Dendritic cell depletion exacerbates acetaminophen hepatotoxicity


  • Potential conflict of interest: Nothing to report.

  • Supported in part by grants from a Liver Scholar Award from the American Liver Foundation (to G.M.), a Society of University Surgeons Junior Faculty Grant (to G.M.), and National Institute of Health Award DK085278 (to G.M.) and CA155649 (to G.M.).


Acetaminophen (APAP) overdose is one of the most frequent causes of acute liver failure in the United States and is primarily mediated by toxic metabolites that accumulate in the liver upon depletion of glutathione stores. However, cells of the innate immune system, including natural killer (NK) cells, neutrophils, and Kupffer cells, have also been implicated in the centrilobular liver necrosis associated with APAP. We have recently shown that dendritic cells (DCs) regulate intrahepatic inflammation in chronic liver disease and, therefore, postulated that DC may also modulate the hepatotoxic effects of APAP. We found that DC immune-phenotype was markedly altered after APAP challenge. In particular, liver DC expressed higher MHC II, costimulatory molecules, and Toll-like receptors, and produced higher interleukin (IL)-6, macrophage chemoattractant protein-1 (MCP-1), and tumor necrosis factor alpha (TNF-α). Conversely, spleen DC were unaltered. However, APAP-induced centrilobular necrosis, and its associated mortality, was markedly exacerbated upon DC depletion. Conversely, endogenous DC expansion using FMS-like tyrosine kinase 3 ligand (Flt3L) protected mice from APAP injury. Our mechanistic studies showed that APAP liver DC had the particular capacity to prevent NK cell activation and induced neutrophil apoptosis. Nevertheless, the exacerbated hepatic injury in DC-depleted mice challenged with APAP was independent of NK cells and neutrophils or numerous immune modulatory cytokines and chemokines. Conclusion: Taken together, these data indicate that liver DC protect against APAP toxicity, whereas their depletion is associated with exacerbated hepatotoxicity. (HEPATOLOGY 2011;)

Acetaminophen (APAP) is a widely used over-the-counter analgesic and antipyretic agent. Although usually considered safe at therapeutic doses, at higher doses APAP causes acute liver failure, characterized by centrilobular hepatic necrosis. APAP-induced hepatoxicity is the leading cause of acute liver failure, accounting for nearly 50% of all cases.1-4 The spectrum of APAP-related liver failure ranges from individuals taking an intentional overdose (42%) to accidental overdose (49%).3, 4 In the United States, APAP overdose is a major burden to the healthcare system. It is responsible for over 56,000 emergency room visits, 2,600 hospitalizations, and over 450 deaths from acute liver failure annually.1, 3

APAP is a dose-dependent hepatotoxin. When taken at therapeutic doses, over 90% of APAP is metabolized by gluconylation and sulphation and its metabolites are rapidly excreted in the urine. Of the remaining APAP, approximately 2% is excreted intact in urine, and 5%-9% is metabolized by the cytochrome P450 system to N-acetyl-p-benzo-quinoneimine (NAPQ1), a highly reactive metabolite.1, 5, 6 At therapeutic doses of APAP, hepatic glutathione (GSH), a major intracellular antioxidant, induces the formation of a safely excretable APAP-protein adduct. However, at toxic doses of APAP, GSH becomes overwhelmed and severely depleted in both the cytoplasm and mitochondria.1, 7 Once GSH is depleted, NAPQ1 is able to exert its harmful effects by forming covalent bonds with cellular proteins. Covalent bonding to mitochondrial proteins causes mitochondrial dysfunction by inhibition of the Ca2+-Mg2+-ATPase, resulting in accumulation of cytosolic calcium. This disturbance leads to a decrease in ATP synthesis, disruption of cellular membrane, and eventually necrotic cell death.1, 7-9

