Compound Abrus cantoniensis Extract Alleviates AFB1‐Induced Liver Injury by Regulating PI3K‐AKT Pathway and Gut Microbiota/Butyrate Metabolism Hongjie Hu, Ming cheng, Shiqi Zheng, Huixin Liu, Xiaofang Wei, Wenwen Yang, Hongchun Yang, Liuyang Yuan, Li Jiang, Jie Chen, Linglong Zhao, Yanchao Guo, Hongbin Si Abstract Aflatoxin B1 (AFB1) is a potent hepatotoxin commonly ingested through contaminated food. Abrus cantoniensis Hance, a traditional Chinese medicinal herb, is commonly included in hepatoprotective compound prescriptions, yet its protective role against AFB1‐induced hepatic injury remains to be elucidated. Here, network pharmacology combined with molecular docking suggested that the compound Abrus cantoniensis extract (ACCE) attenuates AFB1‐induced hepatotoxicity via modulation of the PI3K‐AKT signaling pathway. This was validated through integrated transcriptomic, microbiomic, metabolomic, and in vivo/in vitro analyses. In vivo, ACCE ameliorated hepatic and intestinal damage in AFB1‐exposed mice, modulated PI3K‐AKT and TLR4/NF‐κB pathways, reshaped gut microbiota by enriching Lachnospiraceae_NK4A136_group, and increased short‐chain fatty acids (SCFAs), particularly butyrate. In vitro, ACCE activated PI3K‐AKT and suppressed TLR4/NF‐κB signaling in HepG2 cells, effects diminished by pathway inhibitors. Butyrate alone also protected against AFB1‐induced injury via TLR4/NF‐κB inhibition, an effect similarly reversed by inhibitors. Notably, combined treatment with ACCE or its main components and butyrate yielded superior protection. Collectively, these findings suggest that ACCE can alleviate AFB1‐induced liver injury by regulating the PI3K‐AKT pathway and the gut microbiota and its metabolite butyrate. ACCE has a good alleviating effect on AFB1‐induced liver injury, and this alleviating effect is related to the PI3K‐AKT pathway regulated by ACCE and the butyrate metabolism of gut microbiota. Introduction Aflatoxin B1 (AFB1), one of the most potent naturally occurring mycotoxins, widely contaminates food and animal feed globally. It poses severe threats to global food security and public health, causing considerable financial losses to the animal husbandry industry []. The liver bears the brunt of AFB1 toxicity. After metabolic activation by cytochrome P450 enzymes, AFB1 generates highly electrophilic epoxides that bind covalently to nucleic acids and proteins, inducing hepatocyte injury and genetic mutations []. Epidemiological studies have confirmed that AFB1 exposure is a class I carcinogen associated with hepatocellular carcinoma (HCC). Liver injury resulting from AFB1 contamination has become a significant health burden, particularly in developing countries []. The concept of the gut‐liver axis describes a two‐way signaling system between the intestine and the liver, which is critical for maintaining hepatic metabolism and intestinal homeostasis []. Gut microbial metabolites, including short‐chain fatty acids (SCFAs), are carried through the portal vein to the liver, where they are involved in modulating host metabolism and immune responses. An intact intestinal mucosa is regarded as the primary barrier that impedes harmful agents from accessing the liver [, ]. However, exposure to AFB1 can severely disrupt the homeostasis of the gut‐liver axis. AFB1 has been reported to suppress the expression of intestinal tight junction proteins (e.g., ZO‐1, Occludin, and Claudin‐1), thereby weakening the gut barrier, raising mucosal permeability, and inducing dysbiosis of the gut microbiota [, , ]. These changes facilitate the translocation of harmful substances, such as lipopolysaccharides (LPS), into the portal circulation, subsequently activating inflammatory pathways in the liver (e.g., TLR4/NF‐κB signaling pathway), exacerbating liver injury. Therefore, the disruption of the gut‐liver axis is considered one of the key pathways for AFB1‐induced hepatotoxicity. The gut microbiota, as a significant player in the gut‐liver axis, produces SCFAs, particularly butyrate, which can exert anti‐inflammatory actions through suppression of the TLR4/NF‐κB pathway and maintaining intestinal barrier function []. Previous studies have reported that the dysregulation of multiple signaling pathways or targets is involved in AFB1‐induced liver injury, including the PI3K‐AKT, PXR, NRF2, and NLRP3 pathways [, , ]. Among these, the PI3K‐AKT pathway is a core regulator of cell survival, proliferation, and apoptosis. Activation of the PI3K‐AKT pathway promotes hepatocyte survival and counteracts pro‐apoptotic signals, whereas AFB1 exposure suppresses this signaling cascade, thereby enhancing hepatocyte apoptosis and aggravating hepatic injury []. Meanwhile, the TLR4/NF‐κB pathway is a central mediator of inflammatory responses. When the intestinal barrier is compromised, LPS gains access to the portal vein and stimulates hepatic TLR4, thereby initiating NF‐κB activation and driving the expression of pro‐inflammatory cytokines including TNF‐α, IL‐1β, and IL‐6 []. Therefore, hepatoprotective strategies that can simultaneously regulate the PI3K‐AKT and TLR4/NF‐κB pathways may have potential for alleviating AFB1‐induced liver injury. Originating from the Lingnan area of China, Abrus cantoniensis Hance is a geo‐authentic medicinal herb that contains high levels of flavonoids, saponins, and polysaccharides. These bioactive compounds confer multiple therapeutic properties, such as anti‐inflammatory, antioxidant, and immunomodulatory activities []. In traditional Chinese medicine (TCM) practice, Abrus cantoniensis is incorporated into over 100 compound formulations due to its capacity to soothe the liver, promote bile secretion, detoxify, and resolve stasis, with its hepatoprotective efficacy widely acknowledged. However, whether the compound Abrus cantoniensis confers protection against AFB1‐induced hepatic injury remains unknown, and the associated mechanisms likewise require further investigation. Therefore, the current investigation was designed to comprehensively explore the hepatoprotective efficacy and mechanistic basis of the compound Abrus cantoniensis extract (ACCE) against AFB1‐related hepatic injury. To achieve this goal, we employed a combination of various research methods, including network pharmacology, molecular docking, transcriptomics, microbiomics, and targeted metabolomics, alongside in vivo and in vitro research systems []. We strive to reveal the mechanisms by which ACCE alleviates AFB1‐induced liver injury from multiple angles and levels, providing substantial experimental evidence for its development as a hepatoprotective drug. Results ACCE Effectively Alleviates AFB1‐Induced Liver Injury To clarify the effect of ACCE treatment on AFB1‐induced liver injury, mouse body weights and liver indices in each group were recorded (Figure ). Prior to AFB1 exposure, body weights did not differ significantly among groups. After 7 and 14 days of exposure, mice in the AFB1 group showed a slower increase in body weight compared to other groups, although these differences were not statistically significant. The liver index in the AFB1 group was significantly higher than in the NC group, whereas intervention with ACCE and silymarin effectively reversed this increase. H&E staining revealed that liver tissues from the NC group exhibited normal hepatic structures. Conversely, the AFB1 group displayed disrupted hepatic plate arrangement, hepatocytes of varying sizes, ballooning degeneration with vacuoles of varying dimensions (black arrows), as well as hepatocyte necrosis and nuclear fragmentation or dissolution (red arrows) (Figure ). In comparison, liver tissues from ACCE‐L, ACCE‐H, and silymarin groups showed partial cellular disorganization, significantly attenuated vacuolar degeneration, and improved nuclear morphology. The DMSO control and ACCE‐alone groups exhibited intact hepatic cord arrangements and normal cell morphology, without significant differences from the NC group, indicating that neither 2% (v/v) DMSO nor ACCE alone induced liver injury. Serum activities of liver