Natriuretic peptide receptor guanylyl cyclase-A pathway counteracts glomerular injury evoked by aldosterone through p38 mitogen-activated protein kinase inhibition

Guanylyl cyclase-A (GC-A) signaling, a natriuretic peptide receptor, exerts renoprotective effects by stimulating natriuresis and reducing blood pressure. Previously we demonstrated massive albuminuria with hypertension in uninephrectomized, aldosterone-infused, and high salt-fed (ALDO) systemic GC-A KO mice with enhanced phosphorylation of p38 mitogen-activated protein kinase (MAPK) in podocytes. In the present study, we examined the interaction between p38 MAPK and GC-A signaling. The administration of FR167653, p38 MAPK inhibitor, reduced systolic blood pressure (SBP), urinary albumin excretion, segmental sclerosis, podocyte injury, and apoptosis. To further investigate the local action of natriuretic peptide and p38 MAPK in podocytes, we generated podocyte-specific (pod) GC-A conditional KO (cKO) mice. ALDO pod GC-A cKO mice demonstrated increased urinary albumin excretion with marked mesangial expansion, podocyte injury and apoptosis, but without blood pressure elevation. FR167653 also suppressed urinary albumin excretion without reducing SBP. Finally, we revealed that atrial natriuretic peptide increased phosphorylation of MAPK phosphatase-1 (MKP-1) concomitant with inhibited phosphorylation of p38 MAPK in response to MAPK kinase 3 activation, thereby resulting in decreased mRNA expression of the apoptosis-related gene, Bax, and Bax/Bcl2 ratio in cultured podocytes. These results indicate that natriuretic peptide exerts a renoprotective effect via inhibiting phosphorylation of p38 MAPK in podocytes.

reactivation of renin-angiotensin-aldosterone system through positive feedback 3 , and the induction of reactive oxygen species (ROS) and activation of mitogen-activated protein kinases (MAPKs) in the kidney 4 . The site of action of aldosterone remains obscure, as aldosterone has been reported to exert its effect on podocytes 5 , mesangial cells 6 , endothelial cells 7 , and tubular cells 8 . In addition, aldosterone elevates blood pressure and regulates intraglomerular pressure 8 , which both contributes to the complex pathophysiology in aldosterone-induced renal injury.
Natriuretic peptides consist of atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP), and C-type natriuretic peptide (CNP). ANP and BNP exert physiological effects such as diuresis and vasodilation through guanylyl cyclase-A (GC-A)/natriuretic peptide receptor-A (NPR-A) 9 . We previously demonstrated massive albuminuria, mesangial expansion, segmental sclerosis, and podocyte-injury with the elevation of systolic blood pressure in uninephrectomized, aldosterone-infused, and high salt diet-fed (ALDO) systemic GC-A knockout (KO) mice 10 . In the previous model, reduced blood pressure with the use of hydralazine fails to ameliorate renal injury; however, blockade of the renin angiotensin system by an angiotensin receptor blocker (ARB) or antioxidant tempol reduces glomerular injury 10 . Although our previous report may indicate that natriuretic peptides protect podocytes from aldosterone-induced renal injury, the precise site of action of natriuretic peptides/GC-A system, and the mechanism underlying amelioration of aldosterone-induced glomerular injury by natriuretic peptides have yet to be elucidated. In our previous study, we also demonstrated prominent phosphorylation of p38 MAPK in podocyte of ALDO GC-A KO mice 10 . Activation of p38 MAPK has been shown to be stimulated by UV light, heat, inflammation, and growth factors 11 . We and others have demonstrated that phosphorylation of p38 MAPK is activated in clinical glomerulonephritis, as well as in experimental podocyte injury glomerulopathy 10,12,13 . Blockade of p38 MAPK pathway reportedly ameliorates adriamycin nephropathy 12 and crescentic glomerulonephritis 14 in mice. However, the precise role of p38 MAPK in podocyte injury has yet to be elucidated. In addition, the mechanism underlying the renoprotective action of natriuretic peptides on podocytes remains to be clarified.
In the present study, we generated podocyte-specific (pod) GC-A conditional KO (cKO) mice and then examined the effects of p38 MAPK in systemic and pod GC-A cKO mice administered aldosterone, uninephrectomy, and a high salt diet.