Although toxic metabolites of APAP account for the primary hepatic insult, the liver's innate immune system has also been shown to play a major role in APAP-induced liver injury in what is akin to a “two-hit” mechanism. That is, although GSH depletion and the resulting toxic metabolites are prerequisites for APAP hepatotoxicity, there is evidence that the severity of liver injury may depend on subsequent downstream participation of inflammatory mediators.1, 10-17 Natural killer (NK) and natural killer T-cell (NKT) activation have been purported to be a crucial component in the progression of APAP-induced hepatotoxicity.12 Hepatic NK and NKT cells are a major source of interferon gamma (IFN-γ), which has been shown to mediate hepatocyte apoptosis, leukocyte infiltration, as well as cytokine and chemokine production in APAP-induced liver injury.11 However, more recent evidence suggests that NK cells are less critical to APAP toxicity.15 Kupffer cells have also been shown to contribute to APAP-mediated hepatotoxicity. Michael et al.16 showed that mice treated with gadolinium chloride, a deactivator of macrophages, had dramatically decreased APAP-induced liver injury. Kupffer cells are thought to exacerbate liver injury by increasing the synthesis of oxygen free radicals.16 However, there is also evidence to the contrary. Ju et al.13 found that following depletion of Kupffer cells, APAP-induced liver injury was exacerbated. The mechanism was purported to be related to decreased expression of several hepatoregulatory cytokines, including interleukin-10 (IL-10), which functions to limit inducible nitric oxide synthase expression and peroxynitrite-induced liver injury.13 The role of neutrophils in APAP-induced hepatotoxicity is also controversial. Liu and Kaplowitz18 demonstrated that depletion of neutrophils using RB6-8C5 protected mice against APAP-induced liver injury, as evidenced by reduced serum alanine aminotransferase (ALT) levels, decreased centrilobular necrosis, and improved survival. In contrast, Lawson et al.14 demonstrated that neutrophils contribute to the removal of necrotic debris rather than directly affecting the pathogenesis of APAP-induced injury. Thus, it appears that secondary induction of the hepatic innate immune system likely plays a part in APAP-induced liver injury; however, the precise role of its components remains incompletely understood.

Dendritic cells (DCs) are the principal antigen-presenting cells in lymphoid organs and in the periphery, including the liver, and initiate both innate and adaptive immune responses.19, 20 We and others have shown that liver DCs are characterized by their immaturity and more commonly mediate tolerance rather than immunogenicity.21-24 DCs are primarily responsible for hepatic, oral, and portal venous tolerance22, 23 and have been purported to play a role in sundry other aspects of hepatic tolerance including the acceptance of liver allografts with minimal immune suppression and the frequent deposition of hepatic tumor metastases.24 However, whereas in their steady-state liver, DCs have tolerogenic properties, there is emerging evidence for a reversal of their immune-phenotype after hepatic insult. We recently reported that in chronic liver fibrosis DC mature in vivo, become highly immunogenic, and govern intrahepatic inflammation by way of production of tumor necrosis factor alpha (TNF-α), thereby modulating activation of NK cells and liver T cells.25 Our preliminary investigations also showed that DC become proinflammatory after APAP challenge. Based on these data, we postulated that in acute liver injury induced by APAP, DCs play a central role in exacerbating secondary hepatic injury. Our results confirm an important role for DCs in APAP, but suggest that rather than worsening liver insult, DCs are protective.


ALT, alanine aminotransferase; APAP, acetaminophen; APAP-DC, acetaminophen-treated with depletion of dendritic cells; AST, aspartate aminotransferase; DAMP, damage-associated molecular patterns; DC, dendritic cells; Flt3L, FMS-like tyrosine kinase 3 ligand; GSH, glutathione; NAPQ1, N-acetyl-p-benzo-quinoneimine; NPC, nonparenchymal cells.

Materials and Methods

Animals and In Vivo Models.

