Potential conflict of interest: Nothing to report.
This study was supported, in part, by National Institutes of Health (NIH) funds R01 AA020518-01, R21 AA017421, P20 RR021940, and P20 RR016475 from the IDeA Networks of Biomedical Research Excellence program of the National Center for Research Resources (to W.X.D.). H.J. was supported by NIH R01 DK070195.
Autophagy can selectively remove damaged organelles, including mitochondria, and, in turn, protect against mitochondria-damage–induced cell death. Acetaminophen (APAP) overdose can cause liver injury in animals and humans by inducing mitochondria damage and subsequent necrosis in hepatocytes. Although many detrimental mechanisms have been reported to be responsible for APAP-induced hepatotoxicity, it is not known whether APAP can modulate autophagy to regulate hepatotoxicity in hepatocytes. To test the hypothesis that autophagy may play a critical protective role against APAP-induced hepatotoxicity, primary cultured mouse hepatocytes and green fluorescent protein/light chain 3 transgenic mice were treated with APAP. By using a series of morphological and biochemical autophagic flux assays, we found that APAP induced autophagy both in the in vivo mouse liver and in primary cultured hepatocytes. We also found that APAP treatment might suppress mammalian target of rapamycin in hepatocytes and that APAP-induced autophagy was suppressed by N-acetylcysteine, suggesting APAP mitochondrial protein binding and the subsequent production of reactive oxygen species may play an important role in APAP-induced autophagy. Pharmacological inhibition of autophagy by 3-methyladenine or chloroquine further exacerbated APAP-induced hepatotoxicity. In contrast, induction of autophagy by rapamycin inhibited APAP-induced hepatotoxicity. Conclusion: APAP overdose induces autophagy, which attenuates APAP-induced liver cell death by removing damaged mitochondria. (HEPATOLOGY 2012)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Acetaminophen (APAP) is a safe, effective analgesic at therapeutic doses. However, an overdose can cause liver injury and even liver failure in animals1 and in humans.2 The mechanism of APAP-induced liver injury includes the formation of a reactive metabolite (N-acetyl-p-benzoquinone imine; NAPQI), which depletes glutathione (GSH) and binds to cellular proteins.1 The only clinically used antidote, N-acetylcysteine (NAC), acts as a prodrug for GSH synthesis and is, therefore, most effective when administered very early, when it prevents protein binding of NAPQI;3 however, later mechanisms of protection have also been identified.4
Although protein binding is a critical initiating event in the pathophysiology, it is insufficient to cause the massive cell injury typical of APAP overdose.5 Beginning with the identification of mitochondrial protein binding,6 inhibition of mitochondrial respiration,7 adenosine triphosphate (ATP) depletion,8 and the occurrence of a selective mitochondrial oxidant stress8 during APAP hepatotoxicity, the concept emerged that mitochondrial dysfunction and damage is a central propagation event responsible for cell necrosis. The oxidant stress leads to mitochondrial peroxynitrite formation, which causes mitochondrial DNA damage and nitration of mitochondrial proteins9 and triggers membrane permeability transition and subsequent cell necrosis.10, 11 Given this central role of mitochondria in the pathophysiology and the fact that not all mitochondria are affected at the same time, it is feasible that removal of damaged mitochondria might be beneficial.
Macroautophagy (hereafter referred to as autophagy) is a bulk intracellular degradation system that is mainly responsible for the degradation of long-lived proteins and other cellular contents. Autophagy involves the formation of double-membrane autophagosomes, which enwrap cytoplasm and organelles and then fuse with lysosomes, thus degrading the enveloped contents. This process is tightly regulated and highly inducible. More than 30 Atg genes have been defined to participate in autophagy or autophagy-related processes.12, 13 Under stress conditions, such as nutrient starvation, autophagy is induced largely as a result of the inhibition of mammalian target of rapamycin (mTOR) complex 1, a kinase complex that works as a nutrient sensor to initiate autophagy by activating ULK1/Atg1. Then, two ubiquitin-like conjugation systems, including Atg7 (E1-like), Atg3, and Atg10 (E2-like) and the Atg5-Atg12-Atg16 complex, promote the conjugation of light chain 3 (LC3), a mammalian homolog of yeast Atg8, with phosphatidylethanolamine (PE). The PE-conjugated form of LC3 (LC3-II) translocates to the autophagosomal membrane and promotes the formation a double-membrane autophagosome. In addition, Beclin 1/Atg6 forms a complex with VPS34, VPS15, and Atg14. VPS34 is a class III phosphoinositide 3-kinase (PI3K) required for autophagy induction. 3-Methyladenine (3-MA), a widely used autophagy inhibitor, inhibits the type III PI3K and autophagosome formation.14 Chloroquine (CQ), a clinically used antimalarial drug, suppresses autophagy by increasing lysosomal pH.15, 16
Autophagy is usually activated as a survival mechanism in response to an adverse environment, such as the deprivation of nutrients or growth factors.17 However, autophagy may also be involved in the pathogenesis of a number of human diseases.18, 19 Most important, autophagy can eliminate damaged mitochondria and maintain mitochondrial homeostasis (i.e., mitophagy).20 We have recently shown that removal of damaged mitochondria by mitophagy reduces alcohol-induced liver injury.21 Although mitochondrial damage is a key cellular event that contributes to APAP-induced hepatotoxicity, it is not known whether APAP can modulate autophagy in hepatocytes. Therefore, the aim of the current investigation was to assess the occurrence of autophagy in APAP-induced liver cell injury in vivo and in vitro and evaluate its potential pathophysiological relevance. We found that APAP induced autophagy both in the mouse liver and in primary cultured mouse hepatocytes. Moreover, pharmacological suppression of autophagy exacerbated APAP-induced liver injury, whereas induction of autophagy protected against APAP-induced liver injury.
