused GC-A Knockout Mice

Analyses on glomerular expression of extracellular matrix (ECM)–related genes (Figure 7A) revealed that TGF-β1 mRNA was enhanced with aldosterone both in wild-type and in GC-A knockout mice. Such upregulation was significantly reduced with spironolactone or tempol, and tended to decrease with olmesartan treatment. Similar tendency was observed in connective tissue growth factor, fibronectin, and collagen 1 and 4 mRNA expressions (Figure 7A and Supplemental Figure 3).

Figure 7: Glomerular mRNA expression at 4 weeks after aldosterone administration and ROS. (A, B) Real-time RT-PCR analyses of TGF-β1 (Tgfb1), connective tissue growth factor (Ctgf), fibronectin (Fn1), and gp91phox (Cybb) are shown. GAPDH mRNA expression is used as the control. (C) Immunohistochemical study for 8-OHdG. Aldosterone-infused wild-type mice (WT aldo) did not show the increase of 8-OHdG. Strong staining of 8-OHdG was detected in vehicle-infused GC-A knockout mice (GC-A KO veh), and aldosterone infusion (GC-A KO aldo) augmented 8-OHdG staining mainly at podocytes and tubular cells. Staining of 8-OHdG was reduced by administration with spironolactone (GC-A KO aldo+spi), olmesartan (GC-A KO aldo+olm), or tempol (GC-A KO aldo+temp), but not with hydralazine (GC-A KO aldo+hyd). (D) Quantification of the mean intensity of staining for 8-OHdG. WT veh, n=5; WT aldo, n=8; KO veh, n=5; KO aldo, n=8; KO aldo+hyd, n=6; KO aldo+spi, n=5; KO aldo+olm, n=5; and KO aldo+temp, n=5. ^# P<0.05, WT veh versus WT aldo,*P<0.05, **P<0.01, KO veh versus KO aldo, ^† P<0.05, ^†† P<0.01, versus KO aldo, ^¶ P<0.05, KO veh versus WT veh. WT, wild-type; veh, vehicle; aldo, aldosterone; KO, knockout; hyd, hydralazine; spi, spironolactone; olm, olmesartan; temp, tempol.

We next examined the expression of NADPH oxidase 2 (Nox-2) or gp91phox/Cybb, p22phox/Cyba, and Nox-4, which are essential membrane components of NADPH oxidase.²⁹ Aldosterone infusion upregulated the glomerular expression of gp91phox/Cybb mRNA, and less potently that of p22phox and Nox-4 (Figure 7B and Supplemental Figure 3). Treatment with spironolactone or tempol significantly reduced them. Immunohistochemical study for 8-hydroxydeoxyguanosine (8-OHdG) showed that aldosterone infusion exhibited strong staining mainly at podocytes and tubular cells only in GC-A knockout mice (Figure 7, C and D). Such upregulation was significantly reduced with spironolactone, olmesartan, or tempol, but not with hydralazine.

Enhanced Phosphorylation of MAPKs in Aldosterone-Infused GC-A Knockout Mice

We found a mild increase of phosphorylated extracellular signal-regulated kinase (ERK) in glomeruli of aldosterone-infused wild-type mice (Figure 8A). Phosphorylation of ERK in glomeruli was pronounced in aldosterone-infused GC-A knockout mice. Double immunostaining revealed that the cells expressing phospho-ERK were also positive for Wilms’ tumor 1 (WT1), a podocyte marker (Figure 8B). Phosphorylation of ERK was reduced by treatment with spironolactone, olmesartan, or tempol, but not with hydralazine (Figure 8, C and D). Essentially similar results were obtained as to phospho-p38 MAPK-positive cells, which were double stained with WT1 (Figure 8, E and F); the phosphorylation of p38 MAPK was reduced with spironolactone or olmesartan, and also with tempol (Figure 8, E, G, and H).

