Potential conflict of interest: Nothing to report.
Drug-induced liver injury (DILI) is a major safety concern in drug development. Its prediction and prevention have been hindered by limited knowledge of the underlying mechanisms, in part the result of a lack of animal models. We developed a mouse model of halothane-induced liver injury and characterized the mechanisms accounting for tissue damage. Female and male Balb/c, DBA/1, and C57BL/6J mice were injected intraperitoneally with halothane. Serum levels of alanine aminotransferase and histology were evaluated to determine liver injury. Balb/c mice were found to be the most susceptible strain, followed by DBA/1, with no significant hepatotoxicity observed in C57BL/6J mice. Female Balb/c and DBA/1 mice developed more severe liver damage compared with their male counterparts. Bioactivation of halothane occurred similarly in all three strains based on detection of liver proteins adducted by the reactive metabolite. Mechanistic investigations revealed that hepatic message levels of tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β); IL-6, and IL-8 were significantly higher in halothane-treated Balb/c mice compared to DBA/1 and C57BL/6J mice. Moreover, a higher number of neutrophils were recruited into the liver of Balb/c mice upon halothane treatment compared with DBA/1, with no obvious neutrophil infiltration detected in C57BL/6J mice. Neutrophil depletion experiments demonstrated a crucial role for these cells in the development of halothane-induced liver injury. The halothane-initiated hepatotoxicity and innate immune response-mediated escalation of tissue damage are consistent with events that occur in many cases of DILI. In conclusion, our model provides a platform for elucidating strain-based and gender-based susceptibility factors in DILI development. (HEPATOLOGY 2006;44:1421–1431.)
Drug-induced liver injury (DILI) represents a major health problem that challenges not only health care professionals but also the pharmaceutical industry and drug-regulatory agencies. In the United States, DILI accounts for as many as 25% of liver failure cases in intensive care units.1 Due to its association with significant patient morbidity and mortality, DILI is currently the most common cause for the withdrawal of drugs from the pharmaceutical market.2 The key to predicting and preventing DILI is a thorough understanding of the underlying mechanisms. Prospective human clinical trials would be impossible given the unpredictable nature and low incidence of DILI. Retrospective studies may be conducted; however, they will not yield mechanistic information on the development of DILI. In vitro experiments may be used as a means to understand isolated events in DILI development, but these experiments are not likely to provide insights into the complex molecular and cellular mechanisms of DILI. The best approach for these purposes is to perform animal studies.
At present, the only widely studied model of DILI is the acetaminophen (APAP)-induced liver injury model, particularly in mice. Although important information on the mechanism(s) of drug-induced acute inflammatory injury has been obtained using this model,3–10 APAP hepatotoxicity in mice or humans does not encompass the complete clinical spectrum nor all possible mechanistic features of DILI. Another recently developed model, the lipopolysaccharide (LPS)-drug idiosyncrasy model, has begun to uncover some interesting mechanistic insights, but its wider application to the broad spectrum of DILI remains unknown.11–15 Therefore, due to the significant variations in the clinical characteristics and mechanisms of DILI development in patients, it is critical to have several animal models.
The inhalation anesthetic halothane is known to cause both mild and severe forms of hepatotoxicity.16 Mild liver injury, with transient increases in serum aminotransferases, occurs in approximately 20% of patients treated with halothane.17 In a much smaller percentage of patients, subsequent re-exposure to halothane could cause massive hepatocyte necrosis, frequently leading to fulminant liver failure (halothane hepatitis). The guinea pig is the only nonclinical animal species studied to date in which halothane can cause mild liver injury without extensive pretreatment or manipulations, such as exposure to hypoxic conditions.18–21 It has been demonstrated that the metabolism of halothane to the reactive metabolite trifluoroacetylchloride (TFA) is essential in causing hepatotoxicity; however, the inherent ability of metabolizing halothane to TFA cannot explain the interstrain and intrastrain dependent susceptibility of guinea pigs to halothane-induced liver injury,22, 23 suggesting that other mechanisms are also important. Moreover, the detailed mechanisms including the molecular and cellular responses of the innate immune system of the liver have not been studied in the guinea pig model, in part because of the high cost of the animals and the limited availability of various immunological tools. In contrast, mouse immunology is far more advanced for research purposes, and the availability of various strains of wild-type and transgenic mice enables investigation of the genetic elements potentially involved in the mechanism(s) of DILI.
