Acute pancreatitis induces FasL gene expression and apoptosis in the liver1,2

Journal of Surgical Research 122, 201–209 (2004) doi:10.1016/j.jss.2004.05.019

Acute Pancreatitis Induces FasL Gene Expression and Apoptosis in the Liver 1,2 Scott F. Gallagher, M.D.,* Jun Yang, M.D.,* Kathryn Baksh, B.S.,* Krista Haines, B.A.,* Heather Carpenter, B.S.,* P. K. Epling-Burnette, Ph.D.† Yanhua Peng, Ph.D.,* James Norman, M.D., F.A.C.S.,* and Michel M. Murr, M.D., F.A.C.S.*,3 *Department of Surgery and †Department of Internal Medicine, James A. Haley Veterans Affairs Medical Center; University of South Florida Health Sciences Center, Tampa, Florida Submitted for publication February 2, 2004

Background. Liver injury is an important prognostic indicator in acute pancreatitis. We previously demonstrated that Kupffer cell-derived cytokines mediate liver injury. In this work, we sought to characterize the role of Fas Ligand (FasL) in liver injury during acute pancreatitis. Methods. Acute pancreatitis was induced in mice using cerulein; serum FasL, AST, ALT, liver FasL, p38MAPK, and caspase-3 were measured. FasL mRNA and protein and its receptor (Fas) were determined in rat Kupffer cells treated with elastase (1 U/ml) to mimic acute pancreatitis. Apoptosis was measured by flow cytometry. Results. Cerulein-induced pancreatitis increased serum AST, ALT, and FasL and up-regulated liver FasL (1315 ⴞ 111 versus 310 ⴞ 164 pg/ml, P ⴝ 0.002 versus sham), while inducing p38-MAPK phosphorylation (P < 0.01 versus sham) and cleavage of caspase-3 (P < 0.04 versus sham); all were attenuated by pretreatment with the Kupffer cell inhibitor, gadolinium (all P < 0.003). In vitro, elastase induced a time-dependent increase in Kupffer cell FasL protein (FasL ⴝ 404 ⴞ 94 versus 170 ⴞ 40, P ⴝ 0.02, versus control), a 100-fold increase in FasL mRNA, and up-regulated Fas (FasL receptor). Gadolinium significantly attenuated the elastase-induced increase in FasL and FasL mRNA (FasL ⴝ 230 ⴞ 20 versus 404 ⴞ 94, P ⴝ 0.01, versus elastase) but had little effect on Fas. Additionally, 1 Presented at the 37th Annual Meeting of the Association for Academic Surgery, November 13–15, 2003, Sacramento, CA. 2 Supported by SSAT Career Development Award (M.M.), Dr. Bob Haines Pancreatitis Research Fund (M.M.), VA Merit Award (J.N.), and VA Merit Award (P.K.E.). 3 To whom correspondence and reprint requests should be addressed at the University of South Florida c/o Tampa General Hospital, PO Box 1289, Tampa, FL 33601. E-mail: [email protected].

elastase-primed Kupffer cell media induced apoptosis in hepatocytes (29 ⴞ 1 versus 16% ⴞ 1%; versus control, P < 0.001). Conclusions. Acute pancreatitis induces liver injury and hepatocyte death while up-regulating FasL, p38MAPK, and caspase-3. Fas is up-regulated within Kupffer cells, suggesting that FasL may autoregulate its production by inducing its originator-cell death. The ability to manipulate interactions between Kupffer cells and hepatocytes may have important therapeutic implications. © 2004 Elsevier Inc. All rights reserved. Key Words: acute pancreatitis; Kupffer cells; Fas ligand; liver injury; hepatocyte apoptosis; p38-MAPK; caspase-3; cell signaling.

INTRODUCTION

Acute pancreatitis is a form of noninfectious, endogenous pancreatic injury that manifests with local, peripancreatic inflammation as well as a systemic inflammatory response and distant organ injury, such as liver injury and acute respiratory distress syndrome [1]. Liver injury is of particular importance as a clinical prognostic indicator and is incorporated into the various scoring systems used to predict the severity of acute pancreatitis. This laboratory has demonstrated that pancreatic enzymes play a major role in extra-pancreatic, macrophage-derived, and organ-specific cytokine production, suggesting a possible link between localized inflammation of the pancreas and the systemic manifestations of pancreatitis in experimental models of liver and lung injury [2– 4]. In that regard, the liver is a unique organ because of its resident macrophages, Kupffer cells, which are the largest population of fixed

201

0022-4804/04 $30.00 © 2004 Elsevier Inc. All rights reserved.

