Artesunate Targets MBL2 to Modulate the TLR4/NRF2/HO‑1 Axis and Ameliorate Podophyllotoxin‑Induced Liver Injury

Abstract

Background

Drug-induced liver injury (DILI) presents a significant clinical challenge with limited preventive and therapeutic options. Artesunate (ART) holds potential for DILI treatment; however, its precise hepatoprotective mechanisms and direct molecular target(s) remain unclear.

Methods

We evaluated ART in a podophyllotoxin (PPT)-induced rat model of DILI. Its direct target was identified through integrated chemical proteomics (pull-down/LC-MS/MS) followed by biophysical validation (SPR, MST) and molecular dynamics simulations. Mechanistic insights from hepatic transcriptomics (RNA-seq) were confirmed by Western blot, and extended with targeted metabolomics and 16S rRNA sequencing of gut microbiota.

Results

ART administration conferred significant hepatoprotection in the PPT-induced rat model, as evidenced by dose-dependent amelioration of key serum markers of liver injury (P < 0.05), attenuation of hepatic oxidative stress, and improved histopathological outcomes. Mechanistically, we identified mannose-binding lectin 2 (MBL2) as a direct, high-affinity target of ART (KD ≈ 3.84 µM). Upon binding to MBL2, ART suppressed the hepatic TLR4/NF-κB inflammatory axis and, in parallel, relieved NF-κB-mediated repression of NRF2, thereby cooperatively activating the NRF2-driven antioxidant defense program. Integrated multi-omics analysis further revealed that ART remodeled the gut microbiota, enriching beneficial genera including Lactobacillus, and that the abundance of these bacteria correlated positively with hepatic levels of antioxidant metabolites (P < 0.05), supporting a functional role of the gut-liver axis in its systemic protective effects.

Conclusion

This study elucidates that ART alleviates DILI through targeting MBL2, thereby dually modulating the NF-κB/NRF2 signaling axis to simultaneously suppress inflammatory responses and enhance endogenous antioxidant defenses. These findings provide a novel and robust pharmacological basis for repositioning ART for the clinical prevention and treatment of DILI.

Graphical Abstract

Introduction

Drug-induced liver injury (DILI) represents a prevalent and serious clinical adverse drug reaction, frequently leading to drug attrition and post-marketing withdrawal.1 In China, its annual incidence is estimated at approximately 23.8 per 100,000 individuals.2 The pathogenesis of DILI is complex, involving multifaceted interactions among direct cytotoxicity, immune-mediated inflammation, and oxidative stress.3 Currently, specific therapeutic options remain limited, with management primarily relying on drug discontinuation and supportive care.4 Consequently, there is an urgent need to develop novel agents that can intervene in these key pathological processes. Notably, many natural bioactive compounds present a “double-edged sword” effect, possessing therapeutic potential alongside inherent hepatotoxicity risks. A prime example is PPT, a potent topoisomerase II inhibitor derived from Podophyllum hexandrum (also known as Sinopodophyllum hexandrum). Its severe dose-limiting hepatotoxicity not only restricts its clinical application but also contributes to incidents of poisoning,5–7 underscoring the critical importance of establishing effective protective strategies against drug-induced hepatotoxicity.

Beyond its well-established antimalarial efficacy, artesunate (ART), a derivative of Artemisia annua, has demonstrated broad pharmacological activities in preclinical studies, including anti-inflammatory, antioxidant, and cell death-modulating effects.8 In models of liver disease, ART shows promise in ameliorating fibrosis and mitigating metabolic injury,9 suggesting its potential to counteract various forms of liver injury, including DILI, by modulating shared stress-response pathways. However, the specific protein target responsible for initiating these protective effects remains unidentified. This knowledge gap confines the mechanistic understanding of ART’s hepatoprotection to phenotypic observations and descriptions of downstream pathways, thereby significantly impeding its further development and clinical translation as a hepatoprotective agent.

Multiple studies have explored the protective effects of artesunate in various liver injury models. In acetaminophen-induced liver injury models, artesunate has been reported to attenuate oxidative stress and apoptosis.10,11 In models of nonalcoholic fatty liver disease, it protects by inhibiting NLRP3 inflammasome activation.12 Moreover, in hepatic ischemia–reperfusion injury models, artesunate alleviates hepatocyte death by promoting autophagy and inhibiting the NLRP3 inflammasome.13 However, despite these phenotypic observations, the direct protein target responsible for artesunate’s hepatoprotective effects remains unknown. This represents a critical knowledge gap in the field. Addressing this gap will help mechanistically explain the previously observed protective phenotypes. It will also provide an important basis for the rational development of artesunate as a therapeutic agent against drug-induced liver injury (DILI).

The central pathophysiological process underlying liver injury involves a vicious cycle between inflammation and oxidative stress. Mannose-binding lectin 2 (MBL2), a key pattern recognition receptor in the liver, plays a complex role in innate immunity. Dysregulation of MBL2 has been linked to various liver pathologies14,15 and, in chemical liver injury, is closely associated with injury severity.16 While MBL2 deficiency exacerbates the inflammatory cascade, its normal function may confer protection, potentially through the negative regulation of the Toll-like receptor 4 (TLR4)/nuclear factor-kappa B (NF-κB) axis. Specifically, MBL2 negatively regulates TLR4 signaling through multiple mechanisms, including competitive binding to pathogen-associated molecular patterns, interference with the assembly of TLR4 and its co-receptor MD-2, and promotion of inhibitory signal transduction.17–19 In drug-induced liver injury (DILI), this negative regulation is particularly important because excessive activation of TLR4/NF-κB is the core event that drives the inflammatory cascade and the vicious cycle of oxidative stress. Crucially, a profound crosstalk exists between NF-κB signaling and nuclear factor erythroid 2-related factor 2 (NRF2), the master regulator of cellular antioxidant defense.18 Under pathological conditions, hyperactivated NF-κB can directly or indirectly repress NRF2 activity,20 thereby simultaneously propagating inflammation and crippling antioxidant capacity, which forms a self-amplifying feedback loop that exacerbates injury. Thus, targeting MBL2 to coordinately modulate both NF-κB and NRF2 signaling represents a promising strategy to disrupt this vicious cycle. Notably, the chemical structure of ART contains multiple polar groups that may form hydrogen bonds and hydrophobic interactions with the carbohydrate recognition domain of MBL2, providing structural plausibility for direct binding.21,22 Therefore, ART binds to MBL2 and stabilizes or enhances its negative regulatory functions, thereby suppressing overactivated TLR4/NF-κB inflammatory signaling. Through the NF-κB–NRF2 crosstalk, this also relieves the repression of NRF2 activity, effectively breaking the vicious cycle between inflammation and oxidative stress and providing a novel conceptual framework for artesunate-mediated hepatoprotection.