Male C57BL/6 (H-2Kb), CD11c-DTR, OT-I (B6.Cg-RAG2tm1Fwa-TgN), and OT-II (B6.Cg-RAG2tm1Alt-TgN) mice (4-8 weeks old) were purchased from Taconic Farms (Germantown, NY) and then bred in-house. Mice were housed in a pathogen-free environment. To induce acute hepatic injury, mice were treated intraperitoneally with 500 μg/g APAP diluted in phosphate-buffered saline (PBS). In selected experiments, mice were treated for 10 days with Flt3L (10 μg; Celldex, Fall River, MA) before APAP challenge. To effect DC depletion, CD11c.DTR mice were treated with a single intraperitoneal dose of diphtheria toxin (4 ng/g; Sigma-Aldrich, St. Louis, MO). In vivo plasmacytoid DC depletion was accomplished using 120G8 (200 μg; Imgenex, San Diego, CA).26 In selected experiments, the novel immune-modulator VAG539 (30 mg/kg, Novartis, Basel, Switzerland) was used to partially inactivate DC.27 To deplete Gr1+ cells, RB6-8C5 was employed (150 μg/day; Monoclonal Antibody Core Facility, Sloan-Kettering Institute, New York, NY). For in vivo NK cell depletions, 100 μL of a 1:5 dilution of anti-asialo GM1 (Wako Chemical, Richmond, VA) was injected intraperitoneally 3 days prior to APAP treatment. In selected experiments, antibodies directed against IFN-α (1 μg, F18, Sigma), TNF-α (200 μg, AB-410, R&D, Minneapolis, MN), IL-6 (200 μg, AB-406, R&D), or MCP-1 (50 μg, AB-479, R&D) were administered in vivo before APAP challenge. Changes in serum liver enzymes, including ALT, aspartate aminotransferase (AST), were determined using the Olympus AU400 Chemistry Analyzer (Center Valley, PA). In survival experiments, animals were euthanized when they were moribund and death was imminent. Animal procedures were approved by the New York University School of Medicine Animal Care and Use Committee.

Cellular Isolation.

Liver DC were isolated as described.25 Briefly, immediate postmortem laparotomy was performed and the portal vein was cannulated and infused with 1% Collagenase IV (Sigma-Aldrich). Hepatectomy was then performed and livers mechanically minced before incubation with Collagenase IV at 37° for 10 minutes. Low-speed (30g) centrifugation was performed to exclude the pelleted hepatocytes followed by high-speed (300g) centrifugation to isolate the hepatic nonparenchymal cells (NPCs). The NPC were then further enriched over a 40% Optiprep (Sigma-Aldrich) density gradient. To purify the DC population, NPCs were incubated with 1 μg of anti-FcγRIII/II (2.4G2, Fc block; Monoclonal Antibody Core Facility, Sloan-Kettering Institute) per 106 cells, labeled with fluorescently conjugated anti-CD11c and anti-MHC II (both BD Biosciences, Franklin Lakes, NJ), and FACS sorted using a MoFlo cell sorter (Beckman Coulter, Fullerton, CA). Splenocytes were prepared by mechanical disruption and specific cellular subgroups were purified by FACS.

T-Cell Proliferation Assays.

For in vitro T-cell proliferation assays, peptide-pulsed DC (3 × 104) were added to CD8+OT-I TCR-transgenic T cells (1 × 105) specific for Ova257-264, or CD4+OT-II TCR-transgenic T cells specific for Ova323-339 in 96-well plates for 48-72 hours before pulsing with 3H-thymidine as described.25 DCs were loaded with the relevant Ova peptide (10 μg/mL; AnaSpec, San Jose, CA) for 90 minutes before co-culture with respective T cells. In selected experiments, VAF347 (5 mM, Novartis), a low-molecular-weight compound that binds the aryl hydrocarbon receptor was used to prevent DC induction of CD4+ T cells.28

Western Blotting and Quantitative Polymerase Chain Reaction (qPCR).

Western blotting was performed as described.29 Livers were minced in PBS with Protease Inhibitor cocktail (Roche, Pleasanton, CA) and homogenized. The whole organ lysate was spun at 500g and the postnuclear supernatant was obtained. After determination of total protein by the Lowry assay, 10% polyacrylamide gels were equiloaded with samples, electrophoresed at 90 V, electrotransferred to PVDF membranes, and probed with primary mouse monoclonal antibodies for HMGB-1 or glutathione (Abcam, Cambridge, UK). Secondary goat antimouse horseradish peroxidase (HRP) was used (1:4,000). Blots were developed by ECL (Thermo Scientific, Asheville, NC). Total GSH from whole liver homogenates was measured using the Glutathione Assay Kit (Cayman Chemicals, Ann Arbor, MI) according to the manufacturer's protocol. For PCR assays, RNA was isolated from pancreas using a Qiagen RNEasy isolation kit (Qiagen, Germantown, MD). qPCR was performed using a standardized preconfigured PCR array (SA Biosciences, Frederick, MD) on the Stratagene MX3000P (Promega, Madison, WI) according to the respective manufacturers' protocols. (See Additional Supporting Methods.)


DC Depletion Exacerbates APAP-Mediated Hepatic Injury.