3-MA, 3-methyladenine; p70S6K, 70-kDa ribosomal protein S6 kinase-1; ActD, actinomycin D; ALT, alanine aminotransferase; APAP, acetaminophen; ATP, adenosine triphosphate; CQ, chloroquine; CYP2E1, cytochrome P450 2E1; Cyto c, cytochrome c; DMSO, dimethyl sulfoxide; EM, electron microscopy; GFP, green fluorescent protein; GSH, glutathione; H&E, hematoxylin and eosin; HMGB1, high-mobility group box 1 protein; HSP 60, heat shock protein 60; IP, intraperitoneal; LC3, light chain 3; mTOR, mammalian target of rapamycin; NAC, N-acetylcysteine; NAPQI, N-acetyl-p-benzoquinone imine; PE, phosphatidylethanolamine; PFA, paraformaldehyde; PI, propidium iodide; PI3K, phosphoinositide 3-kinase; ROS, reactive oxygen species; SEM, standard error of the mean; 4-EBP1, translational initiation factor 4E binding protein-1; TNF-α, tumor necrosis factor alpha.
Materials and Methods
C57BL/6 wild-type and green fluorescent protein (GFP)-LC3 transgenic mice, as well as isolated primary mouse hepatocytes, were used in this study. For in vivo studies, mice were either given saline (intraperitoneal; IP) or APAP (500 mg/kg, IP). Induction or suppression of autophagy was achieved by injection (IP) of rapamycin (2 mg/kg) or CQ (60 mg/kg). For in vitro studies, primary cultured hepatocytes were treated with various concentrations of APAP at different time points. Autophagic flux in APAP-treated cells was determined by quantifying the number of GFP-LC3 puncta, LC3-II levels, number of autophagosomes (electron microscopy; EM), as well as p62 degradation with or without the lysosomal inhibitor, CQ. Necrosis was determined by propidium iodide (PI) staining and immunostaining for high mobility group box 1 protein (HMGB1). Liver injury was assessed by hematoxylin and eosin (H&E) staining of liver sections and serum alanine aminotransferase (ALT) activities.
For additional details on methods, please refer to the Supporting Materials.
APAP Induces Autophagy in Mouse Liver.
Activation of autophagy by APAP was first examined in GFP-LC3 transgenic mice. In agreement with previous reports that APAP induced liver injury,4, 8 APAP treatment induced a significant elevation of serum ALT in GFP-LC3 transgenic mice (Fig. 1A). Interestingly, APAP treatment also significantly increased GFP-LC3 puncta in the liver (Fig. 1B), which represent autophagosomes. Immunoblotting analysis confirmed the increase of the membrane-associated, PE-conjugated form of GFP-LC3 (GFP-LC3-II) in the APAP-treated mouse liver (Fig. 1C). Importantly, EM analysis indicated an increased accumulation of autophagosomes after APAP treatment (Fig. 1D). Interestingly, the double-membrane autophagosomes often had enveloped mitochondria, suggesting that APAP-induced autophagy may help remove the damaged mitochondria caused by APAP (Fig. 1D, panels c,d).
APAP Induces Autophagy in Primary Hepatocytes.
