fax: (81) 58-230-6310.
Liver Biology and Pathobiology
Leflunomide protects from T-cell–mediated liver injury in mice through inhibition of nuclear factor κB
Article first published online: 28 SEP 2004
Copyright © 2004 American Association for the Study of Liver Diseases
Volume 40, Issue 5, pages 1160–1169, November 2004
How to Cite
Imose, M., Nagaki, M., Kimura, K., Takai, S., Imao, M., Naiki, T., Osawa, Y., Asano, T., Hayashi, H. and Moriwaki, H. (2004), Leflunomide protects from T-cell–mediated liver injury in mice through inhibition of nuclear factor κB. Hepatology, 40: 1160–1169. doi: 10.1002/hep.20438
- Issue published online: 14 OCT 2004
- Article first published online: 28 SEP 2004
- Manuscript Accepted: 26 JUL 2004
- Manuscript Received: 3 SEP 2003
Leflunomide is a novel immunosuppressive and anti-inflammatory agent for the treatment of autoimmune disease. The aim of this study was to investigate whether leflunomide protects from liver injury induced by concanavalin A (Con A), a T-cell–dependent model of liver damage. BALB/c mice were injected with 25 mg/kg Con A in the presence or absence of 30 mg/kg leflunomide. Liver injury was assessed biochemically and histologically. Levels of circulating cytokines and expressions of cytokine messenger RNA (mRNA) in the liver and the spleen were determined. Treatment with leflunomide markedly reduced serum transaminase activities and the numbers of dead liver cells. Leflunomide significantly inhibited increases in plasma tumor necrosis factor alpha (TNF-α) and interleukin 2 concentrations, and also reduced TNF-α mRNA expression in the liver after administration of Con A. These findings were supported by the results in which leflunomide administration decreased the number of T lymphocytes infiltrating the liver as well as inhibiting their production of TNF-α. Activation of nuclear factor κB (NF-κB), which regulates TNF-α production, was inhibited in the liver of mice treated with leflunomide, resulting in a reduction of TNF-α production from lymphocytes infiltrating the liver. In conclusion, leflunomide is capable of regulating T-cell–mediated liver injury in vivo and that this event may depend on the decrease of TNF-α production in the liver through inhibition of NF-κB activation caused by leflunomide. (HEPATOLOGY 2004.)
Leflunomide [HWA-486; N-(4-trifluoromethylphenyl)-5-methylisoxazol-4-carboxamide] is an agent that exhibits anti-inflammatory, antiproliferative, and immunosuppressive effects through mechanisms that are not fully understood.1–4 HWA-486 is a prodrug that is rapidly converted in the cell to an active metabolite, N-(4-trifluoromethylphenyl-2,2-cyano-3-hydroxycrotoamide), named A77 1726. This agent has been used to treat autoimmune diseases such as rheumatoid arthritis1 or to prevent rejection in transplantation.5 Early experiments suggest that A77 1726 blocks T-cell proliferation in response to stimulation by anti-CD28 and PMA, anti-CD3, and interleukin (IL)-2.6
Concanavalin A (Con A)–induced hepatitis is considered to be an experimental murine model of human autoimmune hepatitis.7 It is characterized by massive hepatocellular degeneration and lymphoid infiltration in the liver.8 The hepatitis induced by Con A is dependent on T-cell activation, as indicated by nonoccurrence in athymic nude mice, SCID mice, and FK506-pretreated mice.9, 10 Tissue injury caused by Con A administration is limited to the liver.9 The reason for this liver specificity is unknown but may be related to the high affinity of Con A toward the hepatic sinus, and it leads to T-cell activation in the liver.11 Intravenous administration of Con A induces lymphocyte accumulation in the liver through the increased influx of circulating lymphocytes and local proliferation via blastoid formation.12 T-cell activation elicited by Con A results in the elevation of the plasma levels of various cytokines including tumor necrosis factor alpha (TNF-α), interferon gamma (IFN-γ), and IL.12 In particular, TNF-α and IFN-γ released from activated T cells are considered to play critical roles in the development of Con A–induced hepatic injury because passive immunization against these cytokines effectively protects animals from hepatic injury.10, 11, 13, 14
TNF-α is an important mediator of inflammation that contributes to the pathogenesis of a variety of conditions, including rheumatoid arthritis, Crohn's disease, and many types of infections.15 TNF-α is involved in viral hepatitis, alcoholic liver disease, and fulminant hepatitis, and its toxicity is particularly important in the pathophysiology of hepatocytes.16 Binding of TNF-α to TNF receptor-1 (p55) results in the trimerization of its C-terminal cytoplasmic “death domain” and recruitment of some intracellular proteins involved in apoptotic signal transduction.17, 18 Although various cell types are capable of producing TNF-α, macrophages and lymphocytes are its principal source.15 Binding sites for multiple transcription factors, including nuclear factor κB (NF-κB), CCAAT/enhancer binding protein β and c-Jun, have been identified in the proximal promoter of the TNF-α gene. By deletion and mutational analysis, each of these sites has been shown to be capable of contributing to the activation of the TNF-α promoter in macrophages.19–21
In the present study, we investigated whether leflunomide protects from Con A–induced liver injury in vivo and examined the mechanism of leflunomide regarding liver injury.