To evaluate a possible role for DC in either exacerbating or protecting against APAP toxicity, we employed CD11c.DTR mice in which transient DC depletion can be effected (Supporting Fig. 1). Mice were depleted of DC or mock-depleted and then challenged with APAP. At 12 hours mice were sacrificed and the extent of liver injury determined by histopathology. Mice treated with APAP and depleted of DC (APAP-DC) had markedly more extensive centrilobular necrosis compared with controls (Fig. 1A,B). Consistent with these findings, serum liver enzymes were highly elevated in APAP-DC mice (Fig. 1C). Similarly, NPC production of inflammatory mediators after APAP challenge, including MCP-1 and IL-6, were higher in mice depleted of DC (Fig. 1D). DC depletion alone, in the absence of APAP challenge, had no effect (Fig. 1A-D). Similarly, effects of APAP were similar in both CD11c.DTR and WT mice, in the absence of DC depletion (not shown). Notably, APAP metabolism appeared unchanged in APAP-DC mice compared with APAP treatment alone based on tissue glutathione assay and glutathione adduct formation (Supporting Fig. 2A,B). To determine if there was higher systemic toxicity in APAP-challenged mice after DC depletion, we measured serum levels of inflammatory mediators. We found that serum MCP-1, IL-6, and TNF-α were elevated in APAP-DC mice (Fig. 1E,F). However, as expected, organ damage was limited to the liver, as the lungs, kidneys, pancreas, and intestine were histologically normal (Supporting Fig. 3).

Figure 1.

DC depletion exacerbates APAP toxicity. (A) Liver from saline-treated, DC-depleted, APAP, and APAP-DC mice were assessed by hematoxylin and eosin (H&E) staining of paraffin-embedded sections and (B) the percentage of necrotic hepatocyte area determined by counting 10 high-powered fields per mouse. (C) Serum liver enzymes were measured as were (D) NPC production of MCP-1 and IL-6 in cell culture supernatant and (E) serum levels of MCP-1, IL-6, and (F) TNF-α. Experiments were performed in triplicate using 3-5 mice per group (*P < 0.05; **P < 0.01; ***P < 0.001).

To determine whether DC depletion resulted in higher APAP-mediated mortality, mice were treated with APAP and depleted of DC or mock-depleted and then observed for up to 2 weeks. Remarkably, approximately half of APAP-DC mice died within 48 hours of challenge, whereas death was rare in control animals (Fig. 2). There was no further mortality observed in APAP-DC mice after 48 hours from the time of APAP challenge.

Figure 2.

DC depletion exacerbates APAP-related mortality. Mice were treated with saline, DC depletion alone, APAP, or APAP-DC and survival measured according to the Kaplan-Meier method (n = 7-10 mice per group; P < 0.05 for comparison between APAP-DC and other groups).

We next tested whether depletion of specific DC subsets were responsible for the observed higher toxicity associated with APAP-challenge. Mice were depleted exclusively of plasmacytoid DC using 120G8 before APAP challenge. However, plasmacytoid DC depletion did not exacerbate APAP liver toxicity (Supporting Fig. 4). Similarly, we tested whether activation of the aryl hydrocarbon receptor on DC, which inhibits liver DC maturation in vivo and mitigates their ability to induce adaptive Th2 responses27, 30 (Supporting Fig. 5A,B), would also lead in exacerbated injury when administered to APAP-challenged mice. However, VAG539 did not modulate APAP toxicity (Supporting Fig. 5C-E).

DC Immune-Phenotype Is Altered After APAP Challenge.

