and Immunity (2008) 9, 395–404 & 2008 Macmillan Publishers Limited All rights reserved 1466-4879/08 $30.00 www.nature.com/gene

ORIGINAL ARTICLE -20 induced cell death in renal epithelial cells and was associated with acute renal failure

H-H Li1,2, Y-H Hsu3, C-C Wei1,4, P-T Lee5,6, W-C Chen7 and M-S Chang1,2,3,8 1Institute of Basic Medical Sciences, College of Medicine, National Cheng Kung University, Tainan, Taiwan; 2Department of Medical Research, Chi Mei Medical Center, Tainan, Taiwan; 3Institute of Biopharmaceutical Sciences, College of Medicine, National Cheng Kung University, Tainan, Taiwan; 4Department of Medical Technology, Chung Hwa University of Medical Technology, Tainan, Taiwan; 5Institute of Clinical Medicine, College of Medicine, National Cheng Kung University, Tainan, Taiwan; 6Division of Nephrology, Department of Medicine, Kaohsiung Veterans General Hospital, Kaohsiung & National Yang-Ming University, School of Medicine, Taipei, Taiwan; 7Department of Pathology, College of Medicine, National Cheng Kung University, Tainan, Taiwan and 8Department of Biochemistry and Molecular Biology, College of Medicine, National Cheng Kung University, Tainan, Taiwan

Acute renal failure is an abrupt decrease in renal function. Interleukin (IL)-10 inhibits ischemic and cisplatin-induced acute renal failure. We aimed to determine whether IL-20 affects renal tubular epithelial cells and is associated with acute renal failure. We analyzed the expression of IL-20 and its receptor (R) in the kidneys of rats with HgCl2-induced acute renal failure. Reverse transcription-PCR showed upregulated IL-20, and its receptors and immunohistochemical staining showed strongly expressed IL-20 in proximal tubular epithelial cells. We analyzed human proximal tubular epithelial (HK-2) cells, which expressed both IL-20 and its receptors. IL-20 specifically induced mitochondria-dependent apoptosis by activating caspase 9 in HK-2 cells. IL-20 also activated c-Jun N-terminal kinase and extracellular signal-regulated kinase 1/2, the downstream signals implicated in the apoptosis of HK-2 cells. Furthermore, IL-20 upregulated the transcripts of transforming (TGF)-b1, a critical mediator of renal injury. In hypoxic HK-2 cells, IL-20 and IL-22R1 transcripts increased, and IL-20 upregulated IL-1b transcripts. In vivo study further demonstrated that anti-IL-20 antibody reduced the expression of TGF-b1 and IL-1b and the number of damaged tubular cells in the kidneys of rats with acute renal failure. We concluded that IL-20 may be involved in the injury of renal epithelial cells in acute renal failure. Genes and Immunity (2008) 9, 395–404; doi:10.1038/.2008.28; published online 22 May 2008

Keywords: IL-20; renal epithelial cell; apoptosis; necrosis

Introduction Acute renal failure is a rapid drop in the glomerular filtration rate that leads to the accumulation of nitrogen- Interleukin (IL)-20 belongs to the IL-10 family, which ous waste such as serum creatinine and blood urea includes IL-10, IL-19, IL-20, IL-22, IL-24, and IL-26.1 IL-20 nitrogen (BUN). There are many causes of acute renal is preferentially expressed in five major cell types: failure, but the most common is injury of renal tubular epithelial cells, myoepithelial cells, endothelial cells, epithelial cells by ischemia or cytotoxic regents,8 which macrophages and skeletal muscle cells.2 IL-20 induced resulted in their necrosis or apoptosis.9 There are two signal transducer and activator of transcription 3 activa- major apoptotic pathways in human ischemic acute renal tion. IL-20 directly activated signal transducer and failure. The intrinsic pathway requires translocation of activator of transcription 3 either through IL-20 receptor Bax to the mitochondria followed by the release of (R) complex IL-20R1/IL-20R2 or through IL-20 receptor cytochrome c and activation of caspase 9. The extrinsic complex IL-22R1/IL-20R2.3 IL-20 has potent inflamma- pathway is through the activation of caspase 8.10 4,5 6 tory, angiogenic and chemoattractive characteristics. Although animals with HgCl2-induced nephrotoxicity IL-20 targeted , endothelial cells and syno- provide a well-established animal model for studying vial fibroblasts and was associated with ,7 acute renal failure, we do not have a detailed under- rheumatoid arthritis6 and atherosclerosis.5 standing of the toxicological mechanism of this model. One possibility is lipid peroxidation associated with the development of mild necrosis in proximal tubule Correspondence: Professor M-S Chang, Department of Biochemi- epithelial cells.11 Another possibility is mercury-induced stry and Molecular Biology, National Cheng Kung University, ion imbalances leading to renal toxicity.12 Moreover, College of Medicine, Tainan 70428, Taiwan. nephrotoxin-reduced renal blood flow resulted in renal E-mail: [email protected] ischemia.13 This work was supported by a grant from Chi Mei Medical Center, Tainan, Taiwan. Interleukin-10 is a pleiotropic with many Received 24 October 2007; revised 24 January 2008; accepted 20 immunosuppressive effects and a few immunostimula- March 2008; published online 22 May 2008 tory effects.14 Previous studies found that IL-10 IL-20-induced epithelial cell apoptosis H-H Li et al 396 decreased renal injury in animal models treated with and 2 (Figure 2B). We also investigated what kinds of either cisplatin or ischemia.15 IL-10 dramatically inhib- cells expressed rIL-20 in the kidneys of experimental rats. ited both inflammatory, cytotoxic and apoptotic path- rIL-20 was expressed in Henle’s loop and distal tubular ways of renal injury. As IL-20 is a member of the IL-10 epithelial cells in pretreated control rats (Figure 2Ca). family, we investigated whether IL-20 was also involved Two days after experimental rats had been injected with

in the pathogenesis of acute renal failure. HgCl2, rIL-20 was strongly expressed in Henle’s loop, distal and proximal tubular epithelial cells (Figure 2Cc), Results