Figure 8: Enhanced phosphorylation of ERK and p38 MAPK in glomeruli of aldosterone-infused GC-A knockout mice. (A) Immunohistochemical study for phospho-ERK at 4 weeks after aldosterone infusion. (B) Double immunostaining for phospho-ERK (brown) and Wilms’ tumor 1 (WT1, blue). Arrows indicate double positive cells. (C) Western blotting for phospho-ERK and total ERK in glomeruli of mice. (D) Quantification of relative phospho-ERK to total ERK levels analyzed with densitometer. (E) Immunohistochemical study for phospho-p38 MAPK at 4 weeks after aldosterone infusion. (F) Double immunostaining for phospho-p38 MAPK (brown) and WT1 (blue). Arrows indicate double positive cells. (G) Western blotting for phospho-p38 MAPK and total p38 MAPK in glomeruli of mice. (H) Quantification of relative phospho-p38 MAPK to total p38 MAPK levels analyzed with densitometer. n=3, each. ^## P<0.01, WT veh versus WT aldo, **P<0.01, KO veh versus KO aldo, ^† P<0.05, ^†† P<0.01, versus KO aldo, ^¶¶ P<0.01, KO veh versus WT veh. WT, wild-type; veh, vehicle; aldo, aldosterone; KO, knockout; hyd, hydralazine; spi, spironolactone; olm, olmesartan; temp, tempol; WV, wild-type mice with vehicle; WA, wild-type mice with aldosterone; KV, GC-A KO mice with vehicle; KA, GC-A KO mice with aldosterone; KH, GC-A KO mice with aldosterone treated with hydralazine; KS, GC-A KO mice with aldosterone treated with spironolactone; KO, GC-A KO mice with aldosterone treated with olmesartan; KT, GC-A KO mice with aldosterone treated with tempol.

ANP Inhibits Phosphorylation of ERK and p38 MAPK in Cultured Mouse Podocytes

We examined the effect of ANP on phosphorylation of ERK and p38 MAPK in cultured mouse podocytes. Aldosterone caused the phosphorylation of ERK as quickly as 10 minutes after the stimulation and lasted for as long as 30 minutes (Figure 9A). Pretreatment with ANP completely abolished ERK phosphorylation (Figure 9B). The phosphorylation of p38 MAPK was upregulated at 3 hours after aldosterone stimulation, and such induction was significantly inhibited with ANP treatment (Figure 9, C and D).

Figure 9: Aldosterone-induced phosphorylation of ERK and p38 MAPK in cultured mouse podocytes. (A) Time course of phospho-ERK and total ERK protein after administration of aldosterone (1 μM) in cultured podocytes. *P<0.01, versus 0 minutes. n=5. (B) Inhibitory effects of ANP (1 μM) on aldosterone-induced ERK phosphorylation. ANP (aldo+ANP) or vehicle (aldo) was administered 30 minutes before aldosterone or vehicle (cont) stimulation (1 μM). Cells were harvested at 10 minutes after aldosterone stimulation. ^# P<0.01. n=5. (C) Time course of phospho-p38 MAPK and total p38 MAPK protein after administration of aldosterone (1 μM) in cultured podocytes. *P<0.01, versus 0 minutes. n=5. (D) Inhibitory effects of ANP (1 μM) on aldosterone-induced p38 MAPK phosphorylation at 3 hours. Mean ± SEM *P<0.01, versus 0 minutes. ^# P<0.01. n=5. *P<0.01, versus 0 minutes. ^# P<0.01. n=5. (E) Inhibitory effects of MEK inhibitor or p38 MAPK inhibitor on connexin43 (Gja1) mRNA expression in mouse podocytes. U. U0126 (10 μM); SB, SB203580 (10 μM). **P<0.01 versus veh. ^## P<0.01 versus aldosterone. n=3. (F) Glucocorticoid receptor blocker, mifepristone, did not reduce aldosterone-induced connexin43 expression. Spi, spironolactone (10 μM), mifepristone (10 μM). *P<0.05, **P<0.01 versus veh, ^# P<0.05 versus aldosterone. n=3. WT, wild-type; veh, vehicle; aldo, aldosterone; KO, knockout; hyd, hydralazine; spi, spironolactone; olm, olmesartan; temp, tempol.

GC-A knockdown in podocytes showed higher levels of phosphorylated ERK and p38 MAPK than controls (Supplemental Figure 4). Treatment with olmesartan or spironolactone, but not tempol, partially decreased phospho-ERK; cGMP analog strongly reduced phosphorylation of both ERK and p38 MAPK (Supplemental Figure 4). Cultured podocytes stimulated with aldosterone upregulated expression of connexin43 (Gja1), a podocyte injury marker. This upregulation was significantly inhibited by the treatment with MAPK kinase (MEK) inhibitor U0126 or p38 MAPK inhibitor SB203580 (Figure 9E). Finally, we confirmed whether aldosterone action was mediated through a mineralocorticoid receptor. Upregulation of connexin43 mRNA in aldosterone-stimulated podocytes was not reduced by glucocorticoid receptor blocker mifepristone, but by spironolactone (Figure 9F).