Materials and Methods
Animals and Treatment Schedules.
Pathogen-free, male BALB/c mice (7-8 weeks old) were purchased from Japan SLC (Shizuoka, Japan) and maintained for at least 1 week in a 12-hour day/night rhythm with free access to food and water until the day of the experiment. The experiments were conducted in accordance with the institutional guidelines by the Gifu University School of Medicine. Con A (Sigma-Aldrich, St. Louis, MO) (25 mg/kg) was injected intravenously. Leflunomide (HWA-486, Hoechst AG, Weisbaden, Germany) was suspended in 0.5% carboxymethylcellulose at a concentration of 3 mg/mL and administered via oral gavage. BALB/c mice were given 30 mg/kg leflunomide 30 minutes before Con A administration. In control animals, only carboxymethylcellulose solution was given. Recombinant mouse TNF-α (2 μg/mouse), IFN-γ (2 μg/mouse) and IL-6 (1 μg/mouse) (Genzyme, Cambridge, MA) were intravenously administrated into mice. Animals were killed at the indicated time points, and blood samples, spleens, and livers were collected.
A recombinant replication-deficient adenovirus expressing mutant inhibitor κB (Ad5IκB) was constructed as previously described.22, 23 Because of missense mutations at phosphorylation sites where serines 32 and 36 are replaced with alanines, the mutant IκB irreversibly binds to NF-κB and prevents its activation. The viral solution containing 2 × 109 plaque-forming units was injected in mice intravenously. At 72 hours after the viral injection, Con A was injected. Ad5LacZ, which contains the Escherichia coli β-galactosidase gene, was used as a control vector. A p38 mitogen-activated protein kinase (MAPK) inhibitor, SB 239063 (Alexis Biochemicals, Montreal, Canada), was administered at 30 mg/kg via oral gavage 30 minutes before Con A administration.
Transaminase Activity and Plasma Cytokine Levels.
Serum alanine aminotransferase and aspartate aminotransferase activities were measured using a standard clinical automatic analyzer 7600 (Hitachi, Ibaraki, Japan). Plasma TNF-α, IFN-γ, and IL-2 concentrations were assayed using enzyme-linked immunosorbent assay kits (Genzyme Techne, Minneapolis, MN) according to the manufacturer's protocol.
Liver Histology and Terminal Deoxynucleotidyl Transferase-Mediated dUTP Nick End Labeling.
Livers were excised, fixed with 10% buffered formalin, sectioned at a thickness of 5 μm, and stained with hematoxylin-eosin for light microscopic examination. Terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling was performed on 5-μm–thick sections of the liver using a kit (Apop Tag; Intergen, Purchase, NY) according to the manufacturer's instructions.
Electrophoretic Mobility Shift Assay.
Nuclear extracts of the livers and intrahepatic T cells were prepared as described previously.22 NF-κB–binding consensus single-strand oligonucleotide (5′-TAGTTGAGGGGACTTTCCCAGG-3′) was first annealed with the complement oligonucleotide (5′-TGCCTGGGAAAGTCCCCTCAACTA-3′). The annealed DNA fragment was labeled with [α-32P]dCTP (ICN Biochemicals, Costa Mesa, CA) using Klenow DNA polymerase. Nuclear protein (20 μg) was incubated with 2.5 ng of 32P-labeled double-strand oligonucleotide probe for 30 minutes at room temperature. The mixture was electrophoresed on 4% polyacrylamide gels with 0.5× Tris-borate-ethylenediaminetetraacetic acid buffer at 4°C.
