Archives of Medical Research 40 (2009) 79e88

ORIGINAL ARTICLE

Rosiglitazone Attenuates the Severity of Sodium Taurocholate-inducedAcute Pancreatitis and Pancreatitis-associated Lung Injury

Chen Chen,* Sheng Xu,* Wei-Xing Wang, You-Ming Ding, Kai-Huan Yu, Bin Wang, and Xiao-Yan Chen

Department of General Surgery, Renmin Hospital of Wuhan University, Wuhan, PR China

Received for publication July 29, 2008; accepted November 10, 2008 (ARCMED-D-08-00335).Published previously online January 22, 2009.

*These authors con

Address reprint re

Surgery, Renmin Hosp

430060, Hubei Provin

0188-4409/09 $esee fdoi: 10.1016/j.arcm

Background and Aims. In addition to the effect of regulating adipocyte differentiationand insulin sensitivity, peroxisome proliferator activated receptor-g (PPAR-g) ligandsalso exhibit anti-inflammatory effect. However, the mechanisms concerning howPPAR-g ligands affect acute pancreatitis and pancreatitis-associated lung injury havenot been fully elucidated. This study investigated the effect of rosiglitazone, a PPAR-gligand, on acute pancreatitis and pancreatitis-associated lung injury in the rat pancreatitismodel induced by sodium taurocholate.

Methods. Acute pancreatitis was induced by retrograde infusion of 5% sodium taurocholate(1 mL/kg) into the bile-pancreatic duct. Rosiglitazone (6 mg/kg) was administered via thefemoral vein 30 min prior to the infusion of sodium taurocholate. The severity of pancreatitiswas evaluated by serum amylase level, myeloperoxidase activity, and pathology. Pancrea-titis-associated lung injury was evaluated by myeloperoxidase activity, the magnitude ofpulmonary edema and pathology. Intercellular adhesion molecule-1 (ICAM-1) and tumornecrosis factor-a mRNA expression were studied using reverse transcriptase polymerasechain reaction. ICAM-1 protein expression was studied using Western blot analysis.

Results. Prophylactic administration of rosiglitazone attenuated (1) serum amylase level;(2) myeloperoxidase activity of pancreatic and pulmonary tissue; (3) expression of tumornecrosis factor-a and ICAM-1 in pancreas and lung; (4) pancreas and lung pathologicaldamage.

Conclusions. Our study demonstrated that rosiglitazone exerts a protective effect againstsodium taurocholate-induced pancreatic and pulmonary injury. � 2009 IMSS. Pub-lished by Elsevier Inc on behalf of Archives of Medical Research. All rights reserved.

Key Words: Acute pancreatitis, Lung injury, Peroxisome proliferator activated receptor-g ligand,

Cytokine.

Introduction

Acute pancreatitis (AP) is an inflammatory disease of thepancreas. Its clinical course varies from a mild, transient ill-ness to a severe, fatal disease. Most AP patients present asa mild, self-limiting course with little or no complications,and the mortality is low. However, 20e25% of patients willprogress to severe acute pancreatitis, and 60% of severe

tributed equally to this work.

quests to: Wei-Xing Wang, Department of General

ital of Wuhan University, 238 Jiefang Road, Wuhan,

ce, China; E-mail: sate.llite@163.com

ront matter. Crown Copyright � 2009 IMSS. Published by Eed.2008.11.004

acute pancreatis patients will die in the first week from pul-monary complications (1e4).

To date, the exact mechanisms of AP are not entirelyknown. Generally, intra-acinar cell activation of digestive en-zyme zymogens is believed to be the initial event in AP. Oncethe disease process is initiated, inflammatory response is in-voked and results in local pancreatic injury. If this inflamma-tory reaction is marked, it leads to a systemic inflammatoryresponse syndrome and subsequent multiple organ dysfunc-tion syndrome with a mortality rate as high as 40% (1,5). Sys-temic inflammatory response syndrome as well as multipleorgan dysfunction syndrome in acute pancreatitis may besecondary to the excessive activation of leukocytes (6).

lsevier Inc on behalf of Archives of Medical Research. All rights reserved.

80 Chen et al./ Archives of Medical Research 40 (2009) 79e88

Sequestration of leukocytes, particularly neutrophils, playsa critical role in the pathogenesis of both local pancreaticand distant organ injury. In general, leukocyte recruitmentinto inflamed tissue involves sequential leukocyteeendothe-lial cell interaction events including leukocyte rolling, stick-ing, and transendothelial migration. Intercellular adhesionmolecule-1 (ICAM-1), a member of the immunoglobulin su-perfamily, usually expressed at low levels on the surface ofendothelial cells. ICAM-1 is a key player in leukocyteeendothelial cell interaction. It mediates both leukocyte firmadhesion and migration through the endothelium into tissues(7,8). Several studies have demonstrated that ICAM-1 isupregulated during inflammation (7,9,10), and the expres-sion of ICAM-1 is mediated by various inflammatory cyto-kines, particularly by tumor necrosis factor-a (TNF-a)(11). In AP, ICAM-1 is upregulated in both pancreas andlung, and the extent of its expression correlates with the se-verity of organ injury (12e15). Administration of antibodyagainst ICAM-1 ameliorated pancreas and lung injury andsubsequent leukocyte infiltration (14,15).

