Agonist of peroxisome proliferator-activated receptor-c, rosiglitazone, reduces renal injury and dysfunction in a murine sepsis model

Transcription

1 Nephrol Dial Transplant (2005) 20: doi: /ndt/gfh705 Advance Access publication 29 March 2005 Original Article Agonist of peroxisome proliferator-activated receptor-c, rosiglitazone, reduces renal injury and dysfunction in a murine sepsis model Sik Lee 1, Won Kim 1, Kyung Pyo Kang 1, Sang-ok Moon 1, Mi Jeong Sung 1, Duk Hoon Kim 1, Hyung Jin Kim 2 and Sung Kwang Park 1 1 Department of Internal Medicine and 2 Urology, Renal Regeneration Laboratory, Research Institute of Clinical Medicine, Chonbuk National University Medical School, Jeonju, South Korea Abstract Background. Agonists of the peroxisome proliferatoractivated receptor-g may help to regulate inflammation by modulating the production of inflammatory mediators and adhesion molecules. The purpose of this study was to determine the protective effects of rosiglitazone on renal injury in a sepsis model and to explore the mechanism. Methods. In lipopolysaccharide (LPS)-induced mouse sepsis, we examined the effect of rosiglitazone on LPSinduced overproduction of inflammatory mediators, on the expression of adhesion molecules in renal tubular epithelial cells and on renal function. The mechanism of the protective effect was investigated in vitro using human renal tubular epithelial cells. Results. Rosiglitazone significantly decreased serum tumour necrosis factor (TNF)-a and interleukin (IL)- 1b levels during sepsis. The levels of blood urea nitrogen and creatinine were significantly lower in mice pre-treated with rosiglitazone than that in LPS-treated mice. Rosiglitazone reduced the expression of intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1) in tubular epithelial cells and interstitium of LPS-treated mice. Pre-treatment with rosiglitazone reduced the infiltration of macrophages/monocytes in renal tissue. In cultured tubular epithelial cells, rosiglitazone significantly decreased the expression of ICAM-1 and VCAM-1 induced by TNF-a or IL-1b, inhibited the degradation of inhibitor kba (IkBa) and blocked the activation of the p65 subunit of nuclear factor (NF)-kB. Conclusions. These results indicate that pre-treatment with rosiglitazone attenuated the production of TNF-a and IL-1b and reduced adhesion molecule expression Correspondence and offprint requests to: Sung Kwang Park, MD, Department of Internal Medicine, Chonbuk National University Medical School, , Keum-Am Dong, Jeonju, , South Korea. kidney@chonbuk.ac.kr in renal tubular epithelial cells of LPS-treated mice. Rosiglitazone has an anti-inflammatory effect in renal tubular epithelial cells through the inhibition of NF-kB activation. Keywords: adhesion molecules; renal tubular epithelial cells; rosiglitazone; sepsis Introduction Sepsis is the leading cause of death in critically ill patients and leads to multiple organ failure with injuries involving the kidneys, lungs and gastrointestinal tract. The principal pro-inflammatory mediators in the pathology of sepsis are tumour necrosis factor-a (TNF-a) and interleukin-1b (IL-1b), which are released in response to bacterial toxins by monocytes, macrophages and other leukocytes [1]. TNF-a and IL-1b activate nuclear factor (NF)-kB by triggering a signalling pathway that leads to the phosphorylation and consequent degradation of the inhibitor kba (IkBa) [2]. The degradation of IkBa exposes a nuclear localization signal on the NF-kB protein, which then moves into the nucleus and stimulates the transcription of specific genes. The secretion of various proinflammatory cytokines characterizes the first phase of sepsis. The overproduction of pro-inflammatory mediators enhances adhesion molecules and also leads to deleterious effects associated with sepsis, including fever, hypotension, multiple organ failure and shock [3]. Adhesion molecules have a role first in the recruitment and then in the consequent adhesion and transmigration of inflammatory cells. Intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1) are important adhesion molecules that play a role in inflammatory reactions. Although ICAM-1 and VCAM-1 were originally ß The Author [2005]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please journals.permissions@oupjournals.org