Because the absence of DC results in exacerbated APAP-mediated injury, we interrogated the immune-phenotype of DC in animals challenged with APAP. Both the absolute number and fraction of liver DC among hepatic leukocytes did not change after APAP challenge (Fig. 3A,B); however, DC underwent an increase in maturation and alteration in subset composition in acute APAP hepatotoxicity (Fig. 3C). In particular, DC harvested from APAP-injured liver had elevated expression of MHC II and CD86, exhibited a lower B220+ plasmacytoid fraction, and underwent a “myeloid shift” expressing higher CD11b and lower CD8 (Fig. 3C). Conversely, spleen DC phenotype was unchanged after APAP challenge (Fig. 3C). Liver DC also increased their expression of Toll-like receptor (TLR)2, TLR4, TLR7, and TLR9 after APAP challenge (Supporting Fig. 6A). However, there was no measurable increase in selected byproducts of sterile inflammation in APAP-DC liver compared with APAP alone (Supporting Fig. 6B,C). In addition to an altered surface phenotype, the immunogenicity of DC harvested from the APAP injured liver was altered as DC from APAP liver produced higher IL-6, MCP-1, and TNF-α (Fig. 3D) compared with liver DC from saline-treated mice. Furthermore, consistent with their increased TLR expression, liver DC had an exaggerated cytokine response to TLR ligation after APAP toxicity (Fig. 3E). Spleen DC did not produce altered levels of cytokines after APAP challenge (not shown). Despite changes in hepatic DC surface phenotype and cytokine production after APAP challenge, their capacity to stimulate antigen-restricted CD4+ and CD8+ T cells was not enhanced (Fig. 3F).

Figure 3.

DC immune-phenotype is altered after APAP challenge. (A) The total number and (B) fraction of CD11c+ liver NPC per mouse was determined by flow cytometry. (C) Liver or spleen DC expression of surface markers in APAP or saline-treated mice were measured. Median fluorescence indices are listed below histograms. (D) Liver DC production of IL-6, TNF-α, MCP-1, and IL-10 in cell culture supernatant was measured. (E) Liver DC production from IL-6 was also measured after incubation with ligands for TLR9 (CpG ODN 1826), TLR3 (Poly I:C), and TLR4 (LPS). (F) DC induction of antigen-restricted T-cell proliferation was measured using appropriate peptide pulsed DC to stimulate CD4+OT-II and CD8+OT-I T cells. All in vitro experiments performed in triplicate and repeated at least three times (*P < 0.05; **P < 0.01).

DC Expansion Ameliorates the Hepatotoxic Effects of APAP.

Because DC depletion exacerbates APAP-mediated hepatotoxicity, we postulated that expansion of DC populations would mitigate liver injury. To test this, we employed Flt3L, which we have previously shown expands DC populations in vivo more than 10-fold.31, 32 Mice were treated with Flt3L for 10 days before APAP challenge followed by sacrifice at 12 hours. Notably, the maturation level of liver DC in Flt3L-treated mice was similar to controls except for a greater fraction of B220+ plasmacytoid DC in the Flt3L-treated group (Supporting Fig. 7A,B). However, after APAP treatment the fractions of plasmacytoid DCs were similar in both the Flt3L + APAP and in the APAP-only group (Supporting Fig. 7B). Mice treated with Flt3L were partially protected from liver injury, as indicated by reduced serum liver enzymes (Fig. 4A) and histologic measurement of necrosis (Fig. 4B,C). Notably, adoptive transfer of DC to APAP-DC mice did not protect mice from exacerbated toxicity (not shown). However, based on our investigations tracking adoptively transferred DC, it is likely that DC transfer is insufficient for protection, as adoptively transferred DC do not populate the liver at early or late timepoints after administration (Supporting Fig. 8).

Figure 4.

Flt3L treatment protects from APAP injury. Mice were treated with saline, APAP, Flt3L, or APAP+Flt3L. (A) serum liver enzymes as well as (B,C) histological evidence of liver damage were determined (n = 4-5 mice/group; *P < 0.05; **P < 0.01).

DC Depletion Alters the Composition of Hepatic Leukocytes After APAP Challenge.

To determine whether secondary alterations come into effect upon DC depletion that may be, in part, responsible for the exacerbated toxicity in APAP-DC mice, we examined the changes in hepatic leukocyte composition after DC depletion. There was a shift in composition of NPC, including a marked increase in the number of neutrophils in APAP-DC liver compared with APAP treatment alone (Fig. 5).

Figure 5.

Hepatic NPC composition is altered after APAP challenge. The cellular composition of hepatic leukocytes was determined by flow cytometry using an average of 5 mice per group.

Because neutrophils expand in APAP-DC-challenged mice, we postulated that DCs induce neutrophil apoptosis after APAP injury and, conversely, DC depletion would result in increased neutrophil viability. To test this, we measured the fraction of Gr1+CD11b+Annexin V+ cells in the normal liver, APAP-challenged liver, and in the liver of APAP-DC mice. Consistent with our hypothesis, we found that APAP treatment resulted in increased neutrophil apoptosis. Conversely, the fraction of apoptotic neutrophils was sharply decreased upon DC depletion in the context of APAP injury (Supporting Fig. 9A). DC depletion alone in the absence of APAP administration had no effect on the neutrophil apoptotic fraction (not shown).