function markers (ALT, AST, ALP, LDH) were significantly elevated after AFB1 exposure compared to the NC group, indicating liver injury. Treatment with ACCE and silymarin significantly reduced these enzyme activities, with the most notable improvement observed in the ACCE‐H group (Figure ). Regarding oxidative stress, hepatic MDA levels were significantly increased, while activities of SOD and GSH‐Px were significantly reduced in the AFB1 group. Treatment with ACCE‐L, ACCE‐H, and silymarin reversed these changes to varying degrees (Figure ). In addition, the elevated levels of inflammatory factors TNF‐α, IL‐1β, and IL‐6 in liver tissues induced by AFB1 exposure were significantly alleviated by treatment with ACCE‐L, ACCE‐H, and Silymarin. Collectively, these findings indicated that AFB1 exposure caused oxidative and inflammatory liver injury, which was effectively alleviated by ACCE intervention. ACCE Ameliorates AFB1‐Induced Intestinal Injury and Systemic Inflammation To investigate the effects of ACCE treatment on intestinal injury induced by AFB1, ileal tissues were examined by H&E staining (Figure ). Compared with the NC group, mice in the AFB1 group showed significantly decreased villus height and villus‐to‐crypt (V/C) ratio, indicating impaired intestinal absorption. These pathological changes were significantly mitigated in the ACCE‐L, ACCE‐H, and silymarin groups, with the V/C ratio substantially restored (Figure ). In addition, serum levels of LPS were markedly increased following AFB1 exposure, while ACCE and silymarin significantly reduced these levels. Serum inflammatory cytokines, indicative of systemic inflammation, demonstrated consistent changes (Figure ). The expression levels of tight junction proteins (ZO‐1, Occludin, Claudin 1), which are indicative of intestinal mucosal integrity, were found to be down‐regulated by AFB1. Notably, this down‐regulation was significantly reversed following treatment with ACCE‐H (Figure ). Collectively, these results indicated that AFB1 exposure caused intestinal injury, endotoxin translocation into the bloodstream, and systemic inflammation, whereas ACCE intervention effectively attenuated these pathological alterations. Network Pharmacology Combined with Transcriptomics Analysis of the Potential Mechanism by which ACCE Alleviates Liver Injury The composition of ACCE was analyzed using UPLC‐MS/MS (Figure ). Total ion chromatograms (TICs) of pooled quality control (QC) samples in negative and positive ion modes are presented in Figure , respectively (Figure ). A total of 24 main components were identified in ACCE, including salvianolic acid B, neochlorogenic acid, chlorogenic acid, isoschaftoside, and butyl isobutyl phthalate, among others. Detailed information is provided in the supplementary table (Table ). Based on the 24 primary ACCE components identified by UPLC‐MS/MS, target prediction using the SwissTargetPrediction database yielded 396 non‐redundant targets after deduplication. A total of 643 targets associated with AFB1‐induced liver disease (ALD) were retrieved from the GeneCards and OMIM databases. The intersection of component targets and disease‐related targets yielded 100 potential therapeutic targets (Figure ). A “drug‐active component‐potential target” network was constructed using these 100 targets and their corresponding 24 active components, and visualized with Cytoscape 3.10.2 (Figure ). Among the components, []‐gingerol, caffeate, cryptotanshinone, isochlorogenic acid C, salvianolic acid B, and cytarabine ranked highest based on degree values, indicating their potential as key active constituents of ACCE against AFB1‐induced liver injury. The 100 potential targets were further imported into the STRING database to generate a protein‐protein interaction (PPI) network, visualized using Cytoscape 3.10.2 (Figure ). The top eight targets (including TP53, STAT3, MAPK1, and EGFR) occupied central positions in the network, suggesting their critical roles in the regulation of AFB1‐induced liver injury. Additionally, these 100 potential targets underwent GO and KEGG enrichment analyses using the DAVID database (Figure ). GO analysis indicated that ACCE might alleviate AFB1‐induced liver injury through biological processes such as PI3K‐AKT signaling, MAPK cascade, gene expression regulation, and apoptosis modulation. KEGG pathway enrichment further suggested the involvement of multiple pathways, particularly the PI3K‐AKT and MAPK signaling pathways. A drug‐component‐target‐pathway‐disease network was constructed based on the core targets, their corresponding active components, and relevant pathways (Figure ). Among these, five genes (MAPK1, PIK3CA, EGFR, STAT3, and TP53) rank high in Degree values, suggesting that they may serve as key targets. Molecular docking was then performed on the five core targets and their corresponding active components to facilitate virtual screening and predict potential binding modes. The heatmap results revealed that the binding energies between major active components and core targets were all below −4.0 kcal/mol. This value is a commonly used threshold in virtual screening; values below this threshold suggest that there may be good binding activity between the two. This indicates that these components may play a crucial role in alleviating AFB1‐induced liver injury (Figure ). The 12 top‐ranked target–component combinations, based on binding energies, were selected for 3D structural visualization (Figure ). Of course, both molecular docking and network pharmacology analyses are prediction results and do not constitute direct evidence of target‐component binding. To further elucidate the mechanisms underlying the hepatoprotective effects of ACCE against AFB1‐induced injury, mouse liver gene expression profiles were analyzed using RNA‐seq. DEGs between the AFB1 and NC groups, as well as between the ACCE‐H and AFB1 groups, were visualized as volcano plots (Figure ). GO and KEGG enrichment analyses demonstrated that DEGs were primarily enriched in biological processes such as regulation of stress responses, positive regulation of cellular processes, and regulation of inflammatory responses (Figure ). Pathways including Pi3k‐Akt, Mapk, and Toll‐like receptor (TLR) signaling were significantly enriched (Figure ). Gene Set Enrichment Analysis (GSEA) further confirmed significant enrichment of the TLR signaling pathway (Figure ). Additionally, a heatmap of related genes involved in these three pathways was constructed (Figure ). ACCE Restores AFB1‐Induced Gut Microbiota Dysbiosis To explore the effects of ACCE treatment on gut microbiota and the intestinal mucosal barrier, 16S rRNA gene sequencing was performed on mouse intestinal content samples. α‐Diversity analysis revealed decreased values of Sobs, Ace, Chao, and Shannon indices in the AFB1 group, whereas these parameters were reversed by ACCE‐H treatment, with the Shannon index significantly increased (Figure ). β‐Diversity analysis indicated that ACCE‐H treatment corrected the disrupted gut microbiota structure induced by AFB1, making it closer to that observed in the NC group (Figure ). At the phylum level, the relative abundance of Actinomycetota increased following AFB1 exposure (Figure ). At the genus level, significant alterations were observed in the abundances of SCFA‐producing bacteria (e.g., Lachnospiraceae_NK4A136_group, norank_o_Clostridia_UCG‐014, norank_f_Lachnospiraceae) and potentially pathogenic bacteria (e.g., Corynebacterium) in the AFB1 group, whereas ACCE‐H treatment effectively restored their abundances (Figure ). Correspondingly, levels of short‐chain fatty acids (SCFAs), including acetate, propionate, and butyrate, were normalized in the ACCE‐H group, with butyrate concentrations positively correlating with abundances of Lachnospiraceae_NK4A136_group (Figure ). Similarly, peripheral blood butyrate levels were significantly higher in the ACCE‐H group than in the AFB1 group (Figure ). Potential Mechanism of ACCE in Alleviating Liver Injury Based on results from PPI network analysis, molecular