Asiatic acid, purity >98%, was provided by the Chengdu Herbpurify Co., Ltd (Chengdu, China). LPS (Escherichia coli 055:B5), GalN, and dimethyl sulfoxide were purchased from Sigma-Aldrich (St. Louis, MO, USA). Inhibitors of AMPK (Compound C) was offered by Sigma-Aldrich (St. Louis, MO, USA). Fetal bovine serum, Dulbecco’s modified Eagle’s medium, penicillin, and streptomycin were acquired from Invitrogen-Gibco (Grand Island, NY, USA). Antibodies against Nrf2, GCLC, GCLM, HO-1, NQO1, P-AMPK, AMPK, P-PI3K, PI3K, P-AKT, AKT, P-ERK, ERK, PDCD4, P-P65, P65, IκBα, P-IκBα, and β-actin were obtained from Cell Signaling (Boston, MA, USA) or Abcam (Cambridge, MA, USA). Additionally, O2−, H₂O₂, NO, glutathione (GSH), superoxide dismutase (SOD), malondialdehyde (MDA), myeloperoxidase (MPO), and ROS test kits were supplied by Nanjing Jiancheng Bioengineering Institute (Nanjing, China). All other chemicals were offered by Sigma-Aldrich (St. Louis, MO, USA), if not otherwise indicated.

Female BALB/c mice (6–8 weeks), weighing approximately 18–20 g, were purchased from Liaoning Changsheng Technology Industrial Co., LTD (Certificate SCXK2010-0001; Liaoning, China). All animals were housed in a room with temperature at 24 ± 1°C, a 12 h light–dark cycle and relative humidity about 40–80%. Animals were allowed free access to tap water and normal food after feeding several days. All animal experiments were performed according to the guide for the Care and Use of Laboratory Animals, which was published by the US National Institute of Health. This study was reviewed and approved by the Animal Welfare and Research Ethics Committee at Jilin University.

Mice were randomly divided into seven groups: control (PBS) group, AA only (25 mg/kg) group, LPS/d-GalN (L/D, 30 µg/kg and 600 mg/kg) group, and AA (6.25, 12.5, or 25 mg/kg) + (L/D) group and hemin (a HO-1 inducer as a positive group, 30 mg/kg) + (L/D) group, were administered intraperitoneally. In brief, AA (6.25, 12.5, or 25 mg/kg) or hemin (30 mg/kg) was administered intraperitoneally to mice for twice at a 12-h (interval for 12 h), followed by subjected treatment with LPS (30 µg/kg) and d-GalN (600 mg/kg), which is abbreviated as L/D. The survival rates of mice were observed for 48 h after L/D challenge. For other assays, after L/D administration for 3 and 6 h, the animals were euthanized. Subsequently, liver tissues and serum were harvested and used for hematoxylin and eosin (H&E) staining, Western blot assay, and enzyme-linked immunosorbent assay (ELISA).

Liver tissues were immersed in normal 10% neutral buffered formalin and fixed for 48 h, dehydrated in a series of graded ethanol, embedded in paraffin wax, and cut into 5-µm-thick sections. The paraffin-embedded sections were stained with hematoxylin and eosin (H&E) for pathological analysis under a light microscope. The histological changes were evaluated by a point-counting method for severity of hepatic injury using an ordinal scales in accordance with the methods as previous described (23). Briefly, H&E-stained sections were evaluated at 400× magnification by a point-counting method for severity of hepatic injury using an ordinal scale as follows; grade 0: minimal or no evidence of injury, grade 1: mild injury consisting in cytoplasmic vacuolation and focal nuclear pyknosis, grade 2: moderate to severe injury with extensive nuclear pyknosis, cytoplasmic hypereosinophilia, and loss of intercellular borders, and grade 3: severe necrosis with disintegration of hepatic cords, hemorrhage, and neutrophil infiltration.

All mice were euthanized at 3 or 6 h after L/D treatment, liver and blood were collected for biochemical analysis. alanine transaminase (ALT) and aspartate aminotransferase (AST) levels in serum were measured using the corresponding detection kits. In addition, mice liver tissues were homogenized and dissolved in extraction buffer to analyze the MPO, GSH, MDA, and SOD levels according to the manufacturer’s instructions. All results were normalized by the total protein concentration in each sample. For other assays, all mice were sacrificed at 3 or 6 h after L/D treatment, liver tissues were collected for measurement of O2−, NO, H₂O₂, and ROS generation. Mice liver tissues were homogenized and dissolved in extraction buffer to analyze the O2−, NO, H₂O₂, and ROS levels in accordance to the manufacturer’s instructions.

