Potential conflict of interest: Dr. Franklin owns stock in and is an employee of OSI Pharmaceuticals.
Drug-induced liver injury (DILI) is a challenging problem in drug development and clinical practice. Patient susceptibility to DILI is multifactorial, making these reactions difficult to predict and prevent. Clinical observations have suggested that concurrent bacterial and viral infections represent an important risk factor in determining patient susceptibility to developing adverse drug reactions, although the underlying mechanism is not clear. In the present study, we employed the viral RNA mimetic (polyinosinic-polycytidylic acid [polyI:C]) to emulate viral infection and examined its effect on halothane-induced liver injury. Although pretreatment of mice with polyI:C attenuated halothane hepatotoxicity due to its inhibitory effect on halothane metabolism, posttreatment significantly exacerbated liver injury with hepatocellular apoptosis being significantly higher than that in mice treated with polyI:C alone or halothane alone. The pan-caspase inhibitor z-VAD-fmk suppressed liver injury induced by polyI:C/posthalothane cotreatment, suggesting that the increased hepatocyte apoptosis contributes to the exacerbation of liver injury. Posttreatment with polyI:C also caused activation of hepatic Kupffer cells (KCs) and natural killer (NK) cells and upregulated multiple proapoptotic factors, including tumor necrosis factor-α (TNF-α), NK receptor group 2, member D (NKG2D), and Fas ligand (FasL). These factors may play important roles in mediating polyI:C-induced hepatocyte apoptosis. Conclusion: This is the first study to provide evidence that concurrent viral infection can inhibit cytochrome (CYP)450 activities and activate the hepatic innate immune system to proapoptotic factors. DILI may be attenuated or exacerbated by pathogens depending on the time of infection. (HEPATOLOGY 2009;49:215–226.)
Drug-induced liver injury (DILI) accounts for over 50% of liver failure cases in the United States,1 and is also the most common reason for the withdrawal of U.S. Food and Drug Administration-approved drugs from the pharmaceutical market.2 The prediction and prevention of these reactions have been a challenge due to their relatively low incidence, the lack of a diagnostic standard, and the lack of knowledge of the underlying mechanism. It is also difficult to predict which patients will develop DILI to a given drug because patient susceptibility is likely multifactorial.
One such risk factor may be bacterial or viral infection concurrent with drug treatment. It has been observed that when ampicillin is given to patients with acute infectious mononucleosis, the risk of developing extensive maculopapular pruritic rash increases markedly (approaching 100%).3, 4 It has also been reported that 57% to 83% of patients with acquired immune deficiency syndrome who received cotrimoxazole developed adverse reactions such as fever, rash, neutropenia, thrombocytopenia, and hepatitis,5 while the incidence of these reactions are substantially lower in cotrimoxazole-treated patients without human immunodeficiency virus (HIV) infection. A similar increase in the risk of drug-induced hypersensitivity reactions in patients with HIV infection has been noted for several other drugs, including dapsone, carbamazepine, quinolones, and penicillins.6, 7
In terms of the mechanism of DILI, evidence suggests that the formation of chemically reactive metabolites initiates hepatocyte damage, which in turn, activates the innate immune cells, causing inflammation. Bacterial and viral infection can amplify the inflammatory response by further activation of the innate immune system via Toll-like receptors (TLRs).8 TLRs are a family of innate immune receptors that recognize structurally conserved pathogen-associated molecular patterns of microbial origin.9 Such specific microbial products include lipopolysaccharide (LPS), bacterial lipoproteins, peptidoglycan, bacterial DNA, and viral nucleic acids. TLRs are broadly expressed on innate immune cells, including macrophages, dendritic cells, neutrophils, and natural killer (NK) cells.10, 11 TLR recognition of microbial molecules triggers an inflammatory response, including the production of cytokines, chemokines and adhesion molecules. Polyinosinic-polycytidylic acid (polyI:C) is a viral RNA mimetic that induces immune responses similar to a viral infection.12 PolyI:C increases the cytotoxic effect by macrophages and NK cells and activates T cells.13–15 In the mouse liver, polyI:C treatment causes the recruitment and activation of NK cells, a process dependent on Kupffer cells (KCs) release of interleukin (IL)-12.16 However, the effect of polyI:C in modulating DILI has not been studied.