Generating HgCl2-induced acute renal failure in rats

To confirm that the rats in our animal model had HgCl2- induced acute renal failure, we analyzed their renal function. We collected their sera from day 0 to day 8. Serum creatinine levels began to rise on day 1, peaked on day 2 and returned to baseline on day 8 (Figure 1a). BUN levels began to rise on day 1, peaked on day 4 and returned to baseline on day 8 (Figure 1b).

Transcripts of rat (r)IL-20 and its receptors increased in the

kidneys of rats with HgCl2-induced acute renal failure To investigate whether IL-20 is involved in acute renal failure, we sacrificed the rats at different time points and compared the transcripts of rIL-20 and its receptors with those of healthy controls. Reverse transcription-PCR (RT- PCR) revealed that rIL-20, rIL-20R1, rIL-20R2 and rIL- 22R1 had increased in the kidneys of the experimental rats (those with acute renal failure) and had peaked on days 1 and 2 (Figure 2A).

Protein expression of rIL-20 in the kidneys of experimental rats To confirm that the rIL-20 protein level as well as the transcript level had increased, we analyzed the kidney tissue using immunohistochemical staining with anti-IL- 20 antibodies. The expression levels of rIL-20 protein were upregulated and were strongly stained on days 1

Figure 2 Transcript levels and immunohistochemical staining of

IL-20 and its receptors in the kidneys of rats with HgCl2-induced acute renal failure. Experimental rats were sacrificed at different times as indicated (day (D) 1 to D8) after they had been injected

with HgCl2 on D0. (A) Total RNA was isolated from their kidneys for RT-PCR analysis with equal amounts of cDNA and primers specific for rat (r)IL-20, (r)IL-20R1, (r)IL-20R2 and (r)IL-22R1 to amplify the transcripts. b-actin was an internal control. Ctrl, control;

H12, 12 h after injection of HgCl2.(B) Paraffined sections of kidney tissue were stained using anti-IL-20 monoclonal antibody 7E. The reaction was detected using 3-amino-9-ethylcarbazole chromogen stain (red), and the nuclei were counterstained with hematoxylin (blue) (magnification: Â 400). Experimental rats were killed at different times (as indicated) after they had been injected with

HgCl2 on day 0. (C) The rats were killed on (a, b) day 0 and (c, d)

Figure 1 Renal function in rats with HgCl2-induced acute renal day 2. In (a) and (c): P, proximal tubular epithelial cells; D, distal failure. (a) Serum creatinine and (b) BUN levels were detected at tubular epithelial cells; H, Henle’s loop. In (b) and (d): (n), different times (as indicated) after they had been injected with infiltrated immune cells; (k), urothelium in the renal pelvis; (m),

HgCl2 on day 0. Values are means±s.e. (n ¼ 3/group). *Po0.05 smooth muscle cells. The figures represent similar patterns in three compared with controls (day 0). individual specimens.

Genes and Immunity IL-20-induced epithelial cell apoptosis H-H Li et al 397

Figure 3 HK-2 cells expressed the transcripts and of IL-20, IL-20R1, IL-20R2 and IL-22R1. (a) Total RNA was isolated for RT-PCR analysis with primers specific for human (h)IL-20, (h)IL-20R1, (h)IL-20R2 and (h)IL-22R1 to amplify the transcripts. NTC was a non-template control. (b) HK-2 cells were immunocytochemically stained using different monoclonal antibodies (as indicated). Mouse (m)IgG1 was a negative control. The reaction was detected using 3-amino-9-ethylcarbazole chromogen stain (red), and the nuclei were counterstained with hematoxylin (blue).