2 1058 S. Lee et al. identified in endothelial cells, they are also expressed in other cell types, including vascular smooth muscle cells, differentiating skeletal muscle cells, renal and neural epithelial cells and dendritic cells [4]. The peroxisome proliferator-activated receptor-g (PPARg) is a member of the nuclear receptor superfamily of ligand-activated transcription factors [5]. Ligand-activated PPARg binds to a specific DNAbinding site, termed the peroxisome proliferator response element (PPRE), to regulate the transcription of numerous target genes that involve inflammation, cytokine production and adhesion molecule expression [6]. Recently, it has been shown that PPARg is expressed in human monocyte-derived macrophages and that PPARg agonists inhibit monocyte inflammatory cytokine production and macrophage activation [7]. It has also been reported that PPARg agonists have a potent anti-inflammatory effect in human endothelial cells [8]. Previous studies have suggested that ligandactivated PPARg can downregulate NF-kB transcription [9]. Several in vivo studies have shown that various PPARg agonists may participate in the control of inflammation by modulating the production of inflammatory mediators [8]. Nakajima et al. [10] have reported that stimulation of the endogenous PPARg pathway induces anti-inflammatory responses in murine intestine after ischaemia reperfusion injury. Cuzzocrea et al. [11] have shown that rosiglitazone, a PPARg agonist, reduces the development of non-septic shock induced by zymosan in mice and may be useful in the treatment of acute lung inflammation. Decreased renal function in sepsis reduces the success of treatments, including fluid therapy and many drugs. Development of new drugs to treat decreased renal function will reduce mortality and morbidity due to sepsis. However, to date, none of the new therapies for renal injury in sepsis has shown clinical efficacy. In this study, we examined the effects of rosiglitazone on renal tissue injury in a lipopolysaccharide (LPS)-induced sepsis model. In particular, we investigated whether in vivo administration of this drug inhibits the LPSinduced overproduction of host inflammatory mediators and reduces the expression of adhesion molecules in the kidney. Since adhesion molecules have a role in the recruitment of inflammatory cells, we examined whether rosiglitazone inhibits the infiltration of macrophages/monocytes in renal tissue. We also evaluated whether in vitro administration of this drug reduces the expression of adhesion molecules and regulates the expression levels of IkBa and p65, which are induced by TNF-a in an NF-kB-dependent manner in renal tubular epithelial cells. Materials and methods Animal treatment Male ICR mice (Damul, Daejeon, South Korea) were given a standard laboratory diet and water ad libitum and cared for in accordance with both the Guiding Principles in the Care and Use of Animals published by the American Physiological Society and the Home Office Guidance in the Operation of the Animals (Scientific Procedures) Act 1986, published by Her Majesty s Stationary Office, London, UK. The mice were 7 8 weeks of age, weighing g, at the start of the experiments. The mice were divided into four groups: vehicle (saline) (n ¼ 8), rosiglitazone (n ¼ 8), LPS (n ¼ 8) and LPS plus rosiglitazone (n ¼ 8). After a 7 day acclimation period, the mice were administered rosiglitazone (10 mg/kg, donated by Glaxo SmithKline Pharmaceuticals, Seoul, South Korea) by oral gavage once a day for 3 days before the injection of LPS through the tail vein [from Escherichia coli O111:B4, 15 mg/kg, intravenously (i.v.), Sigma Chemical Co., St Louis, MO]. The control group was administered saline. Mice were anaesthetized with pentobarbital (30 mg/kg) and sacrificed by cervical dislocation. Renal tissues were harvested at 2, 4, 6, 12 and 24 h after LPS and/or rosiglitazone treatment. Measurement of serum levels of TNF-, IL-1, creatinine and urea nitrogen Blood samples (0.5 ml) were taken at 0, 1, 2, 3, 4, 6, 12 and 24 h after the administration of LPS or vehicle. Blood samples were collected by cardiac puncture, centrifuged at 7200 g for 3 min in order to obtain serum and immediately used for the study. Serum concentrations of TNF-a and IL-1b were determined by using enzyme-linked immunosorbent assay (ELISA) kits (Endogen, Woburn, MA). In all cases, a standard curve was constructed from standards provided by the manufacturer. Blood urea nitrogen (BUN) and creatinine levels were measured in blood using an autoanalyser (SRL, Tokyo, Japan). Culture of human renal tubular epithelial cells Human renal tissue was obtained from patients