Benzyl alcohol attenuates acetaminophen-induced acute liver injury in a Toll-like receptor-4-dependent pattern in mice


  • Potential conflict of interest: Nothing to report.


Acetaminophen (APAP) toxicity is the most common cause of acute liver failure in industrialized countries. Understanding the mechanisms of APAP-induced liver injury as well as other forms of sterile liver injury is critical to improve the care of patients. Recent studies demonstrate that danger signaling and inflammasome activation play a role in APAP-induced injury. The aim of these investigations was to test the hypothesis that benzyl alcohol (BA) is a therapeutic agent that protects against APAP-induced liver injury by modulation of danger signaling. APAP-induced liver injury was dependent, in part, on Toll-like receptor (TLR)9 and receptor for advanced glycation endproducts (RAGE) signaling. BA limited liver injury over a dose range of 135-540 μg/g body weight or when delivered as a pre-, concurrent, or post-APAP therapeutic. Furthermore, BA abrogated APAP-induced cytokines and chemokines as well as high-mobility group box 1 release. Moreover, BA prevented APAP-induced inflammasome signaling as determined by interleukin (IL)-1β, IL-18, and caspase-1 cleavage in liver tissues. Interestingly, the protective effects of BA on limiting liver injury and inflammasome activation were dependent on TLR4 signaling, but not TLR2 or CD14. Cell-type–specific knockouts of TLR4 were utilized to further determine the protective mechanisms of BA. These studies found that TLR4 expression specifically in myeloid cells (LyzCre-tlr4−/−) were necessary for the protective effects of BA. Conclusion: BA protects against APAP-induced acute liver injury and reduced inflammasome activation in a TLR4-dependent manner. BA may prove to be a useful adjunct in the treatment of APAP and other forms of sterile liver injury. (Hepatology 2014;60:990–1002)




acute liver failure


alanine aminotransferase




benzyl alcohol


body weight


cytochrome p450


damage-associated molecular patterns


dendritic cell


2′,7′-dichlorofluorescein ELISA, enzyme-linked immunosorbent assay




glutathione disulfide


hematoxylin and eosin


hepatic ischemia/reperfusion


high-mobility group box 1






interferon gamma-induced protein 10


c-Jun N-terminal kinase


keratinocyte-derived chemokine




mitochondrial DNA


NACHT, LRR, and PYD domains-containing protein 3


N-acetyl-p-benzoquinone imine


natural killer T


phosphate-buffered saline


polymerase chain reaction


primary mouse hepatocytes


pattern recognition receptors


receptor for advanced glycation endproducts


respiratory control ratio


reactive oxygen species


standard error of the mean




Toll-like receptor


wild type

Acetaminophen (APAP) is a commonly used over-the-counter analgesic and antipyretic medicine used to relieve symptoms of mild inflammatory conditions. APAP overdoses are the leading cause of acute liver failure (ALF) in most industrialized countries. APAP toxicity accounts for approximately 50% of all cases of ALF in the United States and carries a 30% mortality.[1, 2] More than 2,600 hospitalizations and nearly 500 deaths are attributed to APAP in the United States annually.[3] Current treatment for APAP-induced liver failure is limited to N-acetylcysteine, which works mainly by restoration of glutathione levels, and supportive care. For those who progress to liver failure, the only effective treatment is liver transplantation.

In APAP-induced injury, and other forms of sterile liver injury, endogenous damage-associated molecular patterns (DAMPs) are released to activate cellular pattern recognition receptors (PRRs) on both immune and parenchymal cells.[4-11] Release of DAMPs and activation of PRRs have numerous effects, including activation of inflammasome signaling, release of cytokines and chemokines, and cellular injury. Several studies have highlighted the role of different PRRs and inflammasome activation on APAP-induced liver injury.[4] Additional mechanisms of APAP-induced injury have focused on mitochondrial damage and oxidant injury, which contribute to cell necrosis.[12-14]

Benzyl alcohol (BA) is an organic compound and can be produced naturally by plants and exists in many essential oils. It is commonly used as a solvent as well as a precursor to a variety of esters in industries. In the health care field, BA was used as a local anesthetic and is quite frequently added to intravenous medication solutions as a preservative because of its bacteriostatic and antipruritic properties. BA can be oxidized to benzoic acid, conjugated with glycine in the liver, and excreted as hippuric acid. Initial observations found that APAP-induced liver injury was prevented when using a delivery vehicle of a commercial formulation of bacteriostatic saline containing BA. The aim of these investigations was to test the hypothesis that BA protects against APAP-induced liver injury by prevention of mitochondrial injury and/or modulation of danger signaling.