infiltrated immune cells, urothelium in the renal pelvis, immunofluorescence images showed apoptotic cells and smooth muscle cells (Figure 2Cd). (green) (Figure 5c). There are two major apoptotic pathways in human IL-20 and its receptors were expressed in human proximal ischemic acute renal failure; the intrinsic and extrinsic tubular epithelial cells pathways. To investigate the IL-20 pathway through As we found that rIL-20 was overexpressed in the which apoptosis was induced, we used Western blotting kidneys of experimental rats, and that proximal tubular with anti-caspase 8 and 9 antibodies. We found more epithelial cells play an important role in the pathogenesis cleaved caspase 9 in treated than in untreated HK-2 cells of acute renal failure, we used human proximal tubular (Figure 5d) and that the amount was dose-dependent. epithelial cells (HK-2) to perform in vitro study. We found We also found only the precursor form of caspase 8 and that IL-20 and its receptors were expressed in HK-2 cells that the level was not changed after treatment with hIL- at both transcript (Figure 3a) and protein levels 20 (data not shown). The result demonstrated that IL-20 (Figure 3b). induced apoptosis through activation of caspase 9, the mitochondrial pathway. IL-20 dose-dependently and specifically induced cell death in HK-2 cells IL-20 activated c-Jun N-terminal kinase and extracellular As proximal tubular epithelial cell death was a critical signal-regulated kinase 1/2 in HK-2 cells step in the pathogenesis of acute renal failure, we treated To evaluate the downstream signals induced by IL-20 in HK-2 cells with different concentrations of IL-20, or proximal tubular epithelial cells, HK-2 cells were treated mitomycin C as a positive control, for 24 h, and analyzed with IL-20 and cell lysates were analyzed using Western the percentage of cell death using flow cytometry. IL-20 blotting with antibodies specifically against phosphory- dose-dependently induced cell death in HK-2 cells: lated c-Jun N-terminal kinase (JNK), extracellular signal- 5.68 nM of IL-20 induced 34.20% cell death, 11.36 nM regulated kinase (ERK) 1/2, signal transducer and induced 45.87% cell death, and 22.72 nM induced activator of transcription 3 and p38 mitogen-activated 53.33% cell death (Figure 4A). To demonstrate the protein kinase. Phosphorylation of JNK and ERK 1/2 specificity of IL-20-induced cell death, we pre-incubated increased in IL-20-treated HK-2 cells, and maximal IL-20 and anti-IL-20 monoclonal antibody 7E on ice for phosphorylation occurred 10 min after treatment 30 min, and then treated HK-2 cells with it for 24 h. The (Figure 6). We found no phosphorylation of signal 7E antibody completely blocked IL-20-induced cell death transducer and activator of transcription 3 or p38 in HK-2 cells, but treatment with 7E alone had no effect mitogen-activated protein kinase in HK-2 cells after IL-20 in this assay (Figure 4b and C). Figure 4b was the original treatment (data not shown). histogram of Figure 4c. IL-20 upregulated transforming growth factor-b1transcripts Human IL-20 induced mitochondria-dependent apoptosis Increasing evidence indicates that the inflammatory and necrosis of HK-2 cells by activating caspase 9 response plays a major role in acute renal failure.16 To investigate the pathway through which IL-20 induced Injured proximal tubular epithelial cells generate a cell death, we treated HK-2 cells with or without IL-20 number of mediators that potentiate the inflammatory for 18 h, stained them with annexin-V/propidium iodide response, such as proinflammatory and (PI) and then analyzed the percentage of apoptotic and chemotactic cytokines.10 HK-2 cells expressed IL-20 necrotic cells. IL-20 induced 22.24% apoptosis of HK-2 receptors (Figures 3a and b) and may be able to produce cells and 17.88% necrosis of HK-2 cells (Figures 5a and cytokines and in response to IL-20. To test b). To further confirm that apoptotic effect, we treated this possibility, we analyzed the effects of IL-20 on the HK-2 cells with or without IL-20 for 24 h and then did a production of cytokines (transforming growth factor terminal deoxynucleotidyl transferase-mediated uridine (TGF)-b1, -a, IL-6 and IL-1b) and triphosphate nick end labeling assay. Representative chemokines ( chemotactic protein-1, IL-8 and

Genes and Immunity IL-20-induced epithelial cell apoptosis H-H Li et al 398

Figure 4 IL-20 dose-dependently and specifically induced apoptosis in HK-2 cells. (A) HK-2 cells were treated with different reagents for 24 h. (B) and (C) were results of the same experiment shown by histogram and bar. MMC: treated with mitomycin C as a positive control; IL-20: treated with 5.68 nM of IL-20; IL-20 and 7E: 5.68 nM of IL-20 pre-incubated with 6.67 nM of anti-IL-20 monoclonal antibody 7E for 30 min and then added to HK-2 cells for 24 h; 7E: treated with 6.67 nM of 7E alone. Cells were fixed using ethanol, stained using propidium iodide (PI) and analyzed using flow cytometry. The percentages of dead cells (M1 region) are quantified in (C).

regulated on activation normal T cell expressed and secreted) were induced by IL-20 treatment (data not secreted) in HK-2 cells. After HK-2 cells had been shown). incubated with IL-20 for 6 h, we analyzed the induction of transcripts using RT-PCR with primers specific to IL-20 upregulated IL-1b transcripts in hypoxic HK-2 cells cytokines and chemokines. Only TGF-b1 transcript was The injured renal tubular cells generated a number of upregulated (Figures 7a and b). None of the other proinflammatory cytokines. However, we did not find cytokines (tumor necrosis factor-a, IL-6 and IL-1b)or that IL-20 treatment of normoxic HK-2 cells significantly chemokines (monocyte chemotactic protein-1, IL-8 and upregulated proinflammatory cytokines or chemokines regulated on activation normal T cell expressed and other than TGF-b1. Therefore, we speculated that the

Genes and Immunity IL-20-induced epithelial cell apoptosis H-H Li et al 399

Figure 5 IL-20 induced apoptosis and necrosis in HK-2 cells. HK-2 cells were treated with or without human (h)IL-20 for 18 h. Cells were stained with propidium iodide (PI) and annexin-V and then analyzed using flow cytometry. (a) Dot plots are shown. (b) The percentage of apoptotic and necrotic cells are shown in bar. (c) The apoptotic cells were stained using a terminal deoxynucleotidyl transferase-mediated uridine triphosphate nick end labeling assay (green) with immunofluorescent stain. 40,6-Diamidino-2-phenylindole was used to stain cell nuclei (blue). Merge: the left two panels merged (magnification: Â 100). (d) Cleaved caspase 9 with molecular weights of 37 kDa is shown. b-actin was an internal control.

