ASK1 inhibitor treatment suppresses p38/JNK signalling with reduced kidney inflammation and fibrosis in rat crescentic glomerulonephritis

Abstract Activation of p38 mitogen‐activated protein kinase (MAPK) and c‐Jun amino terminal kinase (JNK) is prominent in human crescentic glomerulonephritis. p38 and JNK inhibitors suppress crescentic disease in animal models; however, the upstream mechanisms inducing activation of these kinases in crescentic glomerulonephritis are unknown. We investigated the hypothesis that apoptosis signal‐regulating kinase 1 (ASK1/MAP3K5) promote p38/JNK activation and renal injury in models of nephrotoxic serum nephritis (NTN); acute glomerular injury in SD rats, and crescentic disease in WKY rats. Treatment with the selective ASK1 inhibitor, GS‐444217 or vehicle began 1 hour before nephrotoxic serum injection and continued until animals were killed on day 1 (SD rats) or 14 (WKY rats). NTN resulted in phosphorylation (activation) of p38 and c‐Jun in both models which was substantially reduced by ASK1 inhibitor treatment. In SD rats, GS‐444217 prevented proteinuria and glomerular thrombosis with suppression of macrophage activation on day 1 NTN. In WKY rats, GS‐444217 reduced crescent formation, prevented renal impairment and reduced proteinuria on day 14 NTN. Macrophage activation, T‐cell infiltration and renal fibrosis were also reduced by GS‐444217. In conclusion, GS‐444217 treatment inhibited p38/JNK activation and development of renal injury in rat NTN. ASK1 inhibitors may have therapeutic potential in rapidly progressive glomerulonephritis.

pathways in many cells types in human kidney disease which is most pronounced in rapidly progressive crescentic glomerulonephritis. [4][5][6][7] This increased p38 and JNK activation is evident in infiltrating cells as well as intrinsic kidney cells, including podocytes, mesangial cells, tubular epithelial cells and fibroblasts. [4][5][6][7] Activation of p38 signalling correlates with renal function across a range of glomerulonephridities, with glomerular p38 activation correlating with segmental proliferative and necrotic lesions and interstitial p38 activation correlating with interstitial inflammation and fibrosis. 7 Similarly, across a range of glomerular diseases, JNK activation correlates with glomerulosclerosis, while JNK signalling in the tubulointerstitial compartment correlates with macrophage infiltration, interstitial fibrosis and renal dysfunction. 5,6 Small molecule inhibitors of p38 MAPK can inhibit inflammation and kidney damage in models of antiglomerular basement membrane (GBM) glomerulonephritis. [8][9][10][11] Subsequently, p38 inhibitors have been shown to suppress acute kidney injury, renal inflammation and renal fibrosis across a range of animal models. [12][13][14][15] Similarly, small molecule JNK inhibitors can suppress renal injury in several disease models, including crescentic glomerulonephritis [16][17][18] However, clinical trials evaluating p38 and JNK inhibitors in human inflammatory and fibrotic disorders have failed because of lack of efficacy or side effects such as liver toxicity. 19,20 Therefore, therapeutic strategies that reduce the activation of the p38 and JNK pathways during pathologic stress, while sparing the physiological roles of these enzymes, may lead to greater efficacy and a better risk/benefit profile.
The p38 and JNK enzymes are activated by a cascade of phosphorylation reactions performed by members of the MAPKKK (MAP3K) and MKK enzyme families. There are more than 20 members of the MAP3K family, many of which have the ability to activate the downstream p38 and JNK enzymes in cell culture studies, although few members of this family have been show to activate p38 and JNK in vivo. 21 We have investigated the role of one of these enzymes, apoptosis signal-regulating kinase 1 (ASK1/MAP3K5), as a mechanism of p38/JNK activation in kidney disease based on several reasons. First, ASK1 is kept inactive under normal physiological conditions through binding to the reduced form of thioredoxin and only undergoes auto-activation upon the oxidation and dissociation of thioredoxin during pathological oxidative stress-an important feature in many forms of kidney disease. 3,22 Second, mice lacking the Ask1 gene have a normal phenotype consistent with a lack of involvement of ASK1 in normal physiology. 3,22 Third, p38 and JNK are the only known downstream targets of ASK1 activation thereby providing a potentially selective means to target these pathways. 3,22 Using Ask1 gene-deficient mice, we have shown that activation of p38 MAPK, and to a lesser extent JNK, is dependent upon ASK1 in the obstructed kidney, while administration of a highly selective ASK1 inhibitor (GS-444217) in experimental diabetic kidney disease also suppressed p38 and JNK activation. 23,24 However, it is unknown whether p38 and/or JNK activation depends upon ASK1 in the aggressive renal inflammation seen in crescentic glomerulonephritis. Indeed, the answer to this question is not obvious as ASK1 is not required for acute IL-1 or LPS-induced p38 activation in cultured tubular epithelial cells, 23 and IL-1 contributes to the development of crescentic disease, 25 while LPS exacerbates glomerular injury. 26 This study used the highly selective ASK1 inhibitor, GS-444217, to ask whether: (i) an ASK1 inhibitor can suppress p38 and JNK activation in crescentic glomerulonephritis, and (ii) an ASK1 inhibitor can suppress disease development. We addressed these questions in two well characterized models of nephrotoxic serum nephritis in different rat strains: acute (day 1) glomerular inflammation in outbred Sprague-Dawley (SD) rats and progressive (day 14) crescentic disease in inbred Wistar Kyoto (WKY) rats.