202

JOURNAL OF SURGICAL RESEARCH: VOL. 122, NO. 2, DECEMBER 2004

tissue macrophages [5, 6]. As such, this laboratory has recently demonstrated that pancreatic elastase upregulates Kupffer cell TNF and FasL expression in vitro [7]. Fas ligand is produced by activated lymphocytes such as T-lymphocytes (T-cells) and natural killer (NK) cells while also functioning as an effector of these cytotoxic cells to remove infected (virus and bacteria) cells or neoplastic cells. Although FasL-induced apoptosis is reported to play an important role in parenchymal cell damage in liver disease [8], acute renal failure, and thyroiditis, there is paucity of information regarding the role of Kupffer cell-derived FasL during acute pancreatitis. The purpose of this study was to characterize FasL gene expression and its role in hepatocyte injury during acute pancreatitis. METHODS Animal care was in accordance with the guidelines established by the Department of Laboratory Animal Medicine at the University of South Florida (USF), a facility accredited by the American Association for Accreditation in Laboratory Animal Care. Studies were conducted with approval from the Institutional Animal Care and Use Committee at USF.

Induction of Acute Pancreatitis Acute pancreatitis was induced in male, NIH Swiss Mice (18-20 g) by intraperitoneal injection of cerulein (50 ␮g/kg IP q4 h for 24 h), a cholecystokinin analogue. Control mice received 0.1 ml of 0.9% NaCl intraperitoneally (q4 h for 24 h). Mice were sacrificed at 4, 8, 16, and 24 h after the first cerulein injection (n ⫽ 5, each time point). In another set of experiments, NIH Swiss mice (n ⫽ 15) were pretreated with the Kupffer cell inhibitor, gadolinium chloride (Gd: 10 mg/kg, IVx1; n ⫽ 5), 24 h prior to the induction of acute pancreatitis [2, 3] with cerulein (n ⫽ 10) or injection of saline (n ⫽ 5); subsequently, mice were sacrificed 4 h after the induction of acute pancreatitis. Serum was harvested by heart puncture for determination of liver parenchymal enzymes and FasL (n ⫽ 4). Livers were harvested, processed with TRIzol Reagent (Invitrogen, Carlsbad, CA) for RNA isolation and protein isolation per the package insert instructions [9, 10], and stored at ⫺80°C prior to reverse transcription polymerase chain reaction (RT-PCR) and immunoblotting, respectively. To determine RNA and total protein under precisely the same experimental conditions in addition to the small-scale nature of the experiments and sample sizes, RNA and total protein were isolated from the same samples; this method has been well-established and validated in the literature [9, 10].

Hepatocyte Tissue Cultures Hepatocytes were isolated from male, NIH Swiss mice (20-25 g) by digestion with collagenase as described previously [8, 11]. Livers were perfused in situ through the portal vein with 10 mM HEPES buffered saline (0.15 M NaCl, 0.42 g/l KCl, 0.99 g/l glucose, 2.1 g/l NaCO 3, and 0.19 g/l EDTA) until cleared of blood and then perfused for 10 min with modified HEPES-buffered saline (no EDTA, 3.5 mM CaCl 2, 1% BAS, and 0.025% collagenase). The liver parenchyma was then dispersed manually, filtered through a 70-␮m then through a 200-␮m pore mesh (CellMicroSieve BioDesign Inc. of New York, Carmel, NY), and centrifuged twice for 3 min at 50g to remove nonparenchymal cells. Hepatocytes were plated at a density of approximately 5 ⫻ 10 5 in 6-well primary tissue culture plates (Becton Dickinson Labware, Franklin Lakes, NJ) with Dulbecco’s Modified

Eagle medium (Atlanta Biologics, Atlanta, GA) supplemented with 200 mM L-glutamine (Sigma, St. Louis, MO), penicillin (100 U/ml), streptomycin (100 ␮g/ml), and 10% fetal bovine serum. Hepatocytes were then stored at 37°C in humidified air with 5% CO 2. The medium was replaced, and nonadherent cells were removed. Hepatocyte viability was determined by trypan blue exclusion assay.

Kupffer Cell Tissue Cultures Freshly isolated rat Kupffer cells were provided by Dr. Hide Tsukamoto at the Non-Parenchymal Liver Cell Isolation Core in the USC Research Center for Liver Disease and USC-UCLA Research Center for Alcoholic Liver and Pancreatic Diseases. Briefly, the cells were isolated from male, Sprague–Dawley rats (350 – 450 g) by in situ sequential digestion of the liver with Pronase and collagenase, low-speed centrifugation to separate parenchymal and nonparenchymal cells, and subsequent separation of a Kupffer cell-enriched fraction by discontinuous arabinogalactin gradient centrifugation [12]. Kupffer cells were incubated in Dulbecco’s Modified Eagle medium (Atlanta Biologics, Atlanta, GA) supplemented with 200 mM L-glutamine (Sigma), penicillin (100 U/ml), streptomycin (100 ␮g/ ml), and 10% fetal bovine serum. Cells were kept for 24 h at 37°C in humidified air with 5% CO 2 before any treatment, and nonadherent cells were discarded. Kupffer cell viability was assessed by exclusion of trypan blue.