undergoing nephrectomy because of renal cell carcinoma. All patients gave their informed consent to participate in the study. Within 30 min after nephrectomy, samples of cortex from the normal pole were obtained for extraction of tubular epithelial cells by sieving. Human renal tubular epithelial cells derived from human renal tubules were characterized extensively according to a method described previously [12]. Cells were passaged every 3 4 days in 100 mm dishes (Falcon, Bedford, MA) using Dulbecco s modified Eagle s medium (DMEM)- F12 (Sigma Chemical Co.) supplemented with 10% fetal bovine serum (FBS) (Life Technologies Inc., Gaithersburg, MD), insulin transferrin sodium selenite media supplement (Sigma Chemical Co.), 100 U/ml penicillin and 100 mg/ml streptomycin (Sigma Chemical Co.). Cultures were incubated in a humidified atmosphere of 5% CO 2, 95% air at 37 C. For experimental use, cells were switched to serum-free DMEM- F12 and incubated for 16 h. Application of TNF-, IL-1 and rosiglitazone in cultured cells Human renal tubular epithelial cells were plated onto gelatinized 60 mm dishes in medium containing 10% serum and then switched to medium containing 1% serum with TNF-a (20 ng/ml; Sigma Chemical Co.) or IL-1b (10 ng/ml; Sigma Chemical Co.). In experiments involving rosiglitazone in vitro, rosiglitazone (5 mm; Cayman Chemicals, Ann Arbor,

3 Rosiglitazone reduces renal injury in a sepsis model 1059 MI) or control buffer was added to the cell medium 60 min before TNF-a (20 ng/ml) or IL-1b (10 ng/ml) treatment. The cells were washed, and then given fresh medium containing rosiglitazone and TNF-a or IL-1b. Control samples received buffer in place of rosiglitazone and/or TNF-a or IL-1b. Mann Whitney test. Multiple comparisons were examined for significant differences using analysis of variance (ANOVA), followed by individual comparisons with the Bonferroni post-test. Statistical significance was set at P<0.05. Western blot analyses for ICAM-1, VCAM-1, IB and p65 For the western blot analysis, cytoplasmic and nuclear proteins were prepared from cultured human renal tubular epithelial cells [13]. Samples (renal tissue or cultured human renal tubular epithelial cells) were mixed with sample buffer, boiled for 10 min, separated by SDS PAGE under denaturing conditions and electroblotted to nitrocellulose membranes. The nitrocellulose membranes were blocked by incubation in blocking buffer and incubated with anti- VCAM-1 polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA), with anti-icam-1 monoclonal antibody (Santa Cruz Biotechnology) or with rabbit polyclonal antibody against IkBa or p65 (Santa Cruz Biotechnology). The blots were washed with phosphate-buffered saline (PBS) and incubated with horseradish peroxidase-conjugated secondary antibody. Signals were visualized by chemiluminescent detection according to the manufacturer s protocol (Amersham Pharmacia Biotech, London, UK). The membranes were reprobed with anti-actin antibody to verify equal loading of protein in each lane. All signals were visualized and analysed by densitometric scanning (LAS-1000, Fuji Film, Tokyo, Japan). Immunohistochemical analyses for ICAM-1, VCAM-1 and ED-1 Isolated kidney tissues were fixed by immersion in 4% paraformaldehyde and blocked in paraffin. Tissue sections (5 mm thick) were deparaffinized with xylene and rehydrated with graded ethanols. Endogenous peroxidase was blocked with 3% hydrogen peroxide for 20 min, and the samples were then rinsed with PBS. To obtain an adequate signal, the slides were treated with pepsin at 42 C for 5 min. After treatment with protein-blocking solution, the slides were incubated overnight at 4 C with primary antibodies of VCAM-1, ICAM-1 or ED-1 (a monoclonal IgG 1 antibody to a cytoplasmic antigen present on monocytes and macrophages) (Bioproducts for Science, Indianapolis, IN). The crossreactivity of the primary antibody was visualized using the Vectastain ABC-Elite peroxidase detection system (Vector Laboratories, Burlingame, CA), followed by reaction with diaminobenzidine as chromogen and counterstaining with haematoxylin (Sigma Chemical Co.). Negative controls consisted of substitution of the primary antibody with equivalent concentrations of an irrelevant murine monoclonal antibody of normal rabbit or goat IgG. Evaluation of all slides was performed by an observer who was unaware of the origin of the slides. The number of ED-1-positive cells in each section was calculated by counting the