Materials and Methods


APAP, BA, pyrazole, and glycyrrhizic acid ammonium salt were purchased from Sigma-Aldrich (St. Louis, MO). The MAGPIX kit was from Millipore (Billerica, MA). The interleukin (IL)-18 enzyme-linked immunosorbent assay (ELISA) kit was bought from R&D Systems (Minneapolis, MN). Caspase-1 and high-mobility group box 1 (HMGB1) antibodies (Abs) were from Abcam (Cambridge, MA). Phospho- and total c-Jun N-terminal kinase (JNK) was from Cell Signaling Technology (Beverly, MA). All other reagents without specific notification were purchased from Sigma-Aldrich.


Male C57BL/6 and IL-6 knockout (KO) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). CD14 KO, global Toll-like receptor (TLR)4 knockout, TLR4 flox mice, receptor for advanced glycation endproducts (RAGE) KO, TLR2 KO, TLR9 mutant, Jα281 KO, caspase-1 KO, and NACHT, LRR, and PYD domains-containing protein 3 (NALP3) KO were bred, genotyped, and maintained in the animal facility of the University of Pittsburgh (Pittsburgh, PA). We used Cre-loxp technology to generate hepatocyte, myeloid cell, adipose, and dendritic cell (DC)-specific TLR4 KO mice (Alb-tlr4/, LyzCre-tlr4/, adipose-tlr4/, and CD11c-tlr4/, respectively) and hepatocyte-specific HMGB1 KO mice. All mice were male, 8-12 weeks old, and weighed 20-30 g at the time of the experimental procedures. Animal handling and care was approved by the University of Pittsburgh's Institutional Animal Care and Use Committee.

APAP-Induced Liver Injury Model

Mice were fasted for 15-16 hours, but had free access to water before APAP administration. APAP solution was prepared in warmed saline (37°C) and was injected intraperitoneally (IP) at a dose of 400 mg/kg body weight (b.w.). After 1-24 hours, mice were anesthetized with isofluorane, then blood was harvested by cardiac puncture and liver was harvested by being snap-frozen or fixed in 10% formalin for hematoxylin and eosin (H&E) staining. Blood was centrifuged at 5,000 rpm for 10 minutes.

MAGPIX Multiplex Cytokine Assay and IL-18 ELISA

Serum IL-1β, keratinocyte-derived chemokine (KC), interferon gamma-induced protein 10 (IP-10), and IL-6 levels were measured by the MAGPIX kit (Millipore), according to its standard protocol. IL-18 level was examined by the IL-18 ELISA kit as per the manufacturer's protocol.

Western Blotting

Western blotting analysis was used to protein levels in whole liver, hepatocytes, or serum. Proteins were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred onto a nitrocellulose membrane. The membrane was blocked for 1 hour in phosphate-buffered saline (PBS)-Tween (0.1%) with 5% milk, followed by immunostaining with optimized dilutions of primary Abs in 1% milk in PBS-Tween overnight at 4°C. Horseradish peroxidase–conjugated secondary Abs were given, and membranes were developed with the Super Signal West Pico chemiluminescent kit (Thermo Fisher Scientific, Waltham, MA) and exposed to film.

Alanine Aminotransferase Measurement

Serum alanine aminotransferase (ALT) was measured with HESKA Dri-Chem 4000 (HESKA; slides from Fujifilm, Tokyo, Japan).


Liver samples were fixed in 10% formalin and embedded in paraffin, then cut to 6-μm-thick sections. Tissues were stained with H&E, and slides were assessed for inflammation and tissue damage.

Mitochondrial Function

Mitochondrial function was assessed in liver in the following way. Livers were harvested after the mouse was perfused with cold PBS. Oxygen consumption in fresh whole-liver homogenates were immediately measured using a Clark-type oxygen electrode (Instech Laboratories, Plymouth Meeting, PA) in the presence of succinate (for state 4 measurements) and adenosine diphosphate (for state 3 measurement). Respiratory control ratio (RCR) was calculated as state 3/4. Each condition was run in triplicate within each experiment, and experiments were repeated three times.