To systematically elucidate the mechanism of action of ART, an integrated research framework spanning from target discovery to systemic validation is essential. Affinity-based chemical proteomics employing active molecular probes represents a pivotal technology for identifying direct drug-binding proteins. Candidate targets require validation of their binding characteristics and modes through biophysical methods, supplemented by computational simulations such as molecular docking and dynamics. Mechanistic dissection necessitates the integration of functional omics: transcriptomics can unveil the reprogramming of gene networks and core signaling pathways,23 while targeted metabolomics quantifies changes in downstream functional metabolites, thereby bridging signaling events with phenotypic outcomes.24 Furthermore, research on the “gut–liver axis” is increasingly recognized as vital. Analysis of the gut microbiota contributes to a systematic interpretation of the drug’s holistic effects from the perspective of host–microbe interactions.25

Based on the above background, this study was designed to systematically elucidate the protective effect of ART against PPT-induced liver injury and to dissect its underlying molecular mechanism. We propose the central hypothesis that ART directly engages the hepatic pattern recognition receptor MBL2, thereby coordinately modulating the crosstalk between the TLR4/NF‑κB inflammatory axis and the NRF2 antioxidant defense system, potentially through the involvement of the gut–liver axis. To test this hypothesis, we integrated chemical proteomics, biophysical binding assays, transcriptomic profiling, targeted metabolomics, and gut microbiome analysis to precisely identify the direct target of ART and to map its downstream signaling networks. This study not only provides a pharmacological rationale for repurposing ART as a protective agent against DILI but also offers a methodological paradigm for target discovery and mechanistic dissection of natural products.

Materials and Methods

Reagents and Antibodies

Artesunate (HPLC> 98%, Shanghai Yuanye Bio-Technology Co., Ltd, Cat# B20992); Podophyllotoxin (HPLC> 98%, Shanghai Yuanye Bio-Technology Co., Ltd, Cat# B20477); Silymarin (HPLC> 98%, Shanghai Yuanye Bio-Technology Co., Ltd, Cat# S25549); Sodium carboxymethyl cellulose (Shanghai Aladdin Biochemical Technology Co., Ltd, Cat# C104985); Methanol (CNW Technologies GmbH, Cat# 67–56-1); Acetonitrile (CNW Technologies GmbH, Cat# 75–05-8); Ammonium hydroxide, 20–22% (Fisher Chemical, Cat# 1336–21-6); Acetic acid (Sigma-Aldrich, Cat# 64–19-7); Tris (Sigma-Aldrich, Cat# 6132–04-3); SDS (Sodium dodecyl sulfate) (Sigma-Aldrich, Cat# 151–21-3); N, N’-Methylenebisacrylamide (Sigma-Aldrich, Cat# V900301); Tween 20 (Sigma-Aldrich, Cat# P9416); Dimethyl sulfoxide (DMSO) (Sigma-Aldrich, Cat# V900090); Series S Sensor Chip CM5 (Cytiva, Sweden, Cat# 29149603); Amine Coupling Kit (Cytiva, Sweden, Cat# BR100050); Protein Labeling Kit RED-NHS second generation (Nanotemper Technologies, Germany, Cat# MO-L011).

Animal Experiments

Grouping and Dosing Protocol

After one week of acclimatization under standard conditions, 60 healthy male Sprague-Dawley rats (aged 6–8 weeks, 230 ± 20 g; Animal License No. SCXK (Beijing) 2021–0006) were randomly divided into five groups (n = 12 per group): a control group (CON), a model group (PPT, podophyllotoxin treatment group), a positive control group (POS, treated with silymarin), and two artesunate treatment groups (ART-L, 75 mg/kg; ART-H, 150 mg/kg). To induce hepatotoxicity, rats in the PPT, POS, ART-L, and ART-H groups received intragastric administration of PPT (10 mg/kg) once daily for four consecutive days (Days 1–4). Starting from Day 1 through Day 4 (concurrent with PPT administration), the ART-L and ART-H groups received daily intraperitoneal injections of ART at their respective doses, while the POS group received silymarin by oral gavage. Notably, PPT administration was discontinued after Day 4, whereas ART treatment was continued for an additional three days (Days 5–7), resulting in a total ART treatment duration of 7 consecutive days (Days 1–7). Both the CON and PPT groups received intraperitoneal injections of an equal volume of the vehicle (0.5% carboxymethyl cellulose sodium containing 2% dimethyl sulfoxide) throughout the same period. All experimental procedures were approved by the Institutional Animal Care and Use Committee (Approval No. D‑2025‑027). Detailed information regarding randomization, allocation concealment, blinding, exclusion criteria, mortality handling, exact sampling timeline, vehicle preparation, and the full dosing schedule is provided in the Supplementary Methods, Table S1 and Figure S1.

Sample Collection and Multifaceted Analysis

Upon completion of the dosing regimen, rats were anesthetized via intraperitoneal injection of tribromoethanol (250 mg/kg) and subjected to terminal blood collection from the abdominal aorta. Following euthanasia, the liver was rapidly excised, cleared of connective tissue, rinsed with ice-cold saline, gently blotted dry, and weighed for the calculation of the liver-to-body weight index. Serum samples were analyzed for alanine aminotransferase (ALT), aspartate aminotransferase (AST), high-density lipoprotein (HDL), low-density lipoprotein (LDL), Gamma-glutamyl Transpeptidase (γ-GT), albumin (ALB), total protein (TP), and lactate dehydrogenase (LDH) using an automated biochemical analyzer. Hepatic levels of malondialdehyde (MDA) and glutathione (GSH) were quantified with commercial assay kits according to the manufacturers’ protocols. For histopathological evaluation, liver tissue samples were fixed in 4% paraformaldehyde, processed through graded ethanol and xylene, embedded in paraffin, sectioned at 5-μm thickness, and stained with hematoxylin and eosin (H&E). The stained sections were examined under a light microscope, and liver injuries—including sinusoidal congestion, hepatocellular edema, and necrosis—were evaluated and scored in accordance with the INHAND (International Harmonization of Nomenclature and Diagnostic Criteria for Lesions in Rats and Mice) guidelines.

Pull-Down Combined with Mass Spectrometry Analysis

An ART-biotin chemical probe was synthesized by conjugating ART to ethylenediamine-biotin via HOBT/EDCI coupling. The product was purified and verified by mass spectrometry, showing a dominant ion at [M+Na]⁺ = 675.3. For affinity purification, the probe was pre-immobilized on streptavidin magnetic beads and incubated with rat tissue lysates; beads loaded with biotin alone served as a control. After extensive washing, specifically bound proteins were eluted, separated by SDS-PAGE, and visualized with silver staining. Protein bands exhibiting differential enrichment were excised and subjected to in-gel tryptic digestion. The resulting peptides were desalted and analyzed by nanoLC-MS/MS using data-dependent acquisition with HCD fragmentation. Acquired MS/MS spectra were searched against the UniProt Rattus norvegicus database using MASCOT, allowing for tryptic cleavage, fixed carbamidomethylation, variable oxidation modifications, and a significance threshold of p < 0.05.

