The method for isolating and extracting KAE from Penthorum chinense Pursh is described in our previous study (Du et al., 2020), and the purity was confirmed to be >98% based on HPLC analysis (Figure 1).

We purchased 48 male C57BL/6 mice (8–10 weeks old, weighing 18–22 g) from Hua Fukang Bioscience (Beijing, China) and housed them under controlled light (12-h light/dark cycle) and temperature (22 ± 2°C) conditions with free access to food and water. After 1 week of adaptive feeding, the mice were randomly divided into the following groups (eight mice each): 1) Sham group (Sham); 2) KAE 60 mg/kg group (KAE60); 3) HI/RI group (HI/RI); 4) HI/RI + KAE 15 mg/kg (HI/RI + KAE15); 5) HI/RI + KAE 30 mg/kg (HI/RI + KAE30); 6) HI/RI + KAE 60 mg/kg (HI/RI + KAE60).

Mice in groups 2), 4), 5), and 6) were administered the above dose of KAE by gavage for 7 days, and HI/RI model was performed, with samples collected on day 8. All animal experiments in this study were reviewed and approved by the Animal Care and Use Committee and Ethics Committee of Southwest Medical University.

According to Yuta Abe’s method, nonlethal segmental (70%) liver ischemia was established. Briefly, all mice were anesthetized with sodium pentobarbital (40 mg/kg, i.p) before dissecting their abdomens along the midline. Then, using an atraumatic microvascular clamp, nonlethal segmental (70%) liver ischemia was induced by occlusion of the hepatic artery and portal vein of the left and median lobes. After 60 min of segmental liver ischemia, the clamps were removed to initiate liver reperfusion. The mice were sacrificed with overdose sodium pentobarbital (90 mg/kg, i. p) after 6 h of reperfusion, and plasma and liver tissues were collected for analysis. Additionally, the mice belonging to the Sham group underwent a midline laparotomy incision without microvascular clamp placement.

Normal human hepatocytes QSG-7701 were purchased from Beyotime Biotechnology (Shanghai, China) and cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco; Thermo Fisher Scientific, United States) supplemented with 10% fetal bovine serum (Biological Industries, Beit Haemek, Israel) and 1% penicillin-streptomycin (Solarbio, Beijing, China) in a humidified incubator (Thermo Fisher Scientific, United States; 37°C, 5% CO₂). The H/R model of hepatocytes was established with some modifications to our previous study (Du et al., 2019). Briefly, experimental group hepatocytes were pretreated with various concentrations of KAE for 24 h before H/R procedure, while the control group was treated with the same volume of the KAE vehicle (DMSO; Solarbio, Beijing China) and maintained in the incubator. After that, all hepatocytes were washed twice with warm PBS and replaced with glucose-free and serum-free DMEM (Balanced with 1% O₂, 5% CO₂, and 94% N₂; Procell, Wuhan, China) 1 h before the hypoxia period. Experimental group cells were then cultured under hypoxic conditions (37°C, 1% O₂, 5% CO_2, and balanced N₂) in an InvivO₂ 400 hypoxic workstation (Baker Ruskinn, United Kingdom) for 6 h. Then, both groups of hepatocytes were replaced with fresh warm DMEM in the incubator (37°C, 5% CO₂) for a 4-h reoxygenation period.

Mice serum samples were obtained by centrifuging (4°C, 5,000 rpm, 5 min) blood. An automatic biochemical analyzer (ADVIA 2400 Chemistry System, Siemens, Germany) was used to determine the serum activities of ALT and AST in mice.

The supernatant was collected after the homogenization of the mice liver, and the content of MDA, SOD and GSH were measured according to the kit manufacturer’s instructions (Beyotime, Shanghai, China).

Liver samples from each group of mice were collected immediately after the I/R procedure and fixed in a 4% paraformaldehyde solution for 24 h, followed by dehydration with gradient ethanol, paraffin embedding, and sectioning for H&E and IHC staining of HO-1 (1:500, Proteintech, Wuhan, China).

TUNEL apoptosis assay for mice liver tissue sections was performed according to the kit manufacturer’s instructions (Servicebio, Wuhan, China).

According to the manufacturer’s instructions, cell viability was determined using the cell counting kit 8 (CCK-8; Beyotime, Shanghai, China). Briefly, hepatocytes (3 × 10⁴/well) were inoculated in a 96-well plate, then 10 μl CCK-8 solution was added and cultured routinely. After 2 h, absorbance at 450 nm was measured using a Cytation5 Imaging Reader (BioTek, United States), and cell viability was calculated. Duplicate wells were used in the respective groups and repeated four times.