This report describes the development of the first mouse model of halothane hepatotoxicity. Three common strains of mice varied in their susceptibility to halothane toxicity. Our data suggest that the strain-dependent susceptibility is not due to variations in halothane metabolism, but rather to inherent differences in the innate immune responses within the liver. This model resembles the mild form of halothane-induced liver injury observed in approximately 20% of patients treated with halothane, and it provides a platform for elucidating strain-based and gender-based susceptibility factors potentially important in DILI development.
DILI, drug-induced liver injury; APAP, acetaminophen; LPS, lipopolysaccharide; TFA, trifluoroacetylchloride; UCHSC, University of Colorado Health Sciences Center; i.p., intraperitoneal; ALT, alanine aminotransferase; H/E, hematoxylin and eosin; MDA, malondialdehyde; IL, interleukin; TNF, tumor necrosis factor; iNOS, inducible nitric oxide synthase; G3PDH, glyceraldehyde-3-phosphate dehydrogenase; FITC, fluorescein isothiocyanate; PE, phycoerythrin; KC, Kupffer cells; HO, hemoxygenase; NK, natural killer; TLR, toll-like receptor.
Materials and Methods
Animal Treatment and Assessment of Hepatotoxicity.
Female and male Balb/c, DBA/1, and C57BL/6J mice (8-10 weeks of age) were purchased from The Jackson Laboratory (Bar Harbor, Maine) and kept in the Center for Laboratory Animal Care at the University of Colorado Health Sciences Center (UCHSC) for 1 week before treatment. Animals were injected intraperitoneally (i.p.) with 2 mL of olive oil alone or halothane (30 mmol/kg, Halocarbon Labs Inc., Hackensack, NJ) dissolved in 2 mL of olive oil. At various times, blood was collected by retro-orbital puncture after the mice were anesthetized. Blood samples were allowed to clot at 4°C overnight 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) according to manufacturer instructions. At 6, 12, and 24 hours after halothane administration, the animals were killed and the livers removed. A portion of each excised liver was fixed in 10% formaldehyde overnight before being transferred into 70% ethanol solution, and the remainder was snap-frozen and stored at −80°C for subsequent RNA isolation and immunoblot analyses. Tissue sections were embedded in paraffin and were subsequently mounted onto glass slides and stained with hematoxylin and eosin (H/E) or for use in immunohistochemical detection of neutrophils and malondialdehyde, MDA (Department of Pathology, UCHSC). All animal experiments were performed in accordance with guidelines from the UCHSC Institutional Animal Care and Use Committee.
Detection of TFA-Protein Adducts in the Liver.
The liver tissue samples collected at 12 hours after halothane treatment were homogenized in ice-cold Tris buffer (100 mmol/L, pH 7.5), containing 250 mmol/L sucrose, 2 mmol/L EDTA, and a cocktail of protease inhibitors (1:100, Sigma, St. Louis, MO). Ten micrograms of each homogenate sample was diluted in Laemmli sample buffer (Bio-Rad Laboratories, Hercules, CA) under reducing conditions, boiled for 5 minutes, and resolved on 12% polyacrylamide gels. After transfer to nitrocellulose membranes (Bio-Rad Laboratories), nonspecific binding was blocked with 5% nonfat milk. The blots were probed with a rabbit polyclonal anti-TFA antibody (1:1000, kindly provided by Dr. Lance Pohl, National Institutes of Health, Bethesda, MD) and then incubated with horseradish peroxidase–conjugated goat anti-rabbit IgG antibody (1:2000, Chemicon International Inc., Temecula, CA). Protein signals were visualized using an ECL Plus Western Blotting Detection System (GE Amersham Bioscience, Little Chalfont, UK) and the data were captured using a Storm 860 system (Molecular Dynamics, Sunnyvale, CA).
Total RNA was isolated from 20 mg of frozen liver tissue using RNeasy Mini Kits (Qiagen, Valencia, CA) as described 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 was amplified using Platinum Taq polymerase (Invitrogen) and gene-specific primers (Table 1) for interleukin-6 (IL-6), IL-1β, tumor necrosis factor-α (TNF-α), IL-8, inducible nitric oxide synthase (iNOS), IL-10, hemoxygenase (HO)-1, and glyceraldehyde-3-phosphate dehydrogenase (G3PDH). All PCR products were resolved on 1.5% agarose gels and visualized using ethidium bromide staining. RNA expression levels were determined by normalizing band intensities relative to the levels of G3PDH expression using Band Leader software (version 3.00, TechKnowledge, Tel Aviv, Israel).
