Ebselen Protects Mice Against T Cell-Dependent, TNF-Mediated Apoptotic Liver Injury1

  1. Gisa Tiegs1,
  2. Sabine Küsters1,
  3. Gerald Künstle2,
  4. Hannes Hentze2,
  5. Alexandra K. Kiemer3 and
  6. Albrecht Wendel2
  1. 1Institute of Experimental and Clinical Pharmacology and Toxicology, University of Erlangen-Nürnberg (G.T., S.K.); 2Biochemical Pharmacology, Faculty of Biology, University of Konstanz, D-78457 Konstanz (G.K., H.H., A.W.); 3Institute for Pharmacology, Toxicology and Pharmacy of the Veterinary Faculty, University of München (A.K.K.), Germany
     
    Next Section

    Abstract

    The seleno-organic drug ebselen (2-phenyl-1,2-benzoisoselenazol-3(2H)-one) has glutathione peroxidase-like activity, and inhibits lipoxygenases, oxidative burst of leukocytes, nitric oxide synthases, protein kinases and leukocyte migration. This study elaborates in vivoin mice hitherto unknown immunopharmacological properties of ebselen. The compound was comparatively investigated in two different T cell-dependent hepatic hyperinflammation models and in two alternative models of receptor-activated liver apoptosis. Mice orally pretreated with ebselen were dose-dependently protected from concanavalin A (ConA)-induced liver injury. In livers from ebselen-pretreated mice exposed to ConA, the nuclear antiapoptotic transcription factor NFκB was upregulated. The release of the proinflammatory cytokine tumor necrosis factor-α (TNF) was downregulated, while the ciculating amount of the anti-inflammatory cytokine interleukin-10 (IL-10) was increased. Ebselen protected also from liver injury induced by the superantigen staphylococcal enterotoxin B in galactosamine (GalN)-sensitized mice. Furthermore, ebselen protected the liver and enhanced circulating IL-10 in GalN-sensitized mice treated with recombinant TNF, i.e., the common distal mediator of ConA and SEB-induced hepatotoxicity. The activation of apoptosis-executing proteases, i.e., caspases, was blocked in livers of ebselen-treated mice following TNF receptor, but not following CD95 receptor activation. We propose a novel mechanism for the immunomodulatory properties of the drug and suggest that it might be useful in the therapy of T cell-mediated inflammatory disorders.

    The heterocyclic seleno-organic compound ebselen (2-phenyl-1,2-benzisoselenazol-3(2H)-one) is a drug with anti-inflammatory properties. It mimics the glutathione peroxidase reaction (Müller et al., 1984; Wendel et al., 1984) as well as the phospholipid hydroperoxide glutathione peroxidase reaction (Maiorino et al., 1988),i.e., it catalyzes the reduction of H2O2 as well as organic hydroperoxides with thiols as electron donors. Ebselen also inhibits 5-lipoxygenase activity of neutrophils (Safayhi et al., 1985), which can be explained by withdrawal of the precursor 5-HPETE from the production of the chemotactical and inflammatory mediator leukotriene B4. Interestingly, the mammalian enzyme 15-lipoxygenase was shown to be subject to much stronger inhibition by ebselen than human recombinant 5-lipoxygenase (Scheweet al., 1994).

    The recent pharmacological knowledge on ebselen includes a large variety of different molecular actions: in vitro influence on calcium homeostasis in human platelets (Brüne et al., 1991), inhibition of NADPH oxidase and protein kinase C (Cotgreave et al., 1989), prevention of inositol trisphosphate binding to its receptors (Dimmeler et al., 1991), as well as inhibition of apoptosis of mouse thymocytes (Ramakrishnan et al., 1996). In vivo, inhibition of migration of polymorphonuclear leukocytes and T lymphocytes in rats was observed (Gao and Issekutz, 1993, 1994). The obvious problem, however, is to decide which properties out of the plethora of mechanisms account for the anti-inflammatory effect of ebselen in vivo.

    The mammalian organism reacts to Gram-negative bacteria or to their cell wall components, such as endotoxins (LPS), with fulminant systemic inflammation known as the septic shock syndrome. This overactivation of the innate immune system frequently results in multiorgan failure and death. The liver not only clears bacterial toxins and initiates the acute phase response, but is also one of the main target organs of LPS-toxicity (Tiegs, 1994). Experimentally, septic liver failure can be initiated by the injection of LPS to mice sensitized with the amino sugar GalN that specifically inhibits transcription in hepatocytes. In this model, ebselen protected mice from lethal septic liver failurein vivo (Wendel and Tiegs, 1986), thus demonstrating a tissue-specific anti-inflammatory action. Under the life-threatening pathological condition of endotoxin exposure, the cytokine TNF is systemically released. Overshooting production of TNF has been identified as the cause of receptor-mediated tissue destruction due to apoptotic cell death (Leist et al., 1995). After signal transduction, execution of TNF-induced apoptosis involves the activation of caspases, a novel family of cysteine proteases. Caspase inhibitors prevented the activation of hepatic caspases (Rodriguezet al., 1996) and the inhibitor z-VAD-fmk blocked TNF-induced apoptosis of murine hepatocytes in vivo(Künstle et al., 1997), indicating the causal role of caspases in cytokine-mediated liver injury. As regulatory mediators, also anti-inflammatory cytokines are released under these conditionsin vivo. Among them, the major counterplayer of LPS-inducible TNF is interleukin-10 (IL-10) (Barsig et al., 1995; Louis et al., 1997). Indeed, mice deficient of IL-10 spontaneously develop chronic inflammatory bowel disease (Kühnet al., 1993).