Peroxisome proliferator activated receptors (PPARs) aremembers of the nuclear hormone receptor superfamily ofligand-activated transcription factors that are related toretinoid, steroid and thyroid hormone receptors (16). ThePPAR subfamily consists of three different isoforms:PPAR-a, PPAR-b/d, and PPAR-g, which heterodimerizewith 9-cis-retinoic acid retinoid X receptors (17). PPARsare ligand-activated transcriptions factors that act as heter-odimers with retinoid X receptors, to recognize PPAR re-sponse elements located in the promoter of target genes(18). Natural ligands of PPAR-g include fatty acids, arach-idonic acid metabolites, and prostaglandins. Syntheticligands include certain nonsteroidal anti-inflammatorydrugs and a series of antidiabetic drugs called thiazolidine-diones such as troglitazone, pioglitazone and rosiglitazone.

Originally, PPAR-g shown to play a central role as tran-scriptional mediators in adipocyte differentiation and glu-cose homeostasis (19,20). Recently, several studies havedemonstrated that PPAR-g ligands exert potent anti-inflam-matory properties. For instance, PPAR-g ligands inhibitproinflammatory cytokine production and macrophage acti-vation (21,22), attenuate the severity of arthritis and inflam-matory bowel disease (23,24), and reduce intestinalischemia/reperfusion injury (25,26). In the present investi-gation we demonstrated the protective effect of rosiglita-zone, the most potent and selective PPAR-g ligand ofthiazolidinediones compounds (27), on acute pancreatitisand pancreatitis-associated lung injury.

Materials and Methods

Animals and Reagents

Male Wistar rats weighing 200e250 g (purchased from theCenter of Experimental Animals of Hubei Academy of

Medical Sciences, Wuhan, China) were used. Rats werehoused in cages under a temperature-conditioned(20e22�C) and humidity-controlled (50e52%) environ-ment with a 12-h light/dark cycle. Rats were fed with stan-dard commercial pellets and water ad libitum. The projectwas performed in accordance with the principles of the1983 Declaration of Helsinki and with the approval of thelocal ethics committee of Wuhan University. Rosiglitazonewas obtained from Cayman Chemical Company (Ann Ar-bor, MI). Sodium taurocholate (TC) and dimethyl-sulfoxide(DMSO) were purchased from Sigma-Aldrich Company(St. Louis, MO). TRI Reagent was obtained from Molecu-lar Research Center (Cincinnati, OH). Primers weredesigned and synthesized by Invitrogen Corporation (Carls-bad, CA).

Induction of Acute Pancreatitis

Prior to the experiment, rats were deprived of food, butdrinking water was available ad libitum. AP model wasinduced as the method described previously (28) withminor modifications. Anesthesia was administered by IPinjection of 10% chloraldurat (3 mL/kg) and rats underwentsterile laparotomy. The bile-pancreatic duct was cannulatedthrough the duodenum and AP was induced by a standard-ized pressure-controlled retrograde infusion of 5% sodiumtaurocholate (1 mL/kg) into the bile-pancreatic duct. Afterinfusion, the part of bile-pancreatic duct entering the duo-denum was clipped by a non-invasive vascular clip for 5min. The vascular clip was then removed and the abdomenwas closed.

Experimental Design

In order to obtain the optimal dose of rosiglitazone for pre-venting sodium taurocholate-induced pancreatitis, we per-formed a preliminary study using different doses (0.3 mg/kg, 3 mg/kg, 6 mg/kg and 10 mg/kg) based on previous re-ports that these doses exert potent anti-inflammatory effectsin rat (25,29). Acute pancreatitis was induced by retrogradeinfusion 5% sodium taurocholate into the bile-pancreaticduct. Rats infused with saline instead of sodium taurocho-late were used as control. Rosiglitazone was dissolved inthe vehicle (10% DMSO v/v) and administered via the fem-oral vein 30 min prior to sodium taurocholate infusion. Allrats were sacrificed under chloraldurat anesthesia at 12 h af-ter induction of pancreatitis, a time point at which intra-pancreatic damage had already peaked (28). The effect ofrosiglitazone was evaluated by the levels of serum amylase(AMY) and alanine aminotransferase (ALT). In a secondseries of experiments, the optimal dose of rosiglitazone (6mg/kg) was used. Fifty four male Wistar rats were ran-domly divided into the following groups: (1) sodium tauro-cholate þ vehicle group (TC group, n 5 18). Rats weresubjected to sodium taurocholate-induced AP and receivedthe vehicle for rosiglitazone (10% DMSO, 2 mL/kg)

81Rosiglitazone Attenuates the Severity of Sodium Taurocholate-induced Acute Pancreatitis and Lung Injury

administered via the femoral vein 30 min prior to sodiumtaurocholate. (2) Sodium taurocholateþ rosiglitazone group(ROSI group, n 5 18). This was the same as the sodium taur-ocholate þ vehicle group but rats were administered rosigli-tazone (6 mg/kg) via the femoral vein 30 min prior to sodiumtaurocholate. (3) Salineþ vehicle as the control group (CONgroup, n 5 18). This group was identical to the sodium taur-ocholate þ vehicle group except that saline was injected in-stead of sodium taurocholate and received the vehicle forrosiglitazone (10% DMSO, 2 mL/kg) administered via thefemoral vein 30 min prior to saline.

Six rats from each group were sacrificed under chloral-durat anesthesia at respective time points (3, 6, and 12 h)after the induction of pancreatitis. Rats underwent thoracot-omy by midsternal incision and |3 mL blood was obtainedfrom the left ventricle of the heart. Samples were centri-fuged at 3000 � g for 15 min, and serum was stored at�20�C until analysis. The middle lobe of the right lungwas removed, rinsed with saline, and blotted dry for detect-ing the magnitude of pulmonary edema. The head of thepancreas and the right lower lobe of the lung wereharvested and fixed with 40 g/L formaldehyde, paraffinembedded and continual sections were made. The otherpancreatic and pulmonary tissues were rinsed with saline,blotted dry, snap frozen in liquid nitrogen and stored at�80�C for further use.