| Immunostaining and quantification
Immunoperoxidase staining for CD68+ macrophages and fibrinogen was performed on formalin-fixed paraffin sections. Antigen retrieval using 10 mmol/L sodium citrate was followed by an avidin-biotin peroxidase complex staining method as previously described. 27,28 Immunoperoxidase staining for RP1+ neutrophils and R73+ T cells was performed on frozen sections of tissue fixed in 2% paraformaldehyde-lysine-periodate. Immunofluorescence staining for glomerular deposition of sheep IgG, rat IgG and rat C3 was performed on snap frozen tissues.

| Real time RT-PCR
A section of kidney cortex was snap frozen and stored at À80°C until use. RNA was extracted from frozen tissues using the RiboPure RNA isolation kit (Ambion, Austin,TX, USA). RNA was reverse transcribed with random hexamer primers and the Superscript II kit (Invitrogen, Carlsbad, CA, USA). Real-time RT-PCR was performed on a StepOne machine (Applied Biosystems, Mulgrave, Vic., Australia) as previously described. 17,29 Probes and primers for detection of TNFa, NOS2, KIM1, CD68, MCP-1, MMP-12, CD206, FoxP3, Col I, a-SMA, TGF-b1 have been described previously. 17,29 Primers and probes for MMP-9, CD3e, IL-2, RANTES and 18S ribosomal RNA were purchased from Applied Biosystems. All amplicons were normalized against the 18S RNA internal control (Applied Biosystems).
The relative amount of individual mRNA species was calculated using the comparative Ct method.

| Western blotting
A section of kidney cortex was snap-frozen and stored at À80°C until use. Frozen kidney samples were homogenized in RIPA lysis buffer as previously described. 30 Samples were run on 12% SDS-PAGE gels and then electro-blotted on to nitrocellulose membranes.
Labelled protein bands were detected using the Odyssey Infrared Image Detection system (LICOR). Detection of a-tubulin was used as the loading control. Densitometry analysis was performed as described. The Gel Pro Analyzer program (Media Cybernetics) was used for densitometry analysis.

| Statistical analysis
Data are presented as mean AE standard deviation. Analysis was performed by ANOVA using Tukey's or the Kruskal-Wallis post-test for multiple comparisons. Analyses were performed using GraphPad Prism 6.0 software (San Diego, CA, USA).

| GS-444217 inhibits the development of crescentic disease in progressive NTN in WKY rats
To investigate whether ASK1 inhibition can prevent the development of progressive crescentic glomerulonephritis, we examined the NTN model in susceptible inbred WKY rats. 31,32 Compared to SD rats, induction of NTN in WKY rats causes less severe glomerular injury on day 1, but WKY rats reliably develop crescentic lesions by day 14. 31,32 Induction of NTN in vehicle-treated WKY rats resulted in proteinuria on days 7 and 14, together with significant renal function impairment on day 14 ( Figure 3A,B). This was associated with marked glomerular damage on day 14 NTN featuring hypercellularity, thrombosis, atrophy and crescent formation ( Figure 3D Figure 6H), whereas mRNA levels of CCL5 and FoxP3 expression were not significantly affected by drug treatment (Figure 6G,I).

| DISCUSSION
The current study demonstrated that the ASK1 inhibitor, GS-444217, is effective in suppressing activation of p38 and, to a lesser degree, JNK pathways in the acute and progressive phases of rat models of NTN. This extends previous studies in which Ask1 gene deletion or GS-444217 treatment identified that ASK1 signalling is required for pathologic activation of p38 and, to a lesser degree, JNK pathways in the mouse obstructed kidney and in mouse diabetic nephropathy. 23,24 Thus, in three distinct models of renal injury induced by diverse insults (diabetes, mechanical stretch and immunity), it is possible that oxidative stress-induced activation of ASK1 is The development of glomerular crescents, renal fibrosis and renal impairment is characteristic features in the progression of the rat NTN model. 29,35 The ability of GS-444217 treatment to reduce proteinuria, crescent formation, renal fibrosis and renal impairment in this model is consistent with studies employing p38 or JNK inhibitors. 8,9,11,16,17 Fibrin deposition plays a role in crescent formation in this model, 36 and ASK1/p38 signalling can promote platelet activation in response to pathologic stimuli. 37 Consistent with these findings, both GS-444217 treatment and p38 inhibition significantly reduced glomerular fibrin deposition and crescent formation in rat NTN. 11 Macrophages are essential for crescent formation in the NTN model. 31,38 Blockade of p38 suppressed glomerular macrophage recruitment and renal expression of the monocyte/macrophage chemokine, CCL2, in association with reduced crescent formation. 8,9,11 In contrast, JNK blockade inhibits macrophage activation rather than recruitment while protecting against the development of glomerular crescents and renal impairment in the NTN model. 16,17 Consistent with these studies, we found that GS-444217 treatment reduced macrophage recruitment and macrophage activation. In particular, expression of MMP-9 and MMP-12; enzymes known to facilitate F I G U R E 7 Renal fibrosis in day 14 nephrotoxic serum nephritis (NTN). Disease was induced in groups of 8 WKY rats which were treated with GS-444217 (drug; NTN-D) or vehicle (NTN-V) alone. A-C, Immunostaining for collagen IV: A, normal rat kidney shows collagen IV in the glomerular and tubular basement membranes; B, vehicle-treated disease shows increased deposition of collagen IV in both glomeruli and the interstitium; C, drug treatment reduces shows a mild increase in collagen IV staining compared to normal. D-F, Immunostaining for a-SMA: D, normal rat kidney shows a-SMA staining in vessel walls; E, vehicle-treated disease shows a-SMA+ myofibroblasts within glomeruli, in the periglomerular area and in the interstitium, and; F, drug-treated disease shows only small numbers of a-SMA+ cells in glomeruli and the interstitium. RT-PCR analysis of mRNA levels for: G, collagen I; H, a-SMA; and; I, TGF-b1. Data are mean AE SD. Analysis by one-way ANOVA with Tukey's multiple comparison test AMOS ET AL.
| 4531 macrophage-mediated renal injury and crescent formation in the NTN model, [39][40][41] was reduced by GS-444217. Furthermore, inhibition of pro-inflammatory and pro-fibrotic molecule expression by GS-444217 is likely to be a combination of direct and indirect effects. In vitro studies show a direct role for JNK and p38 signalling in up-regulation of both pro-inflammatory (eg, TNF-a, NOS2, MCP-1, MMP-12) and pro-fibrotic (eg, Col I, a-SMA and TGF-b1) molecules. 6,17,23,34 However, while GS-444217-induced suppression of pro-inflammatory molecule expression is likely to be a direct effect of p38/JNK blockade-especially on day 1 NTN-the reduction in pro-fibrotic molecule expression is most probably an indirect, secondary effect of suppressing renal inflammation.
T cells play an important role in crescent development in the rat NTN model. 42  Animal data combined with human biopsy studies have identified the p38 and JNK signalling pathways as potential therapeutic targets. 22 However, both p38 and JNK inhibitors have failed in clinical trials of nonrenal diseases, in part because of toxicity issues, indicating that complete blockade of p38 or JNK signalling is not desirable. 19,20 In contrast to the foetal lethal phenotype of mice lacking p38 or Jnk1/2 genes, mice lacking Ask1 have a normal phenotype arguing that ASK1 function is not required for normal homoeostatic mechanisms. 3,22 Indeed, the requirement for oxidative stress in ASK1-dependent activation of p38/JNK signalling means that, at least in theory, blocking ASK1 will prevent p38/JNK signalling in some pathologic conditions but should have little effect upon physiological p38/JNK signalling and thereby limit toxicity. ASK1 inhibitor treatment has been reported to be safe and efficacious to reduce liver fibrosis in a phase II trial of nonalcoholic steatohepatitis, 44 while the results of a trial evaluating ASK1 inhibitor treatment in patients with diabetic kidney disease have yet to be reported.
In conclusion, this study demonstrates that administration of an ASK1 inhibitor suppressed activation of the p38 and, to a lesser extent, the JNK signalling pathway in a rat model of crescent glomerulonephritis. ASK1 inhibitor treatment was effective in suppressing both acute glomerular injury and the development of crescentic disease. ASK1 blockade is a potential therapeutic strategy in rapidly progressive glomerulonephritis.