Surface Plasmon Resonance (SPR) Analysis

SPR analysis was employed to quantitatively determine the binding affinity between the molecules. Experiments were performed on a Biacore 1K instrument. Briefly, the surface of a CM5 sensor chip was activated with an EDC/NHS mixture, followed by immobilization of the target protein. The protein was diluted to 40–50 µg/mL in sodium acetate buffer (pH 4.0) and immobilized to achieve a response level of approximately 10,000 response units (RU). The analyte was serially diluted two-fold in running buffer (PBS containing 0.05% Tween 20, pH 7.2–7.4) to generate a concentration series ranging from 0.39 to 200 µM (10 concentrations). All binding assays were conducted at 25°C with a constant flow rate of 30 µL/min. Analytic samples were injected sequentially from the lowest to the highest concentration, with an association phase of 120 s and a dissociation phase of 180 s. Sensorgram data were processed using the Biacore Insight Evaluation Software. Following double referencing, the binding curves were fitted to a 1:1 steady-state equilibrium model to determine the equilibrium dissociation constant (KD).

Microscale Thermophoresis (MST) Assay

The binding affinity between the MBL2 protein and ART was quantitatively assessed using MST on a NanoTemper Monolith NT.115 instrument. Recombinant MBL2 protein was fluorescently labeled using the Monolith Protein Labeling Kit RED-NHS (MO-L011, NanoTemper Technologies). For the binding assay, ART was serially diluted two-fold in PBST buffer to generate 16 concentration points ranging from 1000 μM to 0.0305 μM. Each dilution was mixed with an equal volume of the labeled protein solution (final protein concentration: 20 nM) and incubated at room temperature for 15 min. The mixtures were loaded into NT.115 standard capillaries and measured at 25°C with 20% excitation power and 40% MST power. The experiment was performed in three independent replicates. Binding data were analyzed using MO Affinity Analysis v2.3 software, and the Kd was calculated by fitting the data to a binding curve. Results are presented as mean ± standard error of the mean (SEM) from at least three independent experiments.

Molecular Docking

To investigate the potential binding mode of ART to the MBL2 protein, molecular docking simulations were performed. The three-dimensional structure of MBL2 was obtained from the Protein Data Bank (PDB ID: 1HUP). The three-dimensional structure of ART was retrieved from the PubChem database (CID: 6917864) and subsequently energy-minimized. The active site of MBL2 was predicted using KVFinder. Docking simulations were conducted using AutoDock Vina with a semi‑flexible strategy, wherein the protein receptor was held rigid while the ligand’s rotatable bonds were allowed to rotate freely. The resulting binding poses were clustered and ranked according to their calculated binding affinity (ΔG). The most favorable pose was selected for further analysis. Interaction analysis was performed using Discovery Studio Client 2024 to identify key intermolecular forces contributing to binding stability, and the results were visualized with PyMOL.

Molecular Dynamics Simulations

To assess the binding stability and dynamic interactions of the ART-MBL2 complex, all-atom MD simulations were performed using the optimal binding pose identified from molecular docking. The protein was described using the AMBER99SB-ildn force field, while parameters for ART were generated with the General AMBER Force Field (GAFF). Atomic partial charges for the ligand were assigned using the restrained electrostatic potential (RESP) method. The complex was solvated in a cubic box of TIP3P water molecules. All simulations were performed with the GROMACS software package at 310 K. The system was first energy-minimized to relieve steric clashes. This was followed by a two-step equilibration: 100 ps in the NVT ensemble to stabilize the temperature, and 100 ps in the NPT ensemble to stabilize the pressure and density. Subsequently, a 200 ns production MD run was conducted with a 2 fs integration time step. Short-range van der Waals and electrostatic interactions were truncated at 1.2 nm, and long-range electrostatics were treated with the Particle Mesh Ewald (PME) method. Trajectory coordinates were saved every 10 ps for analysis. For analysis, the final 100 ns of the production trajectory were used. The binding free energy (ΔG_bind) was calculated, and a per-residue energy decomposition was performed using the gmx_MMPBSA method to quantitatively evaluate the binding affinity and identify key residues contributing to the interaction.

Transcriptomic Analysis of Liver Tissue

To elucidate the mechanism of ART at the transcriptional level, transcriptome sequencing of liver tissue was performed. Total RNA was extracted using the TRIzol method and subjected to stringent quality control (RNA Integrity Number, RIN > 7.0). Messenger RNA was enriched with oligo(dT) magnetic beads for the construction of strand-specific libraries. Sequencing was carried out on an Illumina NovaSeq 6000 platform with a paired-end 150 bp (PE150) strategy. Raw reads were processed with Cutadapt to remove adapters and low-quality sequences. The clean reads were then aligned to the reference genome using HISAT2. Transcripts were reassembled and quantified with StringTie based on FPKM (Fragments Per Kilobase of transcript per Million mapped reads) values. Sample reproducibility was assessed by principal component analysis (PCA) and Pearson correlation analysis. Differentially expressed genes (DEGs) were identified using DESeq2/edgeR with thresholds of |log₂(fold change)| > 1.00 and an adjusted P value < 0.05. Finally, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed on the DEGs to systematically uncover the biological processes and pathways associated with ART intervention, thereby elucidating its hepatoprotective mechanism at the transcriptional level. For each group, samples were derived from six independent biological replicates (n = 6 per group).

Targeted Metabolomics Profiling

To further elucidate the hepatoprotective mechanism of ART, targeted metabolomics analysis was performed. Metabolites were separated using an Agilent 1290 UHPLC system equipped with a Waters Atlantis Premier BEH Z-HILIC column and subsequently detected by a SCIEX Triple Quad 6500+ mass spectrometer in multiple reaction monitoring (MRM) mode. Data processing involved absolute quantification of target metabolites using standard curves to calculate their tissue concentrations. To assess data quality and identify differential metabolites, PCA was first performed to examine overall distribution trends and instrumental stability. This was followed by supervised orthogonal partial least squares-discriminant analysis (OPLS-DA) to maximize inter-group separation. Differential metabolites were identified based on a combination of criteria: VIP score > 1.0, absolute fold change > threshold, and statistical significance (P < 0.05) determined by Student’s t-test or the Mann–Whitney U-test, as appropriate. Finally, metabolic pathway enrichment analysis was conducted on the identified differential metabolites to systematically elucidate the key metabolic regulatory networks associated with liver injury repair under ART intervention. For each group, samples were derived from six independent biological replicates (n = 6 per group).