    We intended to examine the pharmacological profile of ebselen in different in vivo models under conditions of pathologically overshooting cytokine release induced by T cell activation. Therefore, we first compared the drug’s actions in the ConA model (Tiegs et al., 1992; Gantner et al., 1995b; Küsterset al., 1996) with the ones in a model where the superantigen SEB, a Gram-positive derived exotoxin, overactivates T lymphocytes in vivo (Marrack et al., 1990) and causes hepatic apoptosis in GalN-sensitized mice (Nagaki et al., 1994; Gantner et al., 1995a). Both ConA as well as SEB induce the release of systemic TNF, IFNγ and various other cytokines. In both models, TNF was identified as a distal mediator of hepatic injury, since neutralization of TNF or knocking out the TNF receptor gene protected mice from liver failure (Gantner et al., 1995b; Miethke et al., 1992; Mizuhara, et al.; Küsters et al., 1997). In contrast to the GalN/SEB-model, in the ConA model, in addition to TNF, also IFNγ release causally contributes to liver injury (Küsters et al., 1996). Direct injection of recombinant murine TNF to mice also resulted in hepatocellular apoptosis under the condition of GalN-induced arrest of hepatic transcription (Tiegs et al., 1989; Leist et al., 1995). Consequently, we chose this latter model to differentiate direct effects of ebselen on TNFrelease initiated by the inflammogens used from those on TNFactions. In the latter case, we studied possible interactions of ebselen with apoptotic post-receptor signalling by examining the expression of the nuclear transcription factor NF-κB and the activation of caspases.

    Previous SectionNext Section

    Methods

    Animals.

    Animals received humane care according to the legal requirements in Germany and were maintained for at least 10 days under controlled conditions (22°C, 55% humidity, and 12 hr day/night rhythm), and were fed a standard laboratory chow (Altromin 1313; Altromin, Lage, Germany). BALB/c mice were 6 to 8 weeks old and weighed 25 to 30 g when taken into the experiment.

    Application regimen and sampling of material.

    All animals were starved overnight and challenged in the morning. A single dose of 25 mg/kg ConA (Sigma Chemical Co., Deisenhofen, Germany) was injected intravenously (i.v.) in a volume of 300 μl pyrogen-free saline. SEB (Sigma Chemical Co., Deisenhofen, Germany) was given intraperitoneally (i.p.) at a single dose of 2 mg/kg. Recombinant murine TNF was injected at a single dose of 10 μg/kg i.v., both in a total volume of 300 μl pyrogen-free saline containing 0.1% human serum albumin. Anti-CD95 was given i.v. in a volume of 300 μl saline containing 0.1% human serum albumin. GalN (Roth, Karlsruhe, Germany) was given i.p. 10 min before SEB or TNF in 200 μl pyrogen-free saline in a dose of 700 mg/kg. Ebselen was suspended in 1% traganth (diluted in pyrogen free saline) and administered orally (p.o.) in a volume of 300 μl in doses of 10, 50, 150 or 600 mg/kg, respectively.

    For determination of circulating TNF, IL-4 and IL-10, blood samples were taken from the tail vein 90 min after challenge. Eight hours after challenge, blood was withdrawn by cardiac puncture into heparinized syringes under lethal nembutal anaesthesia (150 mg/kg i.v.) for determination of circulating IFNγ as well as for the assessment of liver injury by measurement of plasma transaminases and sorbitol dehydrogenase. Plasma samples were stored at −80°C until determination of enzyme activity or measurement of cytokines by ELISA.

    Cytokine determination by ELISA.

    All incubations of the sandwich ELISA were performed in flat-bottom high-binding polystyrene microtiter plates (Greiner, Nürtingen, Germany). For the IL-4 and IFNγ ELISA, specific rat anti-mouse monoclonal antibody pairs (biotinylated detecting mAb), purchased from Pharmingen (Hamburg, Germany), were used. For the measurement of TNF, a protein G+ purified polyclonal sheep anti-mouse TNF capture antibody (protein content 20 mg/ml, in-house preparation) was used, replacing the PharMingen capture mAb; as detection antibody a rabbit anti-mouse TNF antibody was used. Steptavidin-peroxidase (Jackson Immuno Research, West Grove, PA) and the peroxidase chromogen tetramethylbenzidine (Boehringer Mannheim, Mannheim, Germany) were used to detect the immunocomplex. Circulating IL-10 was determined using the IL-10 ELISA kit Intertest-10X (Genzyme Corporation, Cambridge, MA). The detection limits of the TNF, IFNγ, IL-4 and IL-10 ELISA were 10, 50, 10 and 15 pg/ml, respectively.

    Determination of caspase-3-like activity in liver homogenate.

    Livers were shortly perfused with ice-cold PBS under lethal nembutal anaesthesia (150 mg/kg i.v.). Samples from the large anterior lobe were taken and immediatly frozen in liqid N2. Cytosolic extracts were prepared by Dounce homogenization of ∼100 mg liver sample in hypotonic extraction buffer (25 mM HEPES, pH 7.5, 5 mM MgCl2, 1 mM EGTA, 1 mM PEFA-block and pepstatin, leupeptin and aprotinin 1 μg/ml each) and were subsequently centrifuged for 15 min at 13,000 ×g.

    The fluorometric assay was performed on microtiter plates (Greiner, Nürtingen, Germany). Cytosolic extracts (10 μl, 1–2 mg/ml protein) were diluted 1:10 with substrate buffer (50 mM HEPES, pH 7.4, 1% sucrose, 0.1% CHAPS, 10 mM DTT, 50 μM fluorogenic substrate DEVD-AFC, Biomol, Hamburg, Germany). Blanks contained 10 μl extraction buffer and 90 μl substrate buffer. Generation of free AFC at 37°C was determined by measurement at t = 0/t = 30 min using a fluorometer plate-reader (SLT Fluostar, SLT, Crailsheim, Germany) set at an excitation wavelength of 385 nm and an emission wavelength of 505 nm. Protein concentrations of the corresponding samples were estimated (Pierce-Assay, Pierce, Rockford, IL) and the activity was calculated using standards serially diluted (0–5 μM AFC). Control experiments confirmed that the activity was linear with time and with protein concentration under the conditions described above.

    Preparation of nuclear extracts and EMSA.