For cellular ROS determination and fluorescence analysis, we loaded a dichlorofluorescein-diacetate (DCFH-DA) fluorescent probe according to the manufacturer’s instructions (Beyotime, Shanghai, China). The area scan function (3 × 3 reads/well; excitation/emission = 488/525 nm) of the Gen5 software (Vision.3.08; Biotek, United States) was used to calculate the average fluorescence intensity of 6-well plates, and subsequent fluorescence images were captured using a fluorescence microscope (IX73; Olympus, Tokyo, Japan).

According to the manufacturer’s instructions, RIPA lysis buffer (Beyotime, Shanghai, China) was used to extract total protein from liver tissue and hepatocytes. Western blotting was performed as previously described (Lei et al., 2016). The membranes were incubated with primary antibodies against NF-κB/p65 (1:1,000; Proteintech, United States), phospho-NF-κB/p65 (p-p65; 1:1,000; Zen Bioscience, China), TNF-α (1:1,000; Proteintech; United States), IL-6 (1:1,000; Proteintech, United States), IL-10 (1:1,000; Wanleibio, China), Bax (1:2000; Proteintech, United States), Bcl-2 (1:1,000; Proteintech, United States), Nrf2 (1:1,000; Proteintech, United States), HO-1 (1:1,000; Proteintech, United States), β-actin (1:5,000; Proteintech, United States) and then incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (1:5,000; Proteintech, United States). The relative protein expression was analyzed using ImageJ software (NIH, Maryland, United States).

All data analyses were performed using GraphPad Prism v.8.0 (GraphPad Software, San Diego, United States) and presented as mean ± standard deviation. All data were compared using one-way ANOVA and t-test, and a p-value less than 0.05 (p < 0.05) was considered to represent statistically significant results.

In this study, mice were pretreated with different doses of KAE gavage (15, 30, and 60 mg/kg) for 7 days before establishing HI/RI models, as previously described (Figure 2A). We first evaluated the effect of KAE on the liver of mice and observed no significant difference in liver function (p > 0.05) (Figures 2B,C) and tissue structure (Figure 2D) between the KAE60 and Sham groups of mice, confirming that the dose of KAE gavage was not significantly toxic to mice. While KAE pretreatment dose-dependently reduced I/R-induced transaminase elevation, with the most significant effect of KAE at 60 mg/kg (ALT, p < 0.01; AST, p < 0.001) (Figures 2B,C). The Sham and KAE60 groups had normal liver structure and intact liver lobules on morphological and histopathological inspection. In contrast, mice in the HI/RI group showed a larger area of necrosis in the liver, and KAE pretreatment reversed this result (Figure 2D). The area of I/R-induced necrosis was reduced in all KAE pretreatment groups, with the best effect in the KAE60 group. However, the improvement of I/R by KAE 15 mg/kg was less significant (Figure 2D), consistent with serological results (p > 0.05) (Figures 2B,C).

To assess the KAE effect on oxidative stress in the mice HI/RI model, we measured the content of MDA, SOD and GSH in the supernatant of mice liver tissue homogenates as markers of oxidative stress levels (Figures 3A–C). Compared with the Sham group, the MDA content in the liver of mice subjected to I/R injury significantly increased (p < 0.001) (Figure 3A), but the KAE pretreatment group was considerably lower than the HI/RI group, especially the HIRI + KAE60 group (p < 0.001) (Figure 3A). SOD and GSH, components of the antioxidant system in organisms, were significantly reduced by I/R-induced oxidative stress (p < 0.001) (Figures 3B,C), and KAE pretreatment dose-dependently reversed this result, with the best effect at a KAE dose of 60 mg/kg (p < 0.001) (Figures 3B,C). However, KAE pretreatment at 15 mg/kg had almost no effect on oxidative stress (Figures 3A–C).

To investigate the protective effect of KAE against inflammation induced by I/R, we analyzed the protein expression of inflammatory factors and markers by western blotting. I/R-induced significant NF-κB/p65 phosphorylation and increased the expression of pro-inflammatory factors, including TNF-α and IL-6 (p < 0.001) (Figures 4A–D), whereas KAE pretreatment dose-dependently reversed these effects, and the anti-inflammatory effect was most pronounced at a dose of 60 mg/kg (p < 0.001) (Figures 4A–D). In addition, the IL-10 involved in anti-inflammatory was significantly downregulated by I/R injury (p < 0.001) (Figures 4A,E), whereas KAE pretreatment dose-dependently restores some of the effects (Figures 4A,E). The above results indicate that KAE pretreatment inhibits I/R-induced pro-inflammatory factors release by suppressing NF-κB/p65 activation and increasing the expression of anti-inflammatory factors.