16S rRNA Sequencing

To investigate the impact of ART on the gut microbiota, 16S rRNA gene sequencing was performed on colonic content samples. Following total genomic DNA extraction and quality control, the target hypervariable region was amplified, and sequencing libraries were constructed. Paired-end 250 bp (PE250) sequencing was conducted on an Illumina NovaSeq platform. Raw sequencing data were processed through the DADA2 pipeline to generate an amplicon sequence variant (ASV) table. Taxonomic annotation and rarefaction-based normalization were performed using QIIME2. Differences in microbial community structure were then evaluated via alpha diversity indices (eg, Chao1 and Shannon) and beta diversity analyses, including principal coordinate analysis (PCoA), non-metric multidimensional scaling (NMDS), and analysis of similarities (ANOSIM). Compositional profiles, heatmaps, and linear discriminant analysis effect size (LEfSe) were employed to identify differentially abundant taxa between groups. This approach systematically revealed the potential association between gut microbiota structural changes induced by ART intervention and its hepatoprotective effects. To further elucidate the mechanistic basis of ART’s hepatoprotective effect via the gut-liver axis, Spearman rank correlation analysis was performed between the perturbed gut microbiota and the differentially abundant key metabolites in the liver. For each group, samples were derived from seven independent biological replicates (n = 7 per group).

Western Blotting

Rat liver tissues were homogenized in ice-cold lysis buffer containing protease and phosphatase inhibitors (1:10 w/v). After lysis at 4°C for 2 h, the homogenates were centrifuged at 12,000 × g for 20 min. Protein concentrations in the supernatants were determined using a BCA assay. Equal amounts of protein were mixed with 5× SDS loading buffer, denatured, and separated by SDS-PAGE according to the target proteins’ molecular weights. Following electrophoresis, proteins were transferred to PVDF membranes. The membranes were blocked with 5% BSA, then incubated overnight at 4°C with primary antibodies, followed by incubation with HRP-conjugated secondary antibodies. Protein bands were visualized using an ECL substrate, and band intensities were quantified using ImageJ software. For each group, samples were derived from five independent biological replicates (n = 5 per group).

Statistical Analysis

Statistical analyses and graphing were conducted using GraphPad Prism (version 10.0). All data are presented as mean ± standard deviation (SD), with each data point representing an independent biological replicate. For comparisons between two independent groups, an unpaired t‑test was used when data met the assumptions of normality and homogeneity of variance; otherwise, the Mann‑Whitney U-test was applied. For multi‑group comparisons, the Kruskal‑Wallis test was performed, and if a significant overall effect was detected, pairwise comparisons were conducted using Mann‑Whitney U-tests. The sample sizes (transcriptomics: n=6/group; targeted metabolomics: n=6/group; gut microbiota: n=7/group; WB: n=5/group) meet the minimum requirement of n ≥ 5 per group as recommended by the journal’s statistical guidelines.26 All tests were two‑tailed, and P < 0.05 was considered statistically significant. In figures, NS indicates no significant difference, and asterisks denote *P < 0.05, **P < 0.01, ***P < 0.001.

Results

Pharmacodynamic Evaluation of the Hepatoprotective Effect of ART

Pharmacodynamic evaluation demonstrated that ART (75 and 150 mg/kg) exerted significant protective effects against PPT (10 mg/kg)-induced liver injury in rats (Figure 1A). Organ index analysis revealed that ART significantly reversed the organ atrophy observed in the model group (P < 0.001) (Figure 1B). Regarding serum biochemical markers, ART dose-dependently and significantly ameliorated liver function, as evidenced by marked reductions in the activities of ALT (P < 0.001), AST (P < 0.001), LDH (P < 0.001), and γ-GT (P < 0.001). Concurrently, it significantly increased the levels of TP (P < 0.05), ALB (P < 0.001), and HDL (P < 0.001), while decreasing LDL content (P < 0.001). In terms of oxidative stress indicators, ART significantly elevated hepatic GSH levels (P < 0.05) and reduced MDA content (P < 0.05) (Figure 1C–L). Histopathological examination further confirmed that ART markedly alleviated pathological injuries such as hepatocyte edema, hepatic sinusoid compression, and vascular congestion, accompanied by a decrease in pathological scores. In conclusion, ART effectively ameliorated PPT-induced liver injury at multiple levels, including whole-animal status, serum biochemistry, and histopathology (Figure 1M). Its mechanism of action is closely associated with the regulation of oxidative stress and lipid metabolism.

Figure 1.

Pharmacodynamic Evaluation of the Hepatoprotective Effect of ART. (A) ART ameliorates liver injury induced by PPT. (B–L) ART ameliorates the changes in visceral index and biochemical indicators of liver injury induced by PPT: ALT, AST, LDH, γ-GT, TP, ALB, HDL, LDL, GSH, and MDA (Mean ± SD, n=8). (M) Amelioration of PPT-induced liver injury by ART: histopathological examination by HE staining (200×). In this figure legend, nine colored arrows are used to indicate distinct hepatic pathological alterations. Yellow arrows denote hepatocellular vacuolar degeneration (round vacuoles in the cytoplasm). Red arrows indicate vascular congestion (dilated vessels with erythrocyte aggregation). Brown arrows represent sinusoidal dilation or congestion (widened hepatic sinusoids with or without erythrocyte stasis). Green arrows point to connective tissue proliferation (increased fibrous tissue, mainly in the capsule or portal areas). Blue arrows label hepatocellular edema (swollen cells with pale and loose cytoplasm). Gray arrows mark pale pink homogenous material within irregular cavities, occasionally observed in proliferative connective tissue. Black arrows identify multinucleated giant cells, occasionally seen in association with foreign body reaction or inflammation. Orange arrows indicate lymphocyte infiltration (focal accumulation of inflammatory cells). Light green arrows are used for occasional fibrous connective tissue proliferation around the central vein (mild and localized fibrosis). The asterisk symbols denote the following levels of statistical significance: *P < 0.05; **P < 0.01; ***P < 0.001.