Table 1. Primer Sequences and PCR Conditions for Determining the Messenger RNA Levels of Various Mediators
Preparation of Hepatic Leukocytes and Flow Cytometric Analysis.
Hepatic leukocytes were isolated following a described method8 with slight modification. In brief, the animals were anesthetized, and liver tissues were perfused by insertion of a 20 G catheter into the superior vena cava. The liver was perfused in situ at 37°C with Hank's balanced salt solution (HBSS) 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 100 U/mL of heparin (Baxter Healthcare Corp., Deerfield, IL) and centrifuged at 500g for 15 min. The resulting pellet was collected and resuspended in 2 mL of red blood cell lysing buffer (Sigma). After 5 minutes, the cells were washed in 0.6% acid citrate-dextrose (ACD-A, Sigma) solution containing 0.5% BSA. Total hepatic leukocytes were counted by trypan blue exclusion.
Neutrophil populations in freshly isolated hepatic leukocytes were characterized by staining the cells with fluorescein isothiocyanate (FITC)-conjugated anti-Gr1 (RB6-8C5, eBioscience, San Diego, CA) and phycoerythrin (PE)-conjugated anti-CD11b (M1/70, eBioscience) for 30 minutes on ice. To prevent nonspecific binding, all samples were pre-incubated with anti-FcγR II/III antibody (10 μg/mL, clone 93, eBioscience) plus rat serum (1:10, Sigma) for 10 minutes on ice. The cells were analyzed on a FACSCalibur using CellQuest software (BD Biosciences, San Jose, CA). The data were further analyzed using FlowJo software (Tree Star Inc., Ashland, OR).
Isolation of Kupffer Cells.
Kupffer cells (KCs) were isolated following an established method with some modification.24 In brief, mouse livers were perfused in situ with a 0.025% collagenase (Type IV, Sigma) solution. The liver cells were dispersed, and the suspension filtered and centrifuged to pellet the hepatocytes (30g for 3 minutes). The supernatant enriched in nonparenchymal cells was centrifuged at 330g for 5 minutes. The pellet was resuspended in 15 mL of 35% Percoll containing 100 U/mL of heparin and centrifuged at 500g for 15 min. The resulting pellet was resuspended in 3 mL Complete DMEM and incubated in a 6-well plate precoated with DMEM containing 10% fetal bovine serum (Invitrogen). After 2 hours culture at 37°C in a humidified 5% CO2 incubator, the nonadherent cells were removed by washing the plate 3 times with 10 mL DMEM, and the adherent KCs were collected for RNA extraction.
Immunohistochemical Detection of Neutrophils and MDA.
Liver sections were stained with an anti-Gr-1 monoclonal antibody (RB6-8C5, eBioscience) and anti-MDA polyclonal antibodies (provided by Dr. Dennis Petersen, UCHSC) for evaluation of hepatic neutrophil infiltration and lipid peroxidation, respectively. After the sections were deparaffinized and rehydrated, they were subjected to antigen retrieval using Antigen Unmasking Solution (Vector Laboratories Inc., Burlingame, CA). Endogenous peroxidase activity was inhibited by incubating the sections with 0.3% (vol/vol) H2O2 in methanol for 30 minutes. After blocking nonspecific binding with normal rabbit serum or goat serum for 20 minutes, the tissues were incubated with anti-Gr1 Ab or anti-MDA overnight at 4°C. Labeled cells were visualized using the Vectastain avidin-biotin–conjugated alkaline phosphatase kit (Vector Laboratories Inc.). Tissue sections were counterstained using Gills hematoxylin solution.
Depletion of Neutrophils.
Neutrophil depletion was accomplished by the administration of rabbit anti-neutrophil antisera (product number: AIAD31140, Accurate Chemicals and Scientific Co., Westbury, NY). Female Balb/c mice were intravenously injected with 100 μL of either normal rabbit serum (Accurate Chemicals and Scientific Co.) or anti-neutrophil serum (diluted 1:1 in sterile phosphate-buffered saline) 12 hours before and 4 hours after halothane treatment. ALT activities and hepatic neutrophil infiltration were evaluated 24 hours after halothane administration as described above.