Discussion

We investigated the role of the natriuretic peptide/GC-A system in aldosterone-induced renal injury. Although aldosterone administration in wild-type mice resulted in minor glomerular abnormalities with marginal increase of albuminuria, GC-A–deficient mice exhibited accelerated hypertension and severe glomerulopathy with massive proteinuria, indicating that GC-A signaling should normally act to inhibit aldosterone-induced glomerular injury. We must be cautious to interpret BP by tail-cuff manometry, because it is not as precise as direct monitoring. Nevertheless, these actions were not merely BP dependent, and we therefore hypothesized that GC-A signaling works at podocytes in this model. In fact, aldosterone exerted marked podocyte injury, only in the absence of GC-A.

Besides massive proteinuria, aldosterone caused glomerulosclerosis and interstitial fibrosis in a mineralocorticoid receptor-dependent fashion (Figures 2–4). Spironolactone treatment with modest BP reduction (Figure 1) almost completely abolished these abnormalities, suggesting that aldosterone acts perhaps via the receptor on the podocytes and mesangium in addition to that in tubules.^{7–9,11–13} To note, an ARB olmesartan markedly ameliorated proteinuria and glomerular/podocyte injuries (Figures 2–6), suggesting that aldosterone activation of the RAAS in the kidney may have a causative role, especially in the absence of GC-A signaling. Several studies have provided evidence for positive feedback between aldosterone and the RAAS,^5–7 and GC-A knockout mice showed augmented angiotensin-converting enzyme and AT1 mRNA expression.³⁰ Activated AT1 signaling at podocytes would be sufficient to cause proteinuria and glomerular injury.^31,32 Furthermore, the crosstalk between AngII and GC-A signaling was shown in cultured podocytes.³³ These data suggest that lack of GC-A in the kidney should have a critical role in exaggerated activation of the “local” RAAS, which was abrogated by ARB treatment.

Aldosterone increases ROS production in the kidney.^7,8,12 This study reveals that the glomerular expression of gp91phox, a prototype of Nox family,²⁹ was upregulated in aldosterone-infused GC-A knockout mice (Figure 7). Treatment with tempol, a membrane-permeable radical scavenger,⁸ inhibited such increases together with significant amelioration in nephropathy (Figures 2–6), suggesting the importance of ROS in this model. Thus far, little is known on the relationship between natriuretic peptides and ROS in renal injury. It has been reported that ANP counteracts ROS generation in aortic smooth muscle cells,³⁴ and inhibits ROS-induced cell damage via the GC-A/cGMP pathway.³⁵ The role of ROS in renal injury has been extensively studied,³⁶ and redox control is crucial in the pathophysiology of human and rodent diabetic nephropathy.^37,38 In this study, aldosterone caused increased glomerular expression of ROS- and ECM-related genes both in knockout and control mice. However, augmented 8-OHdG, massive proteinuria, and fibrotic changes occurred only in GC-A knockout mice, suggesting that, in wild-type mice, GC-A signaling may hamper intracellular mechanisms downstream of ROS signal against disease progression.

ROS-induced cell injury is mediated in part by the activation of MAPKs.^8,9 p38 MAPK plays a critical role in inflammation,³⁹ cytoskeleton stability,⁴⁰ and podocyte function.⁴¹ We previously showed that inhibition of p38 MAPK markedly ameliorates podocyte injury and proteinuria in rodent models of nephrotic syndrome.⁴¹ In this study, knockout mice with aldosterone exhibited augmented phosphorylation of p38 MAPK in podocytes, which was inhibited by RAAS blockade as well as tempol (Figure 8). In addition, ANP suppressed p38 MAPK phosphorylation, which was increased by GC-A knockdown in cultured podocytes (Figure 9, Supplemental Figure 4). Moreover, upregulation of connexin43, one of the earliest podocyte injury markers,⁴² was induced by aldosterone and reduced by p38 MAPK inhibition (Figure 9E). It is thus conceivable that GC-A deficiency facilitated aldosterone- and ROS-induced activation of p38 MAPK, causing pronounced podocyte damage. It has been reported that ANP inhibits ROS-mediated p38 MAPK activation in lung endothelial cells.⁴³

This study also demonstrated the enhanced ERK phosphorylation at podocytes in aldosterone-infused GC-A knockout mice, which was attenuated with RAAS blockade. We previously showed that natriuretic peptides prevent glomerular ERK activation in anti-GBM GN²⁵ and ameliorate AngII-induced cardiac hypertrophy and fibrosis.⁴⁴In vitro, ANP and cGMP analog inhibited aldosterone-induced ERK phosphorylation (Figure 9), suggesting that natriuretic peptide/GC-A/cGMP pathway can counteract the activation of both MAPK pathways (Supplemental Figure 5).