Transcriptomic Analysis of the Hepatoprotective Effect of ART

To elucidate the protective mechanism of ART (ART-H) against PPT-induced liver injury, liver transcriptomic sequencing was performed on CON, PPT-induced model, and ART-H treatment groups (n=6). Hierarchical clustering analysis visually demonstrated that the gene expression profile of the ART-H group reverted toward that of the CON group (Figure 2A). Differentially expressed genes (DEGs) were identified using the DESeq2 algorithm with thresholds of |log2 fold change (FC)| ≥ 1.0 and an adjusted p-value (padj) < 0.05. Compared with the CON group, the PPT model group exhibited extensive transcriptomic dysregulation, with 3235 DEGs identified. ART-H intervention significantly reversed this abnormal expression pattern: a comparison between the PPT and ART-H groups revealed 1114 DEGs, among which 180 key genes were completely reversed, exhibiting expression levels restored to near-normal trends (Figure 2B and C). Functional enrichment analysis systematically revealed the core molecular network underlying ART’s action. Consistent findings from KEGG and GO analyses indicated that ART significantly suppressed the NF-κB signaling pathway (particularly the NIK/NF-κB signaling transduction) while activating the glutathione metabolic pathway and related antioxidant activity functions, clearly demonstrating its dual synergistic effects of anti-inflammation and antioxidation (Figure 2D and E). Further mechanistic investigation showed that ART exerts its core anti-inflammatory effect by inhibiting the TLR4/NF-κB axis. The key mechanism involves the inhibition of the NF-κB core subunit P65, which alleviates its competition with the antioxidant transcription factor NRF2 for co-activators (such as CBP/p300), thereby relieving the transcriptional repression of NRF2 by NF-κB. This regulation promotes NRF2 nuclear translocation and subsequently upregulates the expression of key downstream antioxidant genes, including heme oxygenase-1 (HO-1) and NAD(P)H quinone dehydrogenase 1 (NQO1). In summary, this transcriptomic study elucidates that the hepatoprotective effect of ART relies on an integrated regulatory network centered on the NF-κB/NRF2 signaling axis. By inhibiting the TLR4/NF-κB inflammatory signaling and relieving the suppression of NRF2, ART not only mitigates oxidative stress but also actively enhances intracellular endogenous antioxidant capacity, synergistically restoring hepatic redox homeostasis and thereby effectively alleviating liver injury. For each group, samples were derived from six independent biological replicates (n = 6 per group).

Figure 2.

Screening of Direct Targets of ART in Ameliorating PPT-Induced Liver Injury. (A) Heatmap Showing Clustering of Differential Genes.(B) Number of up- and down-regulated differentially expressed genes between groups. (C) Dual volcano plot of PPT-upregulated (PPT vs CON) and ART-downregulated (ART vs PPT) genes (P < 0.05, FC > 2) (n = 6). (D and E) KEGG Pathway Enrichment and GO Analysis of Common Differential Genes. The red boxes in this figure indicate the key genes and the significantly enriched GO terms and KEGG pathways identified in this study. These boxes are used for visual guidance only and do not denote statistical significance between groups.

Screening and Validation of Direct Targets of ART in Alleviating Liver Injury

To identify the direct molecular target responsible for the hepatoprotective effect of ART, an integrated strategy combining chemical biology and biophysical approaches was employed. Potential interacting proteins were first screened using a biotinylated ART probe in pull-down assays coupled with liquid chromatography–tandem mass spectrometry (LC-MS/MS), which identified over 1100 candidates from rat tissue extracts. After applying stringent filtering criteria—including spectral intensity, molecular weight range, and functional relevance—MBL2 was selected as a high-confidence candidate. With an approximate molecular weight of 26 kDa, MBL2 is a key initiator of the lectin complement pathway in innate immunity. Its biological role aligns with the previously observed ability of ART to inhibit NF-κB and activate the Nrf2 pathway, suggesting that MBL2 may act upstream in the protective mechanism of ART.

To quantitatively confirm the interaction, SPR and MST assays were performed. SPR analysis demonstrated concentration-dependent binding of ART to immobilized MBL2, yielding a KD of 3.84 μM based on a steady-state affinity model. MST independently verified the specific binding, with a calculated KD of 6.35 μM. Both orthogonal methods consistently indicated a strong, direct interaction between ART and MBL2 (Figure 3A–C).

Figure 3.

Validation of MBL2 as a Direct Binding Target of ART in Ameliorating PPT‑Induced Liver Injury. (A-B) Validation of ART-MBL2 binding affinity by SPR. (C) Validation of ART-MBL2 binding affinity by MST. (D and E) Validation of ART-MBL2 binding affinity by molecular docking. (F–H) Validation of ART-MBL2 binding affinity by molecular dynamics simulation: RMSD values, RMSF values, and free energy landscape.

Further mechanistic insights were obtained through molecular docking and 200‑ns all-atom MD simulations (Figure 3D and E). Trajectory analysis confirmed the stability of the system and revealed a local conformational adjustment in the N‑terminal α‑helix (residues 88–111) of MBL2 upon ART binding, characterized by an inward movement toward the binding pocket. The binding free energy, calculated using the MM/GBSA method, averaged-39.30 kcal/mol, with van der Waals interactions identified as the major driving force. Residue‑wise energy decomposition highlighted key amino acids contributing to binding stability (Figure 3F–H).

In summary, through pull-down/MS target screening, SPR/MST binding validation, and MD simulations, this study provides the first evidence that MBL2 is a direct molecular target of ART. This finding establishes a crucial foundation for understanding how ART alleviates liver injury by modulating MBL2‑mediated immune and inflammatory pathways.

Targeted Metabolomic Analysis of the Hepatoprotective Effect of ART

Targeted metabolomics analysis systematically revealed the metabolic regulatory mechanism by which ART alleviates PPT-induced liver injury. Multivariate statistical analysis demonstrated distinct separation trends among the healthy control, model, and ART treatment groups in both principal component analysis (cumulative contribution rate: 87.26%) and principal coordinate analysis (PERMANOVA P = 0.002), with sample points from the treatment group shifting significantly toward the healthy control group (Figure 4A–D). Analysis of differential metabolites showed that ART significantly reversed the aberrant levels of 21 core metabolites. Among these, the endogenous antioxidant uric acid and L-methionine—a precursor for glutathione synthesis, both closely related to oxidative stress—were significantly up-regulated (P < 0.05). Conversely, key intermediates in energy metabolism, malate and α-D-glucuronic acid, were significantly down-regulated (P < 0.05) (Figure 4E–G). These differential metabolites were significantly enriched in 30 key metabolic and signal transduction pathways, primarily involved in biological processes such as oxidative stress and amino acid metabolism. Several core pathways closely associated with oxidative stress damage were markedly activated. Notably, the enrichment of the cysteine and methionine metabolism pathway, which is crucial for glutathione synthesis, is directly linked to the body’s most important endogenous antioxidant defense system. The significant enrichment of the pyruvate metabolism pathway is also noteworthy, as exogenous pyruvate can directly inhibit the nuclear translocation of NF-κB, reducing the expression of downstream inflammatory factors, while simultaneously enhancing the antioxidant response via the GPX4/Nrf2 axis, thereby indirectly activating Nrf2 signaling (Figure 4H). These results indicate that ART exerts its hepatoprotective effect by modulating the host metabolic network, particularly through mechanisms such as enhancing the glutathione system function, thereby significantly ameliorating the state of oxidative stress.

Figure 4.