Data are presented as mean ± SEM. Comparisons between two groups were carried out using two-tailed Student t test, and between multiple groups using 1-way analysis of variance (ANOVA) with a post hoc test of significance between individual groups. Differences were considered significant when P was less than .05.
Evaluation of Halothane-Induced Liver Injury in Balb/c, DBA/1, and C57BL/6J Mice.
Halothane is known to cause mild hepatotoxicity in approximately 20% of patients.17 To provide an experimental platform for mechanistic studies of this form of toxicity, we set out to develop a model of halothane-induced liver injury in mice. Following i.p. injection of 30 mmol/kg of halothane dissolved in 2 mL of olive oil to female and male Balb/c, DBA/1, and C57BL/6J mice, serum ALT activities revealed that the Balb/c strain was the most susceptible to halothane hepatotoxicity. The ALT levels reached as high as 1,200 IU/L in female Balb/c mice at 24 hours after halothane treatment (Fig. 1A), whereas the reported levels in guinea pigs were less than 500 IU/L.18, 25 DBA/1 mice (ALT levels approximately 200 IU/L) were less susceptible to halothane toxicity compared with Balb/c mice (Fig. 1B). Interestingly, the ALT activities in female DBA/1 mice were sustained at peak levels for 24 to 48 hours after halothane treatment, whereas the ALT levels declined sharply in Balb/c mice (Fig. 1A-B). Compared with their male counterparts, female Balb/c and DBA/1 mice developed increased liver injury (Fig. 1A-B). Thus, female mice were used in further mechanistic studies. Moreover, no significant liver toxicity was observed in C57BL/6J mice treated with halothane compared with oil-treated control mice (Fig. 1C).
Even though a relatively large volume of olive oil was used to dissolve halothane, the oil itself did not cause elevation of ALT levels above the baseline (Fig. 1). Immunohistochemical detection of MDA within the liver was performed to evaluate whether the olive oil could cause lipid peroxidation. However, no lipid peroxidation was evident in mice treated with olive oil alone compared with naïve mice (data not shown). The halothane product used in these experiments also contained 0.01% (vol/vol) thymol. To ensure that thymol itself did not cause liver injury, we treated mice with 0.01% thymol (Fisher Scientific, Fair Lawn, NJ) in 2 mL of olive oil. ALT levels were not increased compared with that of mice treated with oil alone (data not shown).
In line with the elevated ALT activities, histological evaluation of the liver sections obtained from female Balb/c mice at 24 hours after halothane treatment revealed focal necrosis in the centrilobular areas (Fig. 2C-D). Although massive cellular infiltration was not evident, there appeared to be some infiltrating leukocytes near the central vein. This observation is consistent with previous data reported in the guinea pig model.21 No histopathological differences were observed between naïve and olive oil-treated mice (Fig. 2A-B).
Analysis of TFA-Protein Adducts.
Halothane is bioactivated within the liver to its reactive metabolite, TFA, which has been shown to covalently modify a number of hepatic proteins.26 The formation of TFA and TFA-protein adducts are thought to be a prerequisite for the subsequent development of liver injury and, perhaps, TFA-specific immune responses. To investigate whether the metabolic differences among the three strains of mice might account for the disparities in their susceptibility to halothane-induced hepatotoxicity, immunoblotting was performed using anti-TFA polyclonal antibodies. However, detection of TFA-protein adducts 12 hours after halothane administration demonstrated no significant differences in either the patterns or the levels of liver adducts formed in Balb/c, DBA/1, and C57BL/6J mice (Fig. 3). These data suggest that the strain-dependent susceptibility is not due to variations in the amount of halothane absorbed into the liver after i.p. injection, halothane metabolism to its TFA-reactive metabolite, or the pattern of TFA-protein adducts.
Hepatic Inflammatory Responses Induced by Halothane Administration.
Studies of APAP-induced liver injury in mice suggested that drug (or metabolite)-induced stress and/or damage of hepatocytes may trigger the activation of the innate immune system within the liver and that inflammatory responses can escalate tissue damage.3, 8, 27 Consequently, we investigated whether halothane administration initiates hepatic inflammation, and whether the severity varies among the three strains of mice. Hepatic mRNA expression of the pro-inflammatory mediators TNF-α, IL-1β, IL-6, and iNOS were determined by RT-PCR. Expression levels of these mediators were not changed at 6 hours (data not shown) but were significantly increased at 12 hours after halothane treatment in female Balb/c mice compared with that of naïve or oil-treated mice (Fig. 4) as well as DBA/1 and C57BL/6J mice (Fig. 5). In contrast, hepatic expression levels of these pro-inflammatory mediators were not altered by halothane treatment in DBA/1 and C57BL/6J mice, except for TNF-α levels, which were increased in DBA/1 mice upon halothane treatment (Fig. 5).