This study provides an idea that suppression of GC-A signaling would be a potential risk for proteinuria under high aldosterone state. Impaired GC-A signaling can often be seen in chronic heart failure (natriuretic peptide resistance), in which RAAS activation would inevitably occur both systemically and locally.⁴⁵ Such conditions are well recognized as potentially harmful to the kidney, and supplementation of ANP or BNP could become a therapeutic option against disease progression.

In summary, this study reveals that aldosterone causes massive proteinuria and podocyte injury in the absence of GC-A signaling with the activation of ROS and MAPKs, and that RAAS blockade and ROS inhibition could ameliorate these abnormalities. These findings suggest that local inhibition of the RAAS and oxidative stress in podocytes may be a novel mechanism involved in the pleiotropic and renoprotective properties of endogenous natriuretic peptide/GC-A system.

Concise Methods

Reagents and Antibodies

Aldosterone was obtained from Sigma Aldrich (St. Louis, MO). Reagents used were hydralazine (Sigma Aldrich), spironolactone (Sigma Aldrich), olmesartan (a gift from Daiichi Sankyo Pharmaceutical, Tokyo, Japan), tempol (Sigma Aldrich), MEK inhibitor U0126 (Cell Signaling Technology, Boston, MA), p38 MAPK inhibitor SB203580 (Cell Signaling Technology), and glucocorticoid receptor blocker mifepristone (Sigma Aldrich). Primary antibodies used for Western blotting and immunohistochemical studies were goat anti-nephrin (R&D Systems, Minneapolis, MN), rabbit anti-podocin (Sigma Aldrich), rabbit anti-p44/42 MAPK, rabbit anti-phospho-p44/42 MAPK, rabbit anti-p38 MAPK, rabbit anti-phospho-p38 MAPK (Cell Signaling Technology), and goat anti-8-OHdG (Millipore, Temecula, CA) antibodies.

Animal Experiments

All animal experiments were approved by the Animal Experimentation Committee of Kyoto University Graduate School of Medicine. Mice deficient in GC-A were produced on 129/SVJ background¹⁹ and then backcrossed with C57BL/6J mice more than 10 times. Male GC-A knockout mice or their wild-type littermates (approximately 28 g) received a left uninephrectomy or sham operation under intraperitoneal pentobarbital anesthesia (at −2 weeks). At 2 weeks after uninephrectomy or sham operation, an osmotic minipump (model 2004; Alzet, Cupertino, CA) was implanted subcutaneously to infuse vehicle or aldosterone (at 0 weeks). All mice were fed a diet containing 6% NaCl. Mice were assigned randomly to treated or untreated groups for 4 weeks: group 1, vehicle (2% ethanol)–infused wild-type mice (n=5); group 2, aldosterone (0.2 μg/kg body weight per minute)–infused wild-type mice (n=8); group 3, vehicle-infused GC-A knockout mice (n=5); group 4, aldosterone-infused GC-A knockout mice (n=8); group 5, aldosterone-infused GC-A knockout mice with hydralazine (60 mg/kg per day, in drinking water) (n=6); group 6, aldosterone-infused GC-A knockout mice with spironolactone (30 mg/kg per day, in drinking water) (n=5); group 7, aldosterone-infused GC-A knockout mice with olmesartan (10 mg/kg per day, in drinking water) (n=5); and group 8, aldosterone-infused GC-A knockout mice with tempol (110 mg/kg per day, in drinking water) (n=5).