number of positively stained cells in 30 glomeruli per slide. Statistical analysis Data are expressed as the mean±sem. Comparisons between two groups were made with the non-parametric Results Rosiglitazone reduces renal dysfunction in LPS-treated mice BUN and serum creatinine were determined after LPS injection and again after 24 h in serum of both rosiglitazone-pre-treated and untreated mice (Figure 1). Levels of BUN and serum creatinine increased at 24 h after LPS injection in both groups, but the levels of BUN and creatinine after LPS injection were significantly lower in mice pre-treated with rosiglitazone. Rosiglitazone decreases the expression of ICAM-1 and VCAM-1 in renal tissue obtained from LPS-treated mice Administration of LPS increased the expression of ICAM-1 and VCAM-1 proteins in a time-dependent manner (Figure 2A). The protein level of ICAM-1 Fig. 1. BUN and creatinine levels in serum after LPS injection. Mice pre-treated with rosiglitazone (ROS; squares) presented a significant decrease in BUN and creatinine values at 24 h after LPS injection compared with untreated mice (diamonds). Values are expressed as the mean±sem. *P<0.05, compared with the value for untreated mice.

4 1060 S. Lee et al. began to increase at 4 h, peaked at 6 h and decreased at 24 h. The protein level of VCAM-1 began to increase at 4 h, peaked at 12 h and decreased at 24 h. Thus, LPS increased the protein levels of ICAM-1 and VCAM-1 in renal tissue. Since the protein levels of both ICAM-1 and VCAM-1 significantly increased at 6 h, we investigated the effects of rosiglitazone on the ICAM-1 and VCAM-1 protein levels in renal tissue obtained from LPS-treated mice at 6 h after LPS injection. Pre-treatment with rosiglitazone significantly decreased the protein levels of ICAM-1 and VCAM-1 in LPS-treated mice (Figure 2B). When compared with renal tissue obtained from vehicle-treated mice, renal tissues obtained from LPStreated mice showed a marked increase of ICAM-1 and VCAM-1 staining in glomerular endothelial cells, the basolateral cell border of proximal and distal tubules, and interstitium. Renal tissues obtained at 6 h from mice administered LPS with rosiglitazone pretreatment demonstrated markedly reduced staining for both ICAM-1 and VCAM-1 in interstitial and tubular epithelial cells when compared with renal tissues obtained from LPS-treated mice (Figure 2C). These data suggested that rosiglitazone inhibits the LPSinduced expression of ICAM-1 and VCAM-1 in renal tissue, especially renal tubular epithelial cells. Rosiglitazone suppresses the infiltration of macrophages/monocytes in renal tissue of LPS-treated mice While we noticed only a few ED-1-positive interstitial cells in the rosiglitazone-only group and control kidneys, the number of interstitial cells stained was significantly increased at 24 h in LPS-treated mice. Pretreatment with rosiglitazone significantly decreased Fig. 2. Western blot analysis and immunohistochemical localization of ICAM-1 and VCAM-1 in LPS-treated mouse renal tissue. (A) Renal tissue samples obtained from LPS-treated mice (15 mg/kg i.v.) were homogenized, and total protein (50 mg) was blotted and probed with anti-icam-1 and anti-vcam-1 antibodies (left panel). The blots were reprobed with an anti-actin antibody. Densitometric analyses are presented as the relative ratio of each protein to actin (right panel). Values are expressed as the mean±sem. *P<0.05 vs time 0. (B) Renal tissue samples obtained from LPS-treated mice pre-treated with rosiglitazone (ROS; 10 mg/kg p.o. for 3 days) were homogenized, and total protein (50 mg) was blotted and probed with anti-icam-1 and anti-vcam-1 antibodies (left panel). The blots were reprobed with an antiactin antibody. Densitometric analyses are presented as the relative ratio of each protein to actin (right panel). Values are expressed as the mean±sem. *P<0.05 vs control buffer (CB), y P<0.05 vs LPS only. (C) (a) Negative control section, (b) vehicle-treated section, (c) LPStreated section, (d) LPS-treated section pre-treated with rosiglitazone (ROS; 10 mg/kg p.o. for 3 days). Note that the expression of ICAM-1 and VCAM-1 is highest in (c). Sections were incubated overnight with primary mouse anti-rat monoclonal antibody directed at ICAM-1 [1:500 (v/v) in PBS] or VCAM-1 [1:400(v/v) in PBS]. The antigen antibody complex was labelled specifically using an avidin biotin immunoperoxidase technique with the chromogen diaminobenzidine. Bars indicate 50 mm. Magnification, 400.