    Nuclear extracts were prepared from frozen liver sections. Briefly, tissue samples were homogenized in 3 ml ice-cold hypotonic buffer A (10 mM HEPES pH 7.9; 10 mM KCl; 0.1 mM EDTA; 0.1 mM EGTA; 1 mM DTT; 0.5 mM phenylmethylsulfonyl fluoride) with a Dounce homogenizer. The homogenate was transferred to a polypropylene centrifuge tube and, after a 10 min incubation on ice, centrifuged at 1000 × g for 10 min at 4°C. The cell pellet was suspended in 1.4 ml of ice-cold buffer A and 90 μl of a 10% solution of Nonidet P-40 were added followed by 10 sec of vigorous vortexing. The suspension was incubated on ice for 10 min and then centrifuged at 12,000 × g for 30 sec at 4°C. The supernatant was removed and the nuclear pellet was extracted with 200 μl of hypertonic buffer B (20 mM HEPES pH 7.9; 0.4 M NaCl; 1 mM EDTA; 1 mM EGTA; 1 mM DTT; 1 mM PMSF) by shaking at 4°C for 30 min. The extract was centrifuged at 12,000 × g for 10 min at 4°C and the supernatant was frozen at –70°C.

    A double-stranded oligonucleotide probe containing a consensus binding-sequence for NF-κB (5′-AGT TGA GGG GAC TTT CCC AGG C-3′) (Promega, Heidelberg, Germany) was 5′ end-labeled with [γ-32P]-ATP (3000 Ci/mmol, Amersham, Braunschweig, Germany) using T4 polynucleotide kinase (Promega, Heidelberg, Germany). Then, 10 μg nuclear protein was incubated in a 15 μl reaction volume containing 10 mM Tris-HCl pH 7.5, 5 × 104 cpm radiolabelled oligonucleotide probe, 2 μg poly(dI·dC), 4% glycerol, 1 mM MgCl2, 0.5 mM EDTA, 50 mM NaCl and 0.5 mM DTT for 20 min at room temperature. Nucleoprotein-oligonucleotide complexes were resolved by electrophoresis in a 4.5% nondenaturing polyacrylamide gel in 0.25 × TBE at 100 V. The gel was autoradiographed with an intensifying screen at –70°C overnight. Specificity of the DNA-protein complex was confirmed by competition with a 100 fold excess of unlabeled NF-κB and AP-2 (5′-GAT CGA ACT GAC CGC CCG CGG CCC GT-3′, Promega, Heidelberg, Germany) sequences, respectively.

    Cell culture experiments.

    Spleens from female BALB/c mice were taken and ground through a nylon cell strainer (diameter 100 μm; Becton Dickinson, New Jersey, USA) into 5 ml Click RPMI medium (Biochrom, Berlin, Germany), centrifuged, and resuspended in 5 ml NH4Cl (7.5 g/l in Tris·HCl buffer, pH 7.2) to lyse the erythrocytes for 3 min at room temperature. Then the cells were washed once in medium and plated into microtiter plates at 106 cells per well. The splenocytes were stimulated with either 2 μg/ml ConA or 100 ng/ml SEB. Ebselen was added 30 min before the stimulus in concentrations of 0.1–30 μM. Incubations were carried out in an incubator run at 5% CO2 and 37°C. 20 hr after addition of the stimulus, cells were incubated with 20 μl MTT (5 mg/ml) for 4 hr, then the cells were lysed with 50 μl 20% SDS/0.02 HCl in the incubator at 37°C and 7% CO2 for 16 hr. MTT reduction was measured as a parameter of proliferation by 550 nm with an ELISA-reader.

    Statistical analysis.

    Graphical data in the figures 1, 2, 4and 5 are expressed as mean values ± S.E.M. and were analyzed by one-way analysis of variance (ANOVA); in case of differences among groups (*P < .05), data were subjected to Dunnett’s multiple comparisons test of the control against all other groups with the program Prism 2.00 (GraphPAD Software, San Diego, CA). Data in figure 3are expressed as mean values ± S.E.M. and were analyzed by the unpaired t test with Welsh’s correction (*P < .05).

    Figure 1
    View larger version:
    Figure 1

    Prevention by ebselen pretreatment of liver injury of mice injected with ConA (a) as assessed by plasma enzyme activities and (b) consequences on the levels of circulating cytokines. Male BALB/c mice were pretreated p.o. with either placebo (1% traganth), 50, 150, or 600 mg/kg ebselen 1 hr before i.v. administration of 25 mg/kg ConA. Plasma TNF and IL-10 concentrations were measured 90 min, plasma activities of ALT, AST, and SDH were determined 8 hr after ConA challenge. Data are expressed as mean values ± S.E.M.;n = 10 for the placebo control group andn = 5 for all other groups; *P < .05vs. placebo control.

    Figure 2
    View larger version:
    Figure 2

    Prevention by ebselen of SEB-induced liver injury in GalN-sensitized mice (a) as assessed by plasma enzyme activities and (b, c) consequences on the levels of circulating cytokines. Mice were pretreated p.o. with either placebo (1% traganth), 10, 50, 150, or 600 mg/kg ebselen 1 hr before i.p. administration of 2 mg/kg SEB. GalN (700 mg/kg) was given i.p. 15 min before SEB. Plasma TNF, IL-10, and IL-4 concentrations were measured 90 min, plasma activities of ALT, AST, and SDH, and circulating IFNγ concentration were determined 8 hr after SEB challenge. Data are expressed as mean values ± S.E.M.;n = 10 for the placebo control group andn = 5 for all other groups; *P < .05vs. placebo control; n.d., not detectable,i.e., below detection limit of the IL-4 ELISA (cf. Materials and Methods).

    Figure 4
    View larger version:
    Figure 4

    Consequences of intravenous injection of rmuTNF on the level of circulating IL-10 in placebo- or ebselen-treated mice. Mice were pretreated p.o. with either placebo (1% traganth) or 600 mg/kg ebselen 1 hr before i.v. administration of 10 μg/kg TNF. Plasma IL-10 concentrations were measured 90 min after TNF challenge. Data are expressed as mean values ± S.E.M.; n = 10 placebo control, n = 9 ebselen-treated; *P < .05 vs. placebo control.