Inhibition of p38 MAPK ameliorates glomerular injury in ALDO systemic GC-A KO mice. All
ALDO mice were uninephrectomized and administered aldosterone and a 6% high salt diet for four weeks. First, we examined p38 MAPK phosphorylation in ALDO wild-type and systemic GC-A KO mice (Fig. 1a). Phosphorylation of p38 MAPK was predominantly detected in podocytes in ALDO wild-type mice (Fig. 1a). ALDO systemic GC-A KO mice demonstrated pronounced phosphorylation of p38 MAPK in podocytes (Fig. 1a), and had elevated systolic blood pressure (SBP) (Fig. 1b). Blockade of the p38 MAPK pathway using FR167653 (a p38 MAPK inhibitor) reduced SBP by 20 mmHg in ALDO systemic GC-A KO mice (Fig. 1b). Urinary albumin excretion in ALDO systemic GC-A KO mice was as high as 9,000 μ g/mgCr; however, administration of FR167653 reduced urinary albumin excretion by 95% (Fig. 1c, Table 1). Body and kidney weights, and blood and urine parameters are presented in Table 1. ALDO systemic GC-A KO mice were found to have renal hypertrophy, and no effect of FR167653 on renal hypertrophy was observed. No significant differences in serum potassium, sodium, urinary potassium, or urinary sodium levels were observed (Table 1).
We examined renal histology at four weeks after aldosterone infusion. A mild mesangial expansion in superficial glomeruli was observed in ALDO wild-type mice ( Supplementary Fig. S1a). In contrast, ALDO systemic GC-A KO mice had marked mesangial expansion with glomerular hypertrophy ( Supplementary Fig. S1a-c). Administration of FR167653 significantly reduced mesangial expansion and glomerular hypertrophy ( Supplementary Fig. S1a-c). In juxtamedullary glomeruli, ALDO systemic GC-A KO mice exhibited severe segmental sclerosis with glomerular hypertrophy (Fig. 1d). Treatment with FR167653 inhibited segmental sclerosis and glomerular hypertrophy in ALDO systemic GC-A KO mice ( Fig. 1d-f). We further examined renal fibrotic changes. ALDO systemic GC-A KO mice exhibited tubular dilatation with protein cast deposition, and these changes were ameliorated by FR167653 (Fig. 1g).
Next, we examined gene expression in glomeruli. TGF-β 1 (Tgfb1) is a profibrotic growth factor and is closely associated with matrix accumulation and renal fibrosis. Analysis of glomerular expression revealed increased Tgfb1 mRNA levels in ALDO systemic GC-A KO mice, and this upregulation was significantly reduced by FR167653 ( Supplementary Fig. S2a). Then, we examined the expression of genes associated with inflammation. We examined gene expression relevant to macrophages: Ccl2 (MCP1) is one of macrophage chemoattractant molecules, and Adgre1 (F4/80) is one of markers of macrophages. Expression of Ccl2 (MCP-1) mRNA was increased in ALDO systemic GC-A KO mice, and administration of FR167653 significantly reduced Ccl2 mRNA levels ( Supplementary Fig. S2b). Consistent with this result, FR167653 also reduced Adgre1 (F4/80) mRNA expression in ALDO systemic GC-A KO mice ( Supplementary Fig. S2c). A similar trend in Cybb (NOX-2) mRNA expression was observed ( Supplementary Fig. S2d).

Inhibition of p38 MAPK reduces podocyte injury in ALDO systemic GC-A KO mice.