Histopathological Examination

Continuous sections of the paraffin-embedded tissue weretaken for pathological examination with hematoxylin-eosinstaining. Morphometric documentation for pancreatic andpulmonary sections under light microscope (Olympus Opti-cal Ltd, Tokyo, Japan) were evaluated by two independentpathologists who were blinded to this experiment. Sectionsof pancreas tissue were scored for the severity of pancrea-titis based on edema, inflammation, vacuolization and ne-crosis according to the scale described by Schmidt et al.(30). Histological evaluation of pulmonary injury was as-sessed using a scale for interstitial and intraalveolar edema,interstitial and intraalveolar leukocyte infiltration, and fi-brosis, as described by Werner et al. (15).

Serum Assay

AMY and ALT were measured using standard techniqueswith an automatic biochemistry analyzer (Olympus OpticalLtd).

Measurement of Tissue Myeloperoxidase Activity

Neutrophil infiltration in pancreas or lung was quantitatedby measuring tissue myeloperoxidase (MPO) activity(31). MPO activity was measured photometrically with3,30, 5,50-tetramethyl benzidine as a substrate (32), andthe reaction was started by adding hydrogen peroxide

(H2O2) to the medium. MPO activity was determined bya spectrophotometric method using a commercial assaykit according to the manufacturer’s instructions (NanjingJiancheng Bioengineering Institute, Nanjing, China).

Measurement of Lung Wet-to-Dry Ratio

The magnitude of pulmonary edema was determined bycalculating the wet/dry ratio from the initial weight of theright lung middle lobe (wet weight) to its weight after des-iccation at 70�C for 24 h (dry weight).

Determination of TNF-a, ICAM-1 mRNA Expression byReverse Transcriptase Polymerase Chain Reaction (RT-PCR)

Total RNAwas extracted from the frozen tissue with TRI Re-agent. Frozen pancreas (or lung) tissue was mechanically ho-mogenized on ice in 1 mL of ice-cold TRI Reagent. TotalRNA was solubilized in RNase-free water, quantified in du-plicate by measuring the optical density (OD) at 260 nm. Pu-rity of RNA was assured by examining the OD260/OD280ratio. Two micrograms of RNA was reverse-transcribed tocomplementary DNA by using Revert Aid First Strand cDNASynthesis Kit (Fermentas, Hanover, MD) according to themanufacturer’s instructions. PCR was performed with theprimers for tumor necrosis factor-a (F: CATGTACCTGG-GAGGAGTCT; R: CTTCAGCATCTCGTGTGTTT; 326bp; NM-012675); intercellular adhesion molecule-1(ICAM-1) (F: CGGTAGAC ACAAGCAAGAGA; R: GCAGGGATTGACCATAATTT; 517 bp; NM-012967); and b-actin(F: CCAACTGGGACGATATGGAG; R: CAGAGGCATA-CAGGGACA AC; 207 bp; NM-031144). PCR was per-formed by using a Gene Cycler (Bio-Rad, Hercules, CA).Amplification steps were: tumor necrosis factor-a (initialdenaturation 94�C 5 min, then 94�C denaturation 30 sec,56�C annealing 30 sec, 72�C extension 1 min for 35 cycles,final extension 72�C 7min); ICAM-1 (initial denaturation94�C 5 min, then 94�C denaturation 30 sec, 56�C annealing30 sec, 72�C extension 1 min for 35 cycles, final extension72�C 7 min); b-actin (initial denaturation 94�C 5 min, then94�C denaturation 30 sec, 54�C annealing 30 sec, 74�Cextension 1 min for 35 cycles, final extension 72�C 7 min).Five-mL PCR products were electrophoresed using 2% aga-rose gel containing ethidium bromide (0.5 mg/mL). Gels werevisualized under UV light and then images were photo-graphed. Band intensity was determined by optical densitywith individual PCR product/b-actin ratios.

Determination of ICAM-1 Protein Expression by WesternBlot Analysis

Frozen pancreas (or lung) tissue was mechanically homog-enized in 1 mL of ice-cold extraction buffer (50 mM Tris-HCl, pH 7.4, 1% NP-40, 0.25% sodium deoxycholate,150 mM NaCl, 1 mM ethylene diamine tetraacetic acid(EDTA), 1 mM phenylmethylsulfonyl fluoride, 0.1% sodium

Table 1. Preliminary study results using different doses of rosiglitazone

Group n AMY (U/L) ALT (U/L)

CON 6 1072.2 � 83.9 95.4 � 10.2

TC 6 5347.5 � 673.5a 256.8 � 32.7a

0.3 mg/kg ROSI 6 5740.0 � 749.4a 226.1 � 23.8a

3 mg/kg ROSI 6 5294.7 � 993.4a 275.3 � 37.7a

6 mg/kg ROSI 6 2824.8 � 776.2a,b 190.5 � 37.4a,b

10 mg/kg ROSI 6 2785.9 � 611.3a,b 226.1 � 25.7a

Rats from different groups were killed 12 h after sodium taurocholate

infusion. Data are presented as mean � standard error.

CON, salineþ vehicle as control group; TC, sodium taurocholateþ vehicle

group; ROSI, sodium taurocholate þ rosiglitazone group.ap !0.01 vs. the CON group.bp !0.01 vs. the TC group.