Cell Culture

Primary mouse hepatocytes (PMHs) were harvested from C57BL/6 mice, as previously describe.[15] Cells were utilized on days 1-3 after harvest. APAP treatment in all hepatocyte experiments was utilized at a concentration of 5 mM, and cells were treated from 0 to 12 hours (Sigma-Aldrich). Cells were maintained at 37°C and 5% CO2.

Polymerase Chain Reaction

Cells were cultured as previously described. Total DNA was isolated from plasma samples using a QIAamp Blood and Mini Kit (Qiagen, Valencia, CA), according to the manufacturer's instructions. Samples were then diluted, and the same amount of total DNA was added to each reaction on each plate. Mitochondrial DNA (mtDNA) was determined using primers for mouse cytochrome c oxidase subunit III (5′-ACCAAGGCCACCACACTCCT-3′ and 5′ACGCTCAGAAGAATCCTGCAAAGAA-3′). To construct standard curves, mitochondria were purified from mouse liver. Purity of mtDNA standards was verified by real-time polymerase chain reaction (PCR) using primers for both mitochondrial genes and nuclear-encoded β-actin. Dilutions of these purified mtDNA samples were prepared. All samples were run in triplicate and repeated three times.

Glutathione Assay

Glutathione (GSH) and the oxidized disulfide dimer, glutathione disulfide (GSSG), were measured from isolated primary hepatocytes or liver using the Glutathione Assay Kit from Cayman Chemical (Ann Arbor, MI), as per the manufacturers instructions, and adjusted for protein concentration.

Reactive Oxygen Species Measurements

Hepatocytes were treated with 5 mM of APAP for 6 hours. Hepatocytes were rinsed in PBS. Cells were loaded with 1 mM of 5- and 6 chloromethyl-20,70-dichlorodihydrofluorescein diacetate (Molecular Probes, Eugene, OR) for 30 minutes. Fluorescence intensity was measured by a fluorescence microplate reader (Spectra Max GeminiEM; Molecular Devices, Sunnyvale, CA) at 37°C for 30 minutes using excitation and emission wavelengths of 490 and 530 nm. The slope of the curve was used to calculate the change in fluorescence intensity over time.

Statistical Analysis

Results are expressed as the mean ± standard error of mean (SEM). SigmaPlot 11.0 (Systat Software, Inc., Point Richmond, CA) was used for the statistical analysis, using one-way analysis of variance with Tukey's posthoc analysis for significance. Significance was established at P < 0.05.


BA Attenuates APAP-Induced Acute Liver Injury

C57BL/6 mice were randomized to treatment with APAP at a dose of 400 mg/kg b.w. or vehicle (0.9% saline) at the same volume. Mice were further randomized to concurrently receive BA (270 mg/kg, IP) or vehicle (0.9% saline) at the same volume. Mice were euthanized at 3-24 hours, and liver injury was assayed. BA reduced liver injury, as determined by serum ALT, at both 3- and 6-hour time points (Fig. 1A). At the 24-hour time point, ALT in the APAP group was 7,219 ± 697 and the BA-treated group was 3,062 ± 424 U/mL (P < 0.05). Consistently, histological analysis at 6 hours revealed significantly less liver necrosis and maintenance of normal architecture in the BA-treated APAP group, compared to APAP alone (Fig. 1B). A dose response of BA (135 μg/g b.w. to 810 μg/g) was performed with concurrent APAP administration. BA doses from 135 to 540 μg/g significantly decreased serum ALT levels, compared to APAP alone (P < 0.05 at each dose). However, all the mice in the 810-μg/g group plus APAP (n = 3) died within 6 hours (Fig. 1C), and experiments at this dose were discontinued. Of note, there was no mortality within 6 hours for APAP alone or the 810-μg/g BA dose in the absence of APAP. Further experiments utilizing 270-μg/g dosing of BA were performed to evaluate a time response in this model. In these studies, BA was administered 24, 12, 6, 4, 2, or 1 hour before APAP administration, concurrently with APAP, or 1, 2, or 3 hours after APAP administration. Liver injury was assayed 6 hours after APAP administration in all groups. BA pretreatment protected against APAP-induced liver injury at all time points, as did concurrent treatment (Fig. 1D). Posttreatment was effective at 1 or 2 hours after APAP therapy, but not at 3 hours, in this model.

Figure 1.