5 Rosiglitazone reduces renal injury in a sepsis model 1061 C ICAM-1 VCAM-1 Negative control (a) Control (b) LPS (c) LPS+ROS (d) Fig. 2. Continued. the number of ED-1-positive interstitial cells (Figure 3). These data indicated that rosiglitazone suppresses the recruitment of inflammatory cells in renal tissue of LPS-treated mice. Rosiglitazone decreases serum TNF- and IL-1 levels in LPS-treated mice We measured serum TNF-a and IL-1b levels at various times after the administration of saline, rosiglitazone, LPS or rosiglitazone þ LPS. We first examined the changes in the serum levels of TNF-a and IL-1b after LPS injection (15 mg/kg, i.v.) in saline-treated mice. Serum TNF-a and IL-1b were not detectable before LPS injection. Serum TNF-a levels were 780± 118 pg/ml at 1 h, 1100±230 pg/ml at 2 h, 430± 80 pg/ml at 3 h and dropped to near basal levels at 4 h. Serum IL-1b levels were 85±20 pg/ml at 1 h, 220±54 pg/ml at 2 h, 185±42 pg/ml at 3 h and 98±28 pg/ml at 4 h. Since the serum levels of TNF-a and IL-1b peaked at 2 h, we measured the serum levels of these cytokines at 2 h after administration of

6 1062 S. Lee et al. Fig. 3. Immunohistochemical localization of ED-1 (macrophages/ monocytes) in renal tissue sections. (A) (a) Negative control section (NC), (b) vehicle-treated section, (c) LPS-treated section, (d) LPStreated section pre-treated with rosiglitazone (ROS; 10 mg/kg p.o. for 3 days). Note that the number of ED-1-stained cells is highest in (c). Sections were incubated overnight with anti-ed 1 antibody (1:2500). Specific labelling of the antigen antibody complex was visualized using an avidin biotin immunoperoxidase technique using the chromogen diaminobenzidine. Bars indicate 50 mm. Magnification, 400. (B) Number of ED-1-positive cells/30 glomeruli in renal cortex. Values are expressed as the mean±sem. *P<0.05 vs control, y P<0.05 vs LPS only. rosiglitazone and/or LPS. Pre-treatment with rosiglitazone significantly reduced the levels of TNF-a and IL-1b (Figure 4). These results suggested that the anti-inflammatory effect of rosiglitazone is associated with decreased systemic serum levels of TNF-a and IL-1b in LPS-treated mice. Rosiglitazone decreases the expression of ICAM-1 and VCAM-1 proteins in renal tubular epithelial cells To explore further the effect of rosiglitazone on the expression of ICAM-1 and VCAM-1 in renal tubular epithelial cells, we used cultures of human renal tubular epithelial cells. The protein levels of ICAM-1 and VCAM-1 in human renal tubular epithelial cells significantly increased in a time-dependent manner after treatment with TNF-a or IL-1b (Figure 5A). We further examined the TNF-a- or IL-1b-induced protein levels of ICAM-1 and VCAM-1 in human renal tubular epithelial cells after treatment with rosiglitazone. While rosiglitazone by itself did not change the protein Fig. 4. Effect of rosiglitazone on serum levels of TNF-a (A) and IL-1b (B) in LPS-treated mice. Mice were given saline orally with or without rosiglitazone (10 mg/kg) for 3 days before LPS injection (15 mg/kg i.v.). Blood samples were collected at 2 h after LPS injection. Values are expressed as the mean±sem. *P<0.05, **P<0.01 vs control buffer (CB) or rosiglitazone (ROS), z P<0.05 vs LPS only. levels of ICAM-1 or VCAM-1, rosiglitazone significantly decreased TNF-a- or IL-1b-induced protein expression of both adhesion molecules (Figure 5B). These results indicated that the anti-inflammatory effect of rosiglitazone is associated with the suppression of TNF-a- and IL-1b-induced adhesion molecules in human renal tubular epithelial cells. Rosiglitazone inhibits degradation of IB and activation of the p65 subunit of NF-B in renal tubular epithelial cells We assessed the effect of rosiglitazone on the expression levels of IkBa and p65, which were induced by TNF-a in an NF-kB-dependent manner. The degradation of IkBa in cytoplasm increased 3.5-fold, and activation of p65 increased 6.2-fold at 30 min of TNF-a treatment in human renal tubular epithelial cells (data not shown). Pre-treatment with rosiglitazone in TNFa-treated renal tubular epithelial cells significantly decreased the degradation of IkBa in cytoplasm and also decreased the activation of p65 (Figure 6A and B). These data suggested that rosiglitazone exhibits an anti-inflammatory effect in renal tubular epithelial cells by inhibiting the NF-kB pathway. Discussion Sepsis is a major cause of death in critically ill patients worldwide. It is a systemic response to serious