Halothane causes mild liver injury in approximately 20% of patients,17 and it leads to severe liver injury in a small percentage of patients, that often develops into fulminant liver failure. It is not known why certain individuals are more susceptible to halothane-induced hepatotoxicity, although several risk factors have been identified, including obesity and the female gender.18 We have recently developed a mouse model of halothane-induced liver injury,19 which provides a platform for elucidating susceptibility factors in DILI development. In the present study, we have employed this model to investigate the effect of polyI:C on halothane-induced hepatotoxicity in mice. Although pretreatment with polyI:C inhibited halothane-induced liver injury, a dramatic augmentation of hepatotoxicity was observed in mice administered with polyI:C after halothane challenge (polyI:C/posthalothane) compared with those treated with polyI:C or halothane alone. The data suggest that posttreatment with polyI:C induced a marked increase in hepatocyte apoptosis through the activation of hepatic KC and NK cells. The findings provide evidence that concurrent infection could decrease or increase patient risk of developing DILI depending on the time of infection.
Female BALB/cByJ mice (7-10 weeks of age) were purchased from The Jackson Laboratory (Bar Harbor, ME) and kept in the Center for Laboratory Animal Care at the University of Colorado Health Sciences Center (UCHSC) for 1 week before treatment. All animal experiments were performed in accordance with guidelines from the UCHSC Institutional Animal Care and Use Committee.
Mice were injected intraperitoneally (i.p.) with polyI:C (GE Healthcare Bio-Science Corp., Piscataway, NJ; 50 μg dissolved in 100 μL of phosphate-buffered saline [PBS]) 12 hours or 7 days prior to or 6 hours after i.p. administration of halothane (Halocarbon Labs Inc., Hackensack, NJ; 30 mmol/kg dissolved in 2 mL of olive oil).19 Control mice were treated with either polyI:C alone or halothane alone.
To deplete NK cells, mice were injected intravenously (i.v.) with anti-AsGM1 polyclonal antibody (100 μL; Wako Chemical USA, Richmond, VA) 2 days prior to halothane treatment.20, 21 Control mice were administered with the same volume of normal rabbit serum. To deplete hepatic KCs, mice were injected i.v. with liposome-entrapped clodronate (liposome/clodronate; Sigma-Aldrich, St. Louis, MO) 2 days prior to halothane treatment.22 Control mice were injected with empty liposomes. For in vivo inhibition of caspase activities, mice were injected i.p. with the general caspase inhibitor N-CBZ-Val-Aal-Asp(O-Me) fluoromethyl ketone, z-VAD-fmk (z-VAD; BACHEM/Peninsula Laboratories, Inc., San Carlos, CA; 200 μg dissolved in 1% dimethyl sulfoxide [DMSO]) 5 hours after halothane treatment. Control mice were treated with 1% DMSO. The time course for various polyI:C posttreatments is summarized in Fig. 1.
Assessment of Hepatotoxicity.
At 15 hours and 24 hours after halothane treatment, mice were anesthetized and blood was collected by retroorbital puncture. Blood samples were allowed to clot at 4°C before sera were prepared by centrifugation at 10,000g for 20 minutes. Serum alanine aminotransferase (ALT) levels were measured using a diagnostic assay kit (Teco Diagnostics, Anaheim, CA) following the manufacturer's instructions. At 24 hours after halothane administration, the animals were sacrificed and the livers removed. Liver sections were fixed in 10% formaldehyde overnight before being transferred into 70% ethanol solution. Paraffin-embedded liver sections were mounted onto glass slides and stained with hematoxylin and eosin (H&E; Department of Pathology, UCHSC).
Terminal deoxynucleotidyl transferase–mediated dUTP nick-end labeling (TUNEL) assays were performed using the TACS TdT DAB in situ apoptosis detection kit (R&D Systems, Inc., Minneapolis, MN) following the manufacturer's instructions. Briefly, liver tissue sections were deparaffinized, rehydrated, and incubated in 50 μL proteinase K for 15 minutes at room temperature. After blocking endogenous peroxidase activity using 0.3% H2O2 in methanol (vol/vol), the tissue sections were incubated with terminal deoxynucleotidyl transferase (TdT) for 1 hour at 37°C. Subsequently, they were incubated with peroxidase conjugated streptavidin for 10 minutes and with diaminobenzidine (DAB) solution for 5 minutes at room temperature. The tissue sections were counterstained using Methyl Green Solution (R&D Systems).