In addition to the elevation of pro-inflammatory mediators, the number of hepatic leukocytes isolated from female Balb/c mice 24 hours after halothane treatment was drastically increased compared with that from naïve or oil-treated mice (Fig. 6A). Analysis of specific populations of leukocytes by flow cytometry demonstrated a significant increase in the number of neutrophils (CD11b+Gr1+) that infiltrated the liver at both 12 and 24 hours (but not 6 hours; data not shown) after halothane treatment compared with that of naïve or oil-treated mice (Fig. 6B). The numbers of T lymphocytes, NK, or NKT cells were not altered after halothane treatment in all three strains of mice (data not shown). Consistent with the observed neutrophil infiltration into the liver, hepatic message levels of IL-8, a chemokine specific for neutrophil recruitment, were significantly increased in Balb/c mice after halothane treatment (Fig. 4).
The number of hepatic leukocytes was also increased in DBA/1 mice upon halothane treatment (10 × 105/liver versus 0.6 × 105/liver in naïve or oil-treated mice); however, the increase was less profound compared with that in Balb/c mice (Fig. 7A). Similar to Balb/c mice, the increased infiltrate appeared to be predominantly neutrophils (Fig. 7B). In contrast, the number of neutrophils did not increase significantly in C57BL/6J mice, while total hepatic leukocytes appeared to be decreased after halothane treatment (5 × 105/liver) compared with the respective naïve or oil-treated mice (10 × 105/liver).
Immunohistochemical staining of neutrophils was performed using an anti-Gr1 antibody. Infiltrated neutrophils were observed in liver sections of Balb/c mice at both 12 and 24 hours after halothane treatment. The cells were found to accumulate around the central venous area where hepatocyte damage occurred (Supplementary Fig. 1B). No positive staining of neutrophils was observed in liver sections obtained from mice treated with olive oil (Supplementary Fig. 1A; Supplementary material is available at: http://interscience.wiley.com/jpages/0270-9139/suppmat/index.html).
Neutrophil Depletion Reduces Halothane-Induced Liver Injury in Balb/c Mice.
Because the data demonstrated that the number of infiltrating neutrophils in the liver correlates with the degree of liver damage observed in the three strains of mice, with the highest number of hepatic neutrophils found in the most susceptible Balb/c strain, we investigated whether neutrophils play a functional role in the development of halothane hepatotoxicity via neutrophil depletion experiments. Upon depletion of neutrophils in halothane-treated female Balb/c mice, the percentage of neutrophils in the total hepatic leukocyte population decreased from 29.2% to 4.5% (Fig. 8A), which is similar to the level observed in naïve mice (Fig. 6). Moreover, a significant decrease in the absolute number of hepatic neutrophils was also observed in neutrophil-depleted mice compared with their similarly halothane-treated control given normal serum (Fig. 8B). These results confirmed that administration of the anti-neutrophil serum successfully prevented halothane-induced neutrophil recruitment into the liver. Importantly, neutrophil depletion also prevented the rise in ALT levels, with an average ALT activity of 114 IU/mL (Fig. 8C). This constituted a 90% decrease relative to halothane-treated mice given normal rabbit serum, which had a mean ALT activity of 1112 IU/mL, similar to that of naïve, halothane-treated mice shown in Fig. 1.
To rule out the possibility that the protection against halothane-induced liver injury by anti-neutrophil serum treatment was due to alterations of halothane metabolism, TFA-protein adduct formation was determined by immunoblot analysis using anti-TFA antisera. No differences in either the patterns or the levels of protein adducts could be observed among mice treated with normal serum, anti-neutrophil serum, or halothane alone (Supplementary Fig. 2).