To further assess the extent of I/R-induced injury in vivo, we analyzed the expression of related proteins by western blotting and performed TUNEL staining on liver tissue sections. Western blotting of apoptosis-related protein expression revealed that I/R significantly upregulated the expression of the pro-apoptotic protein Bax while inhibiting the expression of the anti-apoptotic protein Bcl-2 (p < 0.001) (Figures 5A–C), which was reversed by KAE pretreatment at an optimal dose of 60 mg/kg (p < 0.001) (Figures 5A–C). TUNEL staining was the next section, and our results showed that a large number of TUNEL-positive cells were detected in the liver tissue of the HI/RI group compared with the Sham group (Figure 5D). However, apoptotic hepatocytes in the KAE pretreatment group were significantly lower than those in the HI/RI group, which was particularly significant in the HIRI + KAE60 group (Figure 5D), indicating that KAE pretreatment inhibits I/R-induced hepatocellular apoptosis by downregulating pro-apoptotic protein expression and upregulating anti-apoptotic protein expression.

To investigate the role of KAE in the Nrf2/HO-1 signaling pathway, we analyzed the expression of Nrf2 and HO-1 in total protein extracts from the mice liver by western blotting. Our study showed that I/R injury mildly increased the expression of Nrf2 (p < 0.05) (Figures 6A,B) and dramatically increased the expression of HO-1 (p < 0.001) (Figures 6A,C) compared with the Sham group. Compared with the HI/RI group, expression of Nrf2 and HO-1 were further increased in the KAE pretreatment groups, and the extent of the increase correlated with the KAE dose, the effect was most pronounced when KAE pretreatment dose reached 60 mg/kg (p < 0.001) (Figures 6A,C). IHC staining of HO-1 showed similar results: a minor increase in HO-1-positive cells in the HI/RI group compared with the Sham group, and a significant increase in HO-1-positive cells in the KAE pretreatment group (Figure 6D). These results indicate that Nrf2 was activated under stress conditions and upregulated HO-1 expression to counteract I/R injury, and KAE pretreatment further enhanced the effect.

At the cytological level, we started by pretreating the normal human hepatocyte line QSG-7701 with DMSO and different KAE concentrations for 24 h. Then, cell viability was measured to assess the cytotoxicity of KAE (Figure 7A). Hepatocyte proliferation was significantly inhibited at KAE concentrations up to 20 μM (p < 0.001) (Figure 7A), hence KAE concentrations below 20 μM will be used in subsequent experiments. H/R injury significantly inhibited the proliferation of hepatocytes (p < 0.001) (Figure 7B), and KAE pretreatment restored the injury, with 5 μM KAE being the optimal concentration (p < 0.01) (Figure 7B). Subsequently, western blotting was performed to detect the expression of pro-apoptotic factor Bax and anti-apoptotic factor Bcl-2 in each group of hepatocytes (Figures 7C–E). The results showed that H/R significantly upregulated Bax expression but downregulated the expression of Bcl-2, which was reversed by KAE with an optimal concentration of 5 μM (p < 0.001) (Figures 7C–E). These findings are similar to those obtained in vivo.

To assess the level of H/R-induced oxidative stress in vitro, the DCFH-DA fluorescent probe was used to label ROS. DCFH-DA fluorescent probe was loaded on hepatocytes, and the fluorescence area was compared to measure ROS levels. Pretreatment with a 5 μM KAE concentration reversed H/R-induced fluorescent area increase, but increasing the KAE concentration to 20 μM may have caused more ROS generation (p > 0.05) (Figure 8A). The average fluorescence intensity measurement confirmed the above results (p < 0.001) (Figure 8B). Further, we investigated Nrf2 and HO-1 protein expression in hepatocytes under H/R conditions. Compared with the Sham group, H/R injury increased the expression of total Nrf2 and its downstream HO-1 (p < 0.001) (Figures 8C–E), which was further enhanced by KAE pretreatment, especially at KAE pretreatment concentration of 5 μM (p < 0.001) (Figures 8C–E). The above in vitro results were in good agreement with the in vivo experiments, confirming the validity of KAE pretreatment.