Serum Liver Enzymes and FasL Blood samples were centrifuged at 3000 RPM for 10 min in 1.5-ml microcentrifuge tubes, and serum was collected and stored at ⫺80°C until use. Liver parenchymal enzymes (AST and ALT) were determined using a Kodak Ektachem 700 automated analyzer (Kodak, Rochester, NY) as a measure of acute pancreatitis-related liver injury. Fas Ligand was determined (control, AP, and Gadolinium ⫹ AP; n ⫽ 4 in each group) using a commercially available enzymelinked immunosorbent assay (ELISA, Alexis Biochemical, San Diego, CA) kit for humans.

Liver FasL, p38-MAPK, and Caspase-3 (Immunoblotting) Using TRIzol Reagent total protein was isolated from mice livers from the phenol-ethanol supernatant collected after DNA precipitation with ethanol [10], separated by sodium dodecylsulfatepolyacrylamide gel electrophoresis, and transferred to a nitrocellulose membrane. Nonspecific binding was blocked with 5% of bovine serum albumin (Sigma) for 2 h at room temperature. The membranes were then incubated overnight at 4°C with 1:1000 dilution of anti-Fas Ligand/CD95L primary antibodies (BD Biosciences, San Diego, CA), 1:1000 dilution of polyclonal, phospho-specific p38MAPK (Cell Signaling Technology, Beverly, MA), or 1:3000 dilution of caspase-3 antibodies (PharMingen International, San Diego, CA). Subsequently, the membrane was washed and incubated with horseradish peroxidase-conjugated anti-mouse monoclonal antibody or polyclonal anti-rabbit antibody according to the matching requirement of the primary antibodies (New England Biolabs, Beverly, MA) for 2 h at room temperature. The immunoblot was washed; the bands were detected with an enhanced chemiluminescence kit (LumiGlo, New England Biolabs, Beverly, MA) and were quantified by densitometry using UVP Gel Documentation System (GDS) 8000 (UVP, Upland, CA). For these and all subsequent experiments, immunoblots were repeated in triplicate using positive controls provided by the manufacturers (not shown in the selected gels). Western immunoblot band position was confirmed by the antibody molecular weight for each protein (FasL ⫽ 37 kDa, Fas ⫽ 45 kDa, p38-MAPK ⫽ 43 kDa, and caspase-3 cleavage products ⫽ 17 kDa and 19 kDa), respectively.

203

GALLAGHER ET AL.: KUPFFER CELL-DERIVED FAS LIGAND AND HEPATOCYTE INJURY

Kupffer Cell FasL Gene Expression (RT-PCR) Rat Kupffer cells (2 ⫻ 10 7, purity ⬎98%) were seeded in 100-mm tissue culture dishes and were treated with elastase (1 U/ml, Sigma) with or without pretreatment with gadolinium chloride (0.25 mg/ml, 24 h prior to treatment with elastase). Both elastase and gadolinium chloride have been validated in our laboratory and do not affect the Kupffer cell viability at these doses [2, 3]. We have previously validated the elastase dose based on dose–response experiments using human monocytes (unpublished data), rat macrophages, and fresh Kupffer cells [2]. Specifically, elastase (up to doses of 1 U/ml) does not affect Kupffer cell viability as determined by MTT assay and exclusion of trypan blue, while inducing maximal TNF production [2, 3]. Fas Ligand mRNA was measured by semiquantitative differential RT-PCR . Briefly, the total RNA was isolated by guanidinium thiocyanate-phenol-extraction [9, 10]. Four micrograms of total RNA were primed using oligo(dT) (Gibco, Gaithersburg, MD) and subsequently reversed transcribed with reverse transcriptase (Superscript II, Gibco, Gaithersburg, MD). The resulting cDNA products were coamplified in the presence of rat-specific FasL and BMG (betamicroglobulin) primers for 35 cycles of PCR (FasL; 35 cycles of 94°C, 45 s; 61°C, 60 s; 72°C, 45 s, with 10 min extension time at 72°C) in an UNO-Thermoblock (Biometra, Tampa, FL). The sequence for the FasL primer [7] was sense 5=ATGGAACTGCTTTGATCTCTGG 3= and antisense 5=AGATTCCTCAAAATTGATCAGAG 3= (Life Technologies, Inc. Products; Grand Island, NY). The BMG primer sequence was sense 5=CTCCCCAAATTCAAGTGTACTCTCG3= and antisense 5=GAGTGACGTGTTTAACTCTGCAAGC3= (Ransom Hill Biosciences, Ramona, CA). All primers are known to span at least one intron. The reaction products were separated with electrophoresis in 3% agarose gel containing ethidium bromide and digitally photographed under UV light with the UVP GDS 8000 (UVP; Upland, CA). Band intensity of each sample was determined using UVP GDS 8000 image analysis software, and individual Fas/BMG cDNA ratios were calculated for analysis.