We next evaluated podocyte injury. Electron microscopic analysis revealed that ALDO wild-type mice exhibited slightly widened podocytes foot processes (Fig. 2a,b). ALDO systemic GC-A KO mice demonstrated foot process effacement with irregular thickening of the glomerular basement membrane (GBM), and these changes were improved by FR167653 (Fig. 2a,b). We quantitated the width of foot process and revealed an improvement of foot process effacement by FR167653 (Fig. 2c). Immunofluorescent study demonstrated that reduced nephrin and podocin levels in ALDO systemic GC-A KO mice, and these changes were also ameliorated with administration of a p38 MAPK inhibitor (Fig. 2d,e), indicating that inhibition of the p38 MAPK pathway is closely associated with amelioration of podocyte injury. We next examined glomerular apoptosis. ALDO wild-type mice demonstrated almost no apoptotic cells within the glomeruli (Fig. 3a). In contrast, ALDO systemic GC-A KO mice had a large number of TUNEL-positive cells within the glomeruli and tubules, and this increase in the number of apoptotic cells was significantly inhibited by treatment with FR167653 (Fig. 3a). We also examined glomerular p53 expression, as it has been shown to be closely associated with apoptosis 15 . Expression of p53 in glomeruli was not detected in ALDO wild-type mice but was observed in ALDO systemic GC-A KO mice (Fig. 3b). Inhibition of p38 MAPK reduced glomerular expression of p53 in ALDO systemic GC-A KO mice (Fig. 3b). We next counted the number of podocytes by immunohistochemical analysis for Wilms' tumor 1 (WT1). Administration of FR167653 did not change podocyte numbers in ALDO wild-type mice (Fig. 3c,d). ALDO systemic GC-A KO mice had a significant reduction in the number of podocytes (Fig. 3c,d), and this reduction was reversed by administration of FR167653 (Fig. 3c,d). Phosphorylation of p38 MAPK in glomeruli was augmented in ALDO systemic GC-A KO mice, and this phosphorylation was ameliorated in ALDO systemic GC-A KO mice administered FR167653 ( Supplementary Fig. S3). MAPK kinase 3 (MKK3) is an upstream kinase of p38 MAPK 16 . We examined the phosphorylation of MKK3 in these mice by immunohistochemical study. ALDO systemic GC-A KO mice showed enhanced phosphorylation of MKK3 in glomeruli, and this phosphorylation was reduced with the treatment of FR167653 ( Supplementary Fig. S4). These results indicate that inhibition of p38 MAPK ameliorates podocyte injury and apoptosis.

Deletion of GC-A in podocytes aggravates glomerular injury. Although inhibition of p38 MAPK
in ALDO systemic GC-A KO mice ameliorated podocyte and glomerular injury, the relevance of p38 MAPK inhibition with natriuretic peptides in podocyte injury remains obscure. To reveal the role of p38 MAPK signaling in podocytes of ALDO systemic GC-A KO mice, we generated podocyte-specific (pod) GC-A conditional KO (cKO) mice by crossing Nephrin-Cre mice with GC-A floxed mice, and then administering FR167653. Vehicle (VEH)-treated mice were also uninephrectomized, and administered 6% high salt diet and ethanol alone instead of ethanol-containing aldosterone. GC-A floxed mice without Cre recombinase were used as control mice. VEH control and pod GC-A cKO mice had normal SBP (Fig. 4a). ALDO control mice exhibited an increase in SBP of 10 mmHg at four weeks, and ALDO pod GC-A cKO mice had comparable SBP to ALDO control mice (123.2 ± 2.2 vs. 119.4 ± 2.3 mmHg; Fig. 4a). FR167653 was administered to pod GC-A cKO mice as the activation of p38 MAPK is reportedly involved in glomerular injury in ALDO systemic GC-A KO mice. No change in SBP was observed following the administration of FR167653 in ALDO pod GC-A cKO mice (119.0 ± 2.4 mmHg; Fig. 4a). Body weights, kidney weights and serum and urinary parameters are presented in Table 2. Pod GC-A cKO mice exhibited normal kidney weights comparable with control mice ( Table 2). ALDO control and pod GC-A cKO mice exhibited renal hypertrophy, and administration of FR167653 had no effect on renal hypertrophy in pod GC-A cKO mice (Table 2). Pod GC-A cKO mice had similar serum and urinary potassium levels to control mice, and aldosterone infusion increased urinary potassium levels and decreased serum potassium levels to the same extent in both control and pod GC-A cKO mice (Table 2). There was no significant difference in serum creatinine and urea nitrogen (UN) levels between ALDO control and pod GC-A cKO mice.