docking, KEGG pathway enrichment, and transcriptome sequencing, it was hypothesized that the PI3K‐AKT pathway (associated with cell proliferation) and the TLR4/NF‐κB pathway (associated with inflammation) are critically involved in ACCE‐mediated alleviation of AFB1‐induced liver injury. Western blot analysis demonstrated that ACCE‐H treatment significantly increased protein expression levels of phosphorylated EGFR (p‐EGFR), phosphorylated AKT (p‐AKT), and Bcl‐2 within the PI3K‐AKT pathway, while significantly decreasing the expression of TLR4, MyD88, and phosphorylated IκBα (p‐IκBα) within the TLR4/NF‐κB pathway (Figure ). Immunofluorescence staining of liver sections confirmed these findings, indicating that ACCE‐H treatment activated p‐EGFR and p‐AKT and inhibited the expression of TLR4 and NF‐κB p65 (Figure ). Furthermore, correlation analysis revealed that SCFA levels, particularly butyrate, were significantly negatively correlated with inflammatory factors (e.g., TNF‐α) and serum LPS levels, while positively correlated with the expression levels of intestinal tight‐junction proteins (ZO‐1, Occludin, Claudin 1). Collectively, these data support the hypothesis that the gut‐butyrate‐liver axis may play a pivotal role in ACCE‐mediated alleviation of AFB1‐induced liver injury (Figure ). Validation via Cell Experiments Integrating preliminary experimental findings and existing literature, the following hypothesis was proposed: ACCE acts through two complementary mechanisms. On one hand, its active components directly target hepatocytes, activating the PI3K‐AKT survival pathway and suppressing the TLR4/NF‐κB inflammatory pathway; on the other hand, it modulates gut microbiota to enhance the production of butyrate and other SCFAs, which subsequently enter the liver through the portal vein and synergistically inhibit the TLR4/NF‐κB pathway. Thus, ACCE alleviates AFB1‐induced liver injury through a multi‐target and multi‐pathway mechanism. To validate this hypothesis, further in vitro studies were performed using HepG2 hepatocytes. ACCE Alleviates Hepatocyte Injury via the PI3K‐AKT Pathway To investigate the protective effect and underlying mechanism of ACCE against AFB1‐induced hepatocyte injury, in vitro experiments were conducted using HepG2 cells. CCK‐8 assay results showed that ACCE exhibited no cytotoxicity at concentrations ranging from 15.63 to 500 μg/mL. The half‐maximal inhibitory concentration (IC50) of AFB1 was 50 μg/mL, and ACCE dose‐dependently alleviated AFB1‐induced cellular injury (Figure ). Pretreatment with the broad‐spectrum PI3K inhibitor LY294002 (safe concentration: 0–12.5 μm) significantly attenuated the protective effect of ACCE. Compared with the AFB1‐only group, cells treated with ACCE showed significantly enhanced cell viability, decreased levels of inflammatory cytokines and liver function markers, reduced Caspase‐3 levels, and a decreased proportion of late‐apoptotic cells. However, these beneficial effects were partially reversed following pretreatment with LY294002 (Figure ). Furthermore, expression levels of key proteins within the PI3K‐AKT pathway (e.g., p‐AKT) were significantly elevated in the AFB1+ACCE group compared to the AFB1 group, but this elevation was abolished by LY294002 pretreatment (Figure ). Collectively, these findings indicate that ACCE mitigates AFB1‐induced apoptosis by activating the PI3K‐AKT signaling pathway in hepatocytes. Butyrate Alleviates Hepatocyte Inflammation via the TLR4/NF‐κB Pathway To further validate the anti‐inflammatory effects of butyrate and elucidate its underlying mechanisms in hepatocytes, an in vitro ‘AFB1+LPS’ combined stimulation model was established to mimic a “leaky gut” scenario, whereby hepatocytes are concurrently exposed to the exogenous toxin AFB1 and endotoxin LPS. The TLR4 inhibitor TAK‐242 was applied as a pretreatment agent. Preliminary tests determined that safe concentration ranges for butyrate and TAK‐242 in HepG2 cells were 0–1.25 mm and 0–0.025 μm, respectively (Figure ). Combined exposure to AFB1 and LPS significantly decreased cell viability compared with either agent alone, suggesting an additive hepatotoxic effect. Butyrate administration alleviated this combined injury in a dose‐dependent manner (Figure ). After TAK‐242 pretreatment, cell viability in the butyrate‐treated group showed no significant difference from the TAK‐242‐alone group, both of which were significantly higher than the AFB1+LPS group (Figure ). Additionally, groups treated with butyrate alone, TAK‐242 alone, or their combination exhibited significant improvements in inflammatory cytokine levels, liver function markers, Caspase‐3 levels, and apoptosis rates compared to the AFB1+LPS group, with no significant differences among these three treatment groups (Figure ). In addition, both butyrate and TAK‐242 treatments significantly suppressed the expression of key proteins within the TLR4/NF‐κB pathway (Figure ). These results collectively indicate that butyrate exerts anti‐inflammatory and hepatoprotective effects by inhibiting the TLR4/NF‐κB signaling pathway. Synergistic Effect of ACCE and Butyrate in Alleviating Hepatocyte Injury To investigate the synergistic effect of ACCE and butyrate in alleviating hepatocyte injury, a combination of half‐dose ACCE (125 μg/mL) and half‐dose butyrate (0.313 mm) was applied. The results demonstrated that, compared with the corresponding half‐dose single‐agent groups, the combination treatment significantly increased cell viability and produced more pronounced reductions in inflammatory cytokine levels, liver function markers, Caspase‐3 levels, and apoptotic cell proportions, with a protective effect comparable to that of the full‐dose ACCE group (Figure ). Regarding the regulation of key proteins in the PI3K‐AKT and TLR4/NF‐κB pathways, the combination group also exhibited stronger regulatory effects than the single‐agent half‐dose groups (Figure ). These findings indicate that ACCE and butyrate exert synergistic effects in suppressing liver inflammation and reducing hepatocyte apoptosis. Validation of Synergistic Effects among Functional Components of ACCE Based on UPLC‐MS/MS analysis and network pharmacology predictions, three major components with high abundance in ACCE and strong binding affinity to core targets, caffeate, cryptotanshinone, and 6‐gingerol (all of which can be absorbed into the blood), were selected to evaluate their synergistic effects in alleviating liver injury [, , ]. Cell viability assays determined their safe concentration ranges to be 0–12.5 μm, 0–3.125 μm, and 0–12.5 μm, respectively. Concentrations that effectively alleviated AFB1‐induced hepatocyte injury (12.5 μm, 3.125 μm, and 12.5 μm) were defined as effective concentrations (ECs), whereas concentrations showing minimal or no protective effects when used individually (6.25 μm, 1.563 μm, and 3.125 μm) were defined as subeffective concentrations (SCs). A physical mixture of the three compounds at their SCs (subeffective concentration mixture, SCM) was prepared (Figure ). The results showed that ACCE and the ECs of the three compounds significantly enhanced cell viability, whereas SC treatments alone showed no significant effect. In contrast, SCM treatment significantly increased cell viability (Figure ). Similarly, SCM treatment markedly reduced inflammatory cytokine levels, liver function markers, Caspase‐3 levels, and apoptotic cell proportions, and modulated the expression of key proteins in the PI3K‐AKT and TLR4/NF‐κB pathways, whereas SC treatments alone showed no significant effects (Figure ). These findings indicate that the principal active components of ACCE exert synergistic effects in alleviating hepatocyte injury. Cross‐Dimensional Synergy between Functional Components and Butyrate in Alleviating Hepatocyte Injury To further explore the cross‐dimensional synergy between the main active components of ACCE and the microbial metabolite butyrate in alleviating liver injury, a new subeffective concentration mixture (NSCM) was constructed by combining SCM (SCs of