Blood was obtained from each sample in vivo, centrifuged, collected serum for measurement of the TNF-α, IL-6, and IL-1β secretion using an ELISA kit as the manufacturer’s instructions (BioLegend, Inc., CA, USA), respectively. The optical density from each well was detected at 450 nm.

Liver tissues were collected 3 or 6 h after L/D challenge. Total protein was extracted from the liver tissues using a protein extract kit according to the manufacturer’s protocol. Protein concentrations were tested by the BCA method. Equal amounts of proteins (20 µg) were separated by a 10% SDS-polyacrylamide gel and transferred onto a polyvinylidene difluoride membrane. The membrane was blocked with 5% (w/v) non-fat milk for 2 h. Then, the membrane was incubated with primary antibody and secondary antibody. Finally, the membranes were visualized by the ECL Western blotting detection system in accordance with the manufacturer’s instruction and band intensities were quantified using Image J gel analysis software. All experiments were performed in triplicate.

HepG2 cells were cultured in 12-well plates at the density of 3 × 10⁵ cells/well for 24 h. The plasmids of expressing Cas9 with Nrf2-sgRNA and puromycin resistant gene were co-transfected into HepG2 cells using Viafect transfection reagent (Promega). At 48 h after transfection, cells were added puromycin at a concentration of 2 µg/mL and harvested for immunoblotting analysis with Nrf2 antibody. After 7 days, cells were cultured in a 96-well plates (1 cell/well).

All data referenced above were expressed as the means ± SEM and analyzed using SPSS19.0 (IBM). Comparisons between experimental groups were conducted using one-way ANOVA, whereas multiple comparisons were made using the LSD method. Statistical significance was defined as p < 0.05 or p < 0.01.

To investigate whether AA could protect against L/D-induced liver injury in mice, the survival rates of the mice were observed for 48 h after L/D exposure. As shown in Figure 1B, the mice died 7 h after L/D injection, and the survival rate was 0% at 48 h. Conversely, pretreatment with AA (25 mg/kg) effectively increased the survival rate up to 80% and no significant difference compared with hemin (a HO-1 inducer) group, a positive group. In addition, because serum AST and ALT levels were used as well-established marker of hepatic injury (25), the serum ALT and AST levels were measured. Our results indicated that no changes in serum levels of ALT and AST were induced by alone AA at various dosages ranging from 0 to 100 mg/kg, indicating that AA did not exhibit hepatotoxicity (Figure 1C). Meanwhile, we found that the serum ALT and AST levels were significantly increased by L/D administration compared with the control group (p < 0.01). However, this increase was markedly reduced by AA pretreatment, suggesting that AA treatment efficiently protected against L/D-induced FHF (Figures 1G,H) (p < 0.01). Gross and histological examinations of liver tissues were used to evaluate the protective effects of AA on L/D-induced FHF. As illustrated in Figure 1D, gross examination of the livers displayed apparent congestion and hemorrhage in the L/D-induced group mice 6 h after challenge, indicating severe liver injury. Similarly, histological analysis of the mouse liver sections in the L/D group showed obviously disturbed architecture, such as hepatocyte necrosis, hemorrhage, and neutrophil infiltration. However, the L/D-induced liver alterations were effectively relieved by AA treatment based on the liver injury scores (Figures 1E,F).

The inflammatory cytokines TNF-α, IL-6, and IL-1β play vital roles in liver injury. To further explore the anti-inflammatory effects of AA, the effects of AA on TNF-α, IL-6, and IL-1β generation in the sera was measured by ELISA. As shown in Figures 2A–C, L/D effectively increased the secretion of TNF-α, IL-6, and IL-1β (p < 0.01) in the sera, whereas AA treatment inhibited the induction of inflammatory cytokine production by L/D (p < 0.01).

Because the NF-κB and MAPK signaling pathways have been reported to be inflammatory pathways that play imperative roles in mice with L/D-induced FHF, we evaluated the effects of AA treatment on the L/D-induced activation of the NF-κB signaling pathway. As presented in Figures 3A–H, AA treatment remarkably inhibited P65, JNK, ERK, and P38 phosphorylation (p < 0.01) and blocked IκBα phosphorylation and degradation compared to the L/D-challenged group, suggesting the inhibition of inflammatory responses by AA might be partially responsible for blocking the activation of the NF-κB and MAPK signaling pathways.