Kupffer Cell FasL and Fas Receptor (Immunoblotting) Rat Kupffer cells (1 ⫻ 10 7 cells/well, purity ⬎98%) were seeded in 100-mm tissue culture dishes and treated with elastase (1 U/ml; Sigma). Using TRIzol reagent as previously described [10], total protein was isolated in the phenol-ethanol supernatant collected after DNA precipitation with ethanol, separated by sodium dodecylsulfate-polyacrylamide gel electrophoresis, and transferred to a nitrocellulose membrane. Nonspecific binding was blocked with 5% bovine serum albumin (Sigma) for 2 h at room temperature. The membranes were then incubated overnight at 4°C with 1:1000 dilution of anti-Fas Ligand/CD95L and 1:2500 dilution of anti-Fas/ CD95APO-1 primary antibodies (BD Biosciences, San Diego, CA). Subsequently, the membrane was washed and incubated with horseradish peroxidase-conjugated polyclonal anti-rabbit antibody for 2 h at room temperature. The immunoblot was washed, and the bands were detected with an enhanced chemiluminescence kit (LumiGlo, New England Biolabs, Beverly, MA). Kupffer cell FasL and Fas protein (receptor) were quantified by densitometry (UVP, Upland, CA).

Freshly isolated mouse hepatocytes (5 ⫻ 10 5 per well) were seeded in the 6-well FalconTM Cell Culture plates. Twenty-four hours later, the culture medium was carefully removed and replaced with elastasetreated Kupffer cell medium. In previous experiments, elastase (1 U/ml) alone did not affect hepatocyte viability [11, 12]. After 4 h, hepatocytes were washed twice with chilled PBS (pH 7.6), suspended in staining buffer (Clontech; Palo Alto, CA), and then labeled with Annexin-V-FITC (5 ␮L, Clontech) and 7-AAD (10 ␮L, 7-amino-actinomycin D, PharMingen International) following the manufacturer’s protocol. Peak emission of 7-AAD is at 685 nm; hepatocyte apoptosis was measured by multiparameter flow cytometry using an FL3 photomultiplier tube with a 670 nm-long pass filter.

Statistical Analysis All experiments were repeated in at least triplicate as noted above in the methods as well as in the figure legends; results were averaged. All Western immunoblots were repeated in triplicate; controls were measured for each experiment; however, controls are not shown in all figures. Data are mean ⫾ SEM (standard error of the mean). Student’s t test was used to evaluate parametric data. Significance for alpha was set at P ⬍ 0.05.

RESULTS Acute Pancreatitis Induces Liver Injury

The severity of pancreatitis-induced liver injury was quantified measuring biochemical markers of hepatocyte injury (AST and ALT). Cerulein-induced pancreatitis (AP) increased serum AST and ALT levels by at least 2-fold as compared to controls (Table 1). Parenchymal enzymes (AST and ALT) peaked at 4 h (P ⱕ 0.05 versus control), indicating a significant degree of hepatocyte injury. Acute Pancreatitis Increases Serum and Liver FasL

To establish a link between acute pancreatitis and liver injury, serum and liver tissue FasL levels were determined. Cerulein-induced acute pancreatitis (AP) increased serum FasL (0.5 ⫾ 0.05 versus 0.3 ⫾ 0.02 pg/ml; P ⫽ 0.006 versus control). Pretreatment with gadolinium chloride (Gd ⫹ AP) abolished the pancreatitis-induced increase in serum FasL (0.5 ⫾ 0.05 versus 0.3 ⫾ 0.03 ng/ml; P ⫽ 0.02 AP versus Gd ⫹ AP). Similarly, cerulein-induced acute pancreatitis (AP) increased total liver FasL protein 4-fold as compared with sham mice (P ⫽ 0.002, Fig. 1). Pretreatment with gadolinium chloride, an inhibitor of macrophagederived cytokine production, significantly attenuated the pancreatitis-induced increase in liver FasL (P ⫽

Kupffer Cell Medium Induces Hepatocyte Apoptosis (Flow Cytometry) To further test the relationship between elastase-induced, Kupffer cell-derived products and liver injury, the media of elastase-treated Kupffer cells was added to tissue cultures of fresh hepatocytes; apoptosis was measured by flow cytometry. Rat Kupffer cells (1 ⫻ 10 6 per well) were seeded for 24 h and then treated with pancreatic elastase (1 U/ml, Sigma) for 2 h [2, 3]. The medium of the elastase-treated Kupffer cells was collected and centrifuged at 1500 RPM for 10 min. The supernatant was pooled and stored at ⫺80°C for later use.

TABLE 1 AST and ALT Levels During Course of Experiment

AST ALT

Control

4h

8h

16 h

24 h

70 ⫾ 13 32 ⫾ 1

180 ⫾ 37* 71 ⫾ 18†

158 ⫾ 21* 70 ⫾ 7*

158 ⫾ 6* 72 ⫾ 5*

128 ⫾ 21 52 ⫾ 4*

* P ⱕ 0.003 versus control; †P ⬍ 0.05 versus control.