Caspase-3 Activity Assay.
Liver tissue samples were homogenized in ice-cold Tris buffer (100 mM, pH 7.5) containing 250 mM sucrose, 2 mM ethylene diamine tetraacetic acid, and a cocktail of protease inhibitors (1:100; Sigma). Caspase-3 activities were measured by using a fluorogenic substrate Ac-DEVD-AMC (Biomol International, L.P., Plymouth Meeting, PA) as described.23 Reactions were performed at 37°C for 1 hour, and fluorescence intensity was monitored using a Packard FluoroCount plate reader (Packard Instrument, Meriden, CT). Substrate autofluorescence was subtracted from each value and specific activities were calculated based on a standard curve of aminomethyl coumarin (Sigma).
Hepatic Leukocyte Isolation and Flow Cytometric Analysis.
Hepatic leukocytes were isolated following a previously described method, with slight modification.24, 25 Mice were anesthetized and the liver was perfused in situ with Hank's balanced salt solution (HBSS) prewarmed at 37°C for 5 minutes. Single-cell suspensions were filtered through a 100 μm cell strainer (BD Falcon, Bedford, MA) and centrifuged at 300g for 5 minutes. The pellet was resuspended in 15 mL of 35% Percoll (Sigma) containing 50 U/mL of heparin (Baxter Healthcare Corporation, Deerfield, IL) and centrifuged at 500g for 15 minutes. The resulting pellet was collected and resuspended in 1.5 mL of red blood cell lysing buffer (Sigma) for 5 minutes. The cells were then washed in HBSS solution containing 0.6% acid citrate-dextrose (ACD-A, Sigma) and 0.5% BSA. Total viable hepatic leukocytes were counted by trypan blue exclusion.
The NK cell population (DX5+CD3−) in freshly isolated hepatic leukocytes was identified by staining the cells with phycoerythrin (PE)-conjugated anti-pan-NK cells (anti-DX5, clone DX5; eBioscience, San Diego, CA) and allophycocyanin-conjugated anti-CD3 (clone 145-2C11; eBioscience) antibodies for 30 minutes on ice. To prevent nonspecific binding, cells were blocked with normal rat serum (Sigma) and anti-mouse FcγR II/III (clone 93; eBioscience) for 5 minutes at 4°C. In some experiments, the cells were stained with PE-anti-NK receptor group 2, member D (NKG2D) (clone CX5; eBioscience) and PE-anti-Fas ligand (FasL) (clone MFL3; eBioscience) to determine the expression of NKG2D and FasL on hepatic leukocytes. The cells were analyzed on a FACSCalibur cytometer using CellQuest software (BD Biosciences, San Jose, CA). The data were further analyzed using FlowJo software (Tree Star Inc., Ashland, OR).
Tumor Necrosis Factor-Alpha Measurement and Neutralization.
Blood was collected and serum samples were prepared as described above. Tumor necrosis factor-alpha (TNF-α) levels were determined by sandwich enzyme-linked immunosorbent assay using capture and detection antibody pairs according to the manufacturer's instructions (R&D Systems).
In vivo neutralization of TNF-α was accomplished by injecting (i.v.) female BALB/cByJ mice with 200 μg of anti-mouse TNF cV1q antibody (CNTO 2213; kindly provided by Dr. David J. Shealy, Centocor Inc., Radnor, PA) 10 hour posthalothane challenge (4 hour postpolyI:C treatment). Control mice were injected i.v. with rat/mouse immunoglobulin G (IgG) 2a/kappa (CNTO 1322; Centocor Inc.), a direct isotype match of the neutralizing antibody.
Total RNA was isolated from 20 mg of frozen liver tissue using RNeasy Mini Kits (Qiagen, Valencia, CA) as instructed by the manufacturer. RNA (1 μg) was reverse transcribed to cDNA at 42°C for 60 minutes using Superscript III Reverse Transcriptase (Invitrogen, Carlsbad, CA) with oligo(dT) primers (Invitrogen).