Because NK1.1+ cells have also been implicated in the mechanism of APAP-mediated hepatotoxicity,12 we postulated that DC may prevent the activation of NK cells after APAP injury. To test this, hepatic NK1.1+ cells were purified, cocultured with DC, and simultaneously stimulated with phorbol 12-myristate 13-acetate (PMA) + ionomycin. Notably, DC from normal liver further enhanced NK1.1+ cell production of IFN-γ. Conversely, APAP liver DC prevented NK1.1+ cellular activation (Supporting Fig. 9B). Similarly, NK cells treated with PMA + ionomycin and simultaneously cocultured with DC from control livers were potently cytolytic. Conversely, APAP liver DC did not stimulate NK-mediated cytolysis of Yac-1 targets (Supporting Fig. 9C). Taken together, out data suggests that in acute APAP hepatotoxicity, liver DC inhibit neutrophil viability and NK cell activation.

DC Depletion Exacerbates APAP-Mediated Hepatic Injury Independently of Neutrophils, NK Cells, or Inflammatory Mediators.

Both neutrophils and NK cells have been implicated in the pathogenesis of APAP.12, 14, 18 Furthermore, because APAP-DC treated animals experience an expansion of neutrophils and our data shows that DC affect neutrophil viability and NK activation status, we postulated that the exacerbated centrilobular necrosis associated with DC depletion in APAP-challenged animals was secondary to an expanded neutrophil population or activated NK cells, rather than directly related to the absence of DC. To test this, we simultaneously depleted either NK cells or neutrophils in conjunction with DC before APAP administration. However, neither depletion of NK cells or neutrophils, along with DC, mitigated the exacerbated hepatotoxicity associated with DC depletion (Fig. 6), suggesting that the protective effects of DC is not simply secondary to expansion of other leukocyte populations. Similarly, because DC depletion results in elevated serum levels of TNF-α, IL-6, and MCP-1 after APAP administration, we tested whether blockade of these cytokines in vivo would prevent the exacerbated liver injury. However, none of these cytokine blockades protected APAP-DC animals (Supporting Fig. 10). Similarly, IFN-α blockade33 failed to protect APAP-DC animals (Supporting Fig. 10)

Figure 6.

Neutrophil or NK cell depletion do not protect APAP-DC mice. (A,B) CD11c-DTR mice were depleted of DC and simultaneously depleted of either neutrophils or NK cells before challenge with APAP.


There is evidence to suggest that APAP-induced liver toxicity is the result of a “two-hit” mechanism, the first hit being depletion of glutathione, which in turn allows the toxic metabolite NAPQ1 to exert harmful effects by forming covalent bonds with cellular proteins. The second hit is the downstream activation of cells of the innate immune system. Because DC have a central function in liver immunity and inflammation, we postulated a critical role for DC in APAP-mediated toxicity.

Previously, we showed that DC expand 5-fold and undergo a transformation in function from a tolerogenic to an immunogenic role in chronic liver fibrosis.25 We reported that DC contribute to the proinflammatory cascade in liver fibrosis by way of production of TNF-α and subsequent T-cell activation as well as induction of innate immune responses.25 Similar to liver fibrosis, in APAP toxicity DCs are highly proinflammatory, producing elevated levels of IL-6, TNF-α, and MCP-1 (Fig. 3D,E). However, in contrast to chronic liver disease, in acute liver injury as a result of APAP overdose, DC populations remained stable in number. Furthermore, whereas chronic liver injury resulted in the transformation of DC from weak purveyors of tolerance to potent immunogenicity, in the current context DC did not gain enhanced capacity to stimulate CD4+ or CD8+ T cells (Fig. 3F) or NK cells (Supporting Fig. 9B,C). The trigger in the hepatic microenvironment that thrusts DC in certain inflammatory contexts towards immunogenicity is uncertain but may be the key to understanding hepatic tolerance.