At basal levels, urinary albumin excretion was not different between VEH control mice and pod GC-A cKO mice (18.5 ± 3.3 and 14.2 ± 2.6 μ g/mgCr, respectively; Fig. 4b). ALDO control mice had a threefold increase in urinary albumin excretion at four weeks (84.0 ± 6.5 vs. 31.3 ± 3.6 μ g/mgCr; Fig. 4b). Of note, ALDO pod GC-A cKO mice demonstrated a significantly increased urinary albumin excretion by 13-fold compared to ALDO control mice at four weeks (84.0 ± 6.5 vs. 1,188 ± 152 μ g/mgCr; Fig. 4b). These results indicate that deletion of GC-A in podocytes aggravates aldosterone-induced renal injury. Treatment of ALDO pod GC-A cKO mice with FR167653 suppressed urinary albumin excretion by 95% compared with ALDO pod-GC-A cKO mice (80.5 ± 10.5 μ g/mgCr; Fig. 4b), indicating the importance of p38 MAPK pathway in aldosterone-induced podocyte injury. Pod GC-A cKO mice had normal glomerular histology at baseline. In superficial glomeruli, there was no significant difference in glomerular cross-sectional or mesangial areas between ALDO control and ALDO pod GC-A cKO mice ( Supplementary Fig. S5a,b). In juxtamedullary glomeruli, ALDO control mice had mild glomerular hypertrophy and mesangial expansion (Fig. 4c-e). ALDO pod GC-A cKO mice were observed to have marked mesangial expansion and glomerular hypertrophy (Fig. 4c-e). Administration of FR167653 ameliorated glomerular hypertrophy and mesangial expansion (Fig. 4c-e). We examined Npr1 mRNA to examine the deletion efficacy of Npr1 gene. Glomerular mRNA expression of Npr1 (GC-A) in VEH pod GC-A cKO mice was decreased by 40% compared with that in VEH control mice (Fig. 4f).

Deletion of GC-A in podocytes aggravates podocyte injury and its injury is reversed by inhibition of p38 MAPK.
To evaluate podocyte injury, we performed immunofluorescent analyses of juxtamedullary glomeruli for nephrin and podocin. ALDO control mice exhibited no difference in nephrin and podocin expression compared with VEH control mice (Fig. 5a,b). Although VEH pod GC-A cKO mice had normal nephrin and podocin staining, ALDO pod GC-A cKO mice demonstrated decreased staining intensity for nephrin and podocin (Fig. 5a,b). ALDO pod GC-A cKO mice with FR167653 maintained similar nephrin and podocin staining intensities as VEH pod GC-A cKO mice (Fig. 5a,b). We examined podocyte injury in juxtamedullary glomeruli by electron microscopic analysis. VEH pod GC-A cKO mice demonstrated normal foot processes and GBM (Fig. 5c). ALDO pod GC-A cKO mice exhibited mild thickening of the GBM and foot process effacement, and these changes were improved with the administration of FR167653 (Fig. 5c). Next, we examined glomerular p38 MAPK phosphorylation. No glomerular p38 MAPK phosphorylation was observed in VEH pod GC-A cKO mice (Fig. 5d). Administration of aldosterone much increased phosphorylation of p38 MAPK, particularly in podocytes in pod GC-A cKO mice compared to control mice (Fig. 5d). Phosphorylation of p38 MAPK in glomeruli was reduced by administration of FR167653 (Fig. 5d).
We next analyzed glomerular mRNA expression levels. As a previous report showed that natriuretic peptide counteracts aldosterone-induced renal injury 10 , we examined the expression level of Npr1 in this model. Administration of aldosterone upregulated Npr1 expression in both control mice and pod GC-A cKO mice, with no significant difference between the two groups (Fig. 6a). FR167653 treatment reduced Npr1 mRNA expression by 60% (Fig. 6a). We next examined the expression of Tgfb1 and Fn1, because fibronectin (Fn1) is one of main extracellular matrix in the kidney and these gene expression is relevant to extracellular matrix accumulation. The expression  of extracellular matrix-related gene, such as Tgfb1 and Fn1, was enhanced with administration of aldosterone both in control mice and pod GC-A cKO mice, and was ameliorated by treatment with FR167653 (Fig. 6b,c). We examined gene expression relevant to macrophages: Ccl2 (MCP1) is one of macrophage chemoattractant molecules, and Adgre1 (F4/80) is one of markers of macrophages. Expression patterns of Ccl2 and Adgre1 were similar to expression levels of Tgfb1 and Fn1 (Fig. 6d,e). We next examined the expression of oxidative stress-related genes and downstream target genes of p38 MAPK. Aldosterone also upregulated Cybb and Trp53 (p53) mRNA levels in both control mice and pod GC-A cKO mice, and FR167653 ameliorated these changes (Fig. 6f,g).