204

JOURNAL OF SURGICAL RESEARCH: VOL. 122, NO. 2, DECEMBER 2004

FIG. 1. FasL levels are significantly up-regulated in the livers of mice with cerulein-induced acute pancreatitis (AP, n ⫽ 4; *P ⫽ 0.02 AP versus control, n ⫽ 4). Pretreatment with gadolinium chloride (Gd⫹AP, n ⫽ 5) significantly attenuated the pancreatitis-induced upregulation of FasL (**P ⫽ 0.001 Gd⫹AP versus AP). Bar graph represents quantification of the immunoblots by densitometry (immunoblots were repeated in triplicate).

0.001, Gd ⫹ AP versus AP; Fig. 1) suggesting that Kupffer cells are the source of FasL in the liver. Acute Pancreatitis Activates p38-MAPK and Caspase-3

To test whether a component of acute pancreatitisinduced liver injury results from apoptosis, we investigated the activity of two proapoptotic cell signaling systems by measuring p38-MAPK phosphorylation and caspase-3 cleavage. Cerulein-induced acute pancreatitis (AP) significantly up-regulated p38-MAPK phosphorylation (P ⫽ 0.01 versus control, Fig. 2) as well as caspase-3 cleavage in mice livers (P ⫽ 0.043 versus control; Fig. 3). Likewise, pretreatment with gadolinium chloride significantly attenuated the pancreatitis-induced increases in p38-MAPK phosphorylation (P ⬍ 0.001, Gd ⫹ AP versus AP; Fig. 2) as well as caspase-3 cleavage (P ⫽ 0.003 Gd ⫹ AP versus AP, Fig. 3). These data suggest that the inhibition of Kupffer cells attenuates up-regulation of proapoptotic cell signaling systems in the liver. Kupffer Cell FasL Gene Expression

To establish whether Kupffer cells are a source of FasL production in the liver during acute pancreatitis, the time course of FasL gene expression in rat Kupffer cells that were treated with elastase to mimic conditions of acute pancreatitis was determined [2, 3]. Elastase induces a time-dependent increase in FasL pro-

duction as measured by immunoblotting of cell lysates (Fig. 4). FasL protein production peaks 120 min after treatment with elastase and significantly declines by 240 min. Pretreatment with gadolinium chloride attenuates the elastase-induced increases in Kupffer Cell FasL at 30, 60, and 120 min (Fig. 4, gel), which was confirmed by triplicate experiments at the 60 min time point (Fig. 4, bar graph). FasL protein production increases incrementally in response to increasing doses of elastase (0.1–1.0 U/ml) from a baseline of 213 ⫾ 37, to 618 ⫾ 111 (0.1 U/ml), to 680 ⫾ 84 (0.5 U/ml), and to 667 ⫾ 146 (1.0 U/ml) as shown in Fig. 5 (P ⬍ 0.04 all versus control). Kupffer cell FasL mRNA is up-regulated as early as 15 min after treatment with elastase. Similarly, pretreatment with gadolinium chloride significantly attenuates the elastase-induced up-regulation of FasL mRNA (Fig. 6). Kupffer Cell Fas (FasL receptor)

Similarly, we sought to determine whether Fas, the receptor for FasL, is concomitantly up-regulated in Kupffer cells. Elastase up-regulates Fas within Kupffer cells. Similar to FasL, Fas is up-regulated within 15–30 min after treatment with elastase and peaks at 120 min (Fig. 7). Pretreatment with gadolinium chloride has a negligible effect on Kupffer cell Fas expression.

GALLAGHER ET AL.: KUPFFER CELL-DERIVED FAS LIGAND AND HEPATOCYTE INJURY

205

FIG. 2. Cerulein-induced acute pancreatitis (AP, n ⫽ 4) significantly up-regulatesd p38-MAPK in liver homogenates (*P ⫽ 0.01 versus AP versus control, n ⫽ 4). Pretreatment with gadolinium chloride (Gd⫹AP, n ⫽ 5) significantly attenuated the pancreatitis-induced phosphorylation of p38-MAPK (**P ⬍ 0.001, Gd⫹AP versus AP). Bar graph is a quantification of corresponding immunoblots by densitometry (immunoblots were repeated in triplicate).

Elastase-Treated Kupffer Cell Medium Induces Hepatocyte Apoptosis

To confirm that byproducts of elastase-treated Kupffer cells induce liver injury and hepatocyte apoptosis while linking our in vitro and in vivo observations, pooled medium from elastase-treated Kupffer cells increases dual-labeling of fresh hepatocytes with Annexin-V and 7-AAD from 16 ⫾ 1 to 29% ⫾ 1% (P ⬍ 0.001 versus control). DISCUSSION

