https://orcid.org/0000-0001-8027-9571
Figures
Abstract
Isatis indigotica (Banlangen) is a classic Traditional Chinese Medicine herbal remedy with well-documented antiviral and anti-inflammatory properties. However, the molecular mechanisms underlying its therapeutic effects against hepatitis B virus (HBV)-associated hepatic inflammation remain incompletely understood. This study aimed to systematically elucidate the multi-target regulatory mechanisms of Isatis indigotica against HBV-associated hepatic inflammation using network pharmacology and molecular simulation approaches. Bioactive compounds were screened from the TCMSP database (OB ≥ 30%, oral bioavailability; DL ≥ 0.18, drug-likeness). Candidate targets were identified by integrating SwissTargetPrediction with GeneCards/OMIM disease targets. Protein-protein interaction (PPI) network topology analysis identified hub genes. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed using Metascape. Molecular docking and 100 ns all-atom molecular dynamics (MD) simulation were conducted to validate compound-target binding. Seventeen bioactive compounds were identified, yielding 1,023 compound targets. Intersecting with 1,676 HBV disease targets produced 142 candidate genes. PPI network analysis identified AKT1, IL6, TP53, and TNF as hub genes, significantly enriched in NF-κB, JAK-STAT, and TNF signaling pathways (P < 0.01). Molecular docking confirmed favorable binding affinities, with IQ (6-(3-oxoindolin-2-ylidene)indolo[2,1-b]quinazolin-12-one) showing optimal binding to AKT1 (ΔG = −9.35 kcal/mol). MD simulation verified stable binding over 100 ns. This network pharmacology study suggests that Isatis indigotica acts on HBV-associated hepatic inflammation through multi-target synergistic regulation of NF-κB and JAK-STAT signaling pathways, providing mechanistic insights and potential therapeutic targets for HBV management.
Citation: Li X, Wang Y, Cheng H, Chen R, Chen X (2026) Network pharmacology and molecular simulation study on the multi-target gene regulatory mechanism of Isatis indigotica against hepatic inflammation. PLoS One 21(8): e0355930. https://doi.org/10.1371/journal.pone.0355930
Editor: Junzheng Yang, Guangdong Nephrotic Drug Engineering Technology Research Center, Institute of Consun Co. for Chinese Medicine in Kidney Diseases, CHINA
Received: May 20, 2026; Accepted: July 27, 2026; Published: August 13, 2026
Copyright: © 2026 Li et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This work was supported by the Xinjiang Autonomous Region Science and Technology Aid Program (2024E02010) and the Rural Science and Technology Entrepreneurship Action of Science and Technology Commissioners in Xinjiang Uygur Autonomous Region (2025KZ014). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Hepatic inflammation serves as a central pathological mechanism in the progression of various liver diseases, with its molecular basis involving abnormal activation of inflammation-related genes and disruption of regulatory networks [1–3]. In chronic hepatitis B, persistent hepatitis B virus (HBV) infection can induce host immune response disorders, driving aberrant expression of key inflammatory regulatory genes such as AKT1 and TNF, thereby establishing a chronic inflammatory microenvironment that significantly increases the risk of hepatic fibrosis, cirrhosis, and hepatocellular carcinoma [4,5]. Although nucleos(t)ide analogs and pegylated interferon can effectively suppress viral replication, their efficacy in reversing established inflammatory damage and immune exhaustion remains limited [6,7]. Therefore, targeted synergistic regulation of multiple inflammatory gene nodes has emerged as a critical strategy for developing novel anti-hepatic inflammation approaches.
Isatis indigotica Fort. (Banlangen), a classic antipyretic and detoxifying herbal medicine in Traditional Chinese Medicine (TCM), has been demonstrated by modern research to possess significant antiviral, anti-inflammatory, and immunomodulatory activities [8,9]. However, due to the chemical complexity of Isatis indigotica, the traditional “single-component–single-target” research paradigm inadequately explains its synergistic regulatory mechanisms on inflammatory gene networks, thereby constraining deeper mechanistic understanding and clinical translation.