caffeate, cryptotanshinone, and 6‐gingerol) with a subeffective concentration of butyrate. The results demonstrated that ACCE, butyrate alone, and SCM treatment all significantly increased the viability of HepG2 cells after AFB1+LPS exposure, whereas the subeffective concentration of butyrate alone had no significant effect. Notably, NSCM treatment markedly enhanced cell viability (Figure ). Similarly, NSCM significantly reduced inflammatory cytokine levels, liver function markers, Caspase‐3 levels, and apoptotic cell proportions, and effectively regulated the expression of key proteins in the PI3K‐AKT and TLR4/NF‐κB pathways, with stronger effects than those observed with SCM or subeffective butyrate alone (Figure ). These results suggest that the hepatoprotective effects of ACCE arise from both intra‐component synergy among its active constituents and cross‐dimensional synergy between these components and the microbial metabolite butyrate. Butyrate or Cryptotanshinone Can Alleviate AFB1‐Induced Liver Injury in Mice Butyrate and Cryptotanshinone, which have a good protective effect on hepatocytes in cell assays, will be intervened in AFB1 exposed mice to further explore whether they are equally effective in vivo. The results showed that butyrate or Cryptotanshinone could reverse the up‐regulation of AFB1 in liver index, serum liver function indicators (ALT, AST), and inflammatory factors (TNF‐α, IL‐1β, IL‐6) levels in liver tissue. This finding indicated that either butyrate or Cryptotanshinone was effective in alleviating AFB1‐induced liver injury (Figure ). The above in vivo results proved that the hepatoprotective effect of ACCE was related to its main active component, and butyrate, as an important player in the gut‐liver axis, also played an important role in the hepatoprotective process. Discussion According to existing data, AFB1 widely contaminates grain and animal feed globally and accumulates in animals and humans through the food chain, with the liver as its major target []. This contamination inflicts heavy economic losses on the livestock sector and simultaneously endangers human health, including the potential induction of liver cancer []. Currently, it is recognized that AFB1 induces liver injury mainly through two pathways that often occur concurrently. On one hand, AFB1 is metabolized by cytochrome P450 enzymes into epoxide derivatives, leading to oxidative stress, DNA damage, inflammation, and apoptosis, ultimately causing hepatocellular injury. Additionally, the release of damage‐associated molecular patterns (DAMPs) from injured liver tissues amplifies the inflammatory response [, ]. On the other hand, AFB1 compromises intestinal homeostasis by disrupting the mucosal barrier and tight junction proteins, leading to elevated permeability and subsequent LPS translocation to the liver, thereby intensifying inflammatory damage [, , ]. Therefore, novel agents that confer dual protection on both the liver and intestine are urgently needed to effectively counteract AFB1‐induced hepatic injury []. Our findings demonstrate that ACCE significantly alleviated AFB1‐induced hepatic injury and intestinal barrier dysfunction, with its protective effect closely associated with the synergistic actions of its main bioactive components and regulation of gut microbiota. Consistent with previous findings, AFB1 exposure disrupted hepatocyte architecture and caused elevated liver function markers and oxidative stress [, ]. In this study, ACCE significantly decreased serum ALT and AST levels, mitigated oxidative stress, and demonstrated a pronounced hepatoprotective effect. Additionally, the intestinal barrier serves as a critical defense against harmful substances, playing an essential role in maintaining gut homeostasis [, ]. Numerous studies have indicated that AFB1 exposure not only induces liver injury but also impairs intestinal barrier function, with these two aspects closely interconnected. Here, ACCE markedly improved ileal morphological integrity, reduced intestinal permeability, and concurrently decreased serum LPS and inflammatory cytokine levels. These findings suggest that ACCE alleviates liver injury by restoring intestinal barrier function and preventing the systemic entry of detrimental substances such as LPS. Network pharmacology, an essential tool in TCM research, can, to some extent, systematically predict the multi‐component and multi‐target therapeutic mechanisms of TCM formulations [, , ]. By integrating UPLC‐MS/MS analysis with network pharmacology, the present study identified 24 principal active components in ACCE (e.g., caffeate, cryptotanshinone, and 6‐gingerol) and screened 100 potential therapeutic targets related to AFB1‐induced liver injury (such as EGFR, STAT3, and MAPK1). Computer predictions suggested that the active constituents may bind strongly to the identified targets. GO and KEGG pathway analyses indicated that regulation of gene expression, apoptotic processes, and the PI3K‐AKT and MAPK cascades are potentially involved in the hepatoprotective action of ACCE against AFB1‐induced injury. Consequently, it was hypothesized that the protective effect of ACCE might be closely linked to its anti‐apoptotic activity, a mechanism requiring further experimental validation. Integrative multi‐omics analysis is a powerful tool for elucidating drug mechanisms [, , ]. In this study, RNA‐seq data revealed that the DEGs were predominantly enriched in the PI3K‐AKT and Toll‐like receptor pathways, closely aligning with network pharmacology predictions. Results from 16S rRNA gene sequencing indicated that AFB1 exposure perturbed gut microbial homeostasis, whereas ACCE treatment effectively reshaped the microbial community by promoting beneficial taxa, including Lachnospiraceae_NK4A136_group, norank_f_Lachnospiraceae, and norank_o_Clostridia_UCG‐014 [, , ]. Meanwhile, targeted metabolomic analysis demonstrated that levels of SCFAs (acetate, propionate, and butyrate) returned to normal after ACCE administration. Notably, a positive correlation was observed between butyrate levels and the relative abundance of Lachnospiraceae_NK4A136_group, a known SCFA producer. It was thus hypothesized that ACCE exerts its hepatoprotective effect by enriching SCFA‐producing bacteria, especially butyrate‐producing species, thereby elevating intestinal butyrate levels. The consistency between network pharmacology predictions and transcriptomic findings often serves as a critical basis for exploring pharmacological mechanisms [, , ]. Integrating these analyses, the PI3K‐AKT pathway (governing cell proliferation and survival) and the TLR4/NF‐κB pathway (associated with inflammation) were hypothesized as key mechanisms in the protective action of ACCE against AFB1‐induced hepatic injury. This hypothesis was preliminarily supported by alterations in the expression levels of critical pathway proteins. Additionally, multi‐omics data demonstrated that ACCE significantly enriched intestinal SCFA‐producing bacteria and elevated butyrate concentrations. However, whether gut microbiota‐derived butyrate participates in ACCE‐mediated hepatoprotection, and the exact mechanism by which it acts, remained unclear. Previous studies have reported multiple physiological effects of butyrate, including anti‐inflammatory activity, maintenance of intestinal barrier function, and hepatoprotection, with its anti‐inflammatory effects primarily attributed to inhibition of the TLR4/NF‐κB signaling cascade [, , , ]. Notably, in our study, butyrate levels exhibited a significant negative correlation with inflammatory cytokines, serum LPS, and liver function markers, while a positive correlation was observed with intestinal tight junction protein expression. Thus, it was hypothesized that butyrate might serve as a critical signaling molecule connecting gut microbiota to liver protection, exerting its beneficial action by targeting the TLR4/NF‐κB signaling cascade in hepatocytes. Integrating these findings, the following mechanistic hypothesis was proposed for ACCE‐mediated alleviation