BA protects against APAP-induced liver injury. (A) Comparison of serum ALT levels 3 and 6 hours after APAP treatment (400 mg/kg) in C57BL/6 mice with and without BA (270 μg/g, IP). N = 6-9/group, *P = 0.003 (3-hour group) and **P < 0.001 (6-hour group), respectively. (B) Histological comparison at 6 hours after APAP treatment with and without BA. Panels demonstrate representative H&E micrographs in each group. (C) Dose-dependent treatment effects of BA. Data show mean ± SEM; N = 3-7/group. BA was protective at 135-540 μg/g b.w. in APAP treatment. At the dose of 810 μg/g, 3 of 3 mice died within 6 hours of APAP treatment. (D) Effects of pre- (1-24 hours before APAP) and posttreatment (1-3 hours after APAP) with BA (270 mg/kg, IP) on serum ALT levels at 6 hours after APAP dosing. Data are shown as mean ± SEM; N = 5-9/group. *P < 0.05; **P < 0.01.

BA Reduced the Inflammatory Response in Acute APAP Toxicity

APAP toxicity causes significant inflammatory signaling, as demonstrated by cytokine and chemokine responses. APAP-induced cytokine and chemokine responses were confirmed in these experiments, demonstrating increases in IL-6, KC, and IP-10 (Fig. 2A-C). All of these were decreased by BA treatment. HMGB1 is a nuclear protein, which has been shown to be an important mediator of injury in both sterile and nonsterile liver injury, including APAP toxicity.[4, 7-9, 16-21] Consistent with other inflammatory mediators, BA reduced serum HMGB1 in APAP-treated mice, compared to APAP alone, as determined by western blotting (Fig. 2D).

Figure 2.

BA treatment reduced APAP-induced inflammation. (A-C) Serum IL-6, KC-1, and IP-10 are increased, compared to controls, 6 hours after APAP treatment (*P < 0.01; N = 5-7/group). APAP-treated mice that received concurrent BA treatment demonstrated decreased levels of these cytokines and chemokines, compared to APAP alone (#P < 0.01). (D) Concurrent BA treatment (180 or 270 μg/g) reduced serum levels of HMGB-1 protein 6 hours after APAP treatment, as determined by western blotting. Blotting is representative of three independent experiments.

APAP Toxicity Is Dependent, in Part, on TLR9, RAGE, and HMGB1, but not on TLR2 or TLR4

Danger signaling and PRRs are critical pathways in mediating sterile liver injury. Reports in APAP toxicity are somewhat varied,[4, 10] but may depend, in part, on variations in the model as well as timing of individual investigations. The contribution of PRRs and DAMPs were investigated in these studies. Expression of TLR2 or 4 did not influence APAP hepatotoxicity, as determined by serum ALTs (Fig. 3A) or histopathology. However, consistent with previous findings, absence of TLR9 protected against liver injury in this model.[22] Similarly, absence of RAGE expression limited injury as well (Fig. 3A).[23] The influence of expression of TLR4, TLR9, and RAGE expression on HMGB1 serum levels after APAP treatment were also determined by ELISA (Fig. 3B). There was no baseline HMGB1 level difference among control, TLR2−/−, TLR4−/−, TLR9mut, and RAGE−/− mice (data not shown); however, APAP treatment resulted in the release of HMGB1 into the circulation. Serum HMGB1 levels correlated with ALT levels and liver injury, demonstrating decreased levels at a 6-hour time point after APAP in mice that lacked TLR9 and RAGE expression.

Figure 3.

APAP-induced liver injury is dependent, in part, on TLR9, RAGE, HMGB-1, and IL-6 expression. (A) Tg mice for individual PRRs that were either deficient in expression (TLR2, TLR4, or RAGE) or expressed mutant, nonfunctioning receptors (TLR9) or their respective matched controls were treated with APAP and serum ALT levels were determined after 6 hours. TLR9 mutants and RAGE KO mice demonstrated less injury (*P < 0.05), compared to matched control mice, whereas TLR2 and 4 KO mice demonstrated similar levels of liver injury. (B) Serum HMGB-1 levels, as determined by ELISA, were also reduced in TLR9 mutant and RAGE KO mice 6 hours after APAP treatment (*P < 0.05, compared to WT controls). (C) Glycyrrhizin, which poses anti-HMGB-1 properties, protected against APAP-induced increases in serum ALT (*P < 0.05). (D and E) Mice deficient in hepatocyte-specific expression of HMGB-1 were also protected against APAP-induced liver injury, as determined by serum ALT, at 6 hours (*P < 0.05), as well as elevations in serum IL-6 levels (*P < 0.05). (F) Mice deficient in IL-6 expression were also protected against APAP-induced liver injury at 6 hours (*P < 0.05).