The resultant cDNA fragments was amplified for 31 cycles using Platinum Taq polymerase (Invitrogen) and gene-specific primers for TNF-α (sense 5′-TTCTGTCTACTGAACTTCGGGGGATCGGTCC-3′; antisense 5′-GTATGAGATAGCAAATCGGCTGACGGTGTG- GG-3′), NKG2D (sense 5′-GCATTGATTCGTGATCGAAA-3′; antisense 5′-GCCACAGTAGCCCTCTC- TTG-3′), FasL (sense 5′-CACAAATCTGTGGCTACCG-3′; antisense 5′-GCCCATATCTGTCCAGTAG-3′), IL-12 (sense 5′-ATGTGTCCTCAGA- AGCTAAC-3′; antisense 5′-TCCTAGGATCGGACCCTG-3′), and β-actin (sense 5′-TCTTGGGTATGGAATCCTGTGGCA-3′; antisense 5′-ACTCCTGCTT- GCTGATCCACATCT-3′). All polymerase chain reaction (PCR) products were resolved on 1.5% agarose gels and visualized by ethidium bromide staining. mRNA expression levels were determined by normalizing band intensities relative to the levels of β-actin expression using Adobe PhotoShop 6.0.
Immunoblot Analysis of trifluoroacetylchloride-Protein Adducts.
Mice were sacrificed 15 hours after halothane treatment and liver homogenates were prepared as described above. Samples (30 μg) were resolved on 12% polyacrylamide gels and transferred to nitrocellulose membranes (Bio-Rad Laboratories Inc., Hercules, CA). Membranes were blocked with 5% (wt/vol) fat-free milk and probed with a rabbit polyclonal anti-trifluoroacetylchloride (TFA) antisera (1:1,000; gift from Dr. Lance Pohl, National Institutes of Health, Bethesda, MD) overnight at 4°C. Membranes were then incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG antibody (1:2,000; Chemicon International Inc., Temcula, CA) for 1 hour at room temperature. Membranes were then exposed to the ECL Plus Western Blotting Detection System (GE Healthcare Bio-Science Corp.), and data were captured using a Storm 860 system (GE Healthcare Bio-Science)
Data are presented as mean ± standard error of the mean. Two-tailed Student t test was used to compare two groups. Comparisons among multiple groups were performed using one-way analysis of variance with a post hoc test of significance between individual groups. Differences were considered significant when P < 0.05.
PolyI:C Affects Halothane-Induced Liver Injury in Mice.
Clinical evidence suggests that concurrent infection and inflammation may be a risk factor in developing DILI.3–7 To reproduce this phenomenon in animal models and to understand the underlying mechanism of the increased risk, we set out to investigate the effect of polyI:C on halothane-induced hepatotoxicity. Female BALB/cByJ mice were used because they were found to be most susceptible to halothane-induced liver injury in our previous studies.19 Mice were injected i.p. with polyI:C at 12 hours or 7 days prior to or 6 hours after halothane treatment. Control mice were treated with polyI:C or halothane alone. Halothane-induced liver injury, evaluated by measuring serum ALT activities, was significantly attenuated by polyI:C given 12 hours prior to halothane challenge. However, it did not affect halothane hepatotoxicity when administered 7 days before halothane treatment (Fig. 2A). In contrast, polyI:C administered 6 hours after halothane challenge significantly worsened halothane-induced hepatic injury, while polyI:C alone did not cause noticeable tissue damage at the dose administered (Fig. 2B). Histological evaluation of liver sections obtained at 24 hours after halothane treatment revealed much greater levels of hepatocyte necrosis in polyI:C/posthalothane cotreated mice than that in mice treated with halothane alone (Fig. 2C).
Numerous studies have demonstrated that cytochrome (CYP) 450 enzymes are suppressed by interferon inducers, such as polyI:C.26–30 CYP450 2E1 is the predominant isoform involved in halothane metabolism. A previous report described significant reductions of messenger RNA (mRNA) and protein levels of CYP450 2E1 at 6 hours and 12 hours, respectively, after polyI:C treatment of rats.27 Consistent with these report, we found that biotransformation of halothane to its reactive intermediate, TFA, was significantly inhibited by treatment with polyI:C at 12 hours, but not at 7 days prior to (data not shown) or 6 hours after halothane challenge (Fig. 2D). These results suggested that the inhibitory effect of polyI:C on CYP450 is transient, and that the increase in halothane hepatotoxicity caused by polyI:C posttreatment was not due to alterations of CYP450 activities, halothane metabolism to TFA or the formation of TFA protein adducts.