Targeted Metabolomic Analysis of the Hepatoprotective Effect of ART. (A–D) PCA, PCoA, and PLS-DA analyses between groups. (E) Volcano plot of differential metabolites between CON and PPT groups (n = 6). (F) Volcano plot of differential metabolites between PPT and ART_H groups. (G) Heatmap showing common differential metabolites among the three groups. (H) KEGG pathway analysis of differential metabolites. The statistical significance of enriched pathways was assessed using Fisher’s exact test or hypergeometric distribution test, and P-values were adjusted for multiple testing using FDR correction. The red boxes in this figure indicate the key differential metabolites and the significantly enriched KEGG pathways identified in this study. These boxes are used for visual guidance only and do not denote statistical significance between groups.

Alterations in Gut Microbiota Associated with the Hepatoprotective Effect of ART

Analysis of the gut microbiota revealed that ART significantly reversed the intestinal dysbiosis induced by PPT (Figure 5A). Venn diagram analysis demonstrated substantial alterations in ASV composition following high-dose ART intervention (Figure 5B). At the genus level, ART significantly reduced the abundance of Duncaniella (P < 0.05) while increasing the abundances of Enterococcus (P < 0.05), Enteromonas (P < 0.05), and Terrisporobacter (P < 0.05) (Figure 5C). Alpha diversity analysis indicated that ART modulated microbial diversity. Beta diversity analyses, including PCoA (PERMANOVA P = 0.001) and NMDS, confirmed significant separation of the microbial community structures among the three groups (Figure 5D and E). Notably, the microbiota profile of the ART-treated group showed a trend of reversion toward the healthy state. Adonis (R² = 0.436, P = 0.001) and Anosim (R = 0.545, P = 0.001) analyses further validated the significant intergroup differences. Chord diagram visualization of species composition indicated that ART promoted the proliferation of probiotic genera and inhibited the expansion of opportunistic pathogens (Figure 5F). LEfSe analysis identified that ART intervention significantly suppressed potential pathogens associated with intestinal barrier damage (eg, g__Bacteroides_F, LDA = 2.86, P < 0.05; g__Escherichia, LDA = 4.79, P < 0.05) while enriching the probiotic genus g__Lactobacillus (LDA = 4.25, P < 0.05) (Figure 5G and H). These results suggest that ART may exert its hepatoprotective effect by remodeling the gut microbiota structure, inhibiting the proliferation of pathogenic bacteria, and promoting the colonization of protective bacterial communities.

Figure 5.

Analysis of Gut Microbiota Characteristics in ART-Induced Amelioration of PPT-Induced Liver Injury. (A) Rarefaction curve of gut microbiota. (B) Venn diagram of ASVs/OTUs across groups. (C) Bar plot showing the taxonomic composition of samples from three groups. (D and E) PCoA and NMDS showing clustering patterns among three groups. To assess the dissimilarity of microbial community structure, PCoA based on Bray-Curtis dissimilarity was performed. (F) Chord diagram showing the distribution and abundance of differential microbiota across three groups. (G and H) LEfSe was applied to identify differentially abundant microbiota among three groups (n = 7 per group), with the Kruskal–Wallis test threshold set at 0.05 and the logarithmic LDA score threshold at 2.0. Pairwise Wilcoxon rank-sum tests were then performed, and P-values were adjusted using the Benjamini-Hochberg FDR method; (I) Correlation Analysis between Differential Microbiota and Differential Metabolites. To control for multiple comparisons, the P-values were adjusted using the FDR method (Benjamini-Hochberg). The red boxes in this figure indicate the key differential microbial taxa identified in this study. These boxes are used for visual guidance only and do not denote statistical significance between groups. The asterisk symbols denote the following levels of statistical significance: *P < 0.05; **P < 0.01; ***P < 0.001.

Correlation Analysis

To further elucidate the potential mechanism by which ART exerts its hepatoprotective effect via the gut–liver axis, Spearman correlation analysis was performed between the significantly altered gut microbiota and the 21 core differential metabolites (Figure 5I). The results revealed extensive and significant associations between gut microbiota remodeling and systemic host metabolic regulation. Specifically, the abundances of beneficial bacterial genera enriched after ART intervention, such as g__Lactobacillus and g__Enterococcus_B, showed significant positive correlations with several metabolites possessing hepatoprotective potential, including the endogenous antioxidant uric acid and the glutathione synthesis precursor α‑linolenic acid (r > 0, P < 0.05). Conversely, suppressed potential pathogenic genera exhibited an inverse correlation pattern. For instance, g__Escherichia was negatively correlated with L‑asparagine but positively correlated with L‑argininosuccinate (r > 0, P < 0.05), while g__Parasutterella showed a significant negative correlation with L‑methionine (r < 0, P < 0.05), a key precursor for glutathione synthesis. These results collectively suggest that the modulation of gut microbiota structure by ART—particularly the promotion of beneficial bacteria and suppression of potential pathogens—is statistically directly linked to its enhancement of the host hepatic oxidative stress defense system and amelioration of energy metabolism disturbances. This finding further supports that gut microbe‑mediated metabolic regulation is a crucial pathway through which ART exerts its hepatoprotective effects.

Integrated Analysis of Protein and Gene Expression

Western blot analysis revealed coherent, multi-level molecular alterations at the protein level in response to PPT challenge and ART‑H intervention. Compared with the CON group, PPT administration significantly downregulated the hepatic protein expression of MBL2, NRF2, HO‑1, and NQO1, while upregulating TLR4, phosphorylated NF‑κB p65 (p‑NF‑κB p65), and KEAP1 (Figure 6A). Importantly, ART‑H treatment effectively reversed these perturbations, restoring the protein expression of MBL2, NRF2, HO‑1, and NQO1 while suppressing the upregulation of TLR4, p‑NF‑κB p65, and KEAP1 relative to the PPT group. These observations underscore the consistent regulatory effects of ART‑H at the translational level.

Figure 6.

ART Targets MBL2 to Ameliorate Liver Injury via Regulation of the TLR4/NRF2/HO-1 Axis. (A–H) Western blot analysis of MBL2, NRF2, HO-1, NQO1, TLR4, phosphorylated NF-κB p65, and KEAP1 protein expression. All Western Blot experiments were performed with at least five independent biological replicates (n = 5). The asterisk symbols denote the following levels of statistical significance: ***P < 0.001.

Discussion

This study provides evidence that artesunate effectively alleviates PPT-induced liver injury. MBL2 was identified as a potential direct binding target of artesunate. In vivo experiments confirmed that artesunate dose-dependently improved animal survival, liver function, and histopathological outcomes. Mechanistically, artesunate appears to act through interaction with MBL2 to coordinately regulate the NF-κB/NRF2 signaling axis. Specifically, artesunate treatment was associated with upregulation of MBL2 expression, suppression of the TLR4/NF-κB inflammatory pathway, and relief of NF-κB-mediated transcriptional inhibition of NRF2. These changes collectively led to activation of downstream antioxidant responses. Furthermore, systems biology analyses revealed that artesunate remodeled the gut microbiota and enhanced the host antioxidant metabolic network.