We next examined the effects of APAP on primary cultured mouse hepatocytes. Consistent with the in vivo studies, APAP treatment also increased GFP-LC3 puncta in primary cultured mouse hepatocytes in a time- and dose-dependent manner (Fig. 2A,B), indicating a significant accumulation of autophagosomes in APAP-treated hepatocytes. APAP treatment also increased endogenous LC3-II levels. Importantly, p62, an autophagy substrate, which is normally degraded during autophagy, was degraded dramatically by APAP treatment in a time- and dose-dependent manner (Fig. 2C), indicating APAP induces autophagic flux. After 6 hours treatment, all the various concentrations of APAP only mildly increased necrotic cell death (approximately 6%-8%; Supporting Fig. 1). Similar to the in vivo studies, EM studies revealed that there was an increased amount of autophagosomes in APAP-treated cells, compared to the nontreated control cells (Fig. 2D,E). Enlarged photographs revealed that APAP-induced autophagosomes contained intracellular contents, including mitochondria (Fig. 2D, panel c; Supporting Fig. 1B), suggesting APAP induces mitophagy. Indeed, results from western blotting analysis revealed that APAP induced marked mitochondrial protein degradation, including heat shock protein 60 (HSP 60, a mitochondrial matrix protein), cytochrome c (Cyto c, a mitochondrial intermembrane space protein), and Tom20 (an outer mitochondrial membrane protein) (Fig. 2F), further supporting that APAP induced mitophagy in hepatocytes.
Effects of APAP on Autophagic Flux and mTOR in Primary Hepatocytes.
In addition to the determination of the degradation of p62, autophagic flux can also be assessed by using a combination of lysosomal inhibitors, such as CQ.15 We found that in the presence of CQ, the average number of GFP-LC3 dots per cell induced by APAP was further increased than either APAP or CQ treatment alone (Fig. 3A,B), although there was no statistical difference, compared to the CQ-treatment–alone group. APAP-induced degradation of p62 was partially rescued by CQ in the APAP lower dose (2.5 mM) group, but not in the APAP higher dose group (5 mM). However, combined CQ and APAP treatment did not further increase the level of endogenous LC3-II (Fig. 3C). One possible explanation for these observations is likely the increased necrosis when hepatocytes were treated with both CQ and APAP (see below), which may suppress the further induction of autophagy.
Because suppression of mTOR is one central molecular signaling pathway leading to autophagy induction, we next determined whether APAP treatment would affect mTOR activity in primary hepatocytes. We found that APAP treatment significantly decreased the levels of phosphorylated 70-kDa ribosomal protein S6 kinase-1 (p70S6K) and translational initiation factor 4E binding protein-1 (4EBP-1), two well-known downstream phosphorylation targets of mTOR, in a dose-dependent manner (Fig. 3D). Taken together, these data support that APAP induces autophagy and is involved in the suppression of mTOR in hepatocytes.
NAC and 3-MA Suppress APAP-Induced GFP-LC3 Puncta.
To determine whether APAP-induced protein binding and reactive oxygen species (ROS) would contribute to APAP-induced autophagy, we treated hepatocytes with APAP in the presence of NAC. 3-MA, the class-III PI3K inhibitor, which has been shown to suppress autophagy in many models, suppressed APAP-induced GFP-LC3 puncta (Fig. 4A,B). In addition, in the presence of NAC, APAP-induced GFP-LC3 puncta were also completely blocked (Fig. 4A,B). These results suggest that APAP-induced autophagy involves its metabolic activation and ROS production as well as the class-III PI3K. To further determine whether suppression of autophagy would influence APAP-induced ROS production, we treated hepatocytes with APAP in the absence or presence of CQ for 6 hours and stained the cells with 2′7′-dichlorofluorescein diacetate, a fluorescence dye widely used to assess ROS.24 We found that APAP treatment alone increased ROS production, but suppression of autophagy by CQ further enhanced APAP-induced ROS production (Supporting Fig. 2).
APAP Induces Typical Necrosis in Primary Hepatocytes.