Animals were given water ad libitum. BP was measured in conscious mice by the tail-cuff method (MK-2000ST; Muromachi Kikai, Tokyo, Japan) at −2, 0, 1, 2, and 4 weeks.²⁵ For urine measurements, each animal was housed separately in a metabolic cage (Shinano Manufacturing, Tokyo, Japan) at −2, 0, 1, 2, and 4 weeks.⁴⁶ Blood and kidney samples were harvested at 4 weeks. Glomeruli were isolated by the graded sieving method.⁴⁶ Urinary and serum creatinine were measured by the enzymatic method (SRL, Tokyo, Japan).⁴⁶ Serum aldosterone was measured by RIA (SRL). Urinary albumin excretion was assayed with a murine albumin ELISA kit (Exocell, Philadelphia, PA).⁴⁶

Renal Histology and Electron Microscopy

Histologic and electron microscopic examinations were performed as described previously.^46,47 Briefly, kidney sections stained with periodic acid–Schiff were examined by light microscopy (IX-81; Olympus, Tokyo, Japan). The cross-sectional area and the mesangial area in both 10 superficial and 10 juxtamedullary glomeruli were measured quantitatively using a computer-aided manipulator (MetaMorph software; Molecular Devices, Sunnyvale, CA).⁴⁶ The number of sclerotic and all juxtaglomerular glomeruli was counted. The fibrotic area was also measured in Masson’s trichrome-stained kidney sections quantitatively using a computer-aided manipulator (MetaMorph software).⁴⁷ Electron microscopic examination was performed in an electron microscope (H-7600; Hitachi, Tokyo, Japan). GBM thickness and foot process width were measured with Image J software (http://rsbweb.nih.gov/ij/; n=3, each). These procedures were performed by two investigators blinded to the origin of the slides and photos, and the mean values were calculated.

Immunohistochemistry

Immunofluorescence analyses for nephrin and podocin were described previously.^41,46 Briefly, cryostat sections were incubated with goat anti-nephrin antibody or rabbit anti-podocin antibody, and then incubated with FITC-labeled secondary antibodies (Jackson ImmunoResearch, West Grove, PA). Positive area and mean fluorescent intensity were measured with Image J software (n=3, each). Immunohistochemical studies for phospho-ERK, phospho-p38 MAPK, and 8-OHdG were also performed as previously described.⁴¹ Briefly, paraffin-embedded sections were incubated with rabbit anti-phospho-p44/42 MAPK (ERK) antibody, rabbit anti-phospho-p38 MAPK antibody, or goat anti-8-OHdG antibody, and then incubated with horseradish peroxidase–labeled anti-rabbit or anti-goat antibodies (Jackson ImmunoResearch). Mean staining intensity was measured with Image J software (n=3, each). The sections were developed with 3,3′-diaminobenzidine tetrahydrochloride. WT1 immunostaining was performed as described.⁴⁶ For immunohistochemical studies for GC-A, paraffin-embedded, autoclave-heated sections were treated with 3% H₂O₂ and a biotin blocking kit (Vector Laboratories, Burlingame, CA), and were incubated with 10% normal goat serum in PBS. The sections were incubated with rabbit anti-NPR-A (GC-A) antibody (Santa Cruz Biotechnology, Santa Cruz, CA) diluted 1:20 in PBS for 1 hour at room temperature, and then incubated with goat biotin-conjugated anti-rabbit antibody (Vector Laboratories). The sections were processed with LSAB2 streptavidin-HRP (DAKO, Glostrup, Denmark), and developed with 3,3′-diaminobenzidine tetrahydrochloride.

Glomerular RNA Extraction and Real-Time RT-PCR Analyses

Quantitative real-time RT-PCR was performed using the StepOnePlus system (Applied Biosystems, Foster City, CA) as described previously.⁴⁶ TGF-β1, COL1A1, COL4A3, fibronectin, gp91phox/Cybb (Nox-2), p22phox/Cyba, and Nox-4 mRNA expressions were evaluated. Some of primers and probe sets were described elsewhere⁴⁶ and included the following: COL1A1 forward primer, 5′-gtcccaacccccaaagac-3′; COL1A1 reverse primer, 5′-catcttctgagtttggtgatacgt-3′; COL1A1 probe, 5′-FAM-tgctgtgctttctgcccgga-TAMRA-3′; Cybb forward primer, 5′-ggtgacaatgagaacgaagagtatc-3′; Cybb reverse primer, 5′-gagacacagtgtgatgacaattcc-3′; Cybb probe, 5′-FAM-cagccaaccgagtcacggccacatac-TAMRA-3′; Cyba forward primer, 5′-cccctcaccaggaattactacg-3′; Cyba reverse primer, 5′-cactgctcacctcggatgg-3′; Cyba probe, 5′-FAM-ctccacttcctgttgtcggtgcctgc-TAMRA-3′; Nox-4 forward primer, 5′-gcaagactctacacatcacatgtg-3′; Nox-4 reverse primer, 5′-tgctgcattcagttcaaggaaatc-3′; Nox-4 probe, 5′-FAM-tctcaggtgtgcatgtagccgccca-TAMRA-3′; Gja1 forward primer, 5′-ctctccttttcctttgacttcagc-3′; Gja1 reverse primer, gaccttgtccagcagcttcc; and Gja1 probe, 5′-FAM-aaggagttccaccactttggcgtgcc-TAMRA-3′. Expression of each mRNA was normalized with GAPDH mRNA (TaqMan rodent GAPDH control reagents; Applied Biosystems).