of AFB1‐induced liver injury: on one hand, active components of ACCE directly act upon hepatocytes to activate the PI3K‐AKT survival signaling pathway and suppress the TLR4/NF‐κB inflammatory axis; on the other hand, ACCE modulates gut microbiota, enriching SCFAs (particularly butyrate), which subsequently enter the liver via the portal vein and synergistically suppress the TLR4/NF‐κB signaling. Thus, ACCE alleviates liver injury through a multi‐target and multi‐pathway mechanism. This hypothesis was further verified through subsequent cell experiments. In vitro studies visually demonstrated hepatocyte damage induced by AFB1 and the protective effects of ACCE and butyrate at the cellular level []. Upon exposure to AFB1, HepG2 cells exhibited morphological shrinkage, reduced viability, increased inflammatory cytokine levels and liver injury markers, upregulated apoptotic protein expression, and elevated proportions of late‐apoptotic cells. Treatment with ACCE significantly reversed these alterations and inhibited apoptosis; however, pretreatment with the PI3K inhibitor LY294002 weakened ACCE's anti‐apoptotic effects, and the ACCE‐induced increase in PI3K‐AKT pathway proteins was significantly diminished. These results confirmed that ACCE protects hepatocytes from AFB1‐induced apoptosis primarily through PI3K‐AKT pathway activation. To further determine whether butyrate mimics ACCE's anti‐inflammatory effects, a “leaky gut” model was created using co‐exposure to AFB1 and LPS in HepG2 cells, resulting in more severe injury than AFB1 alone. Similar to ACCE, butyrate significantly alleviated this injury and inhibited apoptosis. However, pretreatment with the TLR4 inhibitor TAK‐242 abolished butyrate's anti‐inflammatory effect (the anti‐inflammatory effects of butyrate and TAK‐242 alone were not significantly different). Upon TLR4 blockade, neither the pro‐inflammatory effects of LPS nor the anti‐inflammatory effects of butyrate occurred through the TLR4/NF‐κB axis, as protein levels of this pathway were uniformly lower across all treatment groups relative to the AFB1+LPS combination, further corroborating the pathway's involvement. Therefore, these findings indicate that butyrate, as a key mediator produced through ACCE‐induced modulation of gut microbiota, exerts its hepatoprotective effects predominantly through suppression of the TLR4/NF‐κB axis. As a TCM compound, ACCE is rich in multiple active constituents, and network pharmacology predictions indicated robust binding affinity between these constituents and multiple targets. Therefore, a key question arises: is the hepatoprotection of ACCE against AFB1‐induced hepatic injury driven primarily by a single component, or do multiple components act synergistically? To address this issue, caffeate, cryptotanshinone, and 6‐gingerol, which exhibited high abundance in UPLC‐MS/MS analysis and strong binding affinity to core targets, were selected for further validation. When used individually at SC, the three components had minimal effects on AFB1‐induced elevations in inflammatory cytokines, abnormal liver function markers, Caspase‐3 levels, and apoptotic cell proportions. However, treatment with SCM significantly enhanced these protective effects. Similarly, SCM treatment significantly regulated the expression of PI3K‐AKT and TLR4/NF‐κB pathway proteins. These results suggest that the hepatoprotection of ACCE against AFB1‐induced hepatic injury stems from direct, cooperative regulation of hepatocytes by its multiple active components acting on diverse targets. Moreover, given that butyrate is a key effector molecule mediating the indirect hepatoprotective effect of ACCE, does it exhibit synergy with the primary active components of ACCE? To answer this, a new subeffective concentration mixture (NSCM) was constructed by combining SCM with a subeffective concentration of butyrate. Results revealed that NSCM exerted superior effects compared to SCM or butyrate alone, significantly improving inflammatory cytokines, liver function markers, Caspase‐3 levels, and apoptotic cell proportions. These data confirm that the hepatoprotection of ACCE against AFB1‐induced hepatic injury relies not only on synergy among its active components but also on cross‐dimensional synergy between these components and butyrate. However, this study has limitations in research direction. For example, the liver is a complex organ composed of various cell types, including parenchymal cells (hepatocytes) and non‐parenchymal cells, which encompass hepatic stellate cells, Kupffer cells (resident hepatic macrophages), liver sinusoidal endothelial cells, and cholangiocytes [, , ]. These non‐parenchymal cells are critically involved in AFB1‐induced inflammation, fibrosis, and liver injury. As integral components of the liver's innate immunity, Kupffer cells release pro‐inflammatory mediators (e.g., TNF‐α, IL‐1β, IL‐6), promoting liver inflammation triggered by AFB1. Hepatic stellate cells are central to the process of fibrosis. Our in vitro experiments focused solely on hepatocytes and did not assess the effects of ACCE on non‐parenchymal cells. For instance, it remains unclear whether ACCE can inhibit the activation of Kupffer cells, alleviate macrophage‐mediated inflammatory cascades, or modulate the barrier function of liver sinusoidal endothelial cells. Therefore, the anti‐inflammatory and hepatoprotective effects of ACCE may involve the regulation of multiple liver cell types, and future studies may need to utilize co‐culture systems (such as hepatocytes and Kupffer cells) to further analyze the cell type‐specific actions of ACCE. In addition, AFB1 requires bioactivation by cytochrome P450 enzymes (CYP450), primarily CYP1A2 and CYP3A4, to produce the reactive epoxide AFBO, which are the ultimate hepatotoxic and carcinogenic metabolite []. Consequently, numerous natural products have been shown to mitigate the hepatotoxicity of AFB1 by downregulating the expression and activity of these CYP450 subtypes, thereby reducing the metabolic activation of AFB1 [, ]. Given the complexity of the AFB1 metabolic network and our observation that ACCE has provided strong protective effects in subacute environments, we propose that future studies should systematically evaluate whether ACCE offers additional protective effects through the downregulation of hepatic CYP1A2 and CYP3A4 expression and activity in chronic, low‐dose exposure models. This would offer a more holistic perspective on the multi‐layered protective mechanisms of ACCE. Similar to ACCE, which may act on a variety of cells. Increasing evidence suggests that AFB1‐induced liver injury is also regulated by multiple pathways or targets (e.g., HIPPO, PXR, NRF2 and NLRP3 inflammasome) rather than a single linear event [, , , ]. Our network pharmacology and transcriptomics analysis indicates that the PI3K‐AKT axis is a potential core pathway involved in AFB1 hepatotoxicity, which is consistent with recent network toxicology studies [, ]. This observation prompted us to select the PI3K‐AKT pathway as the core mechanism of ACCE's anti‐apoptotic effect for further investigation. Of course, other previously reported pathways, such as NRF2 and NLRP3, may also contribute to the hepatoprotection of ACCE, though this lies outside the scope of the present mechanistic investigation. Notably, the regulation of the PI3K‐AKT pathway by AFB1 appears to vary across studies. Although our data show that AFB1 alone reduces hepatic p‐AKT levels in mice, which is consistent with several previously reported findings, some investigations have instead observed activation of this signaling axis following AFB1 exposure [, , , ]. The discrepancies in these results may arise from various factors, such as tissue and cell type specificity, differences in AFB1 dosage, and exposure duration, which could lead to divergent regulatory effects of AFB1 on the PI3K‐AKT pathway. Nevertheless, our data consistently demonstrate that under conditions of AFB1 exposure, ACCE can activate this pathway, which remains a robust and functionally