Anti-neutrophil serum treatment causes apoptosis of neutrophils, which are phagocytosed by KCs. To address whether KC activation may be modified by the uptake of apoptotic neutrophils, KC production of pro-inflammatory cytokines (TNF-α and IL-1β) was measured. KCs were isolated and purified from the livers of Balb/c mice treated with normal serum, anti-neutrophil serum, or halothane alone. RT-PCR analyses demonstrated no significant differences in the levels of TNF-α and IL-1β expression by KCs obtained from the aforementioned three groups of mice (Supplementary Fig. 3A).
An alternative explanation for the anti-neutrophil serum-mediated inhibition of halothane-induced liver injury is that the serum treatment may protect the liver by causing up-regulation of hepatoprotective factors independent of the neutrophil depletion. To address this concern, hepatic mRNA expressions of two potent hepatoprotective factors, IL-10 and hemoxygenase-1 (HO-1),4, 28 were determined. The data showed that the mRNA levels of both IL-10 and HO-1 were similar among mice treated with normal serum, anti-neutrophil serum, or halothane alone (Supplementary Fig. 3B). Collectively, these data suggest that hepatic neutrophil recruitment plays an important role in causing and/or exacerbating halothane-induced liver injury in this mouse model.
Advances in understanding of the mechanisms of DILI have been hampered by the scarcity of animal models. APAP-induced liver injury in mice and the recently developed LPS-drug idiosyncrasy model are probably the only reliable models used thus far for investigating the mechanisms of DILI. However, these models alone do not encompass the entire spectrum of the clinical characteristics nor all mechanistic features of DILI development in patients. In this study, we developed another model of DILI, namely, the halothane-induced liver injury model in mice. The data obtained using this model provided novel mechanistic information, including the strain-dependent susceptibility to halothane hepatotoxicity in mice and the important role of the innate immune system, particularly neutrophils, in the pathophysiology of halothane-induced liver injury.
Previous attempts at developing mouse models of halothane-induced liver injury to decipher the complex molecular and cellular mechanisms of this toxicity have been unsuccessful. One of the main reasons for this failure is that halothane could only be administered in a small amount (maximum 10 mmol/kg) to prevent respiratory complications that could cause animal death. Clinically, a large dose of halothane can be given to patients because a respirator is usually used to avoid this problem. In this study, we discovered that a higher dose of halothane (30 mmol/kg in 2 mL olive oil) could be safely administered to mice because the oil creates a slow release depot, thereby decreasing halothane volatility and maximizing its absorption into the liver. Although a large volume of olive oil was used, the oil treatment alone did not appear to cause any effects that would complicate the interpretation of our results. In control mice treated with olive oil alone, we did not observe any hepatotoxicity evaluated by both serum ALT levels and liver histology (Fig. 1 and Fig. 2), nor did we observe any inflammatory responses, including pro-inflammatory cytokine production and hepatic infiltration of inflammatory cells (Fig. 4 and Fig. 6). Further, lipid peroxidation was not detected in mice treated with olive oil (data not shown).
We also found that female Balb/c and DBA/1 mice developed increased liver toxicity compared with their male counterparts (Fig. 1). This is consistent with the female sex being a risk factor in the development of halothane hepatitis in humans, as it has been estimated that the female-to-male ratio in the prevalence of halothane hepatitis is approximately 2:1.29 However, the mechanistic basis for this sex difference in both humans and this mouse model remains to be elucidated. Our data also demonstrate that among the three strains of mice tested, Balb/c mice were most susceptible to halothane-induced hepatotoxicity (Fig. 1). A strain-dependency in the susceptibility to liver injury was recently reported in the APAP-induced hepatotoxicity model in mice.30, 31 Comparisons of the hepatic genome and proteome of susceptible (C57BL/6J) and resistant (SJL) mice to APAP-induced liver injury led to the identification of a number of susceptibility factors, including stress-response proteins and pro-inflammatory mediators, that might explain the strain-based variation.31 It has also been reported that some strains of guinea pigs were more susceptible to halothane-induced liver injury than others, and that the increased susceptibility was not due to the animals' inherent ability to metabolize halothane to TFA.22, 23 We found that the degree and pattern of TFA-modification of liver proteins were similar among Balb/c, DBA/1, and C57BL/6J mice (Fig. 3), suggesting that strain-dependent susceptibility is not due to variations in the amount of halothane absorbed into the liver after i.p. injection, halothane metabolism to its TFA reactive metabolite, or the pattern of TFA-protein adducts. Thus, other mechanisms, such as the innate immune responses to the initial halothane-induced hepatocyte damage, may contribute to the strain variation in susceptibility.