7 Rosiglitazone reduces renal injury in a sepsis model 1063 Fig. 5. Effect of rosiglitazone on ICAM-1 and VCAM-1 protein levels in human renal tubular epithelial cells. (A) Cells were incubated for 0, 2, 4 and 8 h with TNF-a (20 ng/ml; lanes 1 4) or IL-1b (10 ng/ml; lanes 5 8). Total protein (50 mg) from the cells was blotted and probed with appropriate antibodies (left panel). The blots were reprobed with anti-actin antibody. Densitometric analyses are presented as the relative ratio of each protein to actin (right panel). Values are expressed as the mean±sem. *P<0.05 vs time 0. (B) Cells were incubated for 4 h with TNF-a (20 ng/ml) or IL-1b (10 ng/ml). Rosiglitazone (5 mm) or control buffer were added to the cell medium 60 min before TNF-a or IL-1b treatment. Blots and densitometric analyses are as described in (A). Values are expressed as the mean±sem. *P<0.05 vs control buffer (CB) or rosiglitazone (ROS), y P<0.05 vs TNF-a, z P<0.05 vs IL-1b. The data are the averages of four individual experiments. infection and leads to multiple organ dysfunction, hypoperfusion, hypotension and frequently death. Systemic vasodilation, abnormal intrarenal haemodynamics and a complex cascade involving endogenous mediators of inflammation and thrombosis contribute to acute renal failure in sepsis [14]. In this study, we have for the first time demonstrated that a synthetic PPARg agonist, rosiglitazone, has a protective effect on renal injury in an LPS-induced mouse sepsis model. Pre-treatment with rosiglitazone inhibits the LPS-induced systemic increase of TNF-a or IL-1b levels and enhancement of expression of ICAM-1 and VCAM-1. Rosiglitazone also decreases the infiltration of inflammatory cells in kidney that are increased by LPS. In addition, our results show that rosiglitazone blocks TNF-a-mediated IkBa degradation and upregulation of p65 in human renal epithelial cells. These results suggest that rosiglitazone may inhibit activation of NF-kB. Secretion of various cytokines is characteristic of the initial phase of sepsis. Cunningham et al. [15] have reported that mice deficient in TNF receptor 1 (TNFR1 / ) are resistant to LPS-induced renal failure, and that TNFR1 / mice have fewer apoptotic renal cells and fewer neutophils infiltrating the kidney following LPS administration. These observations suggest that TNF-a plays a critical role in LPS-induced renal failure, acting through its receptor TNFR1 in the kidney. Our results show that pre-treatment with rosiglitazone inhibits production of TNF-a and IL-1b that are increased by LPS administration. It is well documented that TNF-a and IL-1b cause activation and translocation of NF-kB [16]. It has also been reported that 15d-PGJ2, an endogenous PPARg agonist, inhibits the activation of NF-kB in vitro by preventing the degradation of IkBa [17] and that ligand-activated PPARg can downregulate NF-kB transcription [9]. The present study has revealed that rosiglitazone inhibits the degradation of IkBa in cytoplasm and activation of the p65 subunit of NF-kB in renal tubular epithelial cells. An important feature of the inflammatory process is the localization of inflammatory cells. Adhesion molecules have a role first in the recruitment, and