significant finding. In addition, butyrate is not limited to inhibiting TLR4/NF‐κB‐mediated inflammation; numerous previous studies have reported that butyrate also acts as a regulator of the PI3K‐AKT pathway. Butyrate attenuates cirrhotic portal hypertension in rats by suppressing HDAC3 and upregulating the PI3K/AKT/eNOS cascade within liver sinusoidal endothelial cells, as previously demonstrated []. A study on a diabetic mouse model explicitly indicated that the gut microbiota metabolite sodium butyrate (NaB) can positively influence bile acid metabolic homeostasis by activating the GPR43‐PI3K‐AKT‐GSK3 signaling pathway []. These findings corroborate the modulatory effect of butyrate on the PI3K‐AKT pathway. We thus propose that ACCE‐regulated, gut‐derived butyrate exerts hepatoprotection through a dual mechanism: it suppresses the TLR4/NF‐κB pathway to alleviate inflammation, and may also modulate the PI3K‐AKT pathway to promote cell survival. However, the more specific relationship of these actions still requires further investigation in the future. Of course, there are some limitations in the experimental design of this study, such as: (1) First, the network pharmacology predictions lack experimental validation regarding the binding affinity of ACCE's major active constituents to the core targets. Therefore, the computational results should only be regarded as a reference for mechanism exploration. (2) Second, the HepG2 cell model selected in this study exhibits lower expression and activity of CYP450 compared to normal human primary hepatocytes, which may not fully simulate the metabolic activation process of AFB1 in human hepatocytes. Importantly, our use of HepG2 cells aims to validate signaling pathways rather than to accurately reproduce the complete metabolism of AFB1, so caution should be exercised when extrapolating the absolute sensitivity of AFB1 toxicity to normal human hepatocytes. Future validations using models with stronger metabolic capabilities, such as HepaRG or primary hepatocytes, will enhance the translational relevance of our findings. (3) Furthermore, while our in vitro inhibitor experiments (LY294002 and TAK‐242) validated the involvement of PI3K‐AKT and TLR4/NF‐κB pathways in ACCE‐mediated protection against AFB1‐induced hepatocyte injury, additional in vivo blockade studies are warranted to establish their causal roles in the intact liver. Therefore, it is essential to validate our findings in future studies using inhibitors or conditional gene knockout mice exposed to AFB1. (4) Lastly, it is important to note that the AFB1 exposure regimen used in this study (0.75 mg/kg bw, intragastric administration for 14 days) represents a widely used subacute/subchronic liver injury model, which has been similarly applied in previous AFB1 toxicology studies []. In contrast, the daily exposure level of AFB1 through dietary intake for residents in different regions of China is approximately 0.0066 to 0.0448 μg/kg.bw []. Thus, the dosage used in this study was significantly higher than the daily exposure level in humans, which aids in inducing reproducible liver injury within a shorter timeframe and aligns with the commonly used framework for subacute/subchronic toxicity assessment. Nevertheless, we acknowledge that this model cannot fully replicate the true scenario of chronic low‐dose exposure in humans; thus, future studies should utilize lower and more environmentally relevant doses, as well as longer exposure regimens, to further validate the clinical translational potential of ACCE. In terms of the prospects and challenges of clinical translation, this study found that ACCE had a hepatoprotective effect against AFB1‐induced liver injury, suggesting its potential future as a feed additive or functional food for mitigating aflatoxin toxicity. However, several key issues need to be addressed before clinical translation. (1) The effective dose of ACCE in mice is 10 g of raw medicine/kg bw/day. Using the standard body surface area conversion formula, this dose translates to approximately 0.81 g/kg bw/day for humans (assuming a 60 kg adult), which amounts to about 48.6 g of raw medicine per day []. For livestock such as poultry or pigs, which are more relevant to AFB1 exposure, the corresponding dose can be determined based on allometric scaling of metabolic body weight. However, these estimates are preliminary and require direct validation in the target species. (2) ACCE is a traditional Chinese medicine compound aqueous extract with a long history of human use, indicating good safety []. In this study, no deaths, behavioral abnormalities, or weight loss were observed in mice treated with ACCE alone (10 g/kg bw/day) for 14 days, suggesting a reasonable safety window at the effective dose. However, chronic toxicity, genotoxicity, and reproductive toxicity studies are currently lacking, and these should be performed before human or livestock trials. (3) ACCE is prepared through a simple boiling method, making it cost‐effective, scalable, and compatible with feed production processes [, ]. Its protective effect against animal liver injury induced by AFB1, a major contaminant in feed, makes it a potential natural feed additive. Nevertheless, the stability of ACCE during feed storage and processing, as well as its impact on feed palatability and animal growth performance, still needs to be evaluated. In summary, while ACCE shows promise as a candidate for alleviating AFB1‐induced liver injury, extensive preclinical and translational research is still required before its application as a feed additive or functional food. Conclusion In summary, ACCE demonstrates hepatoprotective potential against AFB1‐induced liver injury in mice, as revealed by this study. By integrating multi‐omics data with functional experiments, it reveals a novel mechanism by which ACCE alleviates AFB1‐induced liver injury. On one hand, active components of ACCE directly act upon hepatocytes, activating the PI3K‐AKT pathway and mediating anti‐apoptotic and anti‐inflammatory responses. Concurrently, ACCE reshapes gut microbiota composition by enriching butyrate‐producing bacteria, thereby increasing circulating butyrate levels. This butyrate may serve as a pivotal signaling molecule in gut‐liver crosstalk, reaching the liver via systemic circulation and contributing to anti‐inflammatory activity via TLR4/NF‐κB pathway inhibition. Together, these direct (components‐PI3K‐AKT) and indirect (microbiota‐butyrate‐TLR4/NF‐κB) pathways synergistically establish a multi‐target, multi‐dimensional protective network by which ACCE counteracts AFB1‐induced liver injury (Figure ). This study offers experimental support for the alleviation of AFB1‐induced liver injury by ACCE, suggesting its potential value for further research. However, it should be emphasized that the long‐term safety, pharmacokinetic characteristics, and clinical translational potential of ACCE require validation in subsequent studies. Materials and Methods Materials and Reagents The crude herbs of compound Abrus cantoniensis are provided by Hunan Songlingtang TCM Co., Ltd. (Hunan, China), and the quality control complies with the standards of the Chinese Pharmacopoeia. The specific batch numbers and origins are as follows: Abrus cantoniensis (Lot No. 241101, Hunan, China), Artemisia capillaris (Lot No. 240801, Shaanxi, China), Salvia miltiorrhiza (Lot No. 241002, Hebei, China), Cynanchum otophyllum (Lot No. 240802, Anhui, China), Poriacocos (Lot No. 240901, Hunan, China), and Glycyrrhiza uralensis (Lot No. 240902, Inner Mongolia, China). Silymarin (Cat# S873807), Butyrate (Cat# B766935), Cryptotanshinone (Cat# C796209), and 6‐Gingerol (Cat# C796096) were obtained from Shanghai Macklin Biochemical Technology Co., Ltd. (Shanghai, China). Lipopolysaccharide (LPS, Cat# S1732) and Caffeate (Cat# SM1022) were provided by Jiangsu Biyuntian Biotechnology Co., Ltd. (Jiangsu, China). LY294002 (Cat# HY‐10108) and TAK‐242 (Cat# HY‐11109) were supplied by Shanghai MCE (Shanghai, China). Aflatoxin B1 (AFB1, ≥99.8%) was