To determine the nature of the cell death induced by APAP, we loaded the cells with PI, which stains nuclei only after the cellular plasma membranes are permeabilized. At 24 hours after treatment with APAP (10 mM), more than 80% of hepatocytes had PI-positive nuclei, suggesting the permeabilization or disruption of the hepatocytes' plasma membrane. We previously showed that tumor necrosis factor alpha (TNF-α) induces apoptosis in primary hepatocytes in the presence of actinomycin D (ActD).22 It could be observed that the apoptotic cells had shrunk and were often detached with multiple cellular blebbings in TNF-α/ActD-treated cells, whereas APAP-treated hepatocytes were swollen and often still attached (Supporting Fig. 3A, panels b,c). Although most TNF-α/ActD-treated cells eventually showed PI-positive nuclei resulting from the secondary necrosis that occurs during culture conditions, all PI-positive nuclei were fragmented (Supporting Fig. 3A, panel b). In contrast, in APAP-treated cells, all PI-positive nuclei were intact without fragmentation (Supporting Fig. 3A, panels c,d). By quantifying the PI-positive cells with intact nuclei as necrotic cells, we found that APAP-induced necrosis increased in a dose-dependent manner (Supporting Fig. 3B). APAP-induced necrosis was further determined by EM studies. Almost all the cytosolic content and organelles were disrupted with numerous cytosolic vesicles in APAP-treated necrotic cells (Supporting Fig. 3C). During necrotic, but not apoptotic, cell death, HMGB1, a nuclear protein that binds to chromatin, is released from nuclei.23 We determined whether APAP treatment would also induce HMGB1 release from nuclei. In control hepatocytes, HMGB1 is exclusively located in the nuclei, as demonstrated by its colocalization with nuclear staining by Hoechst 33342. However, HMGB1 displayed a marked cytosolic pattern in APAP-treated cells, indicating that APAP induces HMGB1 release (Fig. 5A). The percentage of cells with the release of HMGB1 from the nuclei in APAP-treated cells increased in a dose-dependent manner (Fig. 5B). These results indicate that APAP induced necrosis, but not apoptosis, in primary hepatocytes.
Autophagy Reduces APAP-Induced Hepatotoxicity In Vitro.
Because we observed that both 3-MA and CQ suppressed APAP-induced autophagy, we next determined the cell viability of cells treated with various concentrations of APAP in the presence of 3-MA or CQ. Whereas inhibition of basal autophagy alone by 3-MA or CQ did not affect cell death in hepatocytes (Fig. 6A,B), both 3-MA and CQ significantly increased APAP-induced necrosis, indicating that APAP-induced autophagy is protective against cell death. Interestingly, suppression of autophagy clearly induced a shift in the dose response of APAP-induced necrosis, even in the low dose of APAP (1.25 mM). In contrast, cotreatment of APAP with rapamycin, a mTOR inhibitor and autophagy inducer, significantly inhibited APAP-induced necrosis (Supporting Fig. 3D; Fig. 6C). The protective effects of rapamycin are not the result of the solvent of dimethyl sulfoxide (DMSO), because DMSO alone did not affect APAP-induced necrosis in primary cultured hepatocytes (Fig. 6C). Rapamycin (2 μM) alone decreased the level of phosphorylated p70S6K, as well as increased LC3-II and p62 degradation, suggesting that rapamycin induces autophagy in primary cultured hepatocytes (Fig. 6D). APAP is metabolized via cytochrome P450 2E1 (CYP2E1) in the first 2-3 hours in hepatocytes after its exposure and generates the reactive metabolite, NAPQI.5 To test whether induction of autophagy would still be protective after the initial phase of APAP metabolism, we treated hepatocytes with APAP for 3 hours, then added rapamycin. Even the delayed treatment with rapamycin significantly attenuated APAP-induced necrosis (Fig. 6E). Therefore, induction of autophagy protects against APAP-induced necrosis, whereas inhibition of autophagy further exacerbates it.
Induction of Autophagy Protects Against APAP-Induced Liver Injury In Vivo.
We next determined whether modulation of autophagy would affect APAP-induced liver injury in vivo. C57BL/6 mice were treated with APAP in the absence or presence of CQ and rapamycin for 6 hours. Centrilobular necrosis was evident in livers of mice treated with APAP, which was further exacerbated by CQ treatment, but almost abolished in rapamycin-treated mice (Fig. 7A). Furthermore, CQ treatment also further increased APAP-induced serum ALT levels (Fig. 7B). In agreement with previous reports, DMSO alone (which was used to dissolve rapamycin) also partially reduced APAP-induced liver injury, likely because of its inhibitory effects on CYP2E1.25 However, treatment with rapamycin completely eliminated APAP-induced liver injury (Fig. 7A,C). Because hepatic GSH depletion 2 hours after APAP was similar in DMSO- and rapamycin-treated animals (Supporting Fig. 4), the protection was not caused by inhibition of metabolism. Thus, these data indicate that induction of autophagy protects against APAP-induced liver injury in vivo.
mTOR is an evolutionarily conserved, nutrient-sensing serine/threonine protein kinase that plays a critical role in regulating protein synthesis and autophagy.26, 27 mTOR regulates the phosphorylation of at least two proteins important for translation: p70S6K and 4EBP1. mTOR signaling negatively regulates autophagy, and mTOR suppression by rapamycin contributes to the induction of autophagy,27 which is in agreement with our present study in primary cultured mouse hepatocytes (Fig. 6). We found that APAP decreased the phosphorylation levels of p70S6K and 4EBP1 in hepatocytes (Figs. 3D and 6D), suggesting that APAP may inhibit mTOR. How APAP suppresses mTOR in hepatocytes requires further studies.