Next, we performed TUNEL assays to evaluate podocyte apoptosis. The number of TUNEL-positive cells in the juxtamedullary glomeruli was increased in ALDO pod GC-A cKO mice (Fig. 7a). Treatment with FR167653 completely abolished TUNEL-positive cells in the glomeruli of ALDO pod GC-A cKO mice (Fig. 7a). The number of WT1-positive cells in superficial glomeruli was not different among groups (Supplementary Fig S6). The number of WT1-positive cells in the juxtamedullary glomeruli of VEH pod GC-A cKO mice did not differ from VEH control mice (Fig. 7b,c). Administration of aldosterone decreased the number of WT1-positive cells in pod GC-A cKO mice compared with pod GC-A cKO mice that were not treated with aldosterone, and this decrease tended to be ameliorated by FR167653 (Fig. 7b,c).
Natriuretic peptides inhibit phosphorylation of p38 MAPK via activating MKP-1. We next investigated whether ANP can inhibit expression of apoptosis-related proteins induced by activation of p38 MAPK in cultured human podocytes. ANP was administered to the human podocytes transfected with plasmids expressing wild-type MKK3 (MKK3) or constitutive active MKK3 (CA) (Fig. 8a). Phosphorylation of p38 MAPK was increased with following transfection of MKK3 or CA plasmids, and this increase was abolished not by treatment with ANP in MKK3-transfected podocytes, but by treatment with ANP in CA-transfected podocytes (Fig. 8b,c). Expression of Bax was not significantly different in MKK3-transfected podocytes, but significantly upregulated in cultured podocytes transfected with CA, and this increase was ameliorated only with high dose ANP (Fig. 8d). Consistent with this result, Bcl2 mRNA expression was reduced by transfection with CA, and ANP ameliorated this effect in cultured podocytes (Fig. 8e). As a result, the Bax/Bcl2 ratio was increased following transfection with MKK3 or CA plasmids, and this increase was ameliorated with ANP (Fig. 8f). To confirm the mechanism of ANP-mediated inhibition of p38 MAPK phosphorylation, we examined the phosphorylation status of MAPK phosphatase-1 (MKP-1) which dephosphorylates p38 MAPK, in cultured podocytes (Fig. 8g). Administration of ANP increased phosphorylation of MKP-1 in cultured human podocytes (Fig. 8h,i), indicating that ANP dephosphorylates p38 MAPK via activation of MKP-1.

Discussion
The present study demonstrated that blockade of p38 MAPK ameliorates glomerular injury in ALDO systemic GC-A KO mice. As we previously reported, ALDO wild-type mice show mild albuminuria, a slight increase in mesangial area, and mild podocyte injury; however, ALDO systemic GC-A KO mice show massive albuminuria with prominent mesangial expansion and podocyte injury 10 . One of the differences between ALDO wild-type and ALDO systemic GC-A KO mice was the phosphorylation status of p38 MAPK, because these two types of mice were found to have similar mRNA expression of inflammation, profibrotic, and oxidative stress genes 10 . The present study clearly demonstrated that the blockade of p38 MAPK reduced albuminuria, mesangial expansion, podocyte injury and apoptosis of glomerular cells in ALDO systemic GC-A mice. Previous studies have reported that blockade of p38 MAPK reduces glomerular injury in several experimental glomerulopathy models, including adriamycin nephropathy 12 , complement-mediated glomerular injury 17 , anti-GBM glomerulonephritis 13  activation in response to aldosterone has yet to be fully elucidated, especially in the glomeruli. The present study has unveiled the novel pathway of aldosterone-induced glomerular injury. Aldosterone has been shown to promote apoptosis in many cell types, including podocytes through phosphoinositide-3/Akt, p38 MAPK and p53 pathways in vitro 21 . However, the role of p38 MAPK in aldosterone-induced podocyte apoptosis remains elusive. The present study demonstrates that blockade of p38 MAPK ameliorates apoptosis in podocytes, potentially indicating the mechanism by which MAPK signaling reduces podocyte injury. Interestingly, administration of FR167653 reduced systolic blood pressure in ALDO systemic GC-A KO mice by 20 mmHg, although it had no effect on systolic blood pressure in ALDO wild-type mice. ALDO systemic GC-A KO mice had similar levels of serum aldosterone, and daily urinary sodium excretion, suggesting that the reduction in systolic blood pressure was not due to increased sodium excretion. Other possible explanations are changes in blood vessel tone and cardiac output. Tojo et al. reported that administration of FR167653 to Dahl salt-sensitive rats reduces expression of interleukin-1β and tumor necrosis factor-α in the whole kidney without altering blood pressure 22 . However, Potthoff et al. recently reported that chronic p38 MAPK inhibition improves vascular function and reduces systolic blood pressure in angiotensin II-stimulated hypertension in  23 . Interestingly, they also reported that non-treated mice do not exhibit blood pressure reduction 23 , consistent with our results.