Based upon these findings, the role of HMGB1 was investigated further. Initially, mice were treated with glycyrrhizin (10 mg/kg), a chemical with nonspecific anti-HMGB1 activity.[24] Glycyrrhizin significantly decreased APAP-induced hepatotoxicity (Fig. 3C). The role of HMGB1 as a mediator of liver injury was also determined utilizing mice with hepatocyte-specific knockdown of HMGB1 expression (HC-HMGB1−/−). These mice demonstrated significantly less injury after APAP exposure, compared to wild-type (WT) control mice (Fig. 3D). Likewise, levels of IL-6 as an inflammatory mediator were significantly reduced in HC-HMGB1−/− mice after APAP (Fig. 3E). As an adjunct to these findings, APAP toxicity was determined in ifnb2/ (IL-6 KO) mice. Interestingly, toxicity of APAP was significantly reduced in IL-6 KO mice (Fig. 3F). Together, these data suggest a pathway of TLR9- or RAGE-mediated injury that signals, or is amplified by, hepatocyte parenchymal HMGB1 release and subsequent increases in serum IL-6.

Protection by BA Against APAP Toxicity Is Mediated by TLR4

Given the role of danger signaling in APAP-induced injury, the potential modulation of these signaling pathways by BA to mediate protective effects was investigated. Although liver injury was minimized in TLR9 and RAGE KO mice after APAP, injury was further reduced in both of these mutant mice with BA treatment (Fig. 4A). However, BA failed to protect against liver injury in mice deficient in TLR4 expression, suggesting that BA signaled, at least in part, through this receptor. CD14 has been shown to be a coreceptor of TLR4 in the recognition of lipopolysaccharide,[25] and natural killer T (NKT) cells have been shown to play a vital role in aseptic models, such as hepatic ischemia/reperfusion (HIR).[26] The contribution of this cosignaling molecule and cell type in contributing to hepatoprotection by BA were investigated. CD14 KO mice and mice deficient in NKT cells demonstrated liver injury after APAP treatment and were both protected by BA (Fig. 4A), suggesting that CD14 expression and NKT cells were not critical to injury or BA-induced protection in this model.

Figure 4.

BA protects through TLR4 receptor expression. (A) BA reduced APAP-induced injury in WT, TLR2−/−, TLR9 mutant, RAGE−/−, and Jα281−/− mice, which are mice deficient in NKT cells (*P < 0.05, compared to non-BA, APAP-treated, matched mice). Consistent with previous findings, both TLR9 mutant and RAGE−/− mice had less APAP-induced liver injury, compared to control mice (#P < 0.05). However, BA failed to protect TLR4−/− mice from injury. (B) Cell-type–specific KO of TLR4 demonstrated that BA's protection is dependent on TLR4 expression on cells of myeloid lineage. BA limited liver injury in WT, hepatocyte-specific (Alb-TLR4−/−), DC-specific (CD11c-tlr4−/−), and adipose-cell–specific (adipose-tlr4−/−) TLR4 KO mice; however, it did not limit injury in TLR4 global KOs or myeloid-specific (LyzCre-tlr4−/−) KO mice (*P < 0.05, compared to APAP-treated, BA-treated WT mice). (C) These findings are confirmed with histological analysis by H&E staining. Micrographs are representative findings (10×).

Experiments were performed in cell-type–specific KOs of TLR4 to further investigate BA signaling and protection. For these investigations, we compared BA's protection against APAP toxicity in hepatocyte-specific (Alb-tlr4−/−), DC-specific (CD11c-tlr4−/−), adipose-specific (adipose-tlr4−/−), myeloid cell–specific (LyzCre-tlr4−/−), and TLR4−/− global KO to WT mice. We found that TLR4 absence on hepatocytes, DCs, or adipose cells did not influence BA's protection, which was similarly effective to that noted in WT mice. However, BA failed to protect against APAP-induced liver injury in mice with myeloid cell–deficient expression of TLR4, similar to lack of effect in the global TLR4 KO group (Fig. 4B). Histological results also supported the serum ALT data (Fig. 4C). Furthermore, histological findings are consistent with previous studies demonstrating a significant amount of centrilobular necrosis in the setting of liver injury. Based upon these findings, the ability of BA to protect against liver injury seems to depend, in part, specifically on TLR4 signaling in myeloid cells.