Importantly, PDCD4 negatively or positively mediates inflammatory responses, although this topic is still controversial. Wang et al. (26) discovered that PDCD4 deficiency aggravated colitis via promoting the IL-6/STAT3 pathway in mice. However, Schmid et al. (27) maintained that inflammation could induce the loss of PDCD4, and Wang et al. (8) indicated that PDCD4 upregulation inhibited inflammatory responses by inhibiting NF-κB and MAPK signaling pathway activation in an LPS/d-GalN-induced mouse model. Therefore, PDCD4 may be closely associated with inflammatory responses. Indeed, we found that the PDCD4 protein expression level was dramatically reduced 3 and 6 h after L/D injection (p < 0.05 and p < 0.01, respectively) in our studies. Conversely, AA pretreatment effectively restored PDCD4 protein expression (Figures 3I,J), indicating that inhibition of the inflammatory responses induced by L/D challenge by AA might be partially attributed to the increased PDCD4 protein expression (p < 0.01).

Because oxidative damage is a major factor in mice with L/D-induced FHF, we determined whether AA pretreatment attenuated the L/D-triggered oxidative stress. As shown in Figure 4, L/D not only evidently increased MDA formation (p < 0.05 or p < 0.01), the MPO level (p < 0.01), and ROS generation (i.e., H₂O₂, NO, and O2− production) (p < 0.05 or p < 0.01) but also obviously decreased the SOD (p < 0.05) and GSH (p < 0.01) contents in the livers of the mice. In contrast, AA pretreatment dramatically inhibited these effects induced by L/D (p < 0.05 or p < 0.01). These observations indicated that AA treatment ameliorated hepatic injury by partially lessening L/D-triggered oxidative stress in mice.

Increasing evidence indicates that the Nrf2-mediated signaling pathway is essential for the inhibition of oxidative stress and inflammatory responses in mice with L/D-induced FHF. Therefore, we examined whether the antioxidant and anti-inflammatory activities of AA were related to upregulation of the Nrf2-mediated signaling pathway. The inductions of GCLC, GCLM, HO-1, and NQO1, which are regulated by Nrf2 transcription, have been reported to be a key cellular protective mechanism against the inflammatory response or oxidative stress in various cell types (28, 29). As shown in Figures 5A–F, AA treatment remarkably enhanced Nrf2, GCLC, GCLM, HO-1, and NQO1 protein expression compared with the control group and the L/D-exposed group (p < 0.05 or p < 0.01).

Importantly, previous studies suggested that multiple signaling pathways modulated Nrf2 expression. Our previous study indicated that AA inhibited cellular damage and oxidative stress by regulating Nrf2 signaling dependent upon activation of the AKT and ERK signals in HepG2 cells (24). Accordingly, to investigate the protective mechanism of AA treatment on L/D-induced FHF, the ERK and PI3K/AKT activities were analyzed by Western blotting. Our results showed that AA treatment significantly inhibited L/D-induced phosphorylation of ERK, PI3K, and AKT, indicating that AA-mediated Nrf2 activation was unlikely to be associated with the ERK and PI3K/AKT signaling pathways in vivo (Figure S1 in Supplementary Material). However, AA treatment effectively increased AMPK and GSK3β phosphorylation compared with the L/D-exposed group (p < 0.05 or p < 0.01) (Figures 5G–I). Furthermore, we found that AA obviously induced AMPK and GSK3β phosphorylation and Nrf2 protein expression (p < 0.05 or p < 0.01) in HepG2 cells incubated with AA (6 µM) for different times, whereas these effects were efficiently blocked by compound C (CC, an AMPK inhibitor) (Figure S2 in Supplementary Material). Together, these investigations suggest that the ability of AA to suppress inflammation and oxidative stress in L/D-induced FHF may involve upregulation of the Nrf2-medicated signaling pathway via activation of AMPK/GSK3β in vivo and in vitro.

Based on the above outcomes, we evaluated whether the AA-induced PDCD4 expression was dependent of Nrf2 activation by using WT and Nrf2^−/− HepG2 cells. In the present study, after exposure of HepG2 cells to AA (6 µM) for different times, we found that AA exposure for 1 h obviously enhanced PDCD4 protein expression compared to the unexposed cells (p < 0.05) (Figures 6A,B). Nrf2 protein expression induced by AA was evidently inhibited in the Nrf2^−/− cells compared with the control cells, whereas AA-enhanced PDCD4 protein expression was not suppressed in the Nrf2^−/−HepG2 cells (Figures 6C–H). These results suggested that AA-induced PDCD4 protein expression is independent of Nrf2 activation in HepG2 cells.