To investigate the mechanism by which polyI:C posttreatment potentiates halothane-induced liver injury, hepatocyte apoptosis was examined immunohistochemically using the TUNEL assay. In contrast to very few apoptotic cells detected in the liver of halothane alone treated mice, numerous apoptotic cells were found in liver sections obtained from mice cotreated with polyI:C/posthalothane (Fig. 3A). Furthermore, significantly higher caspase-3 activities were observed in the liver homogenates prepared from polyI:C/posthalothane-treated mice than mice treated with polyI:C alone or halothane alone (Fig. 3B). To examine whether the increased hepatocyte apoptosis contributes to the exacerbation of liver injury observed in polyI:C/posthalothane-treated mice, female BALB/cByJ mice were injected i.p. with a pan-caspase inhibitor, z-VAD-fmk (i.p. 200 μg/mouse, dissolved in 1% DMSO), 5 hours after halothane treatment (1 hour prior to polyI:C administration). Control mice were treated with 1% DMSO. The data demonstrated that while z-VAD-fmk significantly decreased ALT levels in mice cotreated with polyI:C/posthalothane (Fig. 3C), z-VAD-fmk did not affect the level of liver injury in mice treated with halothane alone (data not shown). These data suggest that polyI:C induced caspase-dependent hepatocellular apoptosis, thereby causing an aggravation of liver injury initiated by halothane.
To identify candidate apoptotic factors that could mediate polyI:C/posthalothane-induced hepatocyte apoptosis, hepatic mRNA expression levels of TNF-α, NKG2D and FasL were determined and compared among mice treated with polyI:C, halothane, or polyI:C/posthalothane. The mRNA expressions of TNF-α and NKG2D in the liver were induced by both polyI:C alone and polyI:C/posthalothane cotreatment (Fig. 4A,B). In contrast, hepatic mRNA expression levels of FasL were markedly higher in polyI:C/posthalothane-treated mice compared with those in mice treated with either polyI:C alone or halothane alone (Fig. 4A,B). Liver mononuclear cells (LMNCs), which include NK, NKT, and T cells, are a major cellular source of NKG2D and FasL;20 therefore, these cells were isolated and analyzed by flow cytometry. The expression levels of NKG2D and FasL on LMNCs isolated from polyI:C/posthalothane-treated mice were elevated compared with those in mice treated with polyI:C or halothane alone (Fig. 5A). While the hepatic mRNA level of FasL was dramatically increased in polyI:C/posthalothane cotreated mice, FasL protein level only slightly higher than that in mice treated with polyI:C or halothane alone. This result suggests that, aside from LMNCs, other cells in the liver, such as KCs, may express FasL. Serum levels of TNF-α were also significantly increased in polyI:C/posthalothane-treated mice compared with those in mice treated with halothane alone (Fig. 5B). TNF-α production was not detectable in the sera of mice treated with polyI:C alone. These results indicate that TNF-α, NKG2D, and FasL may be involved in mediating polyI:C/posthalothane-induced hepatocyte apoptosis.
Hepatic KCs and NK Cells Mediate PolyI:C-Induced Increase in Halothane Hepatotoxicity.