Ribonuclease Protection Assay.
For ribonuclease protection assay, frozen liver and spleen tissues from mice were homogenized in ISOGEN (Nippongene, Tokyo, Japan), and total RNA was isolated as described previously.23 Ribonuclease protection assay was performed with the RibaQuant Multi-Probe Rnase Protection Assay System (Pharmingen, San Diego, CA) according to the manufacturer's instructions. Total RNA was hybridized with 32P-labeled probes including cytokine genes synthesized by mCK-3b Multi-Probe template sets (Pharmingen). After hybridization, the samples were treated with ribonuclease and then electrophoresed on 5% polyacrylamide gels with 0.5× Tris-borate-ethylenediaminetetraacetic acid buffer.
Frozen liver tissues were homogenized in lysis buffer (10 mmol/L hydroxyethylpiperazine-N-2 ethanesulfonic acid (pH 7.4), 2 mmol/L ethylenediaminetetraacetic acid, 0.1% CHAPS, 5 mmol/L dithiothreitol, 1 mmol/L phenylmethylsulfonyl fluoride, 10 μg/mL pepstatin A, 10 μg/mL aprotinin, and 20 μg/mL leupeptin). The homogenates were incubated on ice for 1 hour, then centrifuged at 14,000g at 4°C for 20 minutes. The supernatant extracts (200 μg protein) were incubated with each fluorogenic peptide substrate—Ac-DEVD-AMC (Alexis, San Diego, CA) for caspase 3 or Ac-LEHD-AMC (Alexis) for caspase 9—at a final concentration of 50 μmol/L for 2 hours at 37°C. Change in fluorescence of released AMC was monitored using Fluoroskan Ascent FL (ThermoElectron, Helsinki, Finland) at an excitation wavelength of 355 nm and emission wavelength of 460 nm.
Western Blot Analysis.
Frozen liver tissues were homogenized in a radioimmunoprecipitation buffer (50 mmol/L Tris-HCl [pH 8.8], 150 mmol/L NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, and 10 mmol/L ethylenediaminetetraacetic acid) supplemented with a mixture of protease and phosphatase inhibitors (100 μmol/L Na3VO4, 50 mmol/L NaF, 10 mmol/L sodium pyrophosphate, and 5 mmol/L dithiothreitol, 1 mmol/L phenylmethylsulfonyl fluoride, 10 μg/mL pepstatin A, 10 μg/mL aprotinin, and 20 μg/mL leupeptin). Extracted proteins were separated by SDS–polyacrylamide gel electrophoresis and electrophoretically transferred onto nitrocellulose membranes. The membranes were probed with antibodies against p38 and phosphorylated p38 (Cell Signaling Technology, Beverly, MA), then incubated with an anti–rabbit immunoglobulin G HRP-coupled secondary antibody. Detection was performed using an electrochemiluminescence system (Amersham-Pharmacia Biotech, Buckinghamshire, UK). Moreover, p38 MAPK activity was determined using a p38 MAPK Assay Kit (Cell Signaling Technology) according to the manufacturer's protocol.
To isolate intrahepatic leukocytes (IHLs), single-cell suspensions were prepared from the liver that were perfused with phosphate-buffered saline (PBS) via the inferior vena cava and pressed through a 70-μm Cell Strainer (BD Biosciences, Mountain View, CA) as described previously.24 Total IHLs were digested with RPMI 1640 (Life Technologies, Gaithersburg, MD) containing 0.02% collagenase IV (Sigma-Aldrich) and 0.002% Dnase I (Sigma-Aldrich) for 40 minutes at 37°C. Cells were washed with RPMI 1640 and then underlaid with 24% metrizamide (Sigma-Aldrich) in PBS. After centrifugation for 20 minutes at 1,500g, IHLs were isolated at the interface.
T cells were purified via negative selection using a T cell isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's directions.
Single-cell suspension of IHLs was washed in PBS containing 1% bovine serum albumin and 0.02% sodium azide, then incubated for 20 minutes on ice to block FcR. The cells were surface stained with fluorochrome-conjugated monoclonal antibody for 20 minutes on ice. The following antibodies were used: anti-CD3, anti–DX-5, anti–Gr-1, anti-CD11b, and anti-CD40, (BD Pharmingen, San Diego, CA). Samples were acquired on a FACSCalibur flow cytometer (BD Immunocytometry Systems, San Jose, CA) and the data were analyzed using CellQuest software (BD Immunocytometry Systems).