BA Limits APAP-Induced Hepatic Oxidant and Mitochondrial Injury

Many studies have demonstrated that the toxic effects of APAP are mediated, at least in part, secondary to direct hepatocyte toxicity from mitochondrial injury and oxidant stress.[12, 14, 27] The purpose of these experiments was to determine whether BA directly influences these findings. General hepatic oxidant injury was determined by measuring the influence of APAP on hepatic GSH levels and the oxidation of GSH. APAP increased the ratio of GSSG to total GSH at a 6-hour time point, indicative of an oxidant stress (P < 0.05; Fig. 5A). BA limited APAP-induced changes in GSSG/GSH (P < 0.05). Phosphorylation of JNKs has been shown to increase subsequent to APAP treatment and correlate with injury.[12, 28] BA limited APAP-induced JNK phosphorylation, as determined by western blotting of whole-liver homogenates (Fig. 5B; data shown for 3-hour time point). The influence of APAP on hepatic mitochondrial injury was determined. Respiratory control ratio is a marker of overall mitochondrial health and is a ratio of state 3/4 respiration. At 3 hours, APAP resulted in a decrease in hepatic RCR, compared to controls, and this was prevented by BA (Fig. 5C). Excessive mitochondrial injury can lead to the release of mtDNA into the extracellular environment to act as a DAMP and further exacerbate injury, with TLR9 as a potential cell-surface receptor. Levels of plasma mtDNA at a 6-hour time point were determined by PCR for mitochondrial-derived cytochrome c oxidase subunit III. APAP treatment correlated with an increase in plasma levels of mtDNA, whereas BA therapy limited these levels (Fig. 5D).

Figure 5.

BA treatment limits APAP-induced mitochondrial and oxidant stress. (A) BA treatment limits APAP-induced GSH oxidations, as demonstrated by liver homogenate GSSG/GSH percentages at 6 hours (1.74 ± 0.3% vs. 2.7 ± 0.41%; *P < 0.05, compared to control; #P < 0.05, compared to APAP alone). (B) APAP increases hepatic JNK phosphorylation and this is limited by BA treatment (data are representative of three independent experiments). (C) BA limits APAP-induced mitochondrial injury, as demonstrated by a change in respiratory control ratio, compared to unmanipulated mice. Data are from 3 hours after treatment (*P < 0.05, compared to control; #P < 0.05, compared to APAP alone). (D) BA limited plasma levels of mtDNA after APAP treatment at 6 hours, as determined by PCR (15.7 ± 6.1 ng/mL of APAP vs. 5.4 ± 3.6 ng/mL of APAP+BA; *P < 0.05, compared to control; #P < 0.05, compared to APAP alone). Data for all quantitative experiments (A, C, and D) are from 8 mice per group.

The direct influence on hepatocytes was next determined. The influence of APAP (5 mM) on PMHs (C57BL/6) in vitro was determined. APAP treatment led to an increase in reactive oxygen species (ROS) production, as determined by 2′,7′-dichlorofluorescein (DCF) fluorescence (Fig. 6A), and this was limited by BA (1 mg/mL) treatment (P < 0.05). Similar to in vivo studies, APAP increased phosphorylated JNK levels in hepatocytes, whereas this was limited by BA, as determined by western blotting (Fig. 6B). Furthermore, APAP-induced cell death was limited by BA treatment at 8 hours, as determined by trypan blue exclusion (P < 0.05; Fig. 6C). Taken together, these data suggest that BA protects against direct effects of APAP-induced hepatotoxicity and mitochondrial injury. Release of mtDNA and other DAMPs, such as HMGB1, may promote injury response through innate immune-mediated signaling.

Figure 6.