purchased from Chengdu Lingli Biotechnology Co., Ltd. (Chengdu, China). Biochemical assay kits for alanine aminotransferase (ALT, Cat# C009‐2‐1), aspartate aminotransferase (AST, Cat# C010‐2‐1), alkaline phosphatase (ALP, Cat# AKFA018M), lactate dehydrogenase (LDH, Cat# AKCO003M), malondialdehyde (MDA, Cat# A003‐1), glutathione peroxidase (GSH‐Px, Cat# A005‐1), and superoxide dismutase (SOD, Cat# A001‐3) were provided by Nanjing Jiancheng Bioengineering Institute (Nanjing, China) and Beijing Boxbio Science & Technology Co., Ltd. (Beijing, China). Jiangsu Meimian Industrial Co., Ltd. (Jiangsu, China) supplied the ELISA kits for tumor necrosis factor‐α (TNF‐α, Cat# MM‐0122H1), interleukin‐1β (IL‐1β, Cat# MM‐0181H1), and interleukin‐6 (IL‐6, Cat# MM‐0049H1). Antibodies were obtained from Wuhan Servicebio Technology Co., Ltd. and Wuhan Proteintech Biotechnology Co., Ltd. (Wuhan, China). Majorbio Bio‐Pharm Technology Co., Ltd. (Shanghai, China) performed the transcriptomic, microbiomic, and targeted metabolomic analyses. The UPLC‐MS/MS analysis of ACCE was provided by Wuhan Maiwei Metabolomics Biotechnology Co., Ltd. (Wuhan, China). Preparation of ACCE Extract The six medicinal herbs comprising compound Abrus cantoniensis were crushed and thoroughly mixed. Ten times the weight (w/w) of purified water was added, and the mixture was soaked for 30 min, boiled, then simmered gently for 1 h. The decoction was filtered. The residue was re‐extracted with eight times the weight of purified water by gentle simmering for 1 h, followed by filtration. The two filtrates were pooled, centrifuged, and evaporated under reduced pressure to a concentration of 1 g crude drug/mL, yielding the compound ACCE. The concentration was adjusted according to experimental dosage requirements, after which the extract was aliquoted and kept at −20°C, and freshly prepared every 7 days. Any process that might produce microbial fermentation has been minimized. A portion of ACCE was lyophilized into powder (extraction rate of approximately 14.44%) for subsequent cell experiments. UPLC‐MS/MS Analysis of ACCE ACCE was analyzed using UPLC‐MS/MS. After extraction of samples with 70% methanol (containing an internal standard) and filtration of the supernatant, the extracts were subjected to Agilent SB‐C18 column chromatography with a gradient of water/acetonitrile (both with 0.1% formic acid), followed by MS/MS detection in MRM mode on an ESI‐QTRAP system. Metabolites were identified based on an in‐house database and quantified by integrating peak areas. Network Pharmacology and Molecular Docking Analysis Based on UPLC‐MS/MS results, combined with the TCMSP database and previous literature, the potential active components of ACCE were identified. Putative targets for each active component were predicted using the SwissTargetPrediction database. Subsequently, disease targets associated with AFB1‐induced liver disease (AILD) were retrieved from GeneCards and OMIM using search terms including “AFB1”, “liver disease”, and “liver injury”. Overlaps between compound‐related and disease‐related targets were then identified, and a “component‐target” interaction network was constructed with Cytoscape 3.10.2. The overlapping targets were submitted to the STRING database to build a protein‐protein interaction (PPI) network, which was subsequently displayed using Cytoscape 3.10.2. CytoNCA plugin was applied to compute network topology parameters and screen hub targets. These candidates were then uploaded to DAVID for Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses. Results were ranked by p‐value and charted via the Weishengxin Online Platform (https://www.bioinformatics.com.cn/). A “drug‐component‐target‐pathway‐disease” network was constructed by integrating key targets and KEGG pathways. Prior to molecular docking, ligands and target proteins were preprocessed (e.g., removal of water and addition of hydrogens). AutoDock Vina was then used to perform docking simulations, and binding energies were visualized as heatmaps. The interaction modes at the optimal binding sites were displayed using PyMOL. Animals and Experimental Design Male KM mice (4 weeks old, SPF grade, 20 ± 2 g) were obtained from Changsha Tianqin Biotechnology Co., Ltd. (Changsha, China; License No.: SCXK (Xiang) 2019‐0014). The experimental protocol was reviewed and approved by the Animal Ethics Committee of Guangxi University (Approval No.: Gxu‐2025‐231) and followed the National Research Council's Guide for the Care and Use of Laboratory Animals. All mice were allowed one week to acclimatize to the standard laboratory environment before the start of the experiment. The mice were randomly allocated to seven groups (n = 11) and treated daily as follows: the normal control (NC) and AFB1 model groups (AFB1) received purified water by gavage; the vehicle control group (DMSO) received purified water by gavage; the low‐dose and high‐dose ACCE groups (AFB1+ACCE‐L, AFB1+ACCE‐H) received ACCE by gavage at doses of 5 g crude drug/kg body weight (bw) and 10 g crude drug/kg bw, respectively; the positive control group (Silymarin) received silymarin at 100 mg/kg bw by gavage (It has been shown to have a protective effect against AFB1‐induced liver injury) [, ]; and the ACCE‐alone group (ACCE) received ACCE at 10 g crude drug/kg bw by gavage. All groups received a gavage volume of 0.1 mL/10 g bw. Two hours after administration, except for mice in the NC and ACCE groups (gavaged with an equivalent volume of purified water) and the DMSO group (gavaged with an equivalent volume of 2% (v/v) DMSO), all other groups were given AFB1 (0.75 mg/kg bw) by gavage, and the intervention lasted for 14 days to establish an AFB1‐induced, repeatable subacute/subchronic liver injury model (Figure ) [, ]. This dose and duration are widely used in toxicological studies. At the end of the treatment period, the mice were fasted overnight (12 h, water freely available) and weighed. Subsequently, the mice were anesthetized with pentobarbital sodium, and euthanasia was performed by cervical dislocation after blood collection. Liver, ileum, feces, and intestinal contents were promptly collected, snap‐frozen in liquid nitrogen, and stored at −80°C for subsequent analysis. Samples from ≥ 3 mice per group were randomly selected for detection and analysis based on data analysis requirements and objective experimental conditions (e.g., resource limitations). In the in vivo validation experiment, mice were randomized into four groups (n = 6) and treated daily as follows: the normal control group (NC) and the AFB1 model group (AFB1) received pure water via gavage; the butyrate group (AFB1+Butyrate) and cryptotanshinone group (AFB1+Cryptotanshinone) were gavaged with 80 mg/kg and 25 mg/kg of the respective agents, all at 0.1 mL/10 g bw. Two hours later, except for the NC group, which was administered the same volume of pure water, the other groups were given 0.75 mg/kg bw AFB1 via gavage, with the intervention lasting for 14 days. After the treatment, samples were collected, preserved, and analyzed according to the procedures described above. Histopathological Analysis Liver and ileum tissues were fixed in 4% paraformaldehyde (24 h, room temperature), then dehydrated, paraffin‐embedded, and sectioned (3–4 μm). Sections were stained with hematoxylin and eosin (H&E), and images of target areas were captured using an Eclipse Ci‐L microscope (Nikon, Japan). Villus length and crypt depth in the ileum were measured with Image‐Pro Plus 6.0 software. Biochemical Assays and Enzyme‐Linked Immunosorbent Assay (ELISA) Biochemical assays were performed to measure serum liver function biomarkers (ALT, AST, ALP, LDH) and hepatic oxidative stress biomarkers (MDA, GSH‐Px, SOD) in tissue homogenates, following the respective manufacturers’ instructions. Inflammatory cytokines (TNF‐α, IL‐1β, IL‐6) in both serum and liver homogenates were quantified by ELISA. RNA Sequencing (RNA‐seq) Analysis Total RNA was isolated from mouse liver tissues using TRIzol reagent (Invitrogen, USA) according to the manufacturer's protocol. The concentration and purity were assessed on a NanoDrop 2000 (Thermo Fisher, USA), and RNA integrity was evaluated by agarose gel electrophoresis, with RQN values determined on an Agilent 5300. Subsequently, mRNA was enriched with Oligo (dT) magnetic beads, then fragmented, reverse‐transcribed into cDNA, ligated to adapters, and PCR‐amplified to generate the library. The quality‐controlled library was sequenced on an Illumina NovaSeq platform (PE150). Raw data were filtered to remove adapter sequences and low‐quality reads using fastp (v0.20.0). Clean reads were aligned to the mouse reference genome (GRCm39) using HISAT2 (v2.1.0). Illumina RNA‐seq and subsequent bioinformatics analyses were performed by Majorbio Bio‐Pharm Technology Co., Ltd. (Shanghai, China). Differentially expressed genes (DEGs) were identified using thresholds of false discovery rate (FDR) < 0.05 and |log2 (fold change)| ≥ 1 for subsequent functional analyses. 