Furthermore, whereas DCs appear to contribute to the pathologic environment in chronic liver disease, in the current context DCs are protective. This is evidenced by reduced liver enzymes and histologic measurement of necrosis in APAP-treated mice cotreated with Flt3L, which expands DC populations 10-fold. Furthermore, mice depleted of DC had significantly more extensive centro-lobular necrosis (Fig. 1A,B) and increased mortality (Fig. 2) when compared to mock-depleted mice. In addition, APAP-DC mice produced markedly higher serum liver enzyme levels (Fig. 1C) and inflammatory mediators MCP-1, IL-6, and TNF-α (Fig. 1E,F) compared with APAP challenge in the absence of DC depletion. These findings suggest that DCs have a novel protective role in APAP hepatotoxicity. Again, the contextual switch governing DC behavior in diverse states of liver injury remains uncertain.

The role of Kupffer cells in APAP-induced liver toxicity is controversial13, 16, 18 but may perhaps be similar to the role of DC. Ju et al.13 showed an increased susceptibility to APAP-induced liver injury when effectively depleting mice of Kupffer cells with gadolinium chloride. The mechanism of the Kupffer hepatoprotective effect is thought to be related to decreased expression of IL-10, an antiinflammatory cytokine, in Kupffer cell-depleted mice. In our current study, DC did not produce detectible IL-10 after APAP treatment and their depletion did not affect serum or liver IL-10 levels (Fig. 3D). Interestingly, Bamboat et al34 demonstrated that in liver ischemia/reperfusion injury, liver DCs were responsible for IL-10 production by way of TLR9 activation. DC-mediated IL-10 production was shown to suppress monocyte inflammatory function and reduce hepatic injury.34 However, because DC do not produce detectible IL-10 after APAP challenge, it appears that the mechanism responsible for the DC protective role in APAP-induced hepatotoxicity is distinct from that of ischemia/reperfusion liver injury. Our mechanistic investigations further show that the exacerbated liver toxicity associated with DC depletion is independent of associated elevations in neutrophils or inflammatory monocytes, as well as independent of systemic elevations in TNF-α, MCP-1, and IL-6 and unrelated to IFN-α (Supporting Fig. 10).

Our study implicates DC in APAP-induced liver injury as an innate protector that can be compared to the role of DC in sepsis. Sepsis is characterized by an intense hyperinflammatory response designed to eliminate an underlying infectious source.35 However, there is growing evidence that an initial proinflammatory cascade is thought to be followed by an activation of a compensatory antiinflammatory response syndrome that leads to immune suppression and subsequent poorer clinical outcomes.35, 36 DC loss appears to play a significant role in the pathogenesis of sepsis.37 By using cecal ligation and puncture (CLP) surgery as a model for sepsis in rodents, recent investigations have shown that a decline in DC counts occur in conjunction with immune dysfunction, suggesting that DC may even have a protective role against the development of immunosuppression in sepsis.37 The mechanism of DC loss appears to be related to extensive apoptosis of DC during sepsis. This was demonstrated in a study in which mouse spleens showed a significant increase in caspase 3-mediated apoptosis in follicular dendritic cells 36-48 hours after CLP insult compared to controls.37, 38 Furthermore, Scumpia et al.39 showed that DC-deficient mice had increased mortality after CLP. Similarly, in our study, DC depletion exacerbated APAP-mediated hepatic injury and led to a significant increase in mortality.

Sterile inflammation is the activation of the innate immune system and subsequent creation of an inflammatory response in the absence of an infectious source. Inciting stimuli include mechanical trauma, ischemia, and organ-specific toxins such as APAP-induced liver.40, 41 Endogenous signals of tissue injury known as damage-associated molecular patterns (DAMPs) have been shown to activate antigen-presenting cells such as DCs. DAMPs can induce DC maturation by ligating their TLRs. However, our mechanistic studies indicate that, whereas DC express elevated levels of TLRs (Supporting Fig. 6A) and produce exaggerated responses to TLR ligation (Fig. 3E), there was no difference between APAP and APAP-DC liver with regard to the levels of measurable DAMPs including HMGB1 (Supporting Fig. 6B), heat shock proteins (Supporting Fig. 6C), and S100A9 (not shown). Exact understanding of the regulatory role of DC and its interplay with sterile inflammation could be an important step in the development of immune-directed therapy in APAP-induced liver injury and may have implications for other disease processes regulated by immunity and inflammation. Furthermore, the translational potential of this study for the protective role of DC in acute APAP toxicity in humans requires further exact investigation using human specimens.