BA prevents APAP-induced hepatocyte toxicity in vitro. The influence of APAP (5 mM) on primary mouse (C57BL/6) hepatocytes in vitro was determined. (A) APAP treatment led to an increase in ROS) production, as determined by DCF fluorescence and this was limited by BA (1 mg/mL) treatment (*P < 0.05, compared to control; #P < 0.05, compared to APAP alone). Data are mean ± SEM for three independent experiments with each condition performed in quadruplicate. (B) JNK phosphorylation was increased by APAP in vitro and limited by BA, as determined by western blotting (representative of three independent experiments). (C) APAP-induced cell death was limited by BA treatment at 8 hours, as determined by trypan blue exclusion (*P < 0.05, compared to control; #P < 0.05, compared to APAP alone). Data are mean ± SEM for three independent experiments with each condition performed in triplicate.

BA Inhibits Inflammasome Activity in APAP-Induced Mouse ALF

Previous studies have shown that APAP-induced injury is partially mediated by activation of the Nalp3 inflammasome.[6, 22, 29, 30] This perhaps would also link to TLR9 dependence of injury in this model, because activation of Nalp3 can be be downstream of TLR9. Moreover, this could serve as a potential link between release of mtDNA from direct hepatocyte toxicity and signaling to myeloid cells, such as Kupffer cells. Critical to Nalp3 inflammasome signaling is cleavage and activation of caspase-1, which, in turn, cleaves the proforms of IL-1β and IL-18. Consistent with findings of previous investigators, APAP treatment resulted in diminished liver injury in NALP3−/− and caspase-1−/− mice, compared to WT controls, at 6 hours (Fig. 7A). Furthermore, caspase-1 cleavage was increased in APAP-treated mice, compared to controls (data not shown), and BA treatment diminished hepatic cleaved/caspase-1 protein levels, as determined by western blotting (Fig. 7B). Moreover, APAP treatment increased serum IL-1β and IL-18 levels, and cotreatment with BA significantly prevented or reduced levels of these cytokines (Fig. 7C,D).

Figure 7.

BA limits APAP-induced activation of the Nalp3 inflammasome. (A) APAP-induced liver injury was limited in Nalp3−/− (*P < 0.05) and caspase-1−/− (**P < 0.01) mice, compared to WT mice. (B) BA limited caspase-1 cleavage, as demonstrated by western blotting. Blotting is representative of 6 independent mice. (C and D) APAP increased serum levels of IL-1β and IL-18 (*P < 0.05, compared to non-APAP-treated controls), which are both cytokines activated from the proform by cleaved caspase-1. BA treatment limited APAP-induced levels of these cytokines (#P < 0.05, compared to APAP-treated mice). (E and F) BA treatment limited the release if IL-1β and IL-18 in WT mice (*P < 0.05), but not in TLR4−/− mice. (G) Caspase-1 cleavage was limited by BA in WT mice, but not in TLR4−/− mice.

There was no IL-18 or IL-1β baseline difference between TLR4−/− and control mice (data not shown). Interestingly, we found that, compared to the effective reduction in WT mice, BA cotreatment in TLR4−/− mice did not significantly reduce levels of serum IL-18 (Fig. 7E) or IL-1β (Fig. 7F). Moreover, BA, which reduced caspase-1 cleavage in WT mice, failed to reduce cleavage in TLR4−/− mice (Fig. 7G), which further implied that BA's inhibition in hepatic inflammasome signaling was TLR4 dependent. Taken together, these data suggested that BA could protect APAP-induced mouse ALF by decreasing inflammasome signaling through a TLR4-dependent mechanism.


The findings noted in this article highlight the roles of direct hepatocellular toxicity, as well as innate immune-signaling pathways, most notably TLR9, RAGE, Nalp3, and caspase-1, in mediating APAP-induced liver injury. APAP liver toxicity was also associated with release of HMGB1, which was reduced in transgenic (Tg) mice lacking TLR9 and RAGE receptors. Mice with hepatocyte-specific KO of HMGB1 were also protected against liver injury, suggesting a critical role for HMGB1 in this model. Furthermore, BA was found to be a molecule that limits APAP-induced liver injury. The mechanisms of protection of BA include direct effects on hepatocyte injury, as well as through the PRR, TLR4, specifically on myeloid cells.

A small percentage of APAP is metabolized by cytochrome p450s (CYPs) to the potentially toxic intermediate, N-acetyl-p-benzoquinone imine (NAPQI).[31] This can cause mitochondrial injury, leading to production of ROS and stress signaling.[14, 32] The results noted in this article are consistent with these findings, demonstrating mitochondrial injury and oxidant stress. BA had an effect in vivo as well as a direct effect on hepatocytes in vitro. Possible mechanisms underlying these effects include direct influence on APAP metabolism by CYPs. BA can be a product of toluene metabolism by the same CYPs that metabolize APAP to NAPQI.[33] It is possible that there is an inhibitory effect of BA on CYP function. Other potential mechanisms may include other signaling pathways that influence mitochondrial health.