proinflammatory cytokines might be upregulated in and regulated on activation normal T cell expressed and hypoxic cells. We incubated HK-2 cells with IL-20 in a secreted) were induced by IL-20 treatment (data not hypoxic (1% O2) or normoxic (18% O2) chamber for 8 h. shown). All samples were analyzed using RT-PCR with primers specific to cytokines and chemokines. HK-2 cells en- Hypoxia upregulated the transcripts of IL-20 and IL-22R1 dogenously expressed IL-1b, which significantly in- Renal ischemia is a major factor involved in the creased in hypoxic HK-2 cells but not in normoxic cells pathogenesis of human acute renal failure. IL-20 and its (Figures 7c and d). None of the other cytokines (for receptors were upregulated in the kidneys of experi- example, tumor necrosis factor-a and IL-6) or chemo- mental rats, and HgCl2-reduced renal blood flow kines (for example, monocyte chemotactic protein-1, IL-8 resulted in renal ischemia. Therefore, we speculated that

Genes and Immunity IL-20-induced epithelial cell apoptosis H-H Li et al 400 IL-20 and its receptors might be upregulated under the damaged areas of tubular cells and decreased the hypoxia. To test this possibility, we incubated HK-2 protein level of both TGF-b1 and IL-1b in the kidneys of cells in a hypoxic chamber for 48 h and analyzed the rats with acute renal failure (Figure 9). TGF-b1 was transcripts of IL-20 and its receptors. Hypoxia upregu- strongly expressed in the tubular and endothelial cells in

lated the transcripts of IL-20 and IL-22R1, but did not rats with HgCl2-induced acute renal failure, but only significantly upregulate the transcripts of IL-20R1 and weakly expressed in rats treated with 7E. IL-1b showed a IL-20R2 in HK-2 cells (Figure 8). pattern similar to that of TGF-b1, but IL-1b was expressed only in tubular cells, not in endothelial cells. Anti-hIL-20 monoclonal antibody protected rats with This in vivo result is consistent with the in vitro effect of

HgCl2-induced acute renal failure IL-20 on HK-2 cells. To further demonstrate that IL-20 was indeed involved in acute renal failure, rats were treated with both anti-hIL- 20 monoclonal antibody 7E and HgCl2. The damaged Discussion areas of tubular cells and the expression levels of TGF-b1 and IL-1b in the kidneys of rats with acute renal failure This study proposes a new mediator, IL-20, for acute treated with 7E were analyzed. IL-20 antibody reduced renal failure. We confirmed the association of IL-20 with acute renal failure in vivo, and in vitro studies showed that IL-20 induced mitochondria-dependent apoptosis of human renal epithelial cells and upregulated TGF-b1 and IL-1b transcripts, both of which are involved in the pathogenesis of acute renal failure.10 We found that IL-20 and its receptors were over- expressedinthekidneysofratswithacuterenalfailure, which indicated an association between IL-20 and acute renal failure. IL-20 levels showed a trend similar to those of serum creatinine and BUN levels, which implied that IL-20 contributes to the severity of the acute renal failure. We also generated another animal model of ischemia/reperfusion-

induced acute renal failure in addition to HgCl2-induced Figure 6 IL-20 activated JNK and ERK 1/2 signals in HK-2 cells. HK-2 cells were treated with human (h)IL-20 for the indicated renal failure. IL-20 and its receptors were similarly over- times. Cell lysates were collected and the levels of phospho-JNK expressed in the kidneys of these rats (data not shown). and phospho-ERK 1/2 were detected using Western blotting with Interleukin-20 was upregulated in the kidneys of specific antibodies. b-actin was a loading control. experimental rats and expressed strongly in Henle’s

Figure 7 IL-20 upregulated TGF-b1and IL-1b transcripts. (a) HK-2 cells were treated with or without 11.36 nM of human (h)IL-20 for 6 h. Total RNA was isolated for RT-PCR analysis with primers specific for TGF-b1 to amplify the transcript. b-actin was an input control. (b) The relative quantity of PCR products (TGF-b1) was expressed as fold of increase relative to untreated control. (c) Hypoxic (1% oxygen) and normoxic (18% oxygen) HK-2 cells were treated with or without 11.36 nM of IL-20 for 8 h. Total RNA was isolated for RT-PCR analysis with primers specific for IL-1b to amplify the transcript. NTC was the non-template control. b-actin was an input control. (d) The relative quantity of PCR products (IL-1b) was expressed as fold of increase relative to normoxic HK-2 cells. Figure shown was the mean value±s.d. of five experiments.

Genes and Immunity IL-20-induced epithelial cell apoptosis H-H Li et al 401

Figure 8 Hypoxia upregulated the transcripts of IL-20 and IL-22R1. (a) Hypoxic (1% oxygen) and normoxic (18% oxygen) HK-2 cells were incubated for 48 h. Total RNA was isolated for RT-PCR analysis with primers specific for human (h)IL-20, (h)IL-20R1, (h)IL-20R2 and (h)IL- 22R1 to amplify the transcript. NTC was the non-template control. b-actin was an input control. (b) The relative quantity of PCR products was expressed as fold of increase relative to normoxic HK-2 cells. Figure shown was the mean value±s.d. of five experiments.