IHLs were cultured ex vivo for 4 hours in brefeldin A (BD Pharmingen) to allow intracellular cytokines to be sequestered in the Golgi apparatus. Cells were then surface stained with anti–CD-3-FITC, anti–DX-5-PE, anti–Gr-1-FITC, and anti–CD11b-PE monoclonal antibody, washed in fluorescence-activated cell sorting buffer (PBS with 1% fetal calf serum), and fixed in 2% paraformaldehyde for 30 minutes in 25 μL PBS plus 0.5% saponin. Anti–mouse TNF-α–allophycocyanin or isotype control-allophycocyanin monoclonal antibody was added at a final dilution of 1:100, and cells were incubated for 30 minutes at room temperature. Cells were washed and resuspended in 1 mL of fluorescence-activated cell sorting buffer for analysis on a FACScan flow cytometer (BD Immunocytometry Systems).
IHLs and T cells were cultured in RPMI 1640 supplemented with 5% fetal calf serum at a density of 2 ×106 cells/mL and grown at 37°C in 5% CO2. Cells were stimulated with 20 μg/mL Con A alone or after pretreatment for 12 hours with proteasome inhibitor 3 μmol/L MG132 (N-cbz-Leu-Leu-leucinal) (Alexis) or for 2 hours with 10 μmol/L leflunomide metabolite (A77 1726, Hoechst AG). Cells were infected by Ad5IκB at a multiplicity of infection of 1 for 1 hour, cultured in medium for 24 hours, then stimulated with 20 μg/mL Con A. TNF-α concentrations in culture supernatant were determined using a mouse cytokine immunoassay kit (Pierce Biotechnology Inc., Rockford, IL), according to the manufacturer's protocol.
All values in the figures and text are expressed as the mean ± SE. The significance of differences among mean values was evaluated according to the Mann-Whitney U test.
Leflunomide Protects From Acute Liver Injury Induced by Con A.
Injecting Con A to mice caused an increase in serum aminotransferase activity at 8 hours (Fig. 1). At a dose of 30 mg/kg, leflunomide significantly reduced alanine aminotransferase activities (Fig. 1A). Microscopic examination of liver sections from animals administered with Con A revealed severe morphological changes, which are characteristic of apoptosis (Fig. 2C). When leflunomide was pretreated 30 minutes before Con A injection, only focal liver cell death was observed (Fig. 2E). We quantified hepatotoxicity by counting cells positive for terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling. After Con A injection, a marked decrease in the number of cell deaths was observed in the livers of mice pretreated with leflunomide compared with those of mice treated with Con A alone (Fig. 2D, 2F, and 2G).
Leflunomide Blocks Caspases Activities.
The activities of caspase 3–like and caspase 9–like proteases increased gradually after Con A administration as assessed using Ac-DEVD-AMC and Ac-LEHD-AMC peptide. Leflunomide administration significantly reduced caspase 3–like and caspase 9–like activities 12 hours after Con A administration (Fig. 3). These changes of caspase 3 and 9 activities were consistent with those of the cleavages of procaspases 3 and 9 as determined via Western blot analysis (data not shown).
Effect of Leflunomide on Plasma TNF-α, IFN-γ, and IL-2.
Plasma levels of TNF-α, IFN-γ, and IL-2 were determined in Con A–treated mice with or without administration of leflunomide (Fig. 4). At 1 hour after Con A administration, plasma TNF-α reached a peak level, followed by a sharp decrease (Fig. 4A). Plasma IL-2 levels began to increase at 2 hours after Con A administration and reached a peak at 5 hours (Fig. 4C). Administration of leflunomide significantly decreased plasma TNF-α and IL-2 levels following Con A administration. Plasma IFN-γ levels began to comparably increase 1 hour after Con A administration in mice treated or nontreated with leflunomide (Fig. 4B).
Expression of Cytokine Messenger RNA in the Liver and Roles of Cytokines on the Protective Effect of Leflunomide.