Kuhn et al. reported that BNP activates phosphorylation of p38 MAPK in microvascular endothelial cells 24 , a finding contradictory to our results. However, phosphorylation of p38 MAPK is context-dependent rather than ligand-dependent. Phosphorylation of p38 MAPK is prominent in podocytes in ALDO systemic GC-A KO mice; therefore, decreased p38 MAPK activity in response to FR167653 appears to function in the protection of podocytes. However, we were unable to exclude the effects of blood pressure reduction and the role of mesangial cells as well as endothelial cells in ALDO systemic GC-A KO mice administered FR167653. To overcome these limitations, we generated podocyte-specific GC-A cKO mice and administered aldosterone and high salt following uninephrectomy. Recently, Staffel et al. reported that uninephrectomized pod GC-A cKO mice with deoxycorticosterone acetate and high salt diet exhibit similar blood pressure and increased albuminuria compared with control mice, confirming the protective role of natriuretic peptides/GC-A pathway in podocytes 25 . We also constructed pod GC-A cKO mice and demonstrated increased urinary albumin and glomerular injury without blood pressure elevation in ALDO pod GC-A cKO mice. Especially, juxtamedullary glomeruli rather than superficial glomeruli exhibited prominent mesangial expansion and podocyte injury in ALDO pod GC-A cKO mice. We speculate that hemodynamic changes will be the important factor on glomerular injury in ALDO pod GC-A cKO mice, because aldosterone concentration should be same between superficial and juxtamedullary glomeruli. Administration of FR167653 to ALDO pod GC-A cKO mice ameliorated albuminuria, mesangial expansion, loss of nephrin and podocin, and podocyte foot process effacement, suggesting that podocytes are one of the main targets of FR167653. Several studies showed genetic deletion of a specific molecule in podocytes affects mesangial  26,27 , probably because of paracrine effects between podocytes and mesangial cells. However, GC-A was localized in mesangial cells 28 and endothelial cells 29 as well as in podocytes. In addition, urinary albumin excretion of ALDO pod GC-A cKO mice was much less than that of ALDO systemic GC-A KO mice, although ALDO pod GC-A cKO mice showed lower SBP. These finding may indicate the importance of mesangial or endothelial cells in the signaling of natriuretic peptides. This is the limitation of this study.
Aldosterone has been shown to phosphorylate p38 MAPK in many cell types, including vascular smooth muscle cells 30 , cardiomyocytes 31 , neutrophils 32 , peritoneal mesothelial cells 33 , and podocytes 21 . However, the relationship between natriuretic peptides and phosphorylation of p38 MAPK is cell-type and context-dependent. Previous studies have reported that natriuretic peptides activates p38 MAPK in hypoxia-exposed hepatocytes 34 , lung endothelial cells 24 and brown adipocytes 35 . In contrast, other reports, including our previous study, have demonstrated that natriuretic peptides abolish p38 MAPK activation induced by inflammatory stimuli in human umbilical endothelial cells 36 , bovine pulmonary microvascular endothelial cells 37 , fibroblasts 38 and podocytes 10 . The mechanism by which natriuretic peptides reduce phosphorylation of p38 MAPK has yet to be fully evaluated; however, a study has reported that natriuretic peptides augment MKP-1, which dephosphorylates p38 MAPK, in endothelial cells 39 . Consistent with this study, the present study also demonstrated the MKP-1 phosphorylation following treatment of podocytes with ANP. These results indicate that deletion of GC-A enhances phosphorylation of p38 MAPK in the kidney following stimulation with aldosterone.