8 1064 S. Lee et al. Fig. 6. Effect of rosiglitazone on TNF-a-induced activation of p65 and degradation of IkBa in human renal tubular epithelial cells. (A) Cells, either untreated or pre-treated for 30 min with rosiglitazone (5 mm) and then treated with TNF-a (20 ng/ml) for 30 min, were used to prepare cytoplasmic and nuclear extracts, and analysed for p65 by western blot analysis (left panel). Densitometric analyses are presented (right panel). Values are expressed as the mean±sem. *P<0.05 vs control buffer (CB), y P<0.05 vs TNF-a. (B) Cells, either untreated or pre-treated for 30 min with rosiglitazone (5 mm), were incubated with TNF-a (20 ng/ml) for 30 min, and then assayed for IkBa in cytosolic fractions by western blot analysis (left panel). The blots were reprobed with an anti-actin antibody. Densitometric analyses are presented as the relative ratio of each protein to actin (right panel). Values are expressed as the mean±sem. *P<0.05 vs control buffer (CB) or rosiglitazone (ROS), y P<0.05 vs TNF-a. The data are the averages of four individual experiments. then in the consequent adhesion and transmigration of inflammatory cells. This leads to the release of mediators of renal injury, including reactive oxygen species, which contribute to acute renal failure [18]. ICAM-1 and VCAM-1 are important adhesion molecules that play a role in inflammatory cell adhesion. These adhesion molecules are also upregulated by inflammatory cytokines, including TNF-a and IL-1b [19]. In the present study, we have demonstrated that rosiglitazone decreases the expression of ICAM-1 and VCAM-1 in kidney of LPS-treated mice and in human renal tubular epithelial cells. Thus, rosiglitazone exhibits an anti-inflammatory effect in renal tubular epithelial cells by regulating the expression of adhesion molecules. In addition, we have also found that rosiglitazone treatment reduces infiltration of macrophages/monocytes in renal tubular epithelial cells. Consistent with our observations, a very recent study has reported that PPARg agonists such as thiazolidinedione ligands markedly reduce colonic inflammation in a mouse model of inflammatory bowel disease [20]. In this study, we have observed that PPARg protein, which is activated by rosiglitazone, is expressed in kidneys to a similar extent as in the colon (data not shown). These findings have suggested that rosiglitazone has a reno-protective effect by regulating the transcription of TNF-a, IL-1b and adhesion molecules in renal tissue. Although many trials of anti-inflammatory agents in patients with sepsis have been carried out, unfortunately none of these trials has shown clinical efficacy. New therapies to treat sepsis are directed against various substances of the inflammatory cascade, including LPS and pro-inflammatory mediators such as TNF-a and IL-1b. The drugs that downregulate the expression of cell adhesion molecules and proinflammatory mediators are useful for the treatment or prevention of the initial phase of inflammation in sepsis. Our results indicate that the PPARg agonist, rosiglitazone, reduces renal injury in LPS-induced sepsis. Our findings may provide the mechanisms underlying the beneficial effects of PPARg agonists on the reduction of pro-inflammatory cytokines, such as TNF-a and IL-1b, the inhibition of NF-kB activation, the inhibition of ICAM-1 and VCAM-1 expression in renal tubular epithelial cells and the suppression of the infiltration of macrophages/monocytes in kidney. Therefore, rosiglitazone may have a protective effect in maintaining renal function and reducing mortality and morbidity in sepsis. However, the renoprotective effect of this drug in humans needs to be examined. Acknowledgements. We thank Jennifer Macke for editing the text. This work was supported by a grant from the basic research program of the Korean Science and Engineering Foundation (R ) and the National Research Laboratory Program, Ministry of Science and Technology, Republic of Korea. Conflict of interest statement. None declared.