Our understanding of the pathophysiology of acute pancreatitis has evolved significantly as a result of elucidating the role of macrophage-derived cytokines in the systemic progression of acute pancreatitis. We previously demonstrated that pancreatic enzymes, which may gain access to the systemic circulation as a result of inflammatory changes in the pancreas and the retroperitoneum, induce production of proinflammatory cytokines within distant organs such as the liver and lungs [1– 4]. In addition, we have also demonstrated that resident macrophages within the lungs as well as the liver produce large amounts of TNF and FasL in response to stimulation with pancreatic en-

zymes, most significantly elastase [1– 4,7]. However, few data exist regarding physiological concentrations of elastase, or any other potential inciting inflammatory agent, in the portal vein or systemic circulation during acute pancreatitis. Notwithstanding, elastase can be used to mimic the effects of acute pancreatitis in vitro and to map the molecular pathways governing cytokine gene expression in Kupffer cell tissue cultures [2, 4, 7] otherwise impractical to undertake in whole animal models. Liver injury as measured by increases in the serum levels of liver parenchymal enzymes is an important prognostic indicator and as such has been incorporated into Ranson’s criteria and the APACHE-II scoring system. As a potential model to investigate mechanisms of liver injury, we investigated the interaction between Kupffer cells and hepatocytes during acute pancreatitis. Although some consider the liver an amplifier of the systemic response, our study focused on the role of Kupffer cell-derived FasL in the pathogenesis of hepatocyte injury and hepatocyte death. We investigated the possibility that FasL is produced by Kupffer cells because of the pivotal role that FasL plays in the pathogenesis of various diseases and cellular apoptosis [7, 13, 14]. In preliminary investiga-

206

JOURNAL OF SURGICAL RESEARCH: VOL. 122, NO. 2, DECEMBER 2004

FIG. 3. Cerulein-induced acute pancreatitis (AP, n ⫽ 4) activated caspase-3 and increased its cleavage products (*P ⫽ 0.043 AP versus control, n ⫽ 3). Pretreatment with gadolinium chloride (Gd⫹AP, n ⫽ 5) significantly attenuated the pancreatitis-induced activation of caspase-3 and reduces its cleavage products (**P ⫽ 0.003 Gd⫹AP versus AP). Bar graph is a quantification of corresponding immunoblots by densitometry (immunoblots were repeated in triplicate).

tions, we have demonstrated that Kupffer cell-derived FasL gene expression is up-regulated in vitro by pancreatic elastase and that FasL induces hepatocyte apoptosis [7]. The aim of this study was to investigate FasL gene expression in an in vivo model of acute pancreatitis and to further characterize the mechanisms of FasL up-regulation. Cerulein induces a mild-to-moderate form of acute pancreatitis in mice that lends itself to long-term, reliable, and reproducible studies. Cerulein-induced acute pancreatitis induces liver injury as measured by a time-dependent increase in serum levels of AST and ALT. Concomitantly after induction of acute pancreatitis, serum FasL increases significantly and parallels the induction of other cytokines [15]. Interestingly, the cerulein-induced acute pancreatitis increases in serum FasL are attenuated by the Kupffer cell inhibitor gadolinium chloride suggesting that Kupffer cells may be the source of FasL in this model. Therefore, we sought to confirm whether FasL could be produced within the liver-and specifically, from Kupffer cells-during acute pancreatitis using a combination of in vivo and in vitro experiments.

Studying the in vivo model, cerulein-induced acute pancreatitis up-regulates FasL mRNA and FasL protein in mice livers shortly after induction of pancreatitis. Similar to serum FasL, pretreatment with gadolinium chloride significantly attenuates the pancreatitisinduced up-regulation of FasL gene expression, thereby, strongly suggesting that FasL is produced within Kupffer cells. Subsequently, FasL production in rat Kupffer cells was investigated using an in vitro model that mimics pancreatitis [2, 3]. In vitro, pancreatic elastase induces a timedependent increase in FasL mRNA and FasL protein within tissue cultures of rat Kupffer cells. It is very interesting that not only is the cellular machinery to produce FasL up-regulated by elastase but the machinery to produce Fas, the FasL receptor, was upregulated as well within the same time period. In that regards, FasL up-regulation within Kupffer cells exhibits the universal property of autoregulation by inducing Fas up-regulation and therefore apoptosis of its originator immunocompetent cell. In a similar model, gadolinium chloride inhibited the elastase-induced up-regulation of TNF mRNA by at-

GALLAGHER ET AL.: KUPFFER CELL-DERIVED FAS LIGAND AND HEPATOCYTE INJURY

207

FIG. 4. Immunoblot (repeated in triplicate) demonstrates the time course of FasL production by Kupffer cells treated with Elastase (E, n ⫽ 3) at 15, 30, 60, 120, and 240 min. Production of FasL peaks at 120 min and declined by 240 min. Pretreatment with gadolinium chloride (G⫹E, n ⫽ 3) attenuates elastase-induced FasL production at 15, 30, 60, and 120 min. Bar graph represents quantification of triplicate experiments at 60 min by densitometry (*P ⫽ 0.02 E versus control; **P ⫽ 0.04 GD⫹E versus E).