as a target for IL-20 in our in vitro studies. Other cells also expressed IL-20 in the kidneys of experimental rats, which implied that IL-20 may have different roles in these cells. For example, urothelium controls perme- ability, immune responses and cell–cell communica- tion;17 infiltrating cells generate cytokines, chemokines and growth factors and play roles in both injury and regeneration process.18 Thus, IL-20 may have various functions in the pathogenesis of acute renal failure. Previous studies showed that IL-20 targeted keratino- cytes,19 synovial fibroblasts6 and endothelial cells.4 In this study, we found a new target for IL-20: renal tubular epithelial cells. Human tubular epithelial cells expressed both IL-20 and its receptors, which indicated that IL-20 may act on epithelial cells in an autocrine manner. This study also showed that IL-20 induced apoptosis and necrosis in renal epithelial cells. This is the first report of this novel function. Among the members of IL-10 family, IL-19 induced apoptosis of human lung epithelial cells, A549 cells,20 and IL-24 induced apoptosis of many kinds of tumor cells21 and activated nuclear factor-kB to supply survival signals in normal cells.22 Apoptosis and necrosis of renal proximal tubular epithelial cells are important in acute renal failure. Therefore, IL-20 may be significantly involved in the pathogenesis of acute renal failure. Apoptosis and Figure 9 IL-20 antibody reduced tubular damage and the expres- necrosis of tubular epithelial cells were also involved in b b sion of TGF- 1 and IL-1 in the kidneys of rats with HgCl2-induced many renal diseases such as chronic23 and acute renal acute renal failure. Rats were injected with HgCl2 alone, anti-hIL-20 failure,24 renal stone formation25 and autosomal domi- monoclonal antibody 7E alone, or both. The rats were given HgCl2 26 on day 0, and antibody (7E or mouse IgG1) on days À2, À1, 0 and 1. nant polycystic kidney. Therefore, we postulated that (a) The kidney sections were stained with hematoxylin and eosin IL-20 may also be involved in the pathogenesis of other (magnification: Â 400). The damaged tubular cells are indicated by renal diseases characterized by epithelial cell injury. the arrows. (b) Quantitative analysis of the damaged area of tubular Extracellular signal-regulated kinase 1/2 activation is cells. (c) The reaction was detected using 3-amino-9-ethylcarbazole required for apoptosis in renal epithelial cells,27,28 and chromogen stain (red), and the nuclei were counterstained with hematoxylin (blue) (magnification: Â 400). The figures represent JNK cascade activation is considered an important similar patterns in two individual specimens. intermediate of cell apoptosis, especially in apoptosis of tubular cells.29 Therefore, JNK and ERK 1/2 activation by IL-20 in HK-2 cells supports the notion that IL-20 induces loop, distal and proximal tubular epithelial cells, cell death in HK-2 cells. infiltrated immune cells, urothelium in the renal pelvis, Transforming growth factor-b1 is a critical mediator of and smooth muscle cells. We used tubular epithelial cell renal injury. It promotes excessive inflammation and