9 Rosiglitazone reduces renal injury in a sepsis model 1065 References 1. Hotchkiss RS, Karl IE. The pathophysiology and treatment of sepsis. N Engl J Med 2003; 348: Alberts B. Molecular Biology of the Cell. Garland Science, New York; 2002: Thijs A, Thijs LG. Pathogenesis of renal failure in sepsis. Kidney Int 1998; 53 [Suppl 66]: S66 S69 4. Tu Z, Kelley VR, Collins T, Lee FS. IkB kinase is critical for TNF-a-induced VCAM-1 gene expression in renal tubular epithelial cells. J Immunol 2001; 166: Michalik L, Wahli W. Peroxisome proliferator-activated receptors: three isotypes for a multitude of functions. Curr Opin Biotechnol 1999; 6: Hauner H. The mode of action of thiazolidinediones. Diabetes Metab Res Rev 2002; 18: S10 S15 7. Ricote M, Li AC, Willson TM, Killy J, Glass CK. The peroxisome proliferator activated receptor-gamma is a negative regulator of macrophage activation. Nature 1998; 391: Imizumi T, Matsumiya T, Tamo W et al. 15-deoxy- 12,14 - prostaglandin J 2 inhibits CX3CL1/fractalkine expression in human endothelial cells. Immunol Cell Biol 2002; 80: Chung SW, Kang BY, Kim SH et al. Oxidized low density lipoprotein inhibits interleukin 12 production in lipopolysaccharide-activated mouse macrophages via direct interaction between PPARg and NF-kB. J Biol Chem 2000; 276: Nakajima A, Wada K, Miki H et al. Endogenous PPARg mediates anti-inflammatory activity in murine ishemia reperfusion injury. Gastroenterology 2001; 120: Cuzzocrea S, Pisano B, Dugo L et al. Rosiglitazone, a ligand of the peroxisome proliferators-activated receptor-g, reduces the development of nonseptic shock induced by zymosan in mice. Crit Care Med 2004; 32: Chakravorty SJ, Cockwell P, Girdlestone J, Brooks CJ, Savage CS. Fractalkine expression on human renal tubular epithelial cells: potential role in mononuclear cell adhesion. Clin Exp Immunol 2002; 129: Schlezinger JJ, Jensen BA, Mann KK, Ryu HY, Sherr DH. Peroxisome proliferators-activated receptor g-mediated NF-kB activation and apoptosis in pre-cells. J Immunol 2002; 169: Glauser MP, Zanetti G, Baumgartner JD, Cohen J. Septic shock: pathogenesis. Lancet 1991; 338: Cunningham PN, Dyanov HM, Park P et al. Acute renal failure in endotoxemia is caused by TNF acting directly on TNF receptor-1 in kidney. J Immunol 2002; 168: Bauerle PA, Henkel T. Function and activation of NF-kB in the immune system. Annu Rev Immunol 1994; 2: Rossi A, Elia G, Santoro MG. Inhibition of nuclear factor kappa B by prostaglandin A1: an effect associated with heat shock transcription factor activation. Proc Natl Acad Sci USA 1997; 94: Molitoris BA, Marrs J. The role of cell adhesion molecules in ischemic acute renal failure. Am J Med 1999; 106: Donnahoo KK, Meng X, Ayala A, Cain MP, Harken AH, Meldrum DR. Early kidney TNF-a expression mediates neutrophil infiltration and injury after renal ischemia reperfusion. Am J Physiol 1999; 277: Su CG, Wen X, B ST et al. A novel therapy for colitis utilizing PPARg ligands to inhibit the epithelial inflammatory response. J Clin Invest 1999; 104: Received for publication: Accepted in revised form: