Article Text

This article has a correction. Please see:

PDF

Severe hepatic injury in interleukin 18 (IL-18) transgenic mice: a key role for IL-18 in regulating hepatocyte apoptosis in vivo
  1. S Finotto1,
  2. J Siebler1,
  3. M Hausding1,
  4. M Schipp1,
  5. S Wirtz1,
  6. S Klein1,
  7. M Protschka1,
  8. A Doganci1,
  9. H A Lehr2,
  10. C Trautwein3,
  11. R Khosravi-Fahr4,
  12. D Strand1,
  13. A Lohse1,
  14. P R Galle1,
  15. M Blessing5,
  16. M F Neurath1
  1. 1Laboratory of Immunology, I Medical Clinic, University of Mainz, Germany
  2. 2Laboratory of Immunology, I Medical Clinic, University of Mainz, Germany, and Institute of Pathology, Johannes Gutemberg-University of Mainz, Germany
  3. 3Department of Gastroenterology, Medizinische Hochschule, University of Hannover, Germany
  4. 4Beth Israel, Deaconess Medical Center, Division of Pathology, Harvard Medical School, Boston, Massachusetts, USA
  5. 5Laboratory of Immunology, I Medical Clinic, University of Mainz, Germany and Centre for Biotechnology and Biomedicine, Faculty of Veterinary Medicine, University of Leipzig, Leipzig, Germany
  1. Correspondence to:
    Dr S Finotto
    Laboratory of Immunology, I Medical Clinic, Johannes Gutenberg-University of Mainz, Germany; finottomail.uni-mainz.de

Abstract

Background: Interleukin 18 (IL-18) is a cytokine with pleiotropic activity that augments T helper 1 responses and cytotoxic activity of natural killer cells.

Methods: To assess the function of IL-18 in vivo, we generated IL-18 transgenic (IL-18 Tg) mice under the control of a CD2 promoter/enhancer construct.

Results: Macroscopically, IL-18 Tg mice showed reduced relative liver weight compared with wild-type littermates. TUNEL assays demonstrated increased hepatocyte apoptosis, and primary hepatocytes isolated from IL-18 Tg mice exhibited an increased spontaneous apoptosis rate. Furthermore, cross linking of Fas increased significantly the apoptosis rate in hepatocytes isolated from wild- type mice but to a much lesser extent in IL-18 Tg mice, suggesting spontaneous activation of the Fas pathway in the latter mice. In fact, in vivo blockade of Fas signal transduction by an adenovirus overexpressing the dominant negative form of the Fas associated death domain rescued hepatocytes from undergoing apoptosis. Finally, adoptive transfer of CD4+ T cells from IL-18 Tg mice but not from wild-type littermates in SCID mice resulted in severe liver failure with massive periportal fibrosis due to hepatocyte apoptosis.

Conclusion: IL-18 plays a fundamental role in regulating hepatocyte apoptosis. Furthermore, our transgenic model provides a novel tool to study the mechanisms of IL-18 dependent liver injury in vivo.

  • hepatocytes
  • interleukin 18
  • apoptosis
  • nuclear factor κB
  • Fas
  • IL-18, interleukin 18
  • IL-18R, IL-18 receptor
  • IL-18 Tg, IL-18 transgenic
  • NK, natural killer
  • IFN-γ, interferon γ
  • LPS, lipopolysaccharide
  • NFκB, nuclear factor κB
  • TNF, tumour necrosis factor
  • RT-PCR, reverse transcription-polymerase chain reaction
  • SAC, Staphylococcus aureus Cowan’s antigen
  • DEAE, diethylaminoethyl
  • BMP6, bone morphogenetic protein 6
  • adv, adenoviral vectors
  • dnFADD, dominant negative Fas associated death domain
  • β-Gal, β-galactosidase
  • PI, propidium iodide
  • TdT, terminal deoxynucleo-tidyltransferase
  • TUNEL, TdT mediated fluorescein dUTP nick end labelling

Statistics from Altmetric.com

The pleiotropic cytokine interleukin 18 (IL-18) is a member of the IL-1 family that was originally described as an inducer of interferon γ (IFN-γ) production from T cells in the presence of IL-12 (interferon γ inducing factor).1–8 IL-18 is produced by Kupffer cells, macrophages, B cells, and dendritic cells on lipopolysaccharide (LPS) stimulation.9 It has been identified and cloned from cDNA extracted from the liver of mice infected with Propionibacterium acnes. Sequential treatment with heat killed P acnes and LPS induces IL-18 dependent acute liver injury in mice, probably through upregulation of Fas ligand.1 Conversely, treatment with a neutralising antibody to IL-18 protects against such liver injury, and IL-18 deficient mice were resistant to P acnes and LPS induced liver injury, indicating that IL-18 plays a critical role in P acnes and LPS induced liver injury.1,10,11

In common with IL-1β, IL-18 is intracellularly produced as a biologically inactive precursor protein (pro-IL-18), and mature IL-18 is secreted after cleavage of pro-IL-18 by caspase-1, originally designated as IL-1β converting enzyme.12 IL-18 mediates its biological function through binding to a specific receptor complex on the cell surface consisting of two chains: a ligand binding chain termed the IL-18Ralpha chain and a signal transducing chain termed the IL-18Rbeta chain. Whereas the IL-18Ralpha chain has a weak affinity for the ligand, the complete IL-18R complex has high binding affinity for IL-18 and is expressed, for example, on T helper 1 T cells and natural killer (NK) T cells.13–15 Binding of IL-18 to the high affinity IL-18R leads to nuclear factor κB (NFκB) activation through IL-1 receptor associated kinase, the adaptor proteins MyD88, and tumour necrosis factor (TNF) receptor–associated factor 6, and subsequent phosphorylation of IκB via IκB kinases (IKK-1, IKK-2).16–18 In addition, IL-18 triggers NK cell activity and expression of FasL by NK cells.

Furthermore, IL-18 is a coinducer of IL-12 driven T helper 1 T cell differentiation by upregulating IL-12 receptor beta2 chain expression and transactivation of the IFN-γ promoter via the transcription factor AP-1.19 Finally, IL-18 augments proinflammatory immune responses by upregulating the production of various cytokines such as IL-1, IL-6, and TNF.3

The above data prompted us to hypothesise that IL-18 may augment the cytotoxic activity of intrahepatic NK T and NK cells during an immune response. In the present study, we describe generation of IL-18 transgenic (Tg) mice under the control of the CD2 promoter. These mice showed reduced liver weight due to an increase in spontaneous Fas dependent apoptosis of hepatocytes. Furthermore, signal transduction studies in hepatocytes of IL-18 Tg versus wild-type mice showed that NFκB is activated in the nucleus of IL-18 Tg hepatocytes compared with hepatocytes isolated from wild-type mice. Adoptive transfer of CD4+ T cells from IL-18 Tg spleen but not from wild-type littermates in SCID mice resulted in severe liver failure with massive periportal fibrosis and liver inflammation associated with hepatocyte apoptosis. Our data thus suggest a key regulatory role of IL-18 for hepatocyte apoptosis in vivo via a Fas dependent mechanism.

MATERIALS AND METHODS

Vector construct for generation of IL-18 transgenic mice

cDNA of IL-18 was obtained after reverse transcription (RT) (First Strand cDNA synthesis kit for RT-polymerase chain reaction (PCR); Roche Applied Science, Mannheim, Germany) of RNA extracted from the monocytic cell line RAW264.7 activated with LPS (1 μg/ml) (Sigma, St Louis, Missouri, USA) and Staphylococcus aureus Cowan’s antigen (SAC; 0.001%; Cabiochem, Germany). cDNA for IL-18 was then amplified by 33 PCR cycles using specific oligonucleotide primers (5′ CTC- CCC- ACC- TAA- CTT-TGA-TG-3′ and 5′ CCA-GGA-ACA-ATG-GCT-GCC-AT-3′). The PCR product was then subcloned into the polylinker site of the pcDNA 3.1 expression vector (Invitrogen, Heidelberg, Germany) and sequenced. The 595 bp IL-18 cDNA was then subcloned at the SmaI restriction site by blunt end ligation downstream of the CD2 promoter/enhancer construct yielding the CD2-IL-18 plasmid.20

Cell transfection

Transfection experiments were conducted to confirm whether the transgenic construct could induce optimal secretion of mature IL-18 in vitro. The CD2-IL-18 plasmid and the empty vector CD2 were transiently transfected into RAW264.7 and Jurkat T cells. This transfection was carried out using the diethylaminoethyl (DEAE) dextran (GibcoBRL, Karlsruhe, Germany) method, as previously described.21 Briefly, 2.5×107 cells were incubated for one hour with DEAE-dextran and 10g of plasmid at 37°C. Cells were cultured for an additional 48 hours after transfection; then, total RNA and cellular proteins were extracted from the transfected cells. For IL-18 transcript detection, 1 μg of total RNA was reverse transcribed using random primers and RTMV transcriptase (Roche Applied Science). cDNA for IL-18 and β-actin was then amplified by PCR using the primers described above for IL-18.

Western blot analysis was performed using 7.5% polyacrylamide gel under reducing conditions. Recombinant murine IL-18 (75 ng) was loaded as a positive control (R&D System, Wiesbaden, Germany). After protein transfer onto nitrocellulose membranes, IL-18 was detected using a polyclonal antibody made in the goat (0.4 g/ml) (R&D Systems, Wiesbaden, Germany) followed by incubation with the biotinylated antigoat antibody and streptavidin-alkaline phosphatase incubation, and developed using a substrate for alkaline phosphatase (1:30 000) (Dako Diagnostics GmbH, Hamburg, Germany).

Identification of CD2/IL-18 transgenic mice

Based on these results, the fragment obtained after digestion, with the restriction enzymes KpnI and Not I, of the CD2-IL-18 construct containing the IL-18 expression cassette was microinjected into pronuclei of fertilised eggs of FVB/NHSD mice. Mice were maintained under specific pathogen free conditions in isolated cages. Founder mice were born and bred separately to generate F1 progeny, resulting in independent lines. Transgenic integration was confirmed in genomic DNA extracted from the ear or tail by overnight digestion in proteinase K buffer (proteinase K 250 g/ml) at 55°C followed by PCR for IL-18 and the bone morphogenetic protein 6 (BMP6) using the following primers: BMP6: 5′ GTC TGC AGA TGT GTC TGC GTG TCA CG 3′ and 5′ ATG AAT GCC ACC AAC CAC G 3′; IL-18: CD2 upstream 5′ CCC AGC TTT CCC TGA AAG TG 3′ and 5′’ CTC CCC ACC TAA CTT TGA TG 3′ IL-18 downstream primer (see fig 1D).

Figure 1

Design of the construct containing the human CD2 minigene and the mature mouse cDNA for interleukin 18 (CD2-IL-18) used to induce transgenic (Tg) IL-18 overexpression in FVB mice (A). This construct (10 μg), or a construct containing the CD2 minigene alone, was transiently transfected into 2.5×107 Jurkat cells using the diethylaminoethyl-dextran method. Forty eight hours after cell transfection, proteins and RNA were analysed by western blot analysis (C) and reverse transcription-polymerase chain reaction (RT-PCR) (B). Similar data were seen in RAW264.7 cells (not shown). For detection of mature IL-18 protein (18 kDa), 75 ng of recombinant murine IL-18 were used as a positive control. In vitro, mature IL-18 transcription and translation were detected only after transfection with the expression vector CD2/IL-18 but not with the empty control vector CD2. As a control, the original vector pCR3.1 and pCR3.1 with insertion of cDNA for IL-18 (pCR3.1/IL-18) were analysed. PCR was used to analyse transgenic integration in genomic DNA extracted from the ear or tail of IL-18 Tg mice (D) using upstream primers for CD2 and downstream primers for IL-18. The endogenous bone morphogenetic protein 6 gene (BMP6) was used as an internal control for PCR.

Total RNA extraction and RT-PCR

Total liver tissue was homogenised in a Polytron homogeniser and RNA was then extracted by resuspending the homogenised tissue in 1 ml of TRI-REAGENT (Sigma) in accordance with the manufacturer’s instructions. Addition of chloroform was followed by centrifugation, and RNA was precipitated from the aqueous phase using isopropanol. RNA quantification was performed by photometric analysis. Total RNA (2 g) was then reverse transcribed using the first strand cDNA synthesis kit for RT-PCR (AMV; Roche Molecular Biochemicals, Indianapolis, Indiana, USA). The resulting cDNA was then used as a template for PCR. For Fas, we used primer pairs from R&D System (RDP-59-025) and followed the manufacturer’s instructions.

Adenovirus preparation

To generate high virus stocks, 2×108 293 packing cells at 90% confluence were infected at a multiplicity of infection of 5–10 PFU per cell. Infected cells were cultured for 3–5 days until a strong cytopathic effect could be observed and approximately 50% of the cells were detached. Cells were then collected by centrifugation, and viral particles were released by four cycles of freezing in liquid nitrogen and rapid thawing at 37°C. For further purification, the virus preparation was subjected to twofold CsCl banding. CsCl banding and determination of infectivity by viral plaquing were performed according to standard protocols, as described previously. Endotoxin contamination was monitored by the LAL test kit (Chromogenix, Molndal, Sweden), following the manufacturer’s protocol. All virus preparations used for infection experiments were LPS free. Virus preparations were stored at −80°C in 25% glycerol, 10 mM Tris HCl (pH 7.4), 1 mM MgCl2, and 140 mM NaCl. The adenoviral vectors (adv) used in this study (adv β-galactosidase (β-Gal), IκB-AA, and dominant negative Fas associated death domain (dnFADD)) have been described previously.22,23

Hepatocyte isolation: cell cultures and infection experiments with adenovirus

Anaesthetised mice were cannulated at the portal vein and the liver perfused with a 0.05% collagenase (Serva, Heidelberg, Germany) solution (3.9 g NaCl; 0.5 g KCl; 0.7 g CaCl2; HEPES 24 g/l at pH 7.6). The obtained cell suspension was then passed through a sieve and centrifuged. Hepatocytes were plated on collagen coated petri dishes (collagen R solution 0.2 mg/ml in 0.1% acetic acid; Serva) and cultured overnight in supplemented Dulbecco’s Eagle medium (+1% glutamine;1% Na pyruvate; 20 mM HEPES; 0.05% insulin; 6–10 M lithocholic acid) with 10% fetal calf serum. The next day hepatocytes were challenged to undergo apoptosis after cross linking of the Fas receptor with a monoclonal antimouse Fas monoclonal antibody, clone Jo2 (2.5–10 μg/ml per 2×106 hepatocytes) (Pharmingen, San Diego, California, USA). Negative controls (untreated hepatocytes) were incubated with medium alone. Hepatocytes were cultured for an additional 4–24 hours, as indicated. Cells were then washed in cold phosphate buffered saline (PBS) and resuspended in binding buffer. Cells (105) were then incubated with annexin V-FITC and propidium iodide (PI) or annexin V or PI alone as controls, and then analysed within one hour by FACS analysis (annexin V-FITC apoptosis detection kit II; Pharmingen). Morphological changes typical of cell apoptosis, such as membrane blebbing and cell shrinkage, were visualised using an Olympus microscope B×63 (Olympus Optical Co., Hamburg, Germany).

In adenovirus studies using adv-dnFADD or adv-IκBα, hepatocytes were infected with 100 PFU overnight. The next day, hepatocytes were washed twice in PBS and incubated for two hours either with medium containing anti-Fas antibody or with medium alone followed by apoptosis analysis in accordance with the manufacturer’s instructions (annexin V-FITC apoptosis detection kit II; Pharmingen).

In vivo infection experiments with adenoviruses expressing dnFADD and IκBα

IL-18 Tg mice were infected with 1010 PFU of adenoviruses expressing dnFADD, IκB, or β-Gal intraperitoneally. Three days after infection, primary hepatocytes were isolated and cultured overnight on six well plates on collagen, as described above. At this time point, hepatocytes were recovered by scraping with a cell scraper, washed in PBS, and analysed by FACS analysis after staining with annexin V and PI, as specified above.

Isolation and purification of spleen mononuclear cells

Spleen mononuclear cells were isolated from freshly obtained specimens from healthy FVB wild-type or IL-18 Tg mice. Erythrocytes were removed from spleen cell suspensions by hypotonic lysis in ammonium chloride and potassium chloride buffer. Cells were cultured in complete RPMI medium supplemented with 3 mM l-glutamine, 10 mM HEPES buffer, 100 U/ml penicillin-streptomycin, and 10% heat inactivated fetal calf serum in a humidified atmosphere with 5% CO2 in a 37°C incubator in the presence or absence of hamster antibodies against mouse CD3 and CD28 (CD3 clone 145-2C11 2 g/ml; CD28 clone 37.51 0.5 μg/ml; Pharmingen), LPS 1 μg/ml (Sigma ), IFN-γ (10 ng/ml), and IL-12 (5 ng/ml).

ELISA

Concentrations of cytokines released in the cell culture supernatant and in serum were assessed by ELISA using R&D Systems and Pharmingen antibodies and recombinant cytokines on 96 well plates (MaxiSorb; Nunc, Naperville, Illinois, USA), and performed in accordance with the manufacturer’s instructions. IL-18 in cell culture supernatants and serum was detected and quantified using a specific sandwich ELISA (OptEIA system; BD Biosciences, San Diego, California, USA).

Protein extraction and western blot analysis

Tissue proteins were extracted as previously described.24 Briefly, tissue was homogenised in PBS using an electrical homogeniser and protein extracted in the presence of protease inhibitors (in 325 l PBS, 25 μl aprotinin (2.2 mg/ml), 25 μl trypsin inhibitor (5 mg/ml), 25 μl 10% Nonidet, 25 μl sterile water). Protein concentrations were determined by spectrophotometry with a BioRad protein assay (Bradfords’ method) (BioRad Laboratories, Munchen, Germany). The standard curve was performed using 0, 2, 4, 8, and 16 g of bovine serum albumin/ml. Nuclear and cytoplasmic proteins were extracted from primary hepatocytes as follows: the cell pellet was washed with cold PBS twice followed by a 10 minute incubation period in buffer A (10 mM HEPES; 1.5 mM MgCl2; 10 mM KCl; PMSF 0.5 mM) with 0.4% NP40 on ice. After centrifugation, supernatants containing the cytoplasmic proteins were frozen and the pellet was incubated with buffer B (20 mM HEPES; 0.42 M NaCl; 1.5 mM MgCl2; 0.2 mM EDTA; PMSF 0.5 mM, and DTT 1 mM) to extract nuclear proteins. Equal amounts of protein extract were added to the electrophoresis sample buffer (5×). After boiling, proteins were separated by 10% or 15% sodium dodecyl sulphate-polyacrylamide gel electrophoresis, transferred to nitrocellulose membranes, and detected with a specific antibody against nuclear factor κB (NFκB) p65 (1:200) (Santa Cruz, Heidelberg, Germany), polyclonal goat antimouse IL-18 (R&D System), caspase 8 and anti-FasL (Santa Cruz), monoclonal rat antimouse IL-18 (1:500) (R&D), AKT-P (Cell Signaling Technology, New England Biolabs, Frankfurt, Germany), and anti-Fas receptor (Transduction Laboratories, Pharmingen, Germany) at the final concentration suggested by the manufacturers. Specific binding were visualised with the ECL western blotting analysis system after a one hour incubation period with a horseradish peroxidase labelled secondary antibody (Amersham, Germany).

Splenocyte isolation and adoptive transfer of CD62L+ CD4+ T cells in SCID mice

Splenocytes were isolated from wild-type and IL-18 Tg mice by standard methods and in some experiments cultured with antibodies to CD3 and CD28. For the adoptive transfer experiments, CD4+ T cells were directly purified from isolated spleen mononuclear cells, from both wild-type and transgenic IL-18 FVB mice using FITC conjugated monoclonal antibodies (Pharmingen) and anti-FITC immunomagnetic microbeads in a multiparameter magnetic sorter system (MACS; Miltenyi; Bergisch Gladbach, Germany). The resulting CD4+ cells (purity 97%) were further separated by immunomagnetic microbeads into CD62L+ and CD62L T cells. The former cells showed high expression of CD45RB by FACS analysis. Finally 1×106 CD62L+/(CD45RBhigh) CD4+ T cells were transferred intraperitoneally into CB17 SCID mice (Charles River Laboratories, Wilmington, Massachusetts, USA). Reconstituted SCID mice were maintained in isolated ventilated cages for the indicated times and monitored by body weight curves and histological analysis of the liver.

Macroscopic analysis, determination of liver weight, and blood analysis

Organs of transgenic mice from each founder were macroscopically compared with wild-type control mice. The weight of the different organs was then corrected for total body weight. The results are reported as the ratio of the organ weight and total body weight multiplied by 100. Blood samples were taken after ventricular puncture from IL-18 Tg and wild-type control mice. For serum cytokine analysis, blood was coagulated at 4°C for three hours and then centrifuged at 3000 rpm for 10 minutes. The supernatant (serum) was frozen, and subsequently serum cytokine content was quantified after ELISA immunoassay, as described below.

Histological analysis of liver cross sections

Tissues were removed, fixed in 10% buffered formalin, mounted in paraffin, sectioned, and stained with haematoxylin-eosin. Pathological evaluation was performed in a blinded fashion by the same pathologist (HAL).

Analysis of cell apoptosis in liver specimens

Paraffin sections of liver derived from both IL-18 Tg and wild-type mice as well as from SCID reconstituted mice were analysed for apoptosis by terminal deoxynucleotidyltransferase (TdT) treatment that induces incorporation of FITC labelled deoxynucleotides to DNA strand breaks in situ (TUNEL; TdT mediated fluorescein dUTP nick end labelling technique) (in situ cell death detection kit, fluorescein; Roche Molecular Biochemicals). Nuclei were counterstained with DAPI staining (Vector Laboratories, Linaris GmbH, Wertheim, Germany) and analysed by a fluorescence microscope (Olympus Optical Co, Hamburg, Germany). In some experiments, as shown in fig 4D, we detected in situ apoptosis in the liver using the II. ApopTag Peroxidase Kit (Intergen, New York, USA) in accordance with the manufacturer’s instructions.

Statistical analysis

The Student’s t test was used for data analysis. Data are expressed as mean (SEM).

RESULTS

Generation of IL-18 transgenic mice

To obtain insight into the functional immunoregulatory role of IL-18 in vivo, we generated IL-18 transgenic mice. Accordingly, we amplified the murine IL-18 coding region by PCR using RNA from LPS plus SAC stimulated RAW264.7 monocytes and subcloned the IL-18 cDNA (595 bp) into an expression cassette driven by the human CD2 promoter/enhancer using blunt end ligation yielding the CD2-mIL-18 expression vector (fig 1A). This expression construct was then transiently transfected into Jurkat T cells. Mature IL-18 (18 kDa) was detected by western blot analysis in the cell lysate of Jurkat cells transfected with the CD2-mIL-18 construct but not in control T cells transfected with the CD2 vector, suggesting that the vector can express mIL-18 (fig 1C). Consistently, mIL-18 mRNA was detected by RT-PCR (fig 1B) in Jurkat cells transfected with the CD2-IL18 vector but not in cells transfected with the empty CD2 expression cassette.

The CD2-mIL-18 construct was then digested with the restriction enzymes KpnI and NotI and the resulting fragment containing the IL-18 expression cassette was microinjected into the pronuclei of fertilised eggs of FVB/NHSD mice, resulting in the generation of two independent founder lines that showed transgene integration and the same phenotype described below. Transgene integration was confirmed by PCR using simultaneously CD2-IL-18 transgene specific primers that resulted in a 650 bp fragment in transgenic mice and primers to detect BMP6 as the internal control (fig 1D).

Transgenic IL-18 mice exhibit increased IFN-γ and TNF-α levels in serum

To confirm overexpression of mature IL-18 in transgenic mice derived from the two cell lines, we measured IL-18 protein levels in the serum of these mice by ELISA. As shown in fig 2A, IL-18 transgenic mice showed significantly increased mature IL-18 in serum compared with control littermates (p = 0.01). In addition, levels of both IFN-γ and TNF-α were increased in the serum of IL-18 transgenic mice when compared with control mice (fig 2B and 2C, respectively). In contrast, serum levels of IL-4 and IL-6 remained unchanged in IL-18 transgenic mice, suggesting a selective effect of IL-18 overexpression on cytokine production in vivo (data not shown). In addition, splenocytes isolated from IL-18 Tg mice released increased amounts of IL-6 after T cell specific immunological stimuli (anti-CD3/anti-CD28 or anti-CD3/anti-CD28 plus IL-12) (fig 2E). These cells also produced significantly increased amounts of IFN-γ when challenged with LPS plus IFN-γ (fig 2D).

Figure 2

Cytokine expression in the serum of interleukin 18 transgenic (IL-18 Tg) mice. Serum levels of IL-18 in transgenic mice (n = 6) and wild-type littermates were measured using a mIL-18 ELISA kit (A). Interferon γ (IFN-γ) and tumour necrosis factor α (TNF-α) were measured using commercially available kits (B and C, respectively). Cytokines released after 48 hours in the supernatants of splenocytes isolated from wild-type and IL-18 Tg mice are shown. IFN-γ production was significantly increased in IL-18 Tg cells after lipopolysaccharide (LPS) plus IFN-γ challenge (D) whereas IL-6 was increased after stimuli that preferentially induced T cell activation (E). SAC, Staphylococcus aureus Cowan’s antigen.

The liver of IL-18 transgenic mice is reduced in size and weight

IL-18 was originally identified in extracts of mice treated with P acnes plus LPS and appears to control liver injury induced by P acnes plus LPS. In addition, IL-18 knockout mice are resistant to LPS induced liver injury.10 Therefore, we analysed the effect of overexpression of IL-18 on the liver in vivo. Macroscopic analysis of the liver showed that transgenic IL-18 mice had a lighter and smaller liver and a reduced liver/body weight ratio compared with wild-type littermates (p = 0.0003) (fig 3A, B).

Figure 3

Liver injury in interleukin 18 transgenic (IL-18 Tg) mice. Liver weight from IL-18 Tg and wild-type mice was related to total body weight and this ratio was multiplied by 100 (A). ***Significant difference compared with wild-type littermates. The size of the liver of IL-18 Tg mice was reduced compared with that of wild-type littermates (B).

Hepatocytes isolated from IL-18 transgenic mice exhibit increased Fas mediated apoptosis

To investigate whether the observed liver weight reduction in IL-18 transgenic mice was due to increased necrosis or apoptosis of liver cells, we first performed TUNEL assays in the liver of both wild-type and IL-18 Tg mice and found increased apoptotic hepatocytes in the liver of IL-18 Tg mice, indicating that loss of liver weight was due to increased apoptosis of hepatocytes in IL-18 Tg mice (fig 4D). We then isolated and cultured hepatocytes from the liver of both IL-18 Tg mice and wild-type littermates, as specified in the materials and methods sections. After 4 and 24 hours of cell culture, we analysed the apoptosis rate by the annexin V and PI method and found that hepatocytes isolated from IL-18 Tg mice exhibited an increased spontaneous apoptosis rate compared with wild-type littermates (n = 5) (fig 4A).

Figure 4

Hepatocytes isolated from interleukin 18 transgenic (IL-18 Tg) mice exhibited increased spontaneous and Fas mediated apoptosis. Apoptosis in primary hepatocytes was analysed by the annexin V and propidium iodide (PI) method (A). Hepatocytes isolated from IL-18 Tg mice and cultured for 24 hours exhibited increased late apoptosis (annexin V and PI high positive) compared with hepatocytes isolated from wild-type mice (n = 5; *p<0.05). Cross linking of Fas with an antibody directed against Fas (10 μg/107 cells) further induced apoptosis in wild-type mice (A). (B) The livers of IL-18 Tg mice expressed increased amounts of Fas and caspase 8 protein (B; western blot analysis) and Fas RNA (C; polymerase chain reaction analysis) compared with that of wild-type littermates. Ponceau staining showed equal loading. In addition, we found that the liver of SCID mice expressed Fas (C). (D) In situ apoptosis in the liver of IL-18 Tg and wild-type mice was assessed using the TUNEL reaction. Representative views are shown. Increased numbers of apoptotic hepatocytes were detected using a Zeiss Axioskop mot plus microscope (magnification 200×).

In an attempt to identify the mechanism of apoptosis in hepatocytes derived from IL-18 Tg mice, we cross linked the Fas receptor on the surface of hepatocytes with an antibody. Cross linking of Fas receptor on hepatocytes derived from wild-type mice induced a significant increase in hepatocyte apoptosis (fig 4A) and had less effect on the apoptosis rate of hepatocytes derived from IL-18 Tg mice, suggesting spontaneous activation of the Fas pathway in the latter mice in vivo (fig 4A).

To establish the validity of the paradigm that cell death in IL-18 Tg hepatocytes is Fas mediated, we next analysed Fas expression in the liver of both wild-type and IL-18 Tg mice and found that the livers from IL-18 Tg mice contained increased amounts of Fas protein and caspase 8 (fig 4B, 4C). We then infected IL-18 Tg mice with an adenovirus overexpressing the dominant negative Fas associated death domain (adv-dnFADD) protein with the aim of blocking Fas signal transduction in vivo. As shown in table 1, in vivo blockade of Fas led to a significant decrease in hepatocyte cell death (from 67.33% to 29.90%) whereas expression of IκB had little effect. These results are consistent with direct in vivo involvement of the Fas receptor in mediating cell death of IL-18 Tg hepatocytes.

Table 1

Blockade of Fas signal transduction in vivo reduces hepatocyte apoptosis in interleukin 18 transgenic (IL-18 Tg) mice

Increased translocation of NFκB into the nucleus in hepatocytes isolated from IL-18 transgenic mice

To investigate whether expression and activation of NFκB is regulated by Fas pathway activation in IL-18 Tg mice, we analysed cytoplasmic and nuclear NFκB protein before and after Fas cross linking in hepatocyte cultures from wild-type and IL-18 Tg mice. Hepatocytes derived from wild-type mice showed, after Fas cross linking, increased expression of NFκB p65 accompanied by its translocation into the nucleus. In contrast, in hepatocytes derived from IL-18-Tg mice, NFκB p65 was already spontaneously expressed at high levels and translocated into the nucleus (fig 5A). These results suggest that Fas mediated apoptosis in wild-type hepatocytes is accompanied by increased activation of NFκB in hepatocytes derived from wild-type mice whereas hepatocytes from IL-18 Tg mice exhibited evidence of spontaneous activation of the NFκB pathway.

Figure 5

Spontaneous nuclear translocation of nuclear factor κB (NFκB) in hepatocytes isolated from interleukin 18 transgenic (IL-18 Tg) mice (A) and decreased activation of NFκB p65 in hepatocytes isolated from IL-18 Tg mice infected with an adenovirus expressing IκBα (Adv IκBα) (B). Cytoplasmic (C) and nuclear (N) proteins from both IL-18 Tg and wild-type (wt) hepatocytes four hours after cell culture were analysed by western blot analysis for NFκB p65 expression (B). Western blot analysis of nuclear (N) and cytoplasmic (C) proteins isolated from primary hepatocytes isolated from either wild-type or IL-18 Tg mice treated with medium alone or with adenovirus (Adv) expressing dominant negative Fas associated death domain (dnFADD) or IκBα and cross linked at the Fas receptor (B). Activation of the antiapoptotic protein AKT (Phospho-AKT:AkT-P) was decreased after infection with adv-IκBα that caused inactivation of NFκB. There were no changes in Fas ligand (FasL) expression.

Role of NFκB in Fas activated cell death

In vivo blockade of NFκB by infection of transgenic mice with an adenovirus overexpressing IκB caused no change in hepatocyte apoptosis (table 1). In addition, in vitro infection of IL-18 Tg hepatocytes with this adenovirus (adv-IkBα) and cross linking of the Fas receptor with an anti-Fas antibody also did not lead to modification of hepatocyte apoptosis (table 2) although these cells exhibited decreased nuclear levels of NFκB and unchanged levels of FasL (fig 5B). Consistent with these findings, phosphorylated Akt, a prosurvival marker, was found to be decreased after hepatocyte infection with adv-IkBα, suggesting an anti- rather than a proapoptotic role of NFκB in hepatocytes from IL-18 transgenic mice (fig 5B).

Table 2

Blockade of nuclear factor κB activation by infection with adv-IκBα does not increase hepatocyte apoptosis

Reconstitution of SCID mice with CD62Lhigh T cells derived from IL-18 transgenic mice leads to severe liver injury

To investigate whether apoptosis of hepatocytes in IL-18 Tg could be adoptively transferred by IL-18 producing cells, we purified CD62L+ (L-selectin)high T cells from the spleen of IL-18 Tg and wild-type mice to reconstitute SCID mice that expressed detectable levels of Fas in hepatocytes (fig 4C). The body weight of reconstituted SCID mice was monitored every week after cell transfer. Five to six weeks after cell transfer, mice receiving transgenic cells were sacrificed because they had lost more than 20% of their initial body weight and some had died. SCID mice reconstituted with cells derived from IL-18 Tg mice showed elevated levels of serum IL-18 and IFN-γ compared with SCID mice that had been reconstituted with naïve lymphocytes derived from wild-type control littermates (fig 6H). No evidence of colitis was seen. Moreover, the liver of SCID mice reconstituted with cells derived from wild-type mice had a normal architecture with well demarcated portal areas without evidence of tissue injury with normal bile canaliculi. In contrast, the liver isolated from SCID mice that had been reconstituted with cells derived from IL-18 Tg mice had massive periportal fibrosis and inflammation, with extensive disseminated and confluent periportal parenchymal destruction, both in the form of apoptotic bodies and of neutrophil associated piecemeal necrosis (fig 6A–D). The TUNEL reaction on liver sections of reconstituted SCID mice demonstrated many positive hepatocytes in SCID mice which had been reconstituted with cells derived from IL-18 Tg mice (fig 6F,G). Such apoptotic bodies were absent in SCID mice that had been reconstituted with cells derived from wild-type mice (fig 6E).

Figure 6

Reconstitution of SCID mice with T cells from interleukin 18 transgenic (IL-18 Tg) mice induced hepatocytes apoptosis. SCID mice reconstituted with spleen T cells derived from IL-18 Tg mice showed increased IL-18 in the serum, as shown by ELISA (H; ***p<0.01). Paraffin liver sections of SCID mice reconstituted with T cells from wild-type (A, C) and IL-18 Tg (B, D) mice stained with haematoxylin-eosin. The latter livers showed increased periportal fibrosis and inflammation, and apoptosis, as shown after the TUNEL reaction in green (F, G). For comparison, a liver section from SCID mice reconstituted with CD62L positive cells derived from wild-type mice after the TUNEL reaction are shown (E). Nuclei were counterstained in blue with DAPI. IL-18 positive cells are shown in red. Magnification: 20× (A, C); 40× (B, D).

DISCUSSION

We have reported here that transgenic expression of IL-18 under a CD2 promoter/enhancer construct induces hepatocyte apoptosis with consequent reduction of liver size and weight. Hepatocyte apoptosis was at least partially dependent on the Fas pathway, as demonstrated by increased Fas expression in hepatocytes derived from IL-18 transgenic mice and by experiments involving Fas cross linking on isolated hepatocytes. Moreover, in vivo infection with an adenovirus overexpressing the dominant negative form of FADD markedly reduced hepatocyte cell death in IL-18 Tg mice. Although NFκB p65 was spontaneously activated in IL-18 Tg hepatocytes, it did not contribute to induction of hepatocyte apoptosis, as demonstrated by the lack of changes in hepatocyte apoptosis on adenoviral overexpression of IκB. In addition, in vivo overexpression of IκBα after adenovirus infection did not lead to changes in hepatocyte apoptosis in IL-18 Tg mice. Liver injury could be adoptively transferred by splenic T cells from IL-18 Tg spleen into SCID mice, resulting in severe liver failure with massive periportal fibrosis and inflammation associated with hepatocyte apoptosis. In summary, IL-18 Tg mice provide a novel tool to study mechanisms of IL-18 dependent liver injury in vivo.

Compared with peripheral lymphoid organs, the liver has a unique anatomical environment that is closely associated with its immunological function. In addition to the presence of large quantities of resident Kupffer cells, the liver contains conventional T cells, NK cells, and NK-T cells at high frequency. Liver NK cells and NK1.1 T cells exhibit cytotoxicity against virus infected hepatocytes or tumour cells via CD95(Fas), TNF, or perforin dependent pathways.25,26 Interestingly, a new cytokine, denoted IL-18, has been identified that modulates the cytotoxicity of these cells. IL-18 is produced by Kupffer cells and has been shown to be a potent inducer of IFN-γ production by T cells and NK cells and acts as a costimulatory factor for T cell activation.27–29 In addition, IL-18 augments NK cell activity in the spleen and liver, as well as cytotoxicity of T helper 1 cells, probably through an increase in FasL expression, suggesting that its role may be to regulate cytotoxicity against tumours and microbes or virus infected cells. To address the role of IL-18 in the liver under in vivo conditions, in the present study we generated IL-18 Tg mice under the control of the CD2 promoter/enhancer construct. These mice expressed increased mature IL-18 both in serum and in the liver, as demonstrated by ELISA and western blot analysis. Furthermore, augmented IFN-γ release by splenocytes from transgenic mice was found after stimuli that have been shown to induce IL-18 production, such as LPS stimulation. In addition, splenocytes isolated from IL-18 Tg mice and challenged with CD4-T cell specific stimuli (such as anti-CD3/anti-CD28), released increased IL-6 in the supernatant, a cytokine known to be involved in liver regeneration.30

IL-18 Tg mice exhibited a significant reduction in liver size and weight. Primary hepatocytes isolated from IL-18 Tg mice showed augmented Fas expression and exhibited increased ongoing apoptosis compared with wild-type control mice, suggesting increased spontaneous apoptosis in the former compared with the latter. To understand the mechanism underlying the increased hepatocyte apoptosis in IL-18 Tg mice, we cross linked Fas receptor at the cell surface of freshly isolated hepatocytes and analysed cell apoptosis. In vitro cross linking of the Fas receptor increased significantly the apoptosis rate and induced translocation of NFκB into the nucleus in wild-type hepatocytes. In contrast, little or no further increase in the apoptosis rate was detected on Fas cross linking in hepatocytes derived from IL-18 Tg mice (without changes in NFκB p65 activation), suggesting pre-activation of the Fas pathway in these mice in vivo. In addition, FasL, which is expressed on the surface of hepatocytes, was not further increased on primary hepatocytes derived from IL-18 Tg mice, suggesting that IL-18 might induce in vivo FasL expression in T cells31–33 but not hepatocytes.

It has been reported that on IL-18 stimulation, IkBα is degraded and NFκB translocates to the nucleus, suggesting a key role for IL-18 in mediating NFκB activation.34,35 Indeed, we observed constitutive NFκB activation in hepatocytes from IL-18 transgenic mice, indicating an important role for IL-18 in inducing NFκB activation in vivo. The link between Fas and NFκB activation was then investigated with and without Fas receptor cross linking and simultaneous infection of hepatocytes with an adenovirus overexpressing IκB. Under these conditions, NFκB was found to be decreased in the nucleus of infected hepatocytes, and AKT phosphorylation, a marker of cell survival,36,37 was downregulated while FasL and the apoptosis rate remained unchanged, suggesting an antiapoptotic rather than a proapoptotic role for NFκB in hepatocytes isolated from IL-18 Tg mice. These results were also supported by another cohort of experiments where overexpression in vivo of Iκbα with an adenovirus in IL-18 Tg mice for three days did not affect hepatocyte apoptosis of infected mice. These findings are also consistent with the observation that RelA−/− mice die at embryonic day 15 as a result of extensive liver apoptosis.38

In summary, our data indicate that IL-18 is a critical cytokine for the regulation of hepatocyte apoptosis via a Fas dependent mechanism in vivo. The fact that liver injury induced by hepatocyte apoptosis could be adoptively transferred by T cells from transgenic mice to SCID mice suggests that both the lymphoid component and hepatocytes are required for IL-18 induced liver alterations. In fact, as IL-18 is well known to induce FasL expression on NK and CD4+ T cells27 and IL-18 overexpression led to increased expression of Fas on hepatocytes, our data are consistent with a model where Fas/FasL interactions on T cells and hepatocytes induce liver injury in vivo. In addition, expression of both FasL and Fas on hepatocytes derived from IL-18 Tg mice suggests the potential activation of autocrine-paracrine mechanisms of apoptosis in the hepatocytes of these mice, as previously described for liver and biliary cells.39–41 Finally, modulation of NFκB activity in IL-18 Tg mice may allow other transcription factors to facilitate Fas mediated apoptosis. In any case, our data suggest a pivotal role for IL-18 in regulating hepatocyte apoptosis and liver injury in vivo via a Fas dependent pathway. These data are consistent with the recent demonstration of high IL-18 levels in the liver of patients with fulminant hepatic failure30,42 and suggest that overexpression of IL-18 may cause substantial liver injury in vivo.

Acknowledgments

SF and MFN are supported by a DGF and SFB548 grant. MP is supported by the Boehringer Ingelheim Fondation and MH is supported by a DFG fellowship.

REFERENCES

View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • CORRECTION
    BMJ Publishing Group Ltd and British Society of Gastroenterology