Pronounced Hepatoprotective Effects of ART

This study systematically evaluated the hepatoprotective effects of artesunate (75 and 150 mg/kg) against PPT (10 mg/kg)-induced drug‑induced liver injury. The results showed that artesunate significantly reduced mortality, reversed liver atrophy, and dose‑dependently improved serum liver function markers (ALT, AST, γ‑GT) as well as hepatic synthetic function (TP, ALB), indicating effective repair of hepatocellular damage. In addition, artesunate alleviated hepatic oxidative stress (decreased MDA, increased GSH), corrected lipid metabolic disturbances (decreased LDL, increased HDL), and ameliorated histopathological injury.

Artesunate has been reported to exert protective effects in various liver injury models, including acetaminophen‑induced injury,10,11 non‑alcoholic fatty liver disease,12 and hepatic ischemia‑reperfusion injury,13 with its actions mainly attributed to broad‑spectrum activities such as antioxidant, anti‑inflammatory, and autophagy regulation. However, none of these studies have identified the direct protein target through which artesunate mediates its hepatoprotective effects. Our phenotypic findings are consistent with the above reports, and further chemical proteomics screening identified MBL2 as a potential direct binding target of artesunate, providing new molecular‑level evidence for elucidating its hepatoprotective mechanism. The above observations provide a solid phenotypic foundation for subsequent mechanistic investigations.

Chemical Proteomics Identifies MBL2 as a Direct Target of ART

To elucidate the molecular basis of the aforementioned hepatoprotective effects, this study identified MBL2 as a key target in ART-mediated liver protection and further clarified its regulatory mechanism on the NF‑κB/NRF2 signaling axis. Using a biotin-labeled ART probe in pull‑down assays coupled with liquid chromatography–tandem mass spectrometry (LC‑MS/MS) analysis, over 1100 potential interacting proteins were identified from rat tissue samples. Subsequently, through multiple stringent filtering criteria—including spectral scores, molecular weight match, and functional annotation—non‑specific binding proteins were progressively excluded, ultimately establishing MBL2 as a high‑confidence candidate target. In the mass spectrometry analysis, MBL2 exhibited stable and prominent binding signals. Its molecular weight closely matched the experimental detection range, and its known biological functions are intimately associated with the initiation and regulation of inflammatory responses during liver injury, thereby providing a plausible functional context for its role as a direct target of ART.

ART Alleviates Liver Injury by Coordinating the NF-κB/NRF2 Signaling Axis via MBL2

At the mechanistic level, integrating transcriptomic data, this study further links the ART‑MBL2 interaction to the suppression of the NF‑κB/NRF2 signaling pathway. Previous studies have shown that MBL2 can negatively regulate NF‑κB signaling by interfering with TLR4 receptor activation upon binding to pathogen-associated molecular patterns or lipopolysaccharide.17,27 Consistently, this study observed that PPT treatment downregulated MBL2 protein expression in liver tissue, accompanied by a significant increase in the levels of TLR4 and its downstream p‑NF‑κB p65. ART intervention specifically reversed the downregulation of MBL2 and concurrently suppressed the hyperactivation of the TLR4/NF‑κB signaling axis. These changes were consistent at both the protein and transcriptional levels, forming a continuous chain of evidence from “ART‑MBL2 binding” to “inhibition of inflammatory signaling”.

Notably, the inhibition of NF‑κB signaling further triggered transcriptional reprogramming associated with the NRF2 antioxidant pathway. Existing research indicates that persistently activated NF‑κB can inhibit the transcriptional activity of NRF2 through multiple mechanisms, including enhancing KEAP1‑mediated degradation of NRF2 or directly interfering with its transcriptional activation function.28,29 In this study, ART‑mediated NF‑κB suppression showed a clear correlation with increased NRF2 protein levels and activation of its downstream antioxidant genes (HO‑1, NQO1), alongside decreased KEAP1 expression. This suggests that the negative crosstalk between NF‑κB and NRF2 was rebalanced during ART treatment.30,31

The NF-κB/NRF2 signaling axis and its reciprocal regulation are well documented in the context of liver injury and artesunate pharmacology.32 What distinguishes the present study from previous work is the identification of MBL2 as a direct binding protein of artesunate. This offers a potential molecular entry point for understanding how artesunate might modulate these pathways. However, the causal relationship between MBL2 binding and the signaling changes we observed remains to be established. Specifically, while our data show that artesunate treatment correlates with changes in MBL2 expression and NF-κB/NRF2 activity, they do not prove that MBL2 is the obligate mediator of these effects.

Accumulating evidence supports a protective role for MBL2 in liver injury through negative regulation of the TLR4/NF-κB inflammatory axis. In a CCl₄-induced acute liver injury model, MBL-deficient mice exhibited markedly increased hepatocellular apoptosis. Serum and intrahepatic pro-inflammatory cytokine levels (TNF-α, IL-6, IL-1β) were also elevated in these mice. Conversely, liver-specific reconstitution of MBL2 via adeno-associated virus significantly ameliorated liver injury. This reconstitution was accompanied by reduced ALT, LDH, and inflammatory cytokine levels.16 Mechanistically, MBL2 can directly engage the TLR4 ectodomain through its carbohydrate-recognition domain. This binding attenuates LPS binding to the cell surface. It further suppresses NF-κB DNA binding and nuclear translocation. This in turn curtails the production of TNF-α and IL-12.17,33,34 Further demonstrated that MBL suppresses the TLR4/NF-κB pathway in a dose-dependent manner. Collectively, these findings establish MBL2 as a critical negative regulator of inflammatory responses in the innate immune system. Its hepatoprotective function, mediated through inhibition of the TLR4/NF-κB axis, has been validated in multiple independent experimental settings.

These prior observations are consistent with our current findings. In artesunate-treated animals, we likewise observed MBL2 upregulation and concurrent suppression of TLR4/NF-κB signaling. These observations provide biological plausibility for the hypothesis that artesunate may exert its hepatoprotective effects, at least in part, through MBL2. Nevertheless, it must be acknowledged that the cited evidence only supports a role for MBL2 in suppressing TLR4/NF-κB signaling. This evidence does not directly establish that artesunate’s protective efficacy is dependent upon MBL2. Definitive demonstration of this causal relationship will require loss-of-function experiments. For instance, one could examine whether the protective effect of artesunate is abrogated or attenuated in MBL2-deficient cellular or animal models. Thus, while our data support MBL2 as a candidate mediator of artesunate’s hepatoprotective action, its functional necessity remains to be rigorously tested in future genetic intervention studies.