A component of APAP-induced liver failure may be secondary to sterile injury from the innate immune response through DAMPs released from stressed or dying cells to amplify liver damage.[4, 10] DAMPs activate a group of PRRs, including TLRs, nucleotide-binding oligomerization domain (NOD)-like receptors, and RAGE. Specifically, in models of APAP-induced liver injury, Imaeda et al. found that hepatocyte DNA was a trigger of the innate immune response through TLR9 and Nalp3 signaling.[22] This led to the activation of caspase-1 to result in cleavage of proinflammatory cytokines to active IL-1β and IL-18. Inhibition of either TLR9 signaling or Nalp3 activity or genetic depletion markedly attenuated liver injury and improved survival. Findings in this study were consistent with these previous reports. Thus, the products released from injured cells from the toxic affects of APAP can then potentially activate innate immune signaling and inflammation, which may serve to potentially exacerbate or ameliorate injury. The findings in this study and others that show increased plasma levels of mtDNA and HMGB1 suggest that these DAMPS released from injured hepatocytes may then be released to activate immune signaling.

Among the DAMPs in liver injury, HMGB1 has been shown to play a vital role, including in APAP-treated mice.[34] HMGB1 release itself or as a complex with DNA can activate TLR2, 4, and 9. We and others have shown that HMGB1 neutralization or inhibition were protective in HIR, APAP-induced liver injury, or hemorrhagic shock and trauma.[5, 7, 9, 19, 21] In this current study, HMGB1 release was greatly attenuated by BA treatment, and HC-HMGB1 KO mice demonstrated remarkably reduced liver damage after APAP treatment. This suggests that HMGB1 signaling from hepatocytes amplifies liver injury in this model. Moreover, genetic absence of RAGE, which is a receptor of HMGB1, also reduced liver injury.

TLR4 is one of the most important sensors in innate immunity, for both pathogen-associated molecular patterns and DAMPs. Previous studies demonstrate that depletion of TLR4 protected against sterile and nonsterile liver injury. Specifically, investigators have demonstrated, in models of APAP toxicity, that mice-deficient TLR4 signaling were protected,[35, 36] whereas data presented in this article found that TLR4 signaling did not seem to be involved in APAP-induced liver injury. The differences may be attributed to use of KO mice used in this study, whereas previous studies utilized mice with mutated TLR4 receptors. More intriguing, the data presented in this article suggest that TLR4 serves as an essential receptor to mediate the protective effects of BA to limit liver injury and inflammasome activation. Clearly, the role and interplay of danger-signaling pathways and PRRs is complex. Furthermore, the cell-specific KO data in TLR4 implies a complex interplay of different cells in mediating APAP-induced liver injury. In warm HIR injury, the specific absence of TLR4 on hepatocytes and myeloid cells showed reduced liver injury, whereas DC-specific TLR4 KO exacerbated damage.[37] The specifics of injury and the interplay of different cells and signaling pathways are seemingly highly specific to the individual insults that result in liver injury. The influence of BA through TLR4 signaling may be predominantly to prevent amplification of injury that occurs secondary to release of DAMPs from hepatocytes. The influence of BA on TLR4 signaling may negatively regulate other activated innate immune-signaling pathways.

BA has been used as a local anesthetic, and use of BA (5%) is approved by the U.S. Food and Drug Administration in treatment of head lice. High doses of BA have toxicity, including respiratory failure, vasodilation, hypotension, convulsions, and paralysis.[38] Toxicity from BA is a potential concern, and these studies demonstrated a toxic dose in the APAP model in the dose escalation study, suggesting a therapeutic window. BA is a true neutral, water-soluble molecule that readily partitions into lipid bilayers to increase fluidity. BA can inhibit mitochondrial electron transport at several points along the respiratory chain,[39] which may be relevant to the suggested findings in this study of protection against APAP-induced mitochondrial injury.

In summary, our study suggested BA to be a promising reagent for APAP-induced acute hepatic toxicity prevention and treatment. The protective mechanism was, at least in part, mediated by myeloid cell signaling by TLR4 as well as through direct effects on hepatocytes.