It has been suggested that drug-induced stress and/or damage of hepatocytes may trigger inflammatory responses of the innate immune system within the liver. Evidence to support this idea has been obtained from animal models of APAP-induced liver injury3, 8, 27 and the more recent LPS-drug idiosyncrasy model.11–15 In the APAP model, it has been demonstrated that the initial hepatocyte damage caused by the reactive metabolite of APAP can lead to the activation of innate immune cells within the liver, thereby stimulating hepatic infiltration of other inflammatory leukocytes. Activated cells of the hepatic innate immune system, such as KCs, natural killer (NK) cells, and NKT cells, produce a range of inflammatory mediators, including cytokines, chemokines, and reactive oxygen and nitrogen species, that contribute to the progression of liver injury.8, 27, 32 Our findings in this mouse model of halothane-induced liver injury provide another line of evidence for the involvement of the innate immune system in certain instances of DILI. The data demonstrated that, aside from halothane-induced direct toxicity through metabolism to TFA, activation of the innate immune system and induction of inflammatory events play an important role in the propagation of tissue damage. Significant increases in hepatic message levels of the pro-inflammatory cytokines, TNF-α, IL-1β, and IL-6, and the hepatic protoxicant, iNOS, were observed in Balb/c female mice treated with halothane compared with naïve or oil-treated control mice (Fig. 4). Furthermore, hepatic message levels of IL-8, a neutrophil-specific chemokine, were significantly increased in halothane-treated mice compared with that of naïve or oil-treated mice (Fig. 4). This increase in IL-8 correlated with a significant increase in the number of infiltrating neutrophils within the liver of susceptible Balb/c mice following halothane treatment (Fig. 6; Supplementary Fig. 1). Subsequent neutrophil depletion experiments provided strong evidence that these cells play a crucial role in inducing and/or aggravating liver injury initiated by halothane treatment (Fig. 8). This finding is consistent with the results from previous studies, which have demonstrated the important role of neutrophils in a number of experimental models of liver injury, including hepatic ischemia-reperfusion,33 endotoxemia,34, 35 sepsis,36 and drug-induced hepatotoxicity.13, 37, 38 It has been elegantly demonstrated in an endotoxemia model that neutrophil-induced liver injury is mediated by NADPH oxidase and hypochlorous acid formed by myeloperoxidase.34 To definitively demonstrate the role of neutrophils in our mouse model of halothane hepatitis and understand the mechanisms involved, further studies to identify neutrophil-specific oxidant stress are warranted. It is important to note that halothane-induced hepatic infiltration of neutrophils was less profound in DBA/1 and C57BL/6J mice (Fig. 7), suggesting a clear correlation between neutrophil infiltration and liver injury. Similarly, the increase in hepatic message levels of TNF-α, IL-1β, IL-6, IL-8, and iNOS was less dramatic in DBA/1 and C57BL/6J mice compared with that of Balb/c mice (Fig. 6). The differences in inflammatory responses following halothane-induced hepatocyte damage among the three strains correlate with the varying degrees of toxicity observed in these mice. This suggests that the downstream inflammatory response contributes significantly to the overt hepatotoxicity induced by halothane, and that variations in the inflammatory responses among different strains may be an important factor in determining susceptibility.
In summary, we report the development of a mouse model of halothane-induced liver injury. A strain-dependent susceptibility to halothane hepatotoxicity was observed, and the data provided strong evidence that the variations of the innate immune responses in the liver, rather than the degree of halothane metabolism to its TFA reactive metabolite, are an important factor in determining susceptibility. This model provides an experimental system for investigation of key susceptibility factors that may control the development of liver injury in patients treated with halothane and potentially other drugs. Moreover, acute halothane-induced toxicity may be a prerequisite for the subsequent development of severe halothane hepatitis because injury to hepatocytes would allow the release of the antigenic TFA-protein conjugates, and T cells and B cells that recognize these antigens may be activated and produce pathogenic antibodies and cytotoxic T cells in susceptible individuals. Our current model of acute halothane-induced liver damage in mice provides a platform on which a model involving TFA-specific adaptive immune responses and potentially halothane-induced hepatitis might be developed.
We thank Drs. Lance Pohl (NIH, Bethesda, MD) and Dennis Petersen (UCHSC) for their generous gifts of the anti-TFA and anti-MDA antisera.