Cell Cultures

A conditionally immortalized mouse podocyte cell line was provided by Dr. Peter Mundel (University of Miami Miller School of Medicine, Miami, FL) and cultured as described.⁴¹ For time-course experiments, differentiated podocytes were made quiescent in medium that contained 0.1% FBS for 24 hours, and then cells were stimulated with 1 μM aldosterone and further incubated for a period ranging from 5 minutes to 24 hours. The effect of ANP (Peptide Institute, Osaka, Japan) on phosphorylation of ERK and p38 MAPK was studied in the presence of aldosterone. Cells were made quiescent in medium that contained 0.1% FBS for 24 hours, pretreated with 1 μM ANP or vehicle (final 0.005% glucose solution) 30 minutes before stimulation with 1 μM aldosterone or vehicle (final 0.2% ethanol), and then harvested at 10 minutes or 3 hours after stimulation for ERK and p38 MAPK analyses, respectively.

Connexin43 (Gja1) mRNA expression was evaluated by TaqMan PCR. Differentiated podocytes were made quiescent, pretreated with 10 μM U0126 or 10 μM SB203580 for 30 minutes, and then stimulated with 1 μM aldosterone. Cells were harvested at 24 hours after stimulation with a RNeasy Mini kit (Qiagen). For analysis of the glucocorticoid pathway, cells were pretreated with 10 μM mifepristone or 10 μM spironolactone and were harvested at 3 hours after aldosterone (1 μM) stimulation.

Western Blot Analyses

Western blot analysis was performed as described.^46,47 Briefly, isolated glomeruli were homogenized in cold RIPA buffer (50 mM Tris-HCl, 150 mM NaCl, 4 mM EDTA, 1% Nonidet P-40, 0.1% sodium deoxycholate, 10 mM Na₄P₂O₇, 10 mM NaF, 10 μg/ml aprotinin, 2 mM dithiothreitol, 2 mM sodium orthovanadate, and 1 mM PMSF). For cultured cells, cells were lysed with RIPA buffer for ERK detection, or were processed with the AllPrep DNA/RNA/protein Mini kit and then lysed with lysis buffer containing 100 mM Tris-HCl, 3% SDS, 10 mM NaHPO₄, 1% Nonidet P-40, 20 mM EDTA, 10 μg/ml aprotinin, 2 mM dithiothreitol, 2 mM sodium orthovanadate, and 1 mM PMSF for p38 MAPK detection. The homogenates were centrifuged at 15,000 rpm for 15 minutes at 4°C, and the supernatants were treated with NuPAGE sample buffer (Invitrogen, Carlsbad, CA). Western blot analysis was performed as described with some modifications using NuPAGE Bis-Tris gels (Invitrogen).⁴⁷ Filters on isolated cell extracts were incubated with rabbit anti-phospho-p44/p42 MAPK antibody or anti-phospho-p38 MAPK antibody for 1 hour, and immunoblots were developed using horseradish peroxidase-linked donkey anti-rabbit antibodies (Amersham, Arlington Heights, IL) and a chemiluminescence kit (Amersham).

Statistical Analyses

Data are expressed as the mean ± SEM. Statistical analysis was performed using one-way ANOVA. P<0.05 was considered statistically significant.

Disclosures

None.

We gratefully acknowledge Mr. M. Fujimoto and Mr. Y. Sakashita and other laboratory members for technical assistance and Ms. A. Yamamoto for secretarial assistance.

This work was supported in part by research grants from the Japanese Ministry of Education, Culture, Sports, Science, and Technology; the Japanese Ministry of Health, Labour, and Welfare; and the Salt Science Research Foundation.