16S rRNA Gene Sequencing Total genomic DNA was extracted from colon contents using the E.Z.N.A. Soil DNA Kit (Omega Bio‐tek, Norcross, GA, USA). DNA quality, concentration, and purity were assessed by 1% agarose gel electrophoresis and a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). Using the extracted DNA as a template, the bacterial 16S rRNA V3‐V4 region was amplified by PCR (20 μL reaction volume) with universal primers 338F (5’‐ACTCCTACGGGAGGCAGCAG‐3’) and 806R (5’‐GGACTACHVGGGTWTCTAAT‐3’) []. Amplified products were separated by 2% agarose gel electrophoresis, purified using a DNA Gel Recovery Purification Kit (Yuhua, China), and quantified using a Qubit 4.0 fluorometer (Thermo Fisher Scientific, USA). Libraries were constructed using the NEXTFLEX Rapid DNA‐Seq Kit and sequenced on an Illumina NextSeq 2000 platform (Majorbio Bio‐Pharm Technology Co., Ltd., Shanghai, China). Raw sequencing data were submitted to the NCBI SRA database. Following quality control and sequence assembly, sequencing data were denoised with the DADA2 plugin within the Qiime2 pipeline to generate amplicon sequence variants (ASVs) []. Subsequent α‐diversity and β‐diversity analyses were performed. Short‐Chain Fatty Acid Quantification Samples (approximately 20 mg) were extracted using an extraction solution at low temperature. After centrifugation, the supernatant was retained, and metabolites were further extracted with n‐butanol. Standard solutions with various concentrations were prepared and analyzed by GC‐MS under identical conditions. The concentration of target compounds was calculated based on standard curves and converted to actual content in samples. Western Blot Analysis Extraction of total proteins from frozen liver tissues was accomplished with RIPA buffer, while nuclear proteins were obtained using a nuclear extraction kit. Quantification of protein samples was carried out via the BCA method. SDS‐PAGE separation and PVDF membrane (Bio‐Rad) transfer were followed by immunodetection with specific antibodies. Prior to antibody incubation, membranes were blocked, then exposed to primary antibodies and subsequently to matching secondary antibodies. Visualization of target proteins was achieved by standard chemiluminescence detection, and band intensities were quantified with ImageJ software. Detailed antibody information is as follows: Anti‐p‐EGFR (Tyr1110, Bioss, Cat# bs‐5316R, 1:2000), Anti‐EGFR (Servicebio, Cat# GB111504, 1:1000), Anti‐p‐AKT (Ser473, Proteintech, Cat# 66444‐1, 1:5000), Anti‐AKT (Servicebio, Cat# GB15689, 1:1000), Anti‐Bcl‐2 (Servicebio, Cat# GB154830, 1:1000), Anti‐TLR4 (Servicebio, Cat# GB15186, 1:500), Anti‐p‐IκB (Ser32, Servicebio, Cat# GB15212, 1:1000), Anti‐IκB (Servicebio, Cat# GB151509, 1:1000), Anti‐MyD88 (Servicebio, Cat# GB111554, 1:1000), Anti‐p‐NF‐κB p65 (Ser536, Proteintech, Cat# 80379‐2, 1:5000), Anti‐β‐Actin (Servicebio, Cat# GB15001, 1:3000), Anti‐β‐Tubulin (Servicebio, Cat# GB15140, 1:1000), HRP‐labeled goat anti‐mouse IgG (Servicebio, Cat# GB23301, 1:10000), HRP‐labeled goat anti‐rabbit IgG (Servicebio, Cat# GB23303, 1:10000). Immunofluorescence Analysis Paraffin‐embedded liver tissues (sectioned at 5 μm thickness) or HepG2 cells cultured on coverslips were first permeabilized with 0.2% Triton X‐100 and blocked with 5% bovine serum albumin (BSA), then incubated overnight at 4 °C with primary antibodies. Subsequent incubation with fluorescent secondary antibodies was carried out for 1 h at room temperature in the dark, and DAPI was used for nuclear counterstaining. Images were captured using a fluorescence microscope and analyzed quantitatively with Saiviewer software. The relevant antibody information is, primary antibody: Anti‐p‐EGFR (Tyr1110, Bioss, Cat# bs‐5316R, 1:400), Anti‐p‐AKT (Ser473, Proteintech, Cat# 66444‐1, 1:600), Anti‐p‐NF‐κB p65 (Ser536, Proteintech, Cat# 80379‐2, 1:600), Anti‐NF‐κB p65 (Proteintech, Cat# 80979‐1, 1:600), 1:500), Anti‐TLR4 (Proteintech, Cat# 66350–1, 1:500); Secondary antibody: Alexa Fluor 488‐labeled goat anti‐mouse IgG (Servicebio, Cat# GB25301, 1:400) and CY3‐labeled goat anti‐mouse IgG (Servicebio, Cat# GB21301, 1:400) 1:300), Alexa Fluor 488‐labeled goat anti‐rabbit IgG (Servicebio, Cat# GB25303, 1:400), and CY3‐labeled goat anti‐rabbit IgG (Servicebio, Cat# GB21303, 1:300). Cell Culture HepG2 cells, due to their clear genetic background, rapid growth, lack of differentiation requirements, and ease of handling, are commonly used as an in vitro model for AFB1 hepatotoxicity studies, facilitating labor‐intensive cellular experiments [, , , , , , , ]. Seeding of cells into 96‐well plates was carried out at 5 × 104 cells/well, followed by culture in DMEM (10% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin) at 37°C in a 5% CO2 incubator. AFB1, caffeate, cryptotanshinone, and 6‐gingerol were dissolved in DMSO, with the final solvent concentration kept below 0.2% (v/v) in all treatment groups. Cells were allowed to acclimate for 24 h before treatment and subsequently exposed to AFB1, LY294002, TAK‐242, ACCE, butyrate, caffeate, cryptotanshinone, or 6‐gingerol as specified. Cell viability, inflammatory cytokines, liver function markers, apoptosis, and protein expression were respectively determined by CCK‐8 assay, ELISA/biochemical kits, flow cytometry, and Western blotting following the intervention. All cell experiments were conducted with at least three independent biological replicates, each containing a minimum of three technical replicates. The average of the technical replicates was taken as one data point per biological replicate for subsequent statistical analysis. Flow Cytometry Analysis Evaluation of apoptosis was performed using the Annexin V‐FITC/PI kit (Servicebio, Cat# G1511). Following culture, HepG2 cells were collected, washed twice with pre‐cooled PBS, followed by adjusting the cell concentration to 1 × 106 cells/mL using pre‐cooled 1× Binding Buffer. The cell suspension (100 μL) was incubated with 5 μL each of Annexin V‐FITC and PI for 10 min at room temperature in the dark. Stained cells were analyzed on a flow cytometer (Attune Nxt) within 1 h, acquiring 20,000 events per sample. For gating, debris were first excluded by FSC‐A/SSC‐A selection, and quadrants were then defined. Apoptotic cells were classified as Annexin V+/PI− (early) or Annexin V+/PI+ (late), with data processed using FlowJo v10. Statistical Analysis Presentation of data as mean ± SD from ≥3 independent experiments was adopted. Verification of normality and homogeneity of variances was accomplished via Shapiro‐Wilk and Levene's tests, respectively. One‐way ANOVA with Tukey's post hoc correction was applied to datasets fulfilling both assumptions (p > 0.05), whereas Welch's ANOVA with Games‐Howell adjustment was employed for datasets with unequal variances. Importantly, both analytical strategies yielded concordant conclusions for all principal comparisons. SPSS 21.0 (SPSS Inc., Chicago, IL, USA) and GraphPad Prism 10 (GraphPad Software, San Diego, CA, USA) were used for statistical processing, with the threshold for significance fixed at p < 0.05.