PolyI:C can activate a variety of innate immune cells, including macrophages and NK cells,11, 31–36 and these cells represent major sources of the expression of proapoptotic factors. Therefore, we hypothesized that polyI:C activates hepatic KC and NK cells to produce proapoptotic factors, which mediate polyI:C-induced exacerbation of halothane hepatotoxicity. While the number of NK cells (DX5+CD3−) in halothane-treated mice (4%; Fig. 6A) was similar to that in naive mice (5%; data not shown), polyI:C and polyI:C/posthalothane treatments resulted in significant increases in the numbers of DX5+CD3− NK cells within the liver (Fig. 6A). This is consistent with a published study demonstrating that polyI:C treatment of mice causes recruitment and activation of NK cells within the liver.16 To determine the role of NK cells in polyI:C/posthalothane-induced liver injury, female BALB/cByJ mice were injected i.v. with anti-AsGM1 antisera to deplete NK cells (AsGM1 is a glycolipid that functions as a bacterial receptor).20, 21 Control mice were injected with normal rabbit serum. Two days later, both groups of mice were cotreated with polyI:C/posthalothane. A nearly complete depletion of hepatic NK cells in anti-AsGM1-treated mice was confirmed by flow cytometric analysis (Fig. 6B). As a result, NKG2D-expression and FasL-expression on CD3-negative LMNCs, which consisted mainly of NK cells, also disappeared (Fig. 6B). Hepatic mRNA expression levels of NKG2D and FasL were also significantly reduced in NK cell–depleted mice compared with control mice, even though TNF-α levels were unchanged (Fig. 6C). These findings suggest that NK cells represent a major cellular source of NKG2D and FasL, but not TNF-α, in the liver. Compared with control mice, serum ALT activities were significantly decreased in NK cell–depleted mice at 24 hours, but not 15 hours posthalothane treatment (Fig. 7A). NK cell depletion did not affect hepatotoxicity induced by halothane alone (Fig. 7A). These results suggest that polyI:C-induced exacerbation of halothane hepatotoxicity was partially attributable to NK cells.
Macrophages express TLR3 that recognizes polyI:C and mediates its stimulatory effects.11, 31, 35 PolyI:C-induced activation of NK cells in the liver is dependent on IL-12 production by hepatic KCs.16 To determine whether KCs represent another population of hepatic innate immune cells that mediate polyI:C-induced aggravation in halothane hepatotoxicity, KCs were depleted by i.v. injection of liposome/clodronate.22 Control mice were injected with empty liposomes. After 2 days, mice were treated with halothane alone or cotreated with polyI:C/posthalothane and liver injury was assessed by measuring serum ALT activities. Although KC depletion did not affect liver injury induced by halothane alone, the absence of KCs significantly decreased polyI:C/posthalothane-induced ALT levels, at both 15 hours and 24 hours posthalothane treatment, approaching those observed in mice treated with halothane alone by 24 hours (Fig. 7B). KC-depletion also nearly abolished both hepatic mRNA expression and serum levels of TNF-α in mice treated with polyI:C/halothane (Fig. 8A,B). To determine the role of TNF-α in polyI:C-induced aggravation of halothane hepatotoxicity, female BALB/cByJ mice were injected i.v. with an anti-TNF-α neutralizing antibody at 10 hours after halothane treatment (4 hours postpolyI:C challenge), while control mice were treated with an isotype control antibody. TNF-α neutralization significantly reduced ALT levels at 24 hours, but not 15 hours, posthalothane challenge (Fig. 8C).
The data also demonstrated that liposome/clodronate treatment not only depleted KCs, but also significantly inhibited the recruitment of NK cells into the liver (Fig. 9A). Consistent with a previous study demonstrating that the NK cell recruitment and activation induced by polyI:C was dependent on KC stimulation and production of IL-12,16 we observed that hepatic IL-12 message levels diminished in KC-depleted mice (Fig. 9B). These data may explain the greater suppression of polyI:C/posthalothane-induced ALT levels caused by liposome/clodronate treatment than by NK depletion using anti-AsGM1 antibody. In addition, due to the observation that neutralization of TNF-α only abrogated the role of KCs, but not NK cells, the inhibitory effect of TNF-α neutralization on polyI:C/posthalothane-induced liver injury was weaker than that of liposome/clodronate treatment.
Clinical observations have suggested that concurrent bacterial and viral infections may increase patient risk of developing drug-induced adverse reactions,3–7 although the underlying mechanism is not clear. Several studies have demonstrated that cotreatment of rats with LPS and ranitidine, diclofenac, or trovafloxacin could augment hepatotoxicity caused by either LPS or the drug alone.37–41 In these studies, stimulation of the coagulation system and/or increase in the recruitment and activation of neutrophils were observed, which may account for the aggravation in hepatic injury.37, 38, 41 While hepatic innate immune cells are the target of bacteria-induced and virus-induced activation via signaling through TLRs, their role in the mechanism of pathogen-induced increases in susceptibility to DILI has not been investigated. The current study examined the effect of polyI:C, a viral RNA mimetic that binds to TLR3, on halothane-induced liver injury in mice.