Because plasma levels might not specifically reflect cytokine expressions in the liver, messenger RNA (mRNA) expressions of several cytokines in the liver and spleen were measured via ribonuclease protection assay. The expression of TNF-α, IFN-γ, and IL-6 mRNA was not detected constitutively or was only minimally present in whole-organ RNA preparations of the liver (Fig. 5A). The mRNA expressions of these cytokines in the liver increased 1 hour after Con A administration. Administration of leflunomide strongly suppressed hepatic TNF-α, IFN-γ, and IL-6 mRNA induction by Con A treatment.
In the spleen, TNF-α mRNA constitutively rose and did not show a further increase after Con A administration (Fig. 5A). IL-6 and transforming growth factor beta mRNA increased 1 hour after Con A administration, while treatment with leflunomide had little effect.
To clarify which cytokine played a pivotal role in the protective effect of leflunomide, we assessed the effects of TNF-α, IFN-γ, and IL-6 on liver injury in mice treated with Con A plus leflunomide. TNF-α administration increased alanine aminotransferase activity in mice treated with Con A plus leflunomide to a similar extent as mice treated with Con A, but IFN-γ and IL-6 did not (Fig. 5B).
Effects of Leflunomide on NF-κB and MAPK Activation in the Liver.
The above results indicate that leflunomide may protect mice from Con A–induced hepatitis through inhibition of TNF-α production. TNF gene expression can be regulated by transcription factors, including NF-κB and activator protein 1, or activators of MAPKs, p38, p42/44 extracellular signal–regulated kinase, and c-Jun N-terminal kinase–signaling pathways.25 To elucidate the mechanism by which leflunomide protected mice from Con A–induced hepatitis, we investigated whether NF-κB or MAPK signaling pathways are regulated by leflunomide. Electrophoretic mobility shift assay demonstrated early induction of NF-κB activity by Con A in the liver (Fig. 6A). Treatment with leflunomide resulted in a decrease of the induction of NF-κB activity in the liver. In the spleen, NF-κB was constitutively activated and was not affected by either Con A or leflunomide.
Next, we investigated whether MAPK was involved in activation pathways leading to TNF production in Con A–induced hepatitis and leflunomide-regulated MAPK activation in the liver. As shown in Fig. 6B, p38 phosphorylation occurred 1 hour after Con A treatment and gradually decreased up to 5 hours after stimulation. However, pretreatment with leflunomide potently blocked p38 phosphorylation while having no effect on overall levels of nonphosphorylated p38 in the liver (Fig. 6B). Consistent with p38 phosphorylation, p38 activity in the liver was inhibited by leflunomide (Fig. 6B). On the other hand, c-Jun N-terminal kinase and extracellular signal–regulated kinase activation in the liver was induced 1 hour after Con A injection, but leflunomide did not affect their activation (data not shown).
It has been demonstrated that AdIκB is almost exclusively expressed in the liver and inhibits NF-κB DNA binding activity and transcriptional activity in the liver in vivo.22, 23 In mice treated with Ad5IκB, liver injury induced by Con A was inhibited (Fig. 6C). On the other hand, a specific inhibitor of p38 MAPK, SB 239063, did not influence the liver injury. These findings suggest that the anti-inflammatory properties of leflunomide in Con A–induced hepatitis may be mediated by a blockade of NF-κB.
Role of Leflunomide in Con A–Induced Inflammatory Cell Recruitment.
To determine whether leflunomide affects inflammatory cell recruitment into the liver after Con A injection, we analyzed the characteristics of intrahepatic inflammatory cell infiltrate. The absolute number of total IHLs increased 1 hour after Con A injection, but their increase was significantly inhibited by leflunomide pretreatment (Fig. 7). Furthermore, CD3+/DX-5− cells (T cells) increased to the greatest extent after Con A injection, but their increase was significantly inhibited by leflunomide pretreatment. The others, CD3−/DX-5+ cells (natural killer cells), Gr-1+/CD11b− neutrophil population, and Gr-1−/CD11b+ macrophage population also increased in number after Con A injection. In natural killer cells and macrophages, leflunomide pretreatment slightly—but not significantly—inhibited the increase of each population after Con A injection, but it made little difference regarding the number of neutrophils.
Effect of Leflunomide on TNF-α Release From Intrahepatic T Cells.
To determine which cells produced TNF-α after Con A injection and are affected by leflunomide, we stained intrahepatic TNF-α in macrophage (CD11b+) and T-cell (CD3+) subsets 1 hour after Con A injection. As shown in Fig. 8, the T-cell subset expressed higher levels of TNF-α after Con A injection than control, and leflunomide pretreatment decreased TNF-α levels by half. In contrast, TNF-α expression in the macrophage subset was constitutively upregulated to some extent and was not affected by leflunomide pretreatment.
Serial TNF-α Production and NF-κB Activation of Cultured Intrahepatic T Cells.
To clarify whether the effect of leflunomide is upstream or downstream TNF-α, we measured serial TNF-α production and NF-κB activities of intrahepatic T cells cultured in vitro in response to Con A. NF-κB of T cells was activated as early as 10 minutes after the addition of Con A into medium (Fig. 9C), although TNF-α levels in the cells (Fig. 9A) or in the medium (data not shown) were not detected over 20 minutes, and subsequently began to increase. The proteasome inhibitor MG132, which is a well-known NF-κB inhibitor by blocking degradation of IκBα,26–28 was employed for further confirmation of hypothesis that TNF-α production induced by Con A was NF-κB–mediated. MG132 reduced either intracellular TNF-α production 30 and 60 minutes after Con A stimulation (Fig. 9A), or TNF-α levels in the medium (Fig. 9B). In addition, as shown in Fig. 9A, intracellular TNF-α production was decreased when cells were infected with Ad5IκB that selectively inhibits NF-κB activity. At 10 minutes, leflunomide decreased NF-κB activity of cultured intrahepatic T cells stimulated by Con A (Fig. 9D). These data demonstrate that NF-κB activation induced by Con A preceded TNF-α production, that TNF-α production induced by Con A was NF-κB-mediated, and that leflunomide decreases TNF-α production through the direct inhibition of NF-κB activity.
In the present study, we demonstrated that leflunomide protected mice from Con A–induced liver injury and massive liver cell death. Leflunomide is a novel immunosuppressive and anti-inflammatory agent currently being tried for treatment of autoimmune diseases and transplant rejection, although many other immunosuppressive drugs have been reported previously. Allison addressed that their mechanisms of action fell into five groups and classified leflunomide as an inhibitor of de novo pyrimidine synthesis.29 In the liver, administration of dexamethasone, cyclosporine A, or FK506 protects Con A–induced hepatitis in BALB/c mice.9 However, little is known of the effect of leflunomide on liver disease or injury. It has been reported that liver injury caused by Con A is due to apoptotic cell death of hepatocytes.11 In the present study, Con A administration induced liver cell death and caused time-dependent increases in liver caspase 3–like and caspase 9–like activity, suggesting that hepatocyte death depends on caspase cascade activation. Leflunomide clearly prevented liver cell death and the increases of caspase 3–like and caspase 9–like activity induced by Con A (Fig. 3).
Con A–induced hepatitis is thought to be a model of immunologically induced hepatocyte injury. T-cell activation plays a crucial role in the process of Con A–induced hepatitis. T-cell activation elicited by Con A results in the elevation of plasma cytokines, including TNF-α, IFN-γ, IL-1, IL-2, IL-4, and IL-6. Among those cytokines, TNF-α released from activated T cells appears to be a prime suspect in causing liver injury induced by Con A, based on a number of reported observations. Mice pretreated with anti–mouse TNF-α antiserum10 or TNF-α inhibitor30 or mice deficient for TNFR1 and TNFR231 are resistant to Con A–induced hepatitis. IFN-γ is also considered to be involved in the development of Con A–induced hepatitis, because mice pretreated with anti–IFN-γ antiserum14 or IFN-γ–deficient mice32 are resistant to Con A–induced liver injury. These studies suggest that overproduction of TNF-α and IFN-γ in the liver may cause liver injury. In the present study we demonstrated that TNF-α, IFN-γ, and IL-2 plasma levels increased in mice after Con A administration and that leflunomide pretreatment reduced the increases of TNF-α and IL-2 plasma levels (Fig. 4). Leflunomide also inhibited TNF-α mRNA and IFN-γ mRNA expression in the liver upregulated by Con A administration (Fig. 5A). These findings indicated that leflunomide suppressed T-cell activation induced by Con A, inhibited TNF-α and IFN-γ production in the liver, and consequently resulted in the inhibition of liver injury.