    Figure 5
    View larger version:
    Figure 5

    Failure by ebselen to prevent CD95-mediated liver injury. Liver damage was induced by injection of agonistic anti-CD95 antibodies (2 μg/mouse in 300 μl isotonoc saline). Control mice received 300 μl saline only. Ebselen at a dose of 600 mg/kg was administered p.o. 1 hr before challenge. Eight hours after challenge, animals were sacrificed, and (a) plasma enzyme activities of ALT, AST, and SDH or (b) caspase activity in liver homogenate were determined. Data are mean values ± S.E.M. of n = 5 animals per group.

    Figure 3
    View larger version:
    Figure 3

    Prevention by ebselen of TNF-induced liver injury in GalN-sensitized mice (a) as assessed by plasma enzyme activities or (b) by measuring caspase activity in liver homogenate. Mice were pretreated p.o. with either placebo (1% traganth) or 600 mg/kg ebselen 1 hr before i.v. administration of 10 μg/kg rmuTNF. GalN (700 mg/kg) was given i.p. 15 min before TNF. Plasma activities of ALT, AST, and SDH or activity of caspases in liver homogenate were determined 8 hr after TNF challenge. Data are expressed as mean values ± S.E.M.; 3a, n = 15; 3b, n = 5; *P < .05 vs. placebo control.

    Previous SectionNext Section

    Results

    In contrast to untreated control mice where no measurable cytokines were detected in plasma, mice injected with the T cell mitogenic lectin ConA released high amounts of cytokines into the circulation. Among them, the proinflammatory mediator TNF was released very early with peak concentrations at 90 min after injection of the lectin. Under the model conditions and with the doses chosen, ConA-treated animals developed fulminant liver failure within 8 hr, which was quantitated by determination of enzyme activity of alanine aminotransferase (ALT, normal value in mice: ≤35 U/liter), aspartate aminotransferase (AST, normal value ≤ 30 U/l), and sorbitol dehydrogenase (SDH, normal value: ≤10 U/liter) in the plasma. Pretreatment of mice with ebselen 1 hr before injection of ConA significantly protected the animals from liver failure at doses of 150 and 600 mg/kg ebselen (fig. 1a). As a measure of effector cell activation, macrophage- and/or T cell-derived cytokines were determined by ELISA. Plasma samples for TNF, IL-4 and IL-10 were taken 90 min after ConA-challenge, IFNγ was measured 8 hr after ConA-injection (Gantner et al., 1997). ConA-induced TNF release was decreased dose-dependently in the blood of ebselen pretreated animals in comparison to the placebo-treated controls (fig.1b). As TNF concentrations were reduced, the levels of IL-10 in these animals dose-dependently increased (fig. 1b). ConA-induced levels of circulating IFNγ, a T cell-derived proinflammatory mediator, were not significantly affected by ebselen-treatment (4.03 ± 0.41 ng/ml in ConA/placebo controls vs. 3.85 ± 0.43 ng/ml in ebselen-treated mice, n = 5, respectively). However, the plasma concentration of IL-4, a specific T cell cytokine, was significantly reduced following pretreatment with 600 mg/kg ebselen (103 ± 9 pg/ml, n = 5, P < .01vs. ConA/placebo controls: 326 ± 18 pg/ml,n = 5).

    The potency of ebselen to protect against ConA-induced liver failure prompted us to investigate whether the drug was able to prevent hepatic damage initiated by another T cell-stimulus,i.e., SEB. This superantigen also overactivates lymphocytes to release cytokines in vivo and causes liver damage in GalN-sensitized mice (Gantner et al., 1995a). Pre-treatment with ebselen dose-dependently prevented liver failure, providing significant protection when given in a dose of 50 mg/kg or greater (fig. 2a). In parallel to this effect, ebselen significantly decreased the concentrations of circulating TNF and IFNγ (fig. 2b). As in the ConA-induced liver injury, SEB challenged mice pretreated with ebselen also released enhanced amounts of IL-10 into the circulation compared to placebo controls (fig. 2c). No significant influence on IL-4 release was observed, except when the highest dose of ebselen was given (fig. 2c).

    In order to find out whether the protective effect of ebselen was due to an influence on lymphocyte proliferation, freshly isolated mouse splenocytes were exposed in vitro to either 2 μg/ml ConA or to 20 ng/ml SEB for 20 hr. Neither the proliferation of ConA- nor that of SEB-treated splenocytes was affected by ebselen (0.1–30 μM) incubated 30 min in advance of the stimuli (data not shown).

    In both mouse models of liver injury, either induced by ConA or by SEB, TNF was identified as a central cytokine essential for the outcome of hepatic failure, since pretreatment with anti-TNF antibodies protected the animals in any of these models (Gantner et al., 1995b;Miethke et al., 1992; Mizuhara et al., 1994;Küsters et al., 1997). Hence, we examined whether pretreatment of GalN-sensitized mice with ebselen protected against TNF toxicity as such. Figure 3a demonstrates that ebselen given in a high dose of 600 mg/kg significantly protected GalN-sensitized animals from liver failure induced by intravenous injection of 10 μg/kg recombinant murine TNF. In addition, not only the release of liver specific enzymes upon secondary lysis of hepatocytes was suppressed, but also the activation of apoptosis-executing caspases was prevented (fig. 3b). In contrast, concentrations of 50 to 300 μmol/l of ebselen failed to protect primary mouse hepatocytes against TNF-induced cell lysis in the presence of 400 ng/ml actinomycin D (data not shown).