82 Chen et al./ Archives of Medical Research 40 (2009) 79e88

dodecylsulfate, and 1 mg/mL each of aprotinin and leupep-tin). After incubation on ice for 30 min, the homogenatewas centrifuged at 13,000 � g for 40 min at 4�C, and thesupernatant was stored at �80�C until analyses. Proteinconcentration was determined using the Bradford methodwith bovine serum albumin as a standard. Protein samples(50 mg) were electrophoresed using 8% sodium dodecylsulfate polyacrylamide gels, transferred to a nitrocellulosemembrane. The membrane was blocked with the blockingbuffer (TBS containing 5% nonfat dry milk, 0.1% Tween-20) for 2 h at room temperature and then probed with theprimary antibodies. Dilutions for primary antibodies wereas follows: goat polyclonal anti-rat ICAM-1 (1:600, SantaCruz Biotechnology, Santa Cruz, CA) or rabbit polyclonalanti-rat actin antibody (1:2000; Santa Cruz Biotechnology)overnight at 4�C. The membrane was washed with TBST(TBS containing 0.05% Tween-20) and then incubated withhorseradish peroxidase-conjugated goat anti-rabbit or rabbitanti-goat secondary antibodies (1:5000, Pierce Biotechnology,Rockford, IL) at room temperature for 1 h. After repeatedwashings with TBST, the antibody�antigen complexes weredetected with ECL reagent (Immobilon Western HRP Sub-strate, Millipore Corporation, Bedford, MA).

Statistical Analysis

Data were expressed as mean � standard error. Means ofthe different groups were compared using one-way analysis

Figure 1. Representative hematoxylin-eosin staining sections of pancreas histopath

(D) Comparison of the total pathological score of pancreas. CON, saline þ v

ROSI, sodium taurocholate þ rosiglitazone group. ap !0.01 vs. the CON gr

version of this figure available online at www.arcmedres.com.

of variance. Statistical analysis was performed with theSPSS statistical package (SPSS 13.0 for Windows; SPSSInc., Chicago, IL). A value of p !0.05 was regarded as sta-tistically significant.

Results

Preliminary Study Results

Preliminary study results are shown in Table 1. A moderatedose of rosiglitazone (6 mg/kg) significantly decreasedserum AMY and ALT levels compared with the TC group( p !0.01). However, a lower dose (0.3 mg/kg, 3 mg/kg)failed to improve serum AMY and ALT levels and a higher

ology. (A) CON group at 12 h. (B) TC group at 12 h. (C) ROSI group at 12 h.

ehicle as the control group; TC, sodium taurocholate þ vehicle group;

oup; bp !0.01 vs. the TC group (original magnification � 200). Color

Figure 2. Representative hematoxylin-eosin staining sections of lung histopathology. (A) CON group at 12 h. (B) TC group at 12 h. (C) ROSI group at 12 h.

(D) Comparison of the total pathological score of lung. CON, saline þ vehicle as the control group; TC, sodium taurocholate þ vehicle group; ROSI, sodium

taurocholate þ rosiglitazone group. ap !0.01 vs. the CON group; bp !0.01 vs. the TC group (original magnification �100). Color version of this figure

available online at www.arcmedres.com.

83Rosiglitazone Attenuates the Severity of Sodium Taurocholate-induced Acute Pancreatitis and Lung Injury

dose (10 mg/kg) failed to improve serum ALT level com-pared with the TC group. Baselines of serum AMY andALT levels were determined in rats infused with saline in-stead of sodium taurocholate. Rats preteated with rosiglita-zone (6 mg/kg) in the absence of sodium taurocholaterevealed virtually the same serum AMY and ALT levelsas these baseline levels (data not shown). The results indi-cated that IV administration of rosiglitazone (6 mg/kg) doesnot significantly alter serum AMY level and does not havepotential for hepatotoxicity.

Figure 3. Levels of serum amylase (A) and alanine aminotransferase (B) CON, sali

ROSI, sodium taurocholate þ rosiglitazone group. ap !0.01 vs. the CON group a

Histopathology

Representative histological sections are shown in Figures1Ae1C (pancreas) and Figures 2Ae2C (lung). Sodiumtaurocholate-induced pancreatic damage was evident byincreased edema, inflammatory cell infiltration, vacuoliza-tion and necrosis. Pretreatment with rosiglitazone reducedthe inflammatory changes in pancreas. Histological exami-nation of lung sections of rats with sodium taurocholate-induced pancreatitis showed interstitial and intraalveolar

neþ vehicle as the control group; TC, sodium taurocholateþ vehicle group;

t the same time point. bp !0.01 vs. the TC group at the same time point.

Figure 4. Pancreatic (A) and pulmonary (B) tissue myeloperoxidase activity. CON, saline þ vehicle as the control group; TC, sodium taurocholate þ vehicle

group; ROSI, sodium taurocholate þ rosiglitazone group. ap !0.01 vs. the CON group at the same time point. bp !0.01 vs. the TC group at the same time

point.

84 Chen et al./ Archives of Medical Research 40 (2009) 79e88

edema, interstitial and intraalveolar inflammatory cell infil-tration. Prophylactic administration of rosiglitazonereduced the degree of lung injury. Pathological score ofpancreas and lung injury demonstrated that the total scorein the TC group significantly increased at 12 h comparedwith the CON group ( p !0.01) (Figures 1D and 2D). Inthe ROSI group, the total score of pancreas and lung injurywere significantly lower than in the TC group at 12 h( p !0.01) (Figures 1D and 2D).