tenuating the activation of NF-␬B and its upstream regulators [2, 3]. Therefore, we use gadolinium chloride as a powerful inhibitor of Kupffer cells. In this current study, gadolinium chloride significantly attenuates the elastase-induced up-regulation of FasL, not only suggesting that FasL originates within Kupffer cells but also that FasL and TNF gene expression have similar characteristics, likely sharing common pathways within Kupffer cells. Moreover, induction of acute pancreatitis and the associated up-regulation of FasL is associated with p38-MAPK phosphorylation and caspase-3 cleavage in mice livers, thereby, suggesting that cerulein-induced acute pancreatitis up-regulates proapoptotic pathways in the liver and promotes hepatocyte injury as well as hepatocyte death. Gadolinium chloride significantly attenuates p38-MAPK phosphorylation and caspase-3 cleavage in vivo suggesting that inhibition of Kupffer cell function and Kupffer cell-derived cytokines as well as FasL protein production inhibit the up-regulation of important proapoptotic pathways within the liver.

We previously demonstrated that p38-MAPK and caspase-3 are activated within hepatocytes when exposed to elastase-treated Kupffer cell medium in vitro; therefore, we suspect that the major source of p38MAPK and caspase-3 for the in vivo model is within the hepatocytes. Moreover, elastase-treated Kupffer cell medium induces apoptosis of fresh hepatocytes as measured by dual-labeling with Annexin-V and 7-AAD. These data strongly suggest that Kupffer cell-derived FasL plays a role in liver injury and hepatocyte apoptosis. These findings are consistent with published literature regarding the important role of FasL in hepatocyte apoptosis [13, 14, 16]. Through activation of Fas, which is a member of the TNFR family of receptors, FasL activates FADD (Fas-associated death domain) and unmasks its DED (death effector domain) that subsequently activates the caspase cascade and downstream effector caspases, including caspase-3, that ultimately leads to DNA cleavage and cell apoptosis [17, 18]. Additionally, macrophage-derived cytokines can in-

208

JOURNAL OF SURGICAL RESEARCH: VOL. 122, NO. 2, DECEMBER 2004

FIG. 6. Elastase up-regulated Kupffer cell-derived FasL-mRNA within 30 min of treatment. Pretreatment with gadolinium chloride inhibits the elastase-induced up-regulation of FasL mRNA.

duce in liver injury as demonstrated by our prior experiments including cocultures of hepatocytes with Kupffer cells [7] and cultures in the presence of the media from elastase-treated Kupffer cells as well as experiments by other investigators [17]. Although it is impractical to characterize every Kupffer cell byproduct, we focused the current investigation on FasL because of its pivotal and established role in hepatocyte apoptosis. Although we acknowledge certain limitations of this study, we have ample data on the utility of elastase as a surrogate for acute pancreatitis, particularly as it abbreviates induction and mimics the events of acute pancreatitis. Additionally, the data generated by our in vitro experimental model complements and expands on the in vivo data. Commercial ELISA kits for rat FasL and Fas are not available; however, a human FasL ELISA kit was sub-

stituted to measure serum FasL levels, and the potential issues considered. In additional, using ELISA kits for tissue samples is fraught with technical flaws, especially that the protocol reagents and tissue (as compared to either serum or culture media) samples cause interference with colorimetric determinations as we have observed over the years in our laboratory. Notwithstanding, the combination of FasL mRNA upregulation and the increase in FasL protein within Kupffer cells supports our conclusion that Kupffer cells are a source of FasL during acute pancreatitis. CONCLUSIONS

Our understanding of the pathophysiology of pancreatitis-induced organ injury has evolved with the expanding knowledge of the mechanisms of cytokine

FIG. 5. Immunoblot (repeated in triplicate) shows Kupffer cell-derived FasL protein in response to increasing doses of pancreatic elastase (0.1-1.0 U/ml). Gels were quantified by densitometry (all P ⬍ 0.04 versus control, n ⫽ 3).

GALLAGHER ET AL.: KUPFFER CELL-DERIVED FAS LIGAND AND HEPATOCYTE INJURY

209

FIG. 7. Elastase (E) up-regulated Kupffer cell Fas (FasL receptor) at 15, 30, 60, 120, and 240 min; Fas expression peaked at 120 min, at which time point pretreatment with gadolinium chloride (G⫹E) had a modest effect on Fas expression.

gene expression in resident macrophages and immunocompetent cells. The novel finding that acute pancreatitis up-regulates the gene expression of FasL and Fas within Kupffer cells warrants further investigation and may shed light on regulatory mechanisms of immunocompetent cells as well as their role in sepsis and other systemic inflammatory conditions. Moreover, further investigation of the complex interactions between Kupffer cells and hepatocytes may have important therapeutic implications.

Haussinger, D. Regulation of CD95 (APO-1/Fas) receptor and Ligand expression by lipopolysaccharide and dexamethasone in parenchymal and nonparenchymal rat liver cells. Hepatology 27: 200, 1998. 9. Chomczynski, P. A reagent for the single-step simultaneous isolation of RNA, DNA and proteins from cell and tissue samples. Biotechniques 15: 536, 1993. 10. Chomczynski, P., and Sacchi, N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162: 156, 1987. 11.