Our data demonstrated that treatment of BALB/cByJ mice with polyI:C 12 hours prior to halothane administration inhibited halothane-induced liver injury. It is known that polyI:C-activated innate immune cells, such as dendritic cells, macrophages, and NK cells, produce proinflammatory cytokines.10, 11, 31–36 These cytokines, including interferons, have been shown to inhibit CYP450 activity.26–30, 42 Consistent with these reports, we found that TFA-protein adduct formation was significantly decreased by polyI:C pretreatment at 12 hours (data not shown). This effect was transient as polyI:C treatment at 7 days prior to halothane challenge did not inhibit halothane metabolism and had no effect on halothane-induced liver injury (Fig. 2A). These data suggest that polyI:C can modulate DILI through its inhibitory effect on CYP450 activities. This observation is consistent with findings that pretreatment of mice with polyI:C significantly decreases acetaminophen (APAP)-induced hepatotoxicity by inhibition of CYP450 2E1.42
In contrast, treatment with polyI:C 6 hours after halothane challenge dramatically enhanced halothane hepatotoxicity. This is evident by the significantly higher serum ALT activities and increased extent of liver histopathological damage in mice cotreated with polyI:C/posthalothane compare with mice treated with halothane alone (Fig. 2B,C). No obvious liver injury was observed in mice treated with polyI:C alone. A previous study reported that polyI:C caused liver damage;16 however, the dose required to induce hepatotoxicity is eight-fold higher than that used in the present study. Our results demonstrated a synergistic effect of halothane and polyI:C posttreatment on the induction of liver damage. PolyI:C administered 6 hours after halothane challenge did not affect halothane metabolism as immunoblot analysis revealed no significant differences in either the patterns or the levels of TFA-protein adduct formation in the livers of mice treated with halothane alone and polyI:C/posthalothane-treated mice (Fig. 2D).
We found that both the number of apoptotic hepatocytes and caspase-3 activities in the liver tissue homogenates were increased in polyI:C/posthalothane treated mice compared with those in mice treated with polyI:C or halothane alone (Fig. 3A,B). Furthermore, in vivo z-VAD-fmk treatment of polyI:C/posthalothane-challenged mice caused a marked decrease in ALT activities approaching the levels detected in mice treated with halothane alone (Fig. 3C). These data provided strong evidence that the increased hepatocyte apoptosis caused by polyI:C/posthalothane treatment contributed to the exacerbation of liver injury. These data also suggested that hepatocytes were “sensitized” by halothane to become more susceptible to polyI:C-induced apoptosis, as neither polyI:C nor halothane alone caused significant apoptosis. Our hypothesis that halothane “sensitizes” hepatocytes to apoptotic stimuli is supported by studies demonstrating that halothane has a profound inhibitory effect on protein synthesis in hepatocytes.43–46 Numerous studies have demonstrated that inhibition of protein synthesis directly renders hepatocytes more susceptible to apoptotic stimuli. Mice sensitized by D-galactosamine, an inhibitor of protein synthesis, are highly susceptible to hepatic injury upon exposure to LPS, CpG, polyI:C, superantigen, and whole bacteria, all of which induce the production of the apoptotic mediator, TNF-α.47–51 Interestingly, hepatic glutathione (GSH) levels were significantly lower in halothane alone-treated mice (7.8 pmol/μg) than in vehicle-treated controls (16.5 pmol/μg). Depletion of hepatocellular GSH levels is associated with CYP-c release from mitochondria and may dictate cellular susceptibility to apoptotic stimuli.52 In this instance, we would expect that APAP, known to deplete hepatic GSH levels, would also “sensitize” hepatocytes to polyI:C-induced injury. However, treating mice with polyI:C/post APAP challenge did not increase APAP-induced liver injury (data not shown). Thus, GSH depletion is not likely involved in halothane-mediated “sensitization” of hepatocytes to polyI:C-induced apoptosis.