In an attempt to elucidate how leflunomide exerts its protective effect, we found that pretreating mice with this agent prevented Con A–induced CD3+ lymphocyte infiltration of the liver (Fig. 7). Together with the prevention of liver T-cell infiltration, leflunomide inhibited TNF-α production from T cells in the liver (Fig. 8). However, it is well established that macrophage–T-cell interactions are a necessary basis of many immune reactions. Gantner et al. demonstrated that Con A–induced crosstalk between lymphocytes and liver macrophages resulted in augmented cytokine release, which finally leads to liver injury upon Con A injection into mice.33 In this study, however, activation and an increase of TNF-α production were observed in T cells, but not in macrophages, and leflunomide influenced T cells alone (Fig. 8). Recently, it has been reported that Vα14 NKT cells function as effector cells to develop Con A–induced hepatitis.7 In their report, IL-4 produced by Con A–activated Vα14 NKT cells played a critical role in these disease developments by augmenting the cytotoxic activity of Vα14 NKT cells in an autocrine fashion. In our study, Con A treatment certainly increased IL-4 plasma level in mice but leflunomide did not significantly reduce the increase of IL-4 plasma level induced by Con A (data not shown). Further studies are required to elucidate the differences between the effects of leflunomide on T cells and those on other inflammatory cells.
Finally, we investigated the mechanisms through which leflunomide reduced TNF-α production in the liver induced by Con A. One important signaling pathway in the induction of TNF expression leads to activation of the transcription factor NF-κB.26, 34, 35 Inhibiting gene expression through the direct modulation of transcription factors may provide exciting potential therapeutic targets, especially if a tissue-specific regulation of TNF-α production is unraveled. Recently, Manna and Aggarwal demonstrated that treatment of a human T-cell line (Jurkat) with leflunomide blocks NF-κB activation mediated by many inflammatory agents, including phorbol ester, TNF-α, lipopolysaccharide, H2O2, okadaic acid, and ceramide.36 Consistent with this report, in the present study leflunomide inhibited NF-κB activation in the liver induced by Con A, but not in the spleen (Fig. 6), supporting the results of cytokine mRNA expression via ribonuclease protection assay. In the present study using cultured intrahepatic T cells, NF-κB activation occurred in response to stimulation by Con A earlier than the increases of TNF-α production; TNF-α production was inhibited by inhibitors of NF-κB, MG132, or Ad5IκB; and leflunomide inhibited the NF-κB activation (see Fig. 9). The in vitro experiment has demonstrated that the regulation of TNF-α production was NF-κB mediated and that leflunomide decreased TNF-α production through the direct inhibition of NF-κB activity. We speculate that leflunomide blocked NF-κB activation in lymphocytes, especially T cells in the liver; subsequently the transcription of TNF-α, the target gene of NF-κB, was inhibited, and finally mice were protected from Con A–induced hepatic injury.
Leflunomide has been reported to inhibit tyrosine kinases and lead to the inhibition of MAPKs.37, 38 We found that p38 phosphorylation occurred after Con A administration and that pretreatment with leflunomide blocked its phosphorylation in the liver (Fig. 6), although the administration of SB 239063, a specific inhibitor of p38, did not show the involvement of p38 MAPK in Con A–induced liver injury. Recent reports have revealed that p38 MAPK exists upstream of TNF-α and regulates its production. p38 inhibitor reduced TNF-α production induced by some variety of stimuli.39–41 Based on these findings and on our results, we speculate that leflunomide might reduce Con A–induced TNF-α production by blocking p38 MAPK activation as well as NF-κB activation and protects from liver injury. Further experiments are required to assess the roles of p38 MAPK in Con A–induced liver injury and TNF-α production.
In summary, we showed that leflunomide protected mice from T-cell–mediated liver injury induced by Con A. This liver injury was due to caspase-dependent liver cell death, mainly through TNF-α. Leflunomide reduced the production of TNF-α by inhibition of NF-κB, which exists upstream of TNF-α.
The authors thank Dr. Yoshiko Banno (Department of Biochemistry, Gifu University School of Medicine, Gifu, Japan) for advice concerning the assay of MAPK activity and Dr. David A. Brenner (Department of Medicine, Columbia University College of Physicians and Surgeons, New York, NY) for providing the adenoviruses.
- 30Disparate roles for TNF-α and Fas ligand in concanavalin A-induced hepatitis. J Immunol 1998; 60: 4082–4089., , , , , , et al.
- 36Immunosuppressive leflunomide metabolite (A77 1726) blocks TNF-dependent nuclear factor-κB activation and gene expression. J Immunol 1999; 2: 2095–2102., .