Analysis of Serum AMY and ALT

Baselines of serum AMY and alanine ALT levels were de-tected in rats of the CON group. Compared with baselinelevels, serum AMY and ALT levels were significantlyincreased at each time point in the TC group ( p !0.01)(Figures 3A and 3B). Peak serum AMY and ALT levelswere seen at 12 h in the TC group. Pretreatment with rosi-glitazone before sodium taurocholate infusion significantlyreduced serum AMY level at 6 and 12 h compared with theTC group ( p !0.01) (Figure 3A). Prophylactic administra-tion of only rosiglitazone reduced the 12-h peak ALT levelcompared with the TC group ( p !0.01) (Figure 3B).

Analysis of MPO Activity

In both pancreatic and pulmonary tissue, MPO activity ofthe TC group increased significantly at each time pointcompared with the CON group ( p !0.01). For prophylac-tic administration, rosiglitazone reduced both pancreaticand pulmonary tissue MPO activity at 6 and 12 h comparedwith the TC group ( p !0.01) (Figure 4).

Figure 5. Lung wet-to-dry ratio. CON, saline þ vehicle as the control

group; TC, sodium taurocholate þ vehicle group; ROSI, sodium taurocho-

late þ rosiglitazone group. ap !0.01, bp !0.05 vs. the CON group at the

same time point; cp !0.01 vs. the TC group at the same time point.

Analysis of Lung Wet-to-Dry Ratio

Wet-to-dry ratio (W/D ratio) reflected the magnitude of pul-monary edema. Peak W/D ratio was seen at 12 h in the TCgroup, and rosiglitazone pretreatment only reduced W/D

ratio at this time point compared with the TC group( p !0.01) (Figure 5).

TNF-a and ICAM-1 mRNA Expression

In pancreatic tissue, peak expression of TNF-a mRNA wasseen 6 h after sodium taurocholate infusion and continuedto 12 h (Figure 6B). However, pancreatic tissue peakexpression of ICAM-1 mRNA was found at 12 h, lagginghours behind TNF-a mRNA upregulation (Figure 6D). Inpulmonary tissue, peak expression of TNF-a and ICAM-1mRNA was seen at 12 h after sodium taurocholate infusion(Figures 7B and 7D). In this part, pancreatic and pulmonarytissue expression of TNF-a and ICAM-1 mRNA of theCON group at 12 h, a time point at which intrapancreaticdamage had already peaked, was used as the baseline. Pre-treatment with rosiglitazone reduced the expression ofTNF-a and ICAM-1 mRNA in both pancreas and lung at12 h after sodium taurocholate infusion compared withthe TC group ( p !0.01) (Figures 6 and 7).

Figure 6. Expression of TNF-a (A) and ICAM-1 (C) mRNA in pancreatic tissue using RT-PCR analysis. CON, saline þ vehicle as the control group; TC,

sodium taurocholate þ vehicle group; ROSI, sodium taurocholate þ rosiglitazone group. M 5 100 bp marker; 1 5 the CON group at 12 h; 2 5 the TC group

at 3 h; 3 5 the TC group at 6 h; 4 5 the TC group at 12 h; 5 5 the ROSI group at 3 h; 6 5 the ROSI group at 6 h; 7 5 the ROSI group at 12 h. b-actin served

as an internal control. Comparison of the expression of TNF-a (B) and ICAM-1 (D) mRNA. ap !0.01 vs. the TC group at 3 h; bp !0.01 vs. the TC group at 6

h; cp !0.01 vs. the TC group at the same time point; dp !0.01 vs. the CON group at the same time point.

85Rosiglitazone Attenuates the Severity of Sodium Taurocholate-induced Acute Pancreatitis and Lung Injury

ICAM-1 Protein Expression

In the similar rat model of AP (15), peak expression ofICAM-1 protein in pancreas occurred at 12 h and continuedto 24 h. Peak expression of ICAM-1 protein in lung also oc-curred at 12 h. Here we show ICAM-1 protein expression inpancreatic (Figure 8A) and pulmonary (Figure 8C) tissue at12 h after induction of AP. Expression of ICAM-1 proteinin pancreatic and pulmonary tissues was significantly high-er in the TC group than in the CON group ( p !0.01). Inthe ROSI group, expression of ICAM-1 protein in pancre-atic and pulmonary tissues was significantly lower than inthe TC group ( p !0.01) (Figures 8B and 8D).

Discussion

PPAR-g activation usually regulates lipid metabolism, glu-cose homeostasis and influences cell proliferation and differ-entiation (33). In addition to these functions, reportsdemonstrated PPAR-g ligands exhibit anti-inflammatory ef-fects by modulating the production of inflammatory media-tors in vitro (21,22) and in vivo (26,29). According to theseresults, PPAR-g ligands may be a therapeutic option for in-flammatory diseases including AP. In our experiment, we

demonstrated that prophylactic administration of rosiglita-zone attenuated (1) serum AMY and ALT level; (2) neutro-phil infiltration; (3) expression of TNF-a and ICAM-1 and(4) pancreas and lung damage. All of these observationsindicate that pretreatment with rosiglitazone amelioratesthe degree of AP and pancreatitis-associated lung injury inrats.

Although the definite mechanisms of how PPAR-g li-gands affect AP and pancreatitis-associated lung injury havenot been completely elucidated, the anti-inflammatory effectof PPAR-g may be attributed to inhibition of proinflamma-tory transcription factors such as nuclear factor-kB (NF-kB) (21). Fahmi et al. (34) demonstrated that PPAR-g ligandsexert anti-inflammatory effect at least at the transcriptionallevel through a PPAR-g-dependent pathway by interferingwith the activation of NF-kB and AP-1. Kelly et al. (35)showed in intestinal epithelial cells that Bacteroides thetaio-tomicron attenuates inflammatory responses by enhancingthe nuclear export of the NF-kB subunit RelA througha PPAR-g-dependent mechanism. In cerulein-treated mice,PPAR-g natural ligand 15-deoxy-D12,14-prostaglandin J2

(15d-PGJ2) was proven to attenuate the severity of APthrough the inhibition of the transcription factor NF-kBactivation (38). Therefore, PPAR-g ligands may exert the

Figure 7. Expression of TNF-a (A) and ICAM-1 (C) mRNA in pulmonary tissue using RT-PCR analysis. CON, saline þ vehicle as the control group; TC 5

sodium taurocholate þ vehicle group; ROSI, sodium taurocholate þ rosiglitazone group. M 5 100-bp marker; 1 5 the CON group at 12 h; 2 5 the TC group

at 3 h; 3 5 the TC group at 6 h; 4 5 the TC group at 12 h; 5 5 the ROSI group at 3 h; 6 5 the ROSI group at 6 h; 7 5 the ROSI group at 12 h. b-actin served

as an internal control. Comparison of the expression of TNF-a (B) and ICAM-1 (D) mRNA. ap !0.01 vs. the TC group at 3 h; bp !0.01 vs. the TC group at 6

h; cp !0.01 vs. the TC group at the same time point. dp !0.01 vs. the CON group at the same time point.

86 Chen et al./ Archives of Medical Research 40 (2009) 79e88

potent anti-inflammatory properties in part through inhibi-tion of NF-kB activation. NF-kB plays a key role in the in-flammatory response in AP, and the intervention of NF-kBactivation eliminates the induced overexpression of inflam-matory cytokines such as TNF-a (36,37).

TNF-a is a critical cytokine and plays a pivotal role inthe pathogenesis of AP. The function of TNF-a includesregulation the expression of inflammatory genes, cell death,and the recruitment and activation of immune cells(39e41). The rise in both tissue and serum TNF-a concen-trations correlates directly with the severity of pancreaticdamage and inflammation in AP (42). Preventing the activ-ity of TNF-a has a beneficial effect on the severity and mor-tality of AP (43). TNF-a participates in the process ofinflammatory cell recruitment by upregulating selectins,ICAM-1 and VCAM-1 (44,46). In this study we showeda 6-h peak for TNF-a expression and 12-h peak forICAM-1 expression in pancreas. This antecedent expres-sion of proinflammatory cytokine correlated with the subse-quent recruitment of inflammatory cells into the pancreas.Prophylactic administration of rosiglitazone markedlyreduced TNF-a expression in both pancreas and lung at12 h after AP. Rollins et al. also demonstrated pretreatmentwith another PPAR-g ligand—troglitazone—completely

abolished the increase of proinflammatory cytokines IL-6and TNF-a mRNA expression in cerulein-induced mice(45).

The infiltration of inflammatory cells into the pancreas isan early and central event in AP that promotes local injuryand systemic complications (46). Sequestration of leuko-cytes, particularly neutrophils, plays a critical role in this pro-cess. ICAM-1 is known to be upregulated in AP and it playsa deleterious role in the development of the disease by neutro-phil recruitment into pancreas and distant organs (13,47).Werner et al. demonstrated a correlation between the extentof ICAM-1 expression and the severity of pancreatic injury.Treatment with monoclonal anti-ICAM-1 antibody de-creased both local pancreatic injury and systemic pulmonaryinjury (15). In ICAM-1 knockout mice, cerulein-induced APand associated lung injury were efficiently prevented, prov-ing the important role for ICAM-1 in the development of pan-creatitis and subsequent lung damage (13). In the presentstudy we demonstrate that ICAM-1 mRNA and protein ex-pression were significantly increased in both pancreatic andpulmonary tissue after sodium taurocholate infusion. Pre-treatment with rosiglitazone markedly attenuated the expres-sion of ICAM-1 mRNA and protein in both pancreas and lungat 12 h after AP. The increase of ICAM-1 expression in

Figure 8. Expression of ICAM-1 protein in pancreatic (A) and pulmonary (C) tissue using Western blot analysis at 12 h of each group. CON, saline þ vehicle

as the control group; TC, sodium taurocholate þ vehicle group; ROSI, sodium taurocholate þ rosiglitazone group. Actin served as an internal control. Com-

parison of the expression of ICAM-1 protein in pancreatic (B) and pulmonary (D) tissue. ap !0.01 vs. the CON group; bp !0.01 vs. the TC group.

87Rosiglitazone Attenuates the Severity of Sodium Taurocholate-induced Acute Pancreatitis and Lung Injury

pancreas and lung was associated with an increase of tissueleukocyte infiltration that was detected by measurement ofMPO activity. MPO activity significantly increased afterAP and markedly reduced by pretreatment with rosiglitazoneat 12-h time point. Our results are consistent with previous re-ports by Cuzzocrea et al. who demonstrated that rosiglita-zone attenuates the severity of acute inflammation throughthe reduction of pulmonary ICAM-1 expression in carra-geenan-treated rats (29) and pancreatic ICAM-1 expressionin cerulein-treated mice (48).

In summary, our study provides evidences that rosiglita-zone, a specific PPAR-g ligand, markedly decreased theseverity of pancreatic and pulmonary injury in AP. Ourfindings suggested the potential role of rosiglitazone againstAP and pancreatitis-associated lung injury. However, thereare several limitations to our study. Because rosiglitazonepretreatment seems unsuitable for use in clinical practice,our further experiments need to confirm the effect of rosi-glitazone treatment in AP and its complications.

AcknowledgmentsWe completed the study at the Institute for Gastroenterology andHepatology, Renmin Hospital of Wuhan University. The authorsthank Mr. Hong Xia and Ms. Li Yu for their technical support dur-ing the study.

References1. Renzulli P, Jakob SM, Tauber M, Candinas D, Gloor B. Severe acute

pancreatitis: case-oriented discussion of interdisciplinary manage-

ment. Pancreatology 2005;5:145e156.

2. Lund H. Tønnesen H, Tønnesen MH, Olsen O. Long-term recurrence

and death rates after acute pancreatitis. Scand J Gastroenterol 2006;41:

234e238.

3. Williams M, Simms HH. Prognostic usefulness of scoring systems in

critically ill patients with severe acute pancreatitis. Crit Care Med

1999;27:901e907.

4. Mann DV, Hershmann MJ, Hittinger R, Glazer G. Multicentre audit of

death from acute pancreatitis. Br J Surg 1994;81:890e893.

5. Bhatia M, Brady M, Shokuhi S, Christmas S, Neoptolemos JP,

Slavin J. Inflammatory mediators in acute pancreatitis. J Pathol

2000;190:117e125.

6. Domınguez-Munoz JE, Carballo F, Garcıa MJ, et al. Monitoring of

serum proteinaseeantiproteinase balance and systemic inflammatory

response in prognostic evaluation of acute pancreatitis. Results of

a prospective multicenter study. Dig Dis Sci 1993;38:507e513.

7. Springer TA. Adhesion receptors of the immune system. Nature 1990;

346:425e434.

8. Pober JS, Cotran RS. The role of endothelial cells in inflammation.

Transplantation 1990;50:537e544.

9. Roebuck KA, Finnegan A. Regulation of intercellular adhesion mole-

cule-1 (CD54) gene expression. J Leukoc Biol 1999;66:876e888.

10. Roebuck KA. Oxidant stress regulation of IL-8 and ICAM-1 gene ex-

pression: differential activation and binding of the transcription factors

AP-1 and NF-kB. Int J Mol Med 1999;4:223e230.

11. Bernot D, Peiretti F, Canault M, Juhan-Vague I, Nalbone G. Upregu-

lation of TNF-a-induced ICAM-1 surface expression by adenylate

cyclase-dependent pathway in human endothelial cells. J Cell Physiol

2005;202:434e441.

12. Hartwig W, Werner J, Warshaw AL, et al. Membrane-bound ICAM-1

is upregulated by trypsin and contributes to leukocyte migration in

acute pancreatitis. Am J Physiol Gastrointest Liver Physiol 2004;

287:G1194eG1199.

13. Frossard JL, Saluja A, Bhagat L, et al. The role of intercellular adhe-

sion molecule 1 and neutrophils in acute pancreatitis and pancreatitis-

associated lung injury. Gastroenterology 1999;116:694e701.

14. Werner J, Hartwig W, Schmidt E, Gebhard MM, Herfarth C, Klar E.

Reduction of local and systemic complications of acute pancreatitis

by monoclonal antibody to ICAM-1. Langenbecks Arch Chir Suppl

Kongressbd 1998;115:725e729.

88 Chen et al./ Archives of Medical Research 40 (2009) 79e88

15. Werner J, Z’graggen K, Fernandez-del Castillo C, Lewandrowski KB,

Compton CC, Warshaw AL. Specific therapy for local and systemic

complications of acute pancreatitis with monoclonal antibodies

against ICAM-1. Ann Surg 1999;229:834e840.

16. Evans RM. The steroid and thyroid hormone receptor superfamily.

Science 1988;240:889e895.

17. Kliewer SA, Umesono K, Noonan DJ, Heyman RA, Evans RM. Con-

vergence of 9-cis retinoic acid and peroxisome proliferator signalling

pathways through heterodimer formation of their receptors. Nature

1992;358:771e774.

18. Issemann I, Prince RA, Tugwood JD, Green S. The peroxisome prolifera-

tor- activated receptor: retinoid X receptor heterodimer is activated by fatty

acids and fibrate hypolipidaemic drugs. J Mol Endocrinol 1993;11:37e47.

19. Chawla A, Schwarz EJ, Dimaculangan DD, Lazar MA. Peroxisome

proliferator activated receptor (PPAR) gamma: adipose-predominant

expression and induction early in adipocyte differentiation. Endocri-

nology 1994;135:798e800.

20. Nolan JJ, Ludvik B, Beerdsen P, Joyce M, Olefsky J. Improvement in

glucose tolerance and insulin resistance in obese subjects treated with

troglitazone. N Engl J Med 1994;331:1188e1193.

21. Ricote M, Li AC, Willson TM, Kelly CJ, Glass CK. The peroxisome

proliferator-activated receptor-gamma is a negative regulator of mac-

rophage activation. Nature 1998;391:79e82.

22. Jiang C, Ting AT, Seed B. PPAR-g agonists inhibit production of

monocyte inflammatory cytokines. Nature 1998;391:82e86.

23. Shiojiri T, Wada K, Nakajima A, et al. PPAR g ligands inhibit nitrotyr-

osine formation and inflammatory mediator expressions in adjuvant-

induced rheumatoid arthritis mice. Eur J Pharmacol 2002;448:231e238.

24. Su CG, Wen X, Bailey ST, et al. A novel therapy for colitis utilizing

PPAR-g ligands to inhibit the epithelial inflammatory response. J Clin

Invest 1999;104:383e389.