Genes and Immunity IL-20-induced epithelial cell apoptosis H-H Li et al 402 extracellular matrix deposition, thereby contributing to approved by the Animal Care and Use Committee of the the progression of renal disease. TGF-b1 is the main animal center in National Cheng Kung University, inducer of epithelial-to-mesenchymal transition in the Tainan, Taiwan. kidney30 and also promotes tubular apoptosis by activating caspases.31 Some studies32 report that TGF-b1 Reagent promoted tissue regeneration followed by acute injury. The human IL-20 protein was purchased from R&D Thus, the induction of TGF-b1 by IL-20 might work as a Systems Inc. (Minneapolis, MN, USA). The generation double-edged sword involved in both the damage and and specificity of anti-hIL-20 antibody, 7E, was per- repair of proximal tubular epithelial cells. formed using a standard protocol as described in a Cisplatin-induced acute renal failure is associated with previous study.7 the increase of cytokines, IL-1b, IL-8 and IL-6, as well as neutrophil infiltration in the kidney.33 IL-20 gene expres- Semiquantitative RT-PCR sion is induced by IL-1b in cultured normal human Total RNA of rat kidneys or HK-2 cells were extracted keratinocytes.34 IL-1b plays a major role in the proin- using Trizol (Invitrogen, Carlsbad, CA, USA) and then flammatory response. Interestingly, we found that IL-20 total RNA underwent reverse transcription using Super- upregulated the transcription of IL-1b only in hypoxic Script II (Invitrogen) according to the manufacturer’s cells and anti-IL-20 antibody reduced the expression of instructions. Different gene transcripts were amplified HgCl2-induced IL-1b in vivo. It is possible that some using PCR with gene-specific primers. Each RT-PCR hypoxia-induced proteins work together with IL-20 or reaction was performed five times. The relative quantity promote the upregulation of IL-1b by IL-20. In addition, of PCR products versus internal b-actin control was the increased IL-22R1 in hypoxic HK-2 cells might make average of five experiments and expressed as mean them more sensitive to IL-20. We found that IL-20 was value±s.d. The forward and reverse primers are listed upregulated in hypoxic cells, and other studies35 showed below (*used in nested PCR; F, forward primer; R, that HgCl2 induced renal ischemia. Therefore, HgCl2- reverse primer; r, primer for rat genes; h; primer for induced hypoxia may contribute to the overexpression of human genes): rIL-20-F1: 50-TGAACTTCCACTTCTG IL-20 in the kidneys of rats with acute renal failure. Renal GTCTC-30; rIL-20-R1: 50-TCTGGGACTCTTGGACAGGA- ischemia is one major factor involved in the pathogenesis 30; *rlL-20-F2: 50-ATGAGAGGCTTTCGTCTTGC-30;*rIL- of human acute renal failure.36 Thus, IL-20 may also be 20-R2: 50-TAACATCTGCTTCATCCATCT-30; rIL-20R1-F1: induced in the kidneys of patients with acute renal 50-TTGCCCACATCTGTCATTGTG-30; rIL-20R1-R1: 50-C failure and injured renal tubular epithelial cells. CCACTGAACTCAGGGTTGT-30; *rIL-20R1-F2: 50-TTCTC The potent inflammatory, angiogenic and chemoat- TGCGATTGGCTACTCA-30; *rIL-20R1-R2: 50-TACGCTGA tractive characteristics of IL-20 have been delineated, as CCTCATCACTGC-30; rIL-20R2-F1: 50-AAAACAGTGTTT has its involvement in many diseases.37 This study found CAGGGCAAG-30; rIL-20R2-R1: 50-GAAAACAGAGACA a novel biological function of IL-20: the induction of CAGCCCT-30; *rIL-20R2-F2: 50-TTTCTCACTGATGGAGC mitochondria-dependent apoptosis in renal epithelial GAC-30; *rIL-20R2-R2: 50-CGGAAGAGACTCTCCVVGC cells by activating caspase 9; it also found that IL-20 is A-30; rIL-22R-F1: 50-TACTCCTTCTCGGGTGCAGT-30; associated with acute renal failure and that IL-20 rIL-22R-R1: 50-TAGGTGAGGTCAGACAGGCT-30; 0 0 antibody significantly protected rats of HgCl2-induced *rIL-22R-F2: 5 -CCAAGCCACCTGTACCTCCT-3 ; *rIL- acute renal failure. The mechanism may be through 22R-R2: 50-AGACACCACCACTTGGGAGT-30;rb-actin-F: reducing tubular cell death, the expression of TGF-b1, 50-GGGMTGGGTCAGMGGACT-30;rb-actin-R: 50-TTTG 0 0 and IL-1b in the kidneys of rats with HgCl2-induced ATGTCACGCACGATTT-3 ; hIL-20-F1: 5 -CCT CCT GCG acute renal failure. All the in vivo results supported the in CCA TTT GCT AA-30; hIL-20-R1: 50-CACTGTGACCAG vitro results. These findings suggest that IL-20 or its CACAAGAC-30; *hIL-20-F2: 50-AAGATCAGCAGCCTC receptor complexes may become new therapeutic targets GCCAA-30; *hIL-20-R2: 50-CAGGTATTGAAGACTGGAG of drugs that block IL-20 activities. These drugs include CT-30; hIL-20R1-F1: 50-AGCTGGACAGCGAAGGAATA- antibodies or small molecules that neutralize IL-20, block 30; hIL-20R1-R1: 50-AGTGTGTGACCAACCACACG-30; the cell-surface receptors of IL-20 or inhibit the intracel- *hIL-20R1-F2: 50-TCTGGTATGTTTTGCCCGTA-30; *hIL- lular signals of IL-20. Our anti-hIL-20 monoclonal anti- 20R1-R2: 50-GCCTGCGACTCCAATAATGT-30; hIL-20R2- body may have such therapeutic potential. F: 50-GAAGTGGCCATTCTGCCTGCC-30; hIL-20R2-R: 50-GGGAATGGCCTCTCCTTGCAC-30; hIL-22R1-F1: 50-A CCCCAGACACGGTCTACAG-30; hIL-22R1-R1: 50-AGG 0 0 Materials and methods CTCCGGAGAAGGAGTAG-3 ; *hIL-22R1-F2: 5 -GCAA ACTCACGGAGCTCTAC-30; *hIL-22R1-R2: 50-CATGATG Animal model of mercuric chloride induced acute renal failure GTGCCAAGGAACTC-30; hIL-1b-F: 50-GTGGCAATGA Male Sprague–Dawley rats (280–320 g) were purchased GGATGACTTG-30; hIL-1b-R: 50-TGGTGGTCGGAGATT from BioLASCO Taiwan Co., Ltd. The rats were given CGTAG-30; hTGF-b1-F: 50-TTA TCTTTTGATGTCACCGG 0 3 mg per kg body weight of HgCl2 in normal saline by AGT-3; hTGF-b1-R: 5 -GTAGTGAACCCGTTGATGTC-3; subcutaneous injection and killed 0, 0.5, 1, 2, 4 and 8 days hb-actin-F: 50-GCTGGAAGGTGGACAGCGAG-30;hb- 0 0 after the HgCl2 injection. Their serum creatinine and actin-R: 5 -TGGCATCGTGATGGACTCCG-3 . BUN levels were analyzed to monitor the renal function. Polymerase chain reaction products were visualized The left kidneys were fixed in 3.7% formaldehyde for on 2% agarose gels containing ethidium bromide. b-actin immunohistochemical staining and the right kidneys was an internal control. The relative quantity of the PCR were stored in liquid nitrogen for RNA isolation. All products versus b-actin were analyzed using the Version procedures were carried out with strict adherence to the WorkssLS Image Acquisition and Analysis Software NIH Guide for the Care and Use of Laboratory Animals and (UVP, Upland, CA, USA).

Genes and Immunity IL-20-induced epithelial cell apoptosis H-H Li et al 403 Immunohistochemical staining Western blotting Anti-IL-20 monoclonal antibody (7E), anti-TGF-b1 Cells were seeded at 1 Â107 cells mlÀ1 in 10-cm dishes (Abcam, Cambridge, UK) and anti-IL-1b (R&D systems, and stimulated by different concentrations of IL-20 for Minneapolis, MN, USA) antibodies were used to per- 24 h. Western blotting with anti-caspase 8 and 9 (Cell form immunohistochemical staining as described Signaling, Danvers, MA, USA) antibodies was done 7 previously. Mouse isotype IgG1 was a negative control according to the manufacturer’s instructions. In addition, for anti-IL-20 antibody. Immunoreactivity was detected the HK-2 cells were seeded at 1 Â106 cells mlÀ1 in 6-cm using a 3-amino-9-ethylcarbazole substrate kit for dishes and starved for 17 h. They were then stimulated peroxidase (DakoCytomation, Carpinteria, CA, USA) by 11.36 nM of IL-20 for the indicated time periods. and nuclei were counterstained with hematoxylin. Western blotting with antibodies specific for phosphory- lated ERK 1/2 and JNK () was done Cell culture according to the manufacturer’s instructions. The HK-2 cell was a cell line of human proximal tubular epithelial cells, which was purchased from American In vivo study Type Culture Collection (CRL-2190). All in vitro experi- To protect the rats from acute renal failure, male ments were done using HK-2 cells between passages 10 Sprague–Dawley rats (200–220 g) were on days À2, À1, and 20. 0 and 1, subcutaneously injected with 3 mg per kg body weight of anti-hIL-20 monoclonal antibody 7E Immunocytochemical staining in phosphate buffer saline, and on day 0, they were The immunocytochemical staining was done as de- subcutaneously injected with 3 mg per kg body weight of scribed previously.6 Briefly, HK-2 cells were grown on HgCl2 in normal saline. Control rats were subcuta- sterile chamber slides, fixed and blocked, and then neously injected with HgCl2 or 7E alone. primary antibodies (anti-IL-20, anti-IL-20R1, anti-IL-20R2 or anti-IL-22R1 monoclonal antibodies) were added. Histology examination After incubation with secondary antibodies, the immu- Tubular damage, defined as ‘the loss of nuclei’, was noreactivity was detected. estimated by counting the percentage of damaged tubules in 20 high-power fields per section. Detection of cell death The renal epithelial cells were seeded at 1 Â106 cells mlÀ1 Statistical analysis in 6-cm dishes and stimulated by different reagents for Significant difference was evaluated with Student’s t-test 24 h. After removing the culture medium, the cells were using a statistical software package in Microsoft Excel. trypsinized and fixed with 50% ethanol. Before flow Results are given as means±s.e.m., and Po0.05 was cytometric analysis, cells were washed three times with considered statistically significant. cold phosphate buffer saline and stained with PI for

10 min. The presence of sub-G0/G1 phase was used as an indicator of cell death. The percentage of cells in sub-G0/ Conflict of Interests

G1 phase was quantitated using WinMDI 2.8 software. The authors declare no conflict of interests. Measurement of apoptosis and necrosis Apoptosis and necrosis were determined using an annexin-V/PI staining kit (BD Biosciences, San Jose, References CA, USA). In brief, cells were treated with different 1 Pestka S, Krause CD, Sarkar D, Walter MR, Shi Y, Fisher PB. reagents for the indicated times and cells were trypsi- Interleukin-10 and related cytokines and receptors. Annu Rev nized and resuspended in 50 ml of binding buffer Immunol 2004; 22: 929–979. containing fluorescein isothiocyanate-conjugated annex- 2 Hsing CH, Ho CL, Chang LY, Lee YL, Chuang SS, Chang MS. in-V and PI. After 15 min, 200 ml of binding buffer was Tissue microarray analysis of interleukin-20 expression. added and the cells were analyzed with a flow Cytokine 2006; 35: 44–52. cytometer. Ten thousand events were recorded from 3 Parrish-Novak J, Xu W, Brender T, Yao L, Jones C, West J et al. each assay. The annexin-V-positive cells were considered 19, 20, and 24 signal through two distinct receptor apoptotic cells, and the annexin-V-negative/PI-positive complexes. Differences in receptor-ligand interactions cells were considered necrotic cells.3 mediate unique biological functions. J Biol Chem 2002; 277: 47517–47523. 4 HsiehMY,ChenWY,JiangMJ,ChengBC,HuangTY,ChangMS. Terminal deoxynucleotidyl transferase-mediated uridine Interleukin-20 promotes angiogenesis in a direct and indirect triphosphate nick end labeling assay manner. Genes Immun 2006; 7: 234–242. The HK-2 cells were seeded on eight-well chamber slides 5 Chen WY, Cheng BC, Jiang MJ, Hsieh MY, Chang MS. IL-20 is and treated with or without 11.36 nM of IL-20 for 24 h. expressed in atherosclerosis plaques and promotes athero- After removing the culture medium, the breaks of DNA sclerosis in apolipoprotein E-deficient mice. Arterioscler strands induced by apoptosis were detected using an in Thromb Vasc Biol 2006; 26: 2090–2095. situ cell death detection kit (Roche, Basel, Switzerland). 6 Hsu YH, Li HH, Hsieh MY, Liu MF, Huang KY, Chin LS et al. Function of interleukin-20 as a proinflammatory molecule in The terminal deoxynucleotidyl transferase-mediated rheumatoid and experimental arthritis. Arthritis Rheum 2006; uridine triphosphate nick end labeling assay was done 54: 2722–2733. 0 according to the manufacturer’s instructions. 4 ,6-Diami- 7 Wei CC, Chen WY, Wang YC, Chen PJ, Lee JY, Wong TW et al. dino-2-phenylindole was used to stain cell nuclei. Finally, Detection of IL-20 and its receptors on psoriatic skin. Clin the slide was visualized with an Olympus microscope. Immunol 2005; 117: 65–72.

Genes and Immunity IL-20-induced epithelial cell apoptosis H-H Li et al 404 8 Rana A, Sathyanarayana P, Lieberthal W. Role of apoptosis of 25 Miyazawa K, Suzuki K, Ikeda R, Moriyama MT, renal tubular cells in acute renal failure: therapeutic implica- Ueda Y, Katsuda S. Apoptosis and its related genes in renal tions. Apoptosis 2001; 6: 83–102. epithelial cells of the stone-forming rat. Urol Res 2005; 33: 9 Lieberthal W, Koh JS, Levine JS. Necrosis and apoptosis in 31–38. acute renal failure. Semin Nephrol 1998; 18: 505–518. 26 Edelstein CL. What is the role of tubular epithelial cell 10 Devarajan P. Update on mechanisms of ischemic acute kidney apoptosis in polycystic kidney disease (PKD)? Cell Cycle 2005; injury. J Am Soc Nephrol 2006; 17: 1503–1520. 4: 1550–1554. 11 Yamada T. [Studies on the mechanisms of renal damages 27 Kim YK, Kim HJ, Kwon CH, Kim JH, Woo JS, Jung JS et al. induced by nephrotoxic compounds]. Nihon Hoigaku Zasshi Role of ERK activation in cisplatin-induced apoptosis in OK 1995; 49: 447–457. renal epithelial cells. J Appl Toxicol 2005; 25: 374–382. 12 Liu XY, Jin TY, Nordberg GF. Increased urinary calcium and 28 Zhuang S, Yan Y, Daubert RA, Han J, Schnellmann RG. ERK magnesium excretion in rats injected with mercuric chloride. promotes hydrogen peroxide-induced apoptosis through Pharmacol Toxicol 1991; 68: 254–259. caspase-3 activation and inhibition of Akt in renal epithelial 13 Chaudhari A, Kirschenbaum MA. Alterations in rabbit renal cells. Am J Physiol 2007; 292: F440–F447. microvascular prostanoid synthesis in acute renal failure. Am J 29 Kunduzova OR, Bianchi P, Pizzinat N, Escourrou G, Physiol 1988; 254: F684–F688. Seguelas MH, Parini A et al. Regulation of JNK/ERK 14 Ho AS, Moore KW. Interleukin-10 and its receptor. Ther activation, cell apoptosis, and tissue regeneration by Immunol 1994; 1: 173–185. monoamine oxidases after renal ischemia-reperfusion. FASEB 15 Deng J, Kohda Y, Chiao H, Wang Y, Hu X, Hewitt SM et al. J 2002; 16: 1129–1131. Interleukin-10 inhibits ischemic and cisplatin-induced acute 30 Hay ED, Zuk A. Transformations between epithelium and renal injury. Kidney Int 2001; 60: 2118–2128. mesenchyme: normal, pathological, and experimentally in- 16 Okusa MD. The inflammatory cascade in acute ischemic renal duced. Am J Kidney Dis 1995; 26: 678–690. failure. Nephron 2002; 90: 133–138. 31 Dai C, Yang J, Liu Y. Transforming growth factor-beta1 17 Lazzeri M. The physiological function of the urothelium— potentiates renal tubular epithelial cell death by a mechanism more than a simple barrier. Urol Int 2006; 76: 289–295. independent of Smad signaling. J Biol Chem 2003; 278: 18 Ghielli M, Verstrepen W, Nouwen E, De Broe ME. Regenera- 12537–12545. tion processes in the kidney after acute injury: role of 32 Basile DP, Rovak JM, Martin DR, Hammerman MR. Increased infiltrating cells. Exp Nephrol 1998; 6: 502–507. transforming growth factor-beta 1 expression in regenerating 19 Blumberg H, Conklin D, Xu WF, Grossmann A, Brender T, rat renal tubules following ischemic injury. Am J Physiol 1996; Carollo S et al. Interleukin 20: discovery, receptor identifica- 270: F500–F509. tion, and role in epidermal function. Cell 2001; 104: 9–19. 33 Faubel S, Lewis EC, Reznikov L, Ljubanovic D, Hoke T, 20 Hsing CH, Chiu CJ, Chang LY, Hsu CC, Chang MS. IL-19 in Somerset H et al. Cisplatin-induced ARF is associated with an involved in the pathogenesis of endotoxic shock. Shock 2008; increase in the cytokines IL-1{beta},IL-18, IL-6 and neutrophil 29: 7–15. infiltration in the kidney. J Pharmacol Exp Ther 2007; 322: 21 Sauane M, Gopalkrishnan RV, Sarkar D, Su ZZ, Lebedeva IV, 8–15. Dent P et al. MDA-7/IL-24: novel cancer growth suppressing 34 Otkjaer K, Kragballe K, Johansen C, Funding AT, and apoptosis inducing cytokine. Cytokine Growth Factor Rev Just H, Jensen UB et al. IL-20 gene expression is induced by 2003; 14: 35–51. IL-1beta through mitogen-activated protein kinase and 22 Aggarwal S, Takada Y, Mhashilkar AM, Sieger K, Chada S, NF-kappaB-dependent mechanisms. J Invest Dermatol 2007; Aggarwal BB. Melanoma differentiation-associated gene-7/ 127: 1326–1336. IL-24 gene enhances NF-kappa B activation and suppresses 35 Girardi G, Elias MM. Evidence for renal ischaemia as a apoptosis induced by TNF. J Immunol 2004; 173: 4368–4376. cause of mercuric chloride nephrotoxicity. Arch Toxicol 1995; 23 Khan S, Cleveland RP, Koch CJ, Schelling JR. Hypoxia induces 69: 603–607. renal tubular epithelial cell apoptosis in chronic renal disease. 36 Brezis M, Epstein FH. Cellular mechanisms of acute ischemic Lab Invest 1999; 79: 1089–1099. injury in the kidney. Annu Rev Med 1993; 44: 27–37. 24 Hauser P, Oberbauer R. Tubular apoptosis in the pathophy- 37 Wei CC, Hsu YH, Li HH, Wang YC, Hsieh MY, Chen WY siology of renal disease. Wien Klin Wochenschr 2002; 114: et al. IL-20: biological functions and clinical implications. 671–677. J Biomed Sci 2006; 13: 601–612.

Genes and Immunity