    Injection of TNF induced an IL-10 response in placebo-treated animals. This TNF-inducible IL-10 was considerably enhanced in animals that had been pretreated with ebselen (fig. 4). Ebselen failed to prevent liver injury caused by stimulation of CD95 (APO-1/Fas) by agonistic antibody injection. Neither the release of plasma transaminases nor the activation of apoptosis-transducing caspases was affected in this model of receptor-mediated liver injury (fig. 5). From this observation, we conclude that protection of mice against cytokine-mediated apoptotic liver injury by ebselen is specific for TNF-induced apoptosis.

    Since mice unable to express the nuclear transcription factor NFκB develop massive apoptotic liver degeneration (Beg et al., 1995), we checked the influence of ebselen on hepatic NFκB activation. Gelshift assays shown in figure6 demonstrate that after eight hours, NFκB was increased 3.2-fold (S.E.M.: ± 1, n = 3) in livers of ebselen-treated mice compared to ConA-treated controls. This finding suggests that the interference of ebselen with a key regulatory pathway of inflammatory response genes might provide a mechanistic interpretation of the pharmacological properties of the drug observed in our different in vivo models.

    Figure 6
    View larger version:
    Figure 6

    Upregulation by ebselen of hepatic NF-κB activation in mice injected with ConA. NF-κB DNA binding activities were determined by EMSA (see materials and methods) performed with nuclear extracts of livers from untreated control animals (C) and livers taken from animals 1 hr (left panel) or 8 hr (right panel) after ConA (25 mg/kg) alone or ConA + ebselen (600 mg/kg). The figure shows one out of three independent experiments.

    Previous SectionNext Section

    Discussion

    The overactivation of the specific immune system is a key phenomenon that underlies the initiation of autoimmune disorders or allograft rejection. In general, primary information on the in vivo potential of drugs directed against these pathophysiological processes can be gained from small animal models of the type described in this study. It is a particular property of the models used here that a hyperinflammatory response of lymphocytes or monocytic cells including hepatic Kupffer cells ultimately results in the release of high amounts of TNF independent of the mode of initiation. When bound to its type I receptor, TNF activates a signaling pathway that destroys liver parenchymal cells by an apoptotic mechanism (Gantner et al., 1995a; Leist et al., 1995; Leist et al., 1994). Clinically, such conditions of overactivation leading to apoptotic tissue destruction may occur, for instance, in pathological autoimmune processes. Although only little functional data for the significance of Th1 cytokines such as TNF and IFNγ for hepatotoxicity are available for human autoimmune liver disease, it has been described that CD4+ T cells represent the predominant population of the liver-infiltrating T cells in autoimmune hepatitis (Löhr et al., 1994; Löhr et al., 1996). Moreover, Th1-like cytokines correlating with elevated transaminase levels have been detected in sections of pathological hepatic tissue (Hussain et al., 1994).

    Here we report that ebselen inhibited hepatic destruction in mouse models of acute liver failure induced either by the polyclonal T cell-stimulus ConA in nonsensitized mice, or by SEB in mice under the condition of hepatic transcription arrest by GalN. Ebselen protected the animals against SEB at a dose of 50 mg/kg ebselen, while a dose three times greater than that was needed to protect against ConA. It also seemed that the prevention of T cell-activation by ebselen was more efficient in the SEB-model than in the ConA model, since ebselen significantly suppressed the production of the T cell-cytokines IFNγ and IL-4. Although high doses of ebselen significantly reduced the release of IL-4 in the ConA-model, the considerable amounts of circulating IFNγ remained unaffected. It is important to note that, in contrast to GalN/SEB-induced hepatotoxicity, ConA-induced hepatic damage is also mediated by IFNγ, which is known to act synergistically with TNF in this model (Küsters et al., 1996). This might explain that higher doses of ebselen were needed to inhibit ConA-induced cytokine production and toxicity than in the SEB model.

    In both models of T cell-mediated liver injury the release of the proinflammatory cytokine TNF was downregulated, whereas the release of the anti-inflammatory mediator IL-10 was enhanced. Since both, TNF and IL-10, are released very early after challenge, it is likely that they affect each other’s expression. Indeed, recently such a mechanism of an autoregulatory feedback loop involving both secreted and membrane-bound forms of IL-10 and TNFα has been observed in vitro (Parry et al., 1997). From observations made in a macrophage-mediated experimental murine model, we concluded that a mutual downregulation of Il-10 and TNFα takes also place in vivo (Barsig et al., 1995). Indeed, the attenuation of ConA- or SEB-induced TNF release by ebselen was accompanied by an augmentation of its counterplayer IL-10. Notably, ebselen also potentiated the release of IL-10 following administration of recombinant TNF. Furthermore, ebselen significantly reduced the release of transaminases and prevented activation of apoptosis-executing caspases. Hence, the drug might have augmented IL-10 production induced by additional TNF-inducible pro-inflammatory mediators. An alternative explanation for the enhanced IL-10 production in the ebselen-protected, TNF-challenged animals can be given by assuming a yet unknown interference with TNF-signaling involved in the regulation of IL-10 release.

    However, in contrast to TNF toxicity, ebselen failed to reduce CD95-mediated activation of caspases and liver destruction. Our interpretation of these experiments is that although ebselen in its free form reacts with protein thiols, it does not lead to inhibition of hepatic caspases in vivo by blocking the active site cysteine of these proteases. This is in accordance with the fact that 99% of plasma ebselen is bound to serum albumin. The differential effects observed here in the two models of receptor-mediated hepatic apoptosis suggest that target cell-directed effects of ebselen are specific for TNF- but not for CD95-induced cell death.

    A direct effect of ebselen on the proliferation of T lymphocytes, which are the source for IFNγ and IL-4, and which are also able to produce TNF, seems unlikely, since ebselen failed to reduce the proliferation of splenocytes in vitro stimulated by either ConA or SEB. Thus it appears that ebselen interferes with the expression of T cell-derived pro-inflammatory cytokines in vivo rather than with proliferation signals.