25. Cuzzocrea S, Pisano B, Dugo L, et al. Rosiglitazone and 15-deoxy-

d12,14-prostaglandin J2, ligands of the peroxisome proliferator-

activated receptor-gamma (PPAR-g), reduce ischaemia/reperfusion

injury of the gut. Br J Pharmacol 2003;140:366e376.

26. Nakajima A, Wada K, Miki H, et al. Endogenous PPAR g mediates

anti-inflammatory activity in murine ischemia-reperfusion injury. Gas-

troenterology 2001;120:460e469.

27. Murphy GJ, Holder JC. PPAR-g agonists: therapeutic role in diabetes,

inflammation and cancer. Trends Pharmacol Sci 2000;21:469e474.

28. Paszkowski AS, Rau B, Mayer JM, Moller P, Beger HG. Therapeutic

application of caspase 1/interleukin-1b-converting enzyme inhibitor

decreases the death rate in severe acute experimental pancreatitis.

Ann Surg 2002;235:68e76.

29. Cuzzocrea S, Pisano B, Dugo L, et al. Rosiglitazone, a ligand of the

peroxisome proliferator-activated receptor-g, reduces acute inflamma-

tion. Eur J Pharmacol 2004;483:79e93.

30. Schmidt J, Rattner DW, Lewandrowski K, et al. A better model of

acute pancreatitis for evaluating therapy. Ann Surg 1992;215:44e56.

31. Bhatia M, Saluja AK, Hofbauer B, et al. Role of substance P and the

neurokinin 1 receptor in acute pancreatitis and pancreatitis-associated

lung injury. Proc Natl Acad Sci USA 1998;95:4760e4765.

32. Trush MA, Egner PA, Kensler TW. Myeloperoxidase as a biomarker

of skin irritation and inflammation. Food Chem Toxicol 1994;32:

143e147.

33. Escher P, Wahli W. Peroxisome proliferator-activated receptors: in-

sight into multiple cellular functions. Mutat Res 2000;448:121e138.

34. Fahmi H, Di Battista JA, Pelletier JP, Mineau F, Ranger P, Martel-

Pelletier J. Peroxisome proliferator-activated receptor g activators in-

hibit interleukin-1b-induced nitric oxide and matrix metalloproteinase

13 production in human chondrocytes. Arthritis Rheum 2001;44:

595e607.

35. Kelly D, Campbell JI, King TP, et al. Commensal anaerobic gut bac-

teria attenuate inflammation by regulating nuclear-cytoplasmic shut-

tling of PPARg and RelA. Nat Immunol 2004;5:104e112.

36. Dunn JA, Li C, Ha T, Kao RL, Browder W. Therapeutic modification

of nuclear factor kappa B binding activity and tumor necrosis factor-

a gene expression during acute biliary pancreatitis. Am Surg 1997;63:

1036e1044.

37. Gukovsky I, Gukovskaya AS, Blinman TA, Zaninovic V, Pandol SJ.

Early NF-kB activation is associated with hormone-induced pancrea-

titis. Am J Physiol 1998;275(6 Pt 1):G1402eG1414.

38. Hashimoto K, Ethridge RT, Saito H, Rajaraman S, Evers BM. The

PPARg ligand, 15d-PGJ2, attenuates the severity of cerulein-induced

acute pancreatitis. Pancreas 2003;27:58e66.

39. Malleo G, Mazzon E, Siriwardena AK, Cuzzocrea S. Role of tumor

necrosis factor-a in acute pancreatitis: from biological basis to clinical

evidence. Shock 2007;28:130e140.

40. Bhatia M, Wong FL, Cao Y, et al. Pathophysiology of acute pancrea-

titis. Pancreatology 2005;5:132e144.

41. Pastor CM, Frossard JL. Are genetically modified mice useful for the

understanding of acute pancreatitis? FASEB J 2001;15:893e897.

42. Norman JG, Fink GW, Franz MG. Acute pancreatitis induces intra-

pancreatic tumor necrosis factor gene expression. Arch Surg 1995;

130:966e970.

43. Denham W, Yang J, Fink G, et al. Gene targeting demonstrates addi-

tive detrimental effects of interleukin 1 and tumor necrosis factor dur-

ing pancreatitis. Gastroenterology 1997;113:1741e1746.

44. Collins T, Read MA, Neish AS, Whitley MZ, Thanos D, Maniatis T.

Transcriptional regulation of endothelial cell adhesion molecules:

NF-kB and cytokine-inducible enhancers. FASEB J 1995;9:899e909.

45. Rollins MD, Sudarshan S, Firpo MA, et al. Anti-inflammatory effects

of PPAR-g agonists directly correlate with PPAR-gamma expression

during acute pancreatitis. J Gastrointest Surg 2006;10:1120e1130.

46. Vonlaufen A, Apte MV, Imhof BA, Frossard JL. The role of inflamma-

tory and parenchymal cells in acute pancreatitis. J Pathol 2007;213:

239e248.

47. Lundberg AH, Granger N, Russell J, et al. Temporal correlation of tu-

mor necrosis factor-a release, upregulation of pulmonary ICAM-1 and

VCAM-1, neutrophil sequestration, and lung injury in diet-induced

pancreatitis. J Gastrointest Surg 2000;4:248e257.

48. Cuzzocrea S, Pisano B, Dugo L, et al. Rosiglitazone, a ligand of the

peroxisome proliferator-activated receptor-g, reduces acute pancreatitis

induced by cerulein. Intensive Care Med 2004;30:951e956.