Hepatocytes express TLR3,53, 54 and polyI:C, through engaging TLR3, can activate the caspase cascade and cause apoptosis of hepatocytes.55 Therefore, it is possible that polyI:C directly targets hepatocytes and stimulates apoptotic death, particularly when the cells are “sensitized” to undergo apoptosis due to halothane pretreatment. Given the strong stimulatory effect of polyI:C on the innate immune system, polyI:C may alternatively induce hepatocyte apoptosis through the activation of hepatic innate immune cells to produce proapoptotic factors. NK cells represent a major target of polyI:C, as treatment of mice with polyI:C causes recruitment and activation of hepatic NK cells.16, 56, 57 A recent study demonstrated that DMSO, an organic solvent often used to facilitate the dissolution of pharmacological compounds, could activate hepatic NK and NKT cells and upregulate their expression of cytotoxic effector molecules, such as interferon-γ and granzyme B.58 We found that polyI:C also upregulates the expression of several proapoptotic molecules, including FasL and NKG2D on hepatic NK cells, rendering these cells more susceptible to the induction of apoptosis.16, 20, 21, 59 The ligands for NKG2D include retinoic acid early-inducible transcript 1 (Rae-1) and mouse UL16-binding protein-like transcript 1 (Mult-1), the cellular expression of which is induced under various stress conditions.60, 61 A recent study demonstrated that the expression of Rae-1 and Mult-1 in hepatocytes were markedly increased upon concanavalin-A (Con A) stimulation in hepatitis B virus transgenic mice, and that NKG2D recognition of Rae-1 or Mult-1 contributed to the sensitivity of these mice to Con A–induced liver injury.62 We found that the hepatic mRNA expression levels of Rae-1 and Mult-1 were increased by halothane and polyI:C/posthalothane treatments (data not shown), suggesting the involvement of NKG2D in mediating the cytotoxicity of NK cells.
We witnessed a significantly higher number of NK cells in the livers of mice treated with polyI:C or polyI:C/posthalothane than in mice treated with halothane alone (Fig. 6A). This observation is consistent with previous reports on polyI:C-mediated recruitment and activation of NK cells.16 Our results demonstrate that administration of anti-AsGM1 antibody selectively depleted NK cells. However, this depletion resulted in only a small, but significant, decrease in ALT levels at 24 hours after halothane treatment and no change in ALT activities at 15 hours (Fig. 7A). While anti-AsGM1 antibody administration significantly reduced hepatic message levels of NKG2D and FasL, TNF-α expression, which was also upregulated in polyI:C/posthalothane-treated mice, remain unchanged (Fig. 6C). Therefore, it is possible that TNF-α, perhaps in conjunction with NKG2D and FasL, is involved in mediating polyI:C-induced hepatocyte apoptosis. In this regard, both hepatic mRNA and serum levels of TNF-α were significantly lower in KC-depleted mice compared with control mice upon polyI:C/posthalothane cotreatment (Fig. 8A,B), suggesting that KC are a predominant source of TNF-α in the liver. Furthermore, depletion of KCs by liposome/clodronate suppressed polyI:C/posthalothane-induced liver injury to the levels approaching that observed in mice treated with halothane alone (Fig. 7B). Administration of anti-TNF-α neutralizing antibody also significantly reduced ALT levels in mice treated with polyI:C/posthalothane (Fig. 8C), however, the degree of reduction was much less than that deserved in response to KC depletion. Our data revealed that polyI:C-induced recruitment of NK cells was significantly inhibited when KCs were depleted (Fig. 9A), perhaps due to the lack of IL-12 production in the absence of KCs (Fig. 9B). We postulate that the strongest inhibition of polyI:C/posthalothane-induced liver injury observed in liposome/clodronate-treated mice was due to depletion of both KCs and NK cells, resulting in the reduction of multiple proapoptotic factors. These results suggest that the combination of TNF-α, NKG2D, and FasL are necessary in mediating polyI:C-induced hepatocyte apoptosis and liver injury.
In summary, our studies demonstrated that in comparison with mice treated with halothane alone, cotreatment of mice with polyI:C/posthalothane caused a profound increase in tissue damage. The exacerbation of liver injury correlated with enhanced hepatocyte apoptosis, which appeared to be a synergistic effect of the coadministration of polyI:C/posthalothane. Our findings also support a model in which polyI:C induced activation of hepatic KCs and NK cells enhances the expression of multiple proapoptotic factors, including TNF-α, NKG2D, and FasL, that mediate polyI:C-induced hepatocyte apoptosis. This is the first study to provide evidence that concurrent infection, through the activation of the hepatic innate immune system and subsequent inflammation, increases patient risk of developing DILI.
We thank Dr. Lance Pohl (NIH, Bethesda, MD) for the generous gift of anti-TFA antisera.