Therefore, from a signaling pathway perspective, ART targets MBL2 to inhibit NF‑κB activation, thereby indirectly relieving its suppression of NRF2 and promoting the initiation of the NRF2‑mediated antioxidant defense program. This results in the coordinated regulation of inflammatory responses and oxidative stress. Liver transcriptomic analysis provided genome‑wide support for this mechanism. Differentially expressed genes and pathway enrichment results were highly concentrated in functional modules related to NF‑κB signaling reversal, enhanced antioxidant activity, and glutathione metabolism—all associated with NRF2 function—further validating the global remodeling effect of the ART‑MBL2‑NF‑κB/NRF2 axis at the transcriptional level. Additionally, changes in a few genes related to lipid metabolism and inflammation interaction (eg, FABP5, PPARγ) suggest that while inhibiting inflammation and oxidative stress, ART may also exert a synergistic regulatory effect on the hepatic metabolic microenvironment, although its specific mechanisms warrant further investigation.35–37

Systems Biology Analysis Reveals the Role of the Gut–Liver Axis in ART-Mediated Protection

To understand the protective effect of ART from a more systemic perspective, this study integrated targeted metabolomics and gut microbiota analyses, jointly revealing the potential mechanisms by which ART exerts synergistic protective effects at the level of distal organs and systemic regulation. Targeted metabolomics confirmed that ART significantly reversed the systemic metabolic dysregulation induced by PPT. The correction of core metabolites, specifically L-methionine and uric acid, directly validated the upstream activation of the NRF2 pathway.38–41 This functional complementation at the metabolite level provided substantial support for hepatic antioxidant defense.

Concurrently, ART effectively remodeled the gut microbial collapse caused by PPT. It suppressed potential pathogenic bacteria associated with intestinal barrier damage and pro-inflammatory responses (eg, Parasutterella, Bacteroides_F)42,43 and enriched the genus Lactobacillus, known for its roles in intestinal barrier protection and anti-inflammatory effects.44 This restoration of the microbiota structure toward a healthier state likely reduced the translocation of gut-derived endotoxins (eg, LPS), thereby helping to alleviate the aberrant activation of the hepatic TLR4/NF-κB inflammatory pathway at its source.45,46

Crucially, statistically significant and biologically logical correlations were observed between gut microbiota structure and host metabolic profiles. The abundance of Lactobacillus was positively correlated with the level of the antioxidant metabolite uric acid, whereas the pathogenic bacterium Parasutterella was negatively correlated with the level of L-methionine, a precursor for glutathione synthesis. These associations strongly suggest that ART’s remodeling of the gut microbiota is not an isolated event but, rather, influences the host’s systemic metabolic status—particularly the NRF2-related antioxidant metabolic network—via the gut–liver axis, while also modulating the inflammatory metabolic environment associated with NF-κB.

Therefore, this part of the study not only extends the evidence for ART’s systemic protection from the dimensions of metabolism and microecology but also concretizes the “gut–liver axis” from a theoretical concept into an observable biological dialogue tightly linked to the core hepatoprotective pathways (NF-κB/NRF2). This work delineates a multilevel, integrated mechanism of action for ART, spanning from its direct molecular target (MBL2) to local hepatic signaling pathways (NF-κB/NRF2), and further extending to the remodeling of distal gut microbiota and systemic metabolism.

Key Innovations of This Study

Overall, this study confirmed in an animal model that ART exhibits significant therapeutic efficacy against PPT-induced acute drug-induced liver injury. The protective effects were manifested as a dose-dependent amelioration of key serum liver injury markers, a marked alleviation of hepatic oxidative stress, and a reduction in histopathological damage. Mechanistically, the innate immune pattern recognition receptor MBL2 was identified and validated as the direct molecular target through which ART exerts its hepatoprotective effects. Furthermore, it was elucidated that by targeting MBL2, ART coordinately and inversely regulates the NF‑κB inflammatory pathway and the NRF2 antioxidant pathway, thereby integrating anti‑inflammatory and antioxidative mechanisms at the molecular level.

Limitations of the Study

Despite the progress achieved, this study has certain shortcomings and limitations. A primary limitation lies in the model system, as the research relied mainly on a PPT-induced liver injury rat model. The generalizability of ART’s protective effect and its MBL2‑dependent mechanism to chronic liver injury or DILI models of other etiologies remains to be verified. Another notable shortcoming is that, although this study provides systematic support for MBL2 as a plausible direct target of ART through chemical proteomics screening, in vivo correlation analyses, and multi‑level signaling validation, functional validation at the cellular level was not performed. Moreover, genetic approaches (eg, gene knockdown or overexpression) in cells or animals to directly demonstrate the necessity of MBL2 for ART’s efficacy are lacking. Consequently, the functional contribution of MBL2 in ART‑mediated hepatoprotection requires further refined verification in subsequent studies.

Conclusion

Daily intraperitoneal injection of ART at 150 mg/kg for 7 consecutive days significantly attenuated PPT‑induced liver injury. The protective effect was associated with artesunate binding to MBL2 and modulation of the TLR4/NF‑κB/NRF2/HO‑1 pathway. Specifically, artesunate treatment correlated with upregulation of MBL2 expression, suppression of TLR4/NF‑κB inflammatory signaling, and activation of the NRF2‑mediated antioxidant defense system. ART comprehensively exerts its hepatoprotective effect through this integrated mechanism, providing a clear pharmacological rationale for its potential application in the prevention and treatment of drug-induced liver injury.

Data Sharing Statement

This study did not generate new unique reagents. Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Chuanxin Liu (15222003775@163.com).

Accession numbers

1. The raw metabolomics data reported in this paper have been deposited in the Open Archive for Miscellaneous Data (OMIX) in National Genomics Data Center (Nucleic Acids Res 2025), China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences (OMIX: OMIX018425) that are publicly accessible at https://ngdc.cncb.ac.cn/omix/release/OMIX018425.

2. The raw sequence data reported in this paper have been deposited in the Genome Sequence Archive (Genomics, Proteomics & Bioinformatics 2025) in National Genomics Data Center (Nucleic Acids Res 2025), China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences (GSA: CRA045722) that are publicly accessible at https://ngdc.cncb.ac.cn/gsa.

3. The raw 16S rRNA sequencing data reported in this paper have been deposited in the Genome Sequence Archive (Genomics, Proteomics & Bioinformatics 2025) in National Genomics Data Center (Nucleic Acids Res 2025), China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences (GSA: CRA046161) that are publicly accessible at https://ngdc.cncb.ac.cn/gsa.

Ethical Statement

All animal experiments conducted in this study were approved by the Institutional Animal Care and Use Committee (IACUC) of the First Affiliated Hospital of Henan University of Science and Technology (Approval No. D-2025-027). The animal license for the experimental animals used in this research was obtained from a certified vendor (License No. SCXK (Beijing) 2021-0006). All procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals (8th edition, National Academies Press, 2011) and with all applicable institutional and national regulations. We gratefully acknowledge the contribution of the laboratory animals that made this study possible.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.