Murr, M. M., Yang, J., Fier, A., Kaylor, P., and Norman, J. G. Pancreatic-associated ascites induces hepatocyte death independent of local cytokine production. J. Surg. Res. 106: 308, 2002.

12.

Yang, J., Fier, A., Carter, G., Liu, G., Epling-Burnette, P. K., Bai, F., Loughran, T., Mastorides, S., Norman, J. G., and Murr, M. M. Liver injury during acute pancreatitis: the role of pancreatitis-associated ascitic fluid (PAAF), p38-MAPK and Caspase-3 in inducing hepatocyte apoptosis. J. Gastrointest. Surg. 7: 200, 2003.

13.

Ken-ichiro, S., Yashiro, S., Takashi, O., Nobuhiko, K., Hisaya, A., and Hiroyasu, N. Protection against Fas-mediated and tumor necrosis factor receptor 1-mediated liver injury by blockade of FADD without loss of nuclear factor-kB activation. Ann. Surg. 234: 681, 2001.

14.

Muzio, M., Chinnaiyan, A. M., Kischkel, F. C., O’Rouke, K., Shevchenko, A., Ni, J., Scaffidi, C., Bretz, J. D., Zhang, M., Gentz, R., Mann, M., Krammer, P. H., Peter, M. E., and Dixit, V. M. FLICE, a novel FADD-homologous ICE/CED-3-like protease, is recruited to the CD95 (Fas/APO-1) death-inducing signaling complex. Cell 85: 817, 1996.

15.

Norman, J., Fink, G., Denham, W., Yang, J., Carter, G., Sexton, C., Falkner, J., Gower, W., Franz, M. Tissue-specific cytokine production during experimental acute pancreatitis: A probable mechanism for distant organ dysfunction. Dig. Dis. Sci. 42: 1783, 1997.

16.

Kimura, S., and Tsukamoto, H. Cytokine gene expression by Kupffer cells in experimental alcoholic liver disease. Hepatology 21: 1304, 1995.

17.

Deaciuc, I. V., Fortunato, F., D’Souza, N. B., Hill, D. B., Schmidt, J., Lee, E. Y., and McClain, C. J. Modulation of Caspase-3 activity and Fas Ligand mRNA expression in rat liver cells in vivo by alcohol and lipopolysaccharide. Alcohol Clin. Exp. Res. 23: 349, 1999.

18.

Hori, Y., Takeyama, Y., Udea, T., Shinkai, M., Takase, K., and Kuroda, Y. Macrophage-derived transforming growth factorbeta induces hepatocellular injury via apoptosis in rat severe acute pancreatitis. Surgery 127: 641, 2000.

ACKNOWLEDGMENTS We thank the NIAAA-supported Non-Parenchymal Liver Cell Core (R24 AA12885) for providing isolated Kupffer cells for tissue cultures.

REFERENCES 1. 2.

3.

4.

5.

6.

7.

8.

Zyromski, N., and Murr, M. M. Evolving concepts in the pathophysiology of acute pancreatitis. Surgery 133: 235, 2003. Murr, M. M., Yang, J., Fier, A., Kaylor, P., Mastorides, S., and Norman, J. G. Pancreatic elastase induces liver injury by activating cytokine production within Kupffer cells via NF-kB. J. Gastrointest. Surg. 6: 474, 2002. Murr, M. M., Yang, J., Fier, A., Gallagher, S., Carter, G., Gower, W., and Norman, J. G. Regulation of TNF gene expression in Kupffer cells during acute pancreatitis: the role of p38-MAPK, ERK 1/2, SAPK/JNK and NF-kB. J. Gastrointest. Surg. 7: 20, 2003. Jaffray, C., Yang, J., Carter, G., Mendez, C., and Norman, J. Pancreatic elastase activates pulmonary nuclear factor kappa-B and inhibitory kappa-B, mimicking pancreatitisassociated adult respiratory distress syndrome. Surgery 128: 225, 2000. Gloor, B., Todd, K., Lane, J., Lewis, M., and Reber, H. Hepatic Kupffer cell blockade reduces mortality of acute pancreatitis in mice. J. Gastrointest. Surg. 2: 430, 1999. Zhu, X., Zellweger, R., Ayala, A., and Chaudry, I. Cytokine gene expression in Kupffer cells following hemorrhage. Cytokine 8: 430, 1996. Yang, J., Gallagher, S. F., Haines, K., Epling-Burnette, P. K., Bai, F., Gower W. R., Jr., Mastorides, S., Norman, J. G., and Murr, M. M. Kupffer cell-derived Fas Ligand plays a role in liver injury and hepatocyte death. J. Gastrointest. Surg. 8: 166-174, 2004. Muschen, M., Warskulat, U., Douillard, P., Gilbert, E., and