In addition to the mTOR-signaling pathway, we also found that cotreatment of NAC with APAP almost completely suppressed APAP-induced GFP-LC3 puncta (Fig. 4). APAP is metabolized mainly by the CYP2E1 isoform of CYP to NAPQI, which depletes intracellular GSH and covalently binds to proteins, including many mitochondrial proteins. This triggers mitochondrial damage and production of ROS.5 NAC treatment promotes GSH synthesis in the cytosol and supports the recovery of mitochondrial GSH levels and, therefore, improves the scavenging capacity for NAPQI and for mitochondrial ROS.3, 4 The inhibitory effects of NAC on APAP-induced autophagy suggest that the metabolic activation of APAP is required for APAP-induced autophagy. Accumulating evidence now supports that ROS from damaged mitochondria can induce autophagy. For example, inhibition of mitochondrial respiration by mitochondrial electron transport chain inhibitors, such as rotenone (complex I inhibitor) or thenoyl trifluoroacetone (complex II inhibitor), induces autophagy.28, 29 We also found that carbonyl cyanide m-chlorophenylhydrazone, a mitochondrial uncoupler, also induces autophagy in various mammalian cell lines. In the latter case, autophagy is also suppressed by NAC.29 Although the exact mechanisms are still unclear, it is proposed that ROS may induce autophagy by modulating the action of Atg4B on LC3. Atg4B is a cysteine protease that not only cleaves the full-length LC3 to generate LC3-I, but also delipidates LC3-II from the autophagosomal outer membrane to allow the recycling of LC3. ROS can modulate Atg4B activity by directly oxidizing a specific cysteine residue, Cys81, on Atg4B.30 It remains to be studied how exactly NAC suppresses APAP-induced autophagy.
It is well known that NAC can protect against APAP-induced liver injury,3, 4 and we also confirmed that NAC suppressed APAP-induced necrosis and liver injury (data not shown). Why would NAC suppress APAP-induced autophagy, but also protect against APAP-induced liver injury? Early treatment of NAC results in rapid GSH formation, which enhances the capacity to scavenge NAPQI and thus prevents all upstream-initiating events, such as protein binding and mitochondrial ROS formation.3, 4 Therefore, NAC prevents toxicity and also eliminates the stress that triggers the induction of autophagy.
In addition to its nonselective bulk degradation, accumulating evidence now suggests that macroautophagy can be selective. One of the well-studied forms of selective macroautophagy is mitophagy, a process where LC3-positive compartments enwrap mitochondria.20 In the present study, we have found that APAP-induced autophagosomes often contain mitochondria, both in the mouse liver and in primary cultured hepatocytes (Figs. 1D and 2D; Supporting Fig. 1B). Moreover, many mitochondrial proteins were degraded in APAP-treated hepatocytes, supporting the conclusion that APAP-induced autophagy removes damaged mitochondria (Fig. 2F), which are the major source of intracellular ROS during APAP hepatotoxicity (Supporting Fig. 2).4, 8 Timely removal of damaged mitochondria by mitophagy plays an important role in regulating cell death.31
Because we observed that rapamycin almost completely blocked APAP-induced liver injury in mouse livers, it is possible that rapamycin may have affected the metabolic activation of APAP. However, the >90% depletion of hepatic GSH levels in both APAP/rapamycin and APAP/DMSO-treated animals at 2 hours (i.e., before significant cell death; Supporting Fig. 4) suggests that the protection by rapamycin was not caused by inhibition of APAP metabolism, but was the result of induction of autophagy. Although we have demonstrated that rapamycin could be a potential therapeutic drug for treating APAP-induced liver injury, it should also be noted that rapamycin has immunosuppression activity, in addition to inducing autophagy. However, with the recent rapid progress on small-molecule screening for autophagy inducers, future work should examine the protective effects against APAP-induced liver injury using more specific novel autophagy inducers.
In conclusion, we have demonstrated that APAP overdose induces autophagy in both primary cultured mouse hepatocytes and in the mouse liver. Autophagy protects against APAP-induced hepatotoxicity. This protection could be mediated via the removal of damaged mitochondria and thus a reduction to a major intracellular source of ROS production. The cellular and molecular events regulating APAP-induced necrosis and autophagy are summarized in Fig. 8. These findings imply that induction of autophagy could be a novel therapeutic approach to mitigate APAP-induced hepatotoxicity and liver injury.