    A previous paper with an immunopharmacological approach showed that ebselen inhibited the transendothelial migration of human polymorphonuclear leukocytes in vitro (Gao and Issekutz, 1993). A subsequent study demonstrated in a complex experimental set-upin vivo in rats that ebselen inhibited migration of spleen-derived T lymphocyte to sites of inflammation of adjuvant arthritis, dermal inflammation induced by cytokines (IFNγ, TNF) or by cytokine-inducing stimuli (poly I:C, LPS) (Gao and Issekutz, 1994). Even though lymphokine release was not monitored as an activation parameter in that study, the data clearly show that ebselen modulates leukocyte and lymphocyte activation related to the nonspecific as well as to the specific immune responses.

    The manifold pharmacological effects of ebselen can hardly be explained by any of its biochemical actions alone. Neither the so-called antioxidant property of the drug (which would require regeneration of the electron donor to make this mechanism work in vivo) nor the GSH peroxidase-like catalytic activity in the presence of GSH (reviewed in Schewe, 1995) are likely to account for the effects of ebselen on the immune system. However, our finding of an upregulation of NFκB activation by the drug offer a plausible mechanistic interpretation of the novel findings reported here. The interaction with the NFκB system would also allow to explain the specificity of the inhibition of TNF-mediated apoptosis, since the transcription factor is thought to play no central role in CD95 apoptosis.

    Taken together, ebselen seems to unite several pharmacodynamic properties that had not been recognized before (cf. fig.7). On the one hand, it shifts the balance of cytokine release following stimulation of antigen presenting cells, e.g. macrophages or dendritic cells (Gantner et al., 1996; Bhardwaj et al., 1992), and T cells in favour of the enhanced release of the anti-inflammatory cytokine IL-10. On the other hand, it renders protection by an NFκB-dependent mechanism against TNF itself, which is associated with upregulation of IL-10 release. This pharmacological profile suggests that ebselen has a promising potential in the therapy of diseases that are characterized by an initial overactivation of the immune system. Potential indications may include chronic inflammatory conditions such as polyarthrtitis, autoimmune diseases such as Crohn’s disease or Colitis ulcerosa, but also prevention of transplant rejection or preservation of explanted organs.

    Figure 7
    View larger version:
    Figure 7

    Proposed mechanism of the pharmacology of ebselen in experimental T cell-dependent liver injury, displaying potentiation of IL-10 production, inhibition of TNF release, and prevention of TNF toxicity via upregulation of NF-κB transactivation and suppression of caspase activation (MΦ/KC: macrophage/Kupffer cell, HC: hepatocyte).

    Previous SectionNext Section

    Acknowledgments

    We dedicate this work to Professor Leopold Flohé, Braunschweig/Germany on the occasion of his 60th birthday. Since he was the doctoral supervisor of Professor A.W. (his doctoral students are G.K. and H.H.), who in turn was the supervisor of Professor G.T. (her doctoral student is S.K.), this deeply respectful reference of three generations of descendants is addressed to our scientific father and to the grandfather at the same time. The authors wish to thank Dr. G. R. Adolf from Bender & Co., Vienna, Austria, for kindly providing recombinant murine TNF. We are indebted to Dr. Angelika Vollmar, München, for enabling the NF-κB gelshift assays. The excellent technical assistance of Ulla Gehert is gratefully acknowledged.

    Previous SectionNext Section

    Footnotes

    • Send reprint requests to: Dr. Albrecht Wendel, Biochemical Pharmacology, University of Konstanz, PO Box 5560 M 667, D-78457 Konstanz, Germany. E-mail: Albrecht.Wendel{at}uni-konstanz.de

    • 1 This study was supported by Daiichi Seiyaku Co. Ltd., Tokyo, and by the Deutsche Forschungsgemeinschaft, grants We 686/18–1 and Ti 169/4–1.

    • Abbreviations:
      ALT
      alanine aminotransferase
      AST
      aspartate aminotransferase
      ConA
      concanavalin A
      GalN
      d-galactosamine
      IFNγ
      interferon gamma
      IL
      interleukin
      SDH
      sorbitol dehydrogenase
      SEB
      Staphylococcusenterotoxin B
      Th
      T-helper cell
      TNF
      tumor necrosis factor-α
      z-VAD-fmk
      benzoyloxycarbonyl-Val-Ala-Asp-fluoromethylketone.
      • Received December 22, 1997.
      • Accepted July 7, 1998.
    Previous Section
     