Previous studies have reported that aldosterone induces apoptosis in mesangial cells 40 , endothelial cells 41 , renal tubular cells 42 , and podocytes 43 in the kidney. Natriuretic peptides have been documented to have activity against apoptosis in cells in renal cortex and medulla in SHR rats 44 , macrophages 45 , cardiomyocytes in db/db mice 46 , and hepatocytes 47 . Conversely, other reports have shown that ANP induces apoptosis 48,49 . However, the contribution of natriuretic peptides to apoptosis in podocytes remains unclear. In the present study, we demonstrate that natriuretic peptides inhibit Bax expression, leading to decreased levels of apoptosis.
In conclusion, inhibition of p38 MAPK pathway ameliorates glomerular injury in uninephrectomized, high-salt fed, and aldosterone-infused mice by protecting against podocyte injury. These results indicate that p38 MAPK inhibitors may represent promising therapeutic targets against renal injury.  implantation subcutaneously to infuse aldosterone (0.2 μ g/kg body weight per minute) or vehicle. All mice were fed a diet containing 6% NaCl 10 . FR167653 was administered by dissolving it in drinking water 12 . Blood pressure was measured indirectly by using the tail-cuff method (MK-2000ST; Muromachi Kikai, Tokyo, Japan) at − 2, 0, 1, 2, and 4 weeks. Urine samples were collected from metabolic cages (Shinano manufacturing, Tokyo, Japan) for 24 h at − 2, 0, 1, 2, and 4 weeks and urine volumes were measured. Blood and kidney samples were harvested at 4 weeks. Urinary albumin levels were measured with albumin ELISA kits (Exocell, Philadelphia, PA). Serum and urinary Cr, sodium, and potassium levels were measured by enzymatic methods (SRL, Tokyo, Japan).
Construction of GC-A floxed mice and nephrin-promotor driven Cre recombinase transgenic mice were as described previously 53,54 . By crossing GC-A floxed mice (C57BL/6 J background) with Nephrin Cre mice (C57BL/6 J background), we generated pod GC-A cKO mice (Npr1 fl/fl , Nephrin Cre/+ mice) and control littermates (Npr1 fl/fl ,+ /+ ). Male pod GC-A cKO mice or control mice at 12 weeks of age were uninephrectomized, and administered with 6% NaCl diet and aldosterone or vehicle using the same protocol for systemic GC-A KO mice.
Renal histology and electron microscopy. Histological and electron microscopic examinations were performed as described previously 55 . Periodic acid-Schiff (PAS) stained kidney samples were examined by light microscopy. Mesangial areas in juxtamedullary glomeruli were measured using MetaMorph Software (Molecular Devices, Sunnyvale, CA). Electron microscopic examination was performed using an electron microscope (H-7600, Hitachi, Tokyo, Japan).
Immunohistochemistry. Immunofluorescent studies for nephrin and podocin were performed as descried previously 10 . Briefly, cryostat sections were incubated with goat anti-nephrin antibody or rabbit anti-podocin antibody, and then incubated with fluorescein isothiocyanate (FITC)-labeled secondary antibody. Immunohistochemical analysis for phospho-p38 MAPK, WT1, p53, phospho-MKK3 were as previously described with some modifications 10 . Staining for the presence of apoptotic cells was performed using in situ cell death detection kits (Roche, Mannheim, Germany). Paraffin-embedded renal sections were reacted with fluorescent dyed mixed solutions.
Cell culture and transfection. Conditionally immortalized human podocytes were provided by Dr.
Western blotting. Western blotting was performed as previously described 10 . Filters on isolated cell extracts were incubated with rabbit anti-phospho p38 MAPK antibody, rabbit anti-p38 MAPK antibody, rabbit anti-phospho MKP-1 antibody, or mouse anti-GAPDH antibody. Immunoblot were developed using horseradish peroxidase-linked donkey anti-rabbit or anti-mouse antibody and a chemiluminescent kit.