    References

    1. ↵
      1. Barsig J,
      2. Küsters S,
      3. Vogt K,
      4. Volk H-D,
      5. Tiegs G,
      6. Wendel A
      (1995) Lipopolysaccharide-induced interleukin-10 in mice: Role of endogenous tumor necrosis factor-α. Eur J Immunol 25:2888–2893.
      MedlineWeb of Science
    2. ↵
      1. Beg AA,
      2. Sha WC,
      3. Bronson RT,
      4. Gosh S,
      5. Baltimore D
      (1995) Embryonic lethality and liver degeneration in mice lacking the RelA component of NF-kappa B. Nature 376:167–170.
      CrossRefMedline
    3. ↵
      1. Bhardwaj N,
      2. Friedman SM,
      3. Cole BC,
      4. Nisanian AJ
      (1992) Dendritic cells are potent antigen- cells for microbial superantigens. J Exp Med 175:267–273.
      Abstract/FREE Full Text
    4. ↵
      1. Brüne B,
      2. Diewald B,
      3. Ullrich V
      (1991) Ebselen affects calcium homeostasis in human platelets. Biochem Pharmacol 41:1805–1811.
      CrossRefMedline
    5. ↵
      1. Cotgreave LA,
      2. Duddy SK,
      3. Kass GEN,
      4. Thompson D,
      5. Moldéus P
      (1989) Studies on the anti-inflammatory activity of ebselen. Ebselen interferes with granulocyte oxidative burst by dual inhibition of NADPH oxidase and protein kinase C? Biochem Pharmacol 38:649–656.
      CrossRefMedline
    6. ↵
      1. Dimmeler S,
      2. Brüne B,
      3. Ullrich V
      (1991) Ebselen prevents inositol (1,4,5)-trisphosphate binding to its receptor. Biochem Pharmacol 42:1151–1153.
      CrossRefMedline
    7. ↵
      1. Gantner F,
      2. Leist M,
      3. Jilg S,
      4. Germann PG,
      5. Freudenberg MA,
      6. Tiegs G
      (1995a) Tumor necrosis factor-induced hepatic DNA-fragmentation as an early marker of T cell-dependent liver injury in mice. Gastroenterology 109:166–176.
      CrossRefMedlineWeb of Science
    8. ↵
      1. Gantner F,
      2. Leist M,
      3. Lohse AW,
      4. Germann PG,
      5. Tiegs G
      (1995b) Concanavalin A-induced T-cell mediated hepatic injury in mice: the role of tumor necrosis factor. Hepatology 21:190–198.
      CrossRefMedlineWeb of Science
    9. ↵
      1. Gantner F,
      2. Leist M,
      3. Küsters S,
      4. Vogt K,
      5. Volk H-D,
      6. Tiegs G
      (1996) T cell stimulus-induced crosstalk between lymphocytes and liver macrophages results in augmented release of tumor necrosis factor. Exp Cell Res 229:137–146.
      CrossRefMedline
    10. ↵
      1. Gantner F,
      2. Küsters S,
      3. Wendel A,
      4. Hatzelmann A,
      5. Schudt C,
      6. Tiegs G
      (1997) Protection from T cell-mediated murine liver failure by phosphodiesterase inhibitors. J Pharmacol Exp Ther 280:53–60.
      Abstract/FREE Full Text
    11. ↵
      1. Gao J-X,
      2. Issekutz AC
      (1993) The effect of ebselen on polymorphonuclear leukocyte and lymphocyte migration to inflammatory reactions in rats. Immunopharmacol 25:239–251.
      CrossRefMedline
    12. ↵
      1. Gao J-X,
      2. Issekutz AC
      (1994) The effect of ebselen on T-lymphocyte migration to arthritic joints and dermal inflammatory reactions in the rat. Int J Immunopharmacol 16:279–287.
      CrossRefMedline
    13. ↵
      1. Hussain JJ,
      2. Mustafa A,
      3. Gallati H,
      4. Mowat AP,
      5. Mieli-Vergani G,
      6. Vergani D
      (1994) Cellular expression of tumor necrosis factor-α and interferon-γ in the liver biopsies of children with chronic liver disease. J Hepatol 21:816–821.
      CrossRefMedlineWeb of Science
    14. ↵
      1. Kühn R,
      2. Löhler J,
      3. Rennick D,
      4. Rajewsky K,
      5. Müller W
      (1993) Interleukin-10-deficient mice develop chronic enterocolitis. Cell 75:263–274.
      CrossRefMedlineWeb of Science
    15. ↵
      1. Künstle G,
      2. Leist M,
      3. Uhlig S,
      4. Revesz L,
      5. Feifel R,
      6. MacKenzie A,
      7. Wendel A
      (1997) ICE-protease inhibitors block murine liver injury and apoptosis caused by CD95 or by TNF-alpha. Immunol Lett 55:5–10.
      CrossRefMedlineWeb of Science
    16. ↵
      1. Küsters S,
      2. Gantner F,
      3. Künstle G,
      4. Tiegs G
      (1996) Interferon gamma plays a critical role in T cell-dependent liver injury in mice initiated by concanavalin A. Gastroenterology 111:462–471.
      CrossRefMedlineWeb of Science
    17. ↵
      1. Küsters S,
      2. Tiegs G,
      3. Alexopoulou L,
      4. Pasparakis M,
      5. Douni E,
      6. Künstle G,
      7. Bluethmann H,
      8. Wendel A,
      9. Pfizenmaier K,
      10. Kollias G,
      11. Grell M
      (1997) In vivo evidence for a functional role of both TNF receptors and transmembrane TNF in experimental hepatits. Eur J Immunol 27:2870–2875.
      MedlineWeb of Science
    18. ↵
      1. Leist M,
      2. Gantner F,
      3. Bohlinger I,
      4. German PG,
      5. Tiegs G,
      6. Wendel A
      (1994) Murine hepatocyte apoptosis induced in vitro and in vivo by TNFα requires transcriptional arrest. J Immunol 153:1778–1787.
      Abstract
      1. Leist M,
      2. Gantner F,
      3. Bohlinger I,
      4. Tiegs G,
      5. Germann PG,
      6. Wendel A
      (1995) Tumor necrosis factor-induced hepatocyte apoptosis precedes liver failure in experimental murine shock models. Am J Pathol 146:1220–1234.
      Abstract
    20. ↵
      1. Leist M,
      2. Gantner F,
      3. Jilg S,
      4. Wendel A
      (1995) Activation of the 55 kDa TNF receptor is necessary and sufficient for TNF-induced liver failure, hepatocyte apoptosis and nitrite release. J Immmunol 154:1307–1316.
      Abstract
    21. ↵
      1. Löhr HF,
      2. Schlaak JF,
      3. Gerken G,
      4. Fleischer B,
      5. Dienes H-P,
      6. Meyer zum Büschenfelde K-H
      (1994) Phenotypical analysis and cytokine release of liver-infiltrating and peripheral blood T lymphocytes from patients with chronic hepatitis of different etiology. Liver 14:161–166.
      Medline
    22. ↵
      1. Löhr HF,
      2. Schlaak JF,
      3. Lohse AW,
      4. Böcher WO,
      5. Arenz M,
      6. Gerken G,
      7. Meyer zum Büschenfelde K-H
      (1996) Autoreactive CD4+ LKM-specific and anticlonotypic T-cell responses in LKM-1 antibody-positive autoimmune hepatitis. Hepatology 24:1416–1421.
      CrossRefMedlineWeb of Science
    23. ↵
      1. Louis H,
      2. Le Moine O,
      3. Peny MO,
      4. Gulbis B,
      5. Nisol F,
      6. Goldman M,
      7. Deviere J
      (1997) Hepatoprotective role of interleukin 10 in galactosamine/lipopolysaccharide mouse liver injury. Gastroenterology 112:935–942.
      CrossRefMedlineWeb of Science
    24. ↵
      1. Maiorino M,
      2. Roveri A,
      3. Coassin M,
      4. Ursini F
      (1988) Kinetic mechanism and substrate specificity of glutathione peroxidase activity of ebselen (PZ 51). Biochem Pharmacol 37:2267–2271.
      CrossRefMedline
    25. ↵
      1. Marrack P,
      2. Blackman M,
      3. Kushnir E,
      4. Kappler J
      (1990) The toxicity of staphylococcal enterotoxin B in mice is mediated by T cell. J Exp Med 171:455–464.
      Abstract/FREE Full Text
    26. ↵
      1. Miethke T,
      2. Wahl C,
      3. Heeg K,
      4. Echtenachter B,
      5. Krammer PH,
      6. Wagner H
      (1992) T cell-mediated lethal shock triggered in mice by the superantigen Staphylococcal Enterotoxin B: Critical role of tumor necrosis factor. J Exp Med 175:91–98.
      Abstract/FREE Full Text
    27. ↵
      1. Mizuhara H,
      2. O′Neill E,
      3. Seki N,
      4. Ogawa T,
      5. Kusunoki C,
      6. Otsuka K,
      7. Satoh S,
      8. Miwa M,
      9. Senoh H,
      10. Fujiwara H
      (1994) T cell activation-associated hepatic injury: mediation by tumor necrosis factor and protection by interleukin 6. J Exp Med 179:1529–1537.
      Abstract/FREE Full Text
    28. ↵
      1. Müller A,
      2. Cadenas E,
      3. Graf P,
      4. Sies H
      (1984) A novel biologically active seleno-organic compound-I. Glutathione peroxidase-like activity in vitro and antioxidant capacity of PZ 51 (ebselen). (Biochem Pharmacol) 33:3235–3239.
      CrossRefMedlineWeb of Science
    29. ↵
      1. Nagaki M,
      2. Muto Y,
      3. Ohnishi H,
      4. Yasuda S,
      5. Sano K,
      6. Naito T,
      7. Maeda T,
      8. Yamada T,
      9. Moriwaki H
      (1994) Hepatic injury and lethal shock in galactosamine-sensitized mice induced by the superantigen staphylococcal enterotoxin B. Gastroenterology 106:450–458.
      MedlineWeb of Science
    30. ↵
      1. Parry SL,
      2. Sebbag M,
      3. Feldmann M,
      4. Brennan FM
      (1997) Contact with T cells modulates IL-10 production. Role of T cell membrane TNF-α. J Immunol 158:3673–3681.
      Abstract
    31. ↵
      1. Ramakrishnan N,
      2. Kalinich JF,
      3. McClain DE
      (1996) Ebselen inhibition of apoptosis by reduction of hydroperoxides. Biochem Pharmacol 51:1443–1451.
      CrossRefMedlineWeb of Science
    32. ↵
      1. Rodriguez I,
      2. Matsuura K,
      3. Ody C,
      4. Nagata S,
      5. Vassalli P
      (1996) Systemic injection of a tripeptide inhibits the intracellular activation of CPP32-like proteases in vivo and fully protects mice against fas-mediated fulminant liver destruction and death. J Exp Med 184:2067–2072.
      Abstract/FREE Full Text
    33. ↵
      1. Safayhi H,
      2. Tiegs G,
      3. Wendel A
      (1985) A novel biologically active seleno-organic compound-V. Inhibition by ebselen (PZ 51) of rat peritoneal neutrophil lipoxygenase. Biochem Pharmacol 34:2691–2694.
      CrossRefMedline
    34. ↵
      1. Schewe C,
      2. Schewe T,
      3. Wendel A
      (1994) Strong inhibition of mammalian lipoxygenases by the antiinflammatory seleno-organic compound ebselen in the absence of glutathione. Biochem Pharmacol 48:65–74.
      CrossRefMedlineWeb of Science
    35. ↵
      1. Schewe T
      (1995) Molecular actions of ebselen - an antiinflammatory antioxidant. Biochem Pharmacol 26:1153–1169.
    36. ↵
      1. Tiegs G,
      2. Wolter M,
      3. Wendel A
      (1989) Tumor necrosis factor is a terminal mediator in galactosamine/endotoxin-induced hepatitis in mice. Biochem Pharmacol 38:627–631.
      CrossRefMedlineWeb of Science
    37. ↵
      1. Tiegs G,
      2. Hentschel J,
      3. Wendel A
      (1992) T cell-dependent experimental liver injury in mice inducible by concanavalin A. J Clin Invest 90:196–203.
    38. ↵
      1. Tiegs G
      (1994) Immunotoxicology of host-response-mediated experimental liver injury. J Hepatol 21:890–903.
      CrossRefMedlineWeb of Science
    39. ↵
      1. Wendel A,
      2. Fausel M,
      3. Safayhi H,
      4. Tiegs G,
      5. Otter R
      (1984) A novel biologically active seleno-organic compound-II. Activity of PZ 51 in relation to glutathione peroxidase. Biochem Pharmacol 33:3241–3245.
      CrossRefMedline
    40. ↵
      1. Wendel A,
      2. Tiegs G
      (1986) A novel biologically active seleno-organic compound-VI. Protection by ebselen (PZ 51) against galactosamine/endotoxin-induced hepatitis in mice. Biochem Pharmacol 35:2115–2118.
      CrossRefMedline
    « Previous | Next Article »Table of Contents

    This Article

    1. JPET December 1, 1998 vol. 287 no. 3 1098-1104
    1. AbstractFree
    2. » Full Text
    3. Full Text (PDF)

    Classifications

    Responses

    1. Submit a response
    2. No responses published

    Navigate This Article

    Current Issue

    1. November 2009, 331 (2)
    1. Current Issue
    1. Alert me to new issues of JPET