Network pharmacology, by constructing multi-dimensional regulatory networks of “component–target gene–pathway,” aligns with the holistic perspective and systemic regulatory characteristics of TCM, providing a novel strategy for deciphering gene-level mechanisms of complex herbal systems [10]. This integrated strategy has been successfully applied to the study of Food and medicine homology substances. For instance, Liu et al. employed network pharmacology to identify key targets of mulberry leaf extract in metabolic dysfunction-associated fatty liver disease (MAFLD), and then validated their predictions via animal experiments [11]. While several pioneers have mapped the basic network pharmacology landscape of Isatis indigotica against HBV, their findings were limited to static network topologies and generic inflammatory hubs [12]. Such topology-driven approaches are intrinsically biased toward highly interconnected nodes, often yielding recurrent generic signaling hubs across disparate disease models rather than context-specific therapeutic targets [13]. To overcome the limitations of static network inference, molecular docking and molecular dynamics (MD) simulations have emerged as gold-standard computational tools for validating ligand–target binding modes and conformational stability under physiological conditions [14]. Integrating network pharmacology with molecular docking and MD simulations constitutes a powerful in silico pipeline that bridges systems-level target prediction with atomic-level structural evidence [15]. The present work uniquely advances this field by combining network pharmacology with molecular docking and MD simulation, which enables validation of binding modes and stability between bioactive components and target gene products at the protein structural level, offering structural biological evidence for network predictions. Although several network pharmacology studies on Isatis indigotica have reported similar inflammatory pathways, the present study provides three distinct advances: (1) exclusive application of stringent ‘highest confidence’ (score >0.9) PPI filtering to reduce false positives; (2) 100 ns all-atom MD simulation to validate binding stability, which is rarely performed in previous studies; and (3) identification of IQ as a potential AKT1 stabilizer, a hypothesis not previously proposed.
Based on this rationale, the present study integrated network pharmacology-based target gene prediction, molecular docking validation, and MD simulation to systematically identify core bioactive components and key target genes of Isatis indigotica against HBV-associated hepatic inflammation, elucidating its multi-target synergistic anti-inflammatory mechanisms from a gene regulatory network perspective and providing a mechanistic foundation for modern research on Isatis indigotica and anti-hepatitis drug development.
2. Materials and methods
Ethics statement: This study is a purely computational analysis involving no human participants, animal subjects, or biological samples. All data were retrieved from publicly available databases and published literature. Therefore, no ethics committee approval was required.
2.1. Bioactive compound identification and target prediction
Using “Isatis indigotica” as the search keyword, chemical constituents were retrieved from the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP, https://www.tcmsp-e.com/). Screening thresholds were set at OB (oral bioavailability) ≥30% and DL (drug-likeness) ≥0.18, following established criteria for candidate bioactive compound selection. After removing duplicates, candidate bioactive compounds were obtained. SMILES structural formulas were retrieved from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/) and imported into the SwissTargetPrediction server (http://www.swisstargetprediction.ch/) with the organism set to “Homo sapiens.” Targets with prediction probability scores > 0 were selected as potential targets of the respective compounds.
2.2. Identification of HBV-associated disease targets
Using “hepatitis B virus” as the search keyword, disease-associated targets were retrieved from the GeneCards database (https://www.genecards.org/) and the OMIM database (https://www.omim.org/). The union of results from both databases was obtained. For GeneCards, targets with relevance scores above the median value were retained to ensure target quality.
2.3. Identification of intersection targets
The Venny 2.1.0 online tool (https://bioinfogp.cnb.csic.es/tools/venny/) was used to identify the intersection between Isatis indigotica compound targets and HBV disease targets. The resulting intersection targets represent potential therapeutic targets of Isatis indigotica in the context of HBV-associated hepatic inflammation.
2.4. Construction and analysis of compound–target network
Data on Isatis indigotica bioactive compounds and the 142 intersection targets were imported into Cytoscape software (version 3.10.3; https://cytoscape.org/) to construct a visualized compound–target–disease network. Network topology parameters (degree, betweenness centrality, closeness centrality) were calculated using the built-in Network Analyzer plug-in. Core bioactive compounds and key targets were ranked by degree value, with higher values indicating more critical regulatory roles in the network.
2.5. Construction and analysis of protein–protein interaction network
Intersection targets were imported into the STRING database (https://cn.string-db.org/) with the species set to “Homo sapiens” and the minimum interaction confidence threshold set to “highest confidence” (score > 0.9). Isolated nodes were removed prior to protein–protein interaction (PPI) analysis. The resulting network was imported into Cytoscape 3.10.3. Using the CytoNCA plug-in, nodes were ranked by degree centrality to identify core hub targets.
2.6. GO and KEGG pathway enrichment analysis
GO functional enrichment analysis and KEGG pathway enrichment analysis of the 142 intersection targets were performed using the Metascape platform (https://metascape.org/). A significance threshold of P < 0.01 was applied to identify significantly enriched biological processes, cellular components, molecular functions, and signaling pathways. Enrichment results were visualized using the online Bioinformatics platform (https://www.bioinformatics.com.cn/).
2.7. Molecular docking
Three-dimensional crystal structures of core target proteins were downloaded from the Protein Data Bank (https://www.rcsb.org/) and preprocessed using PyMOL for water molecule removal and co-crystallized ligand extraction. Mol2-format structures of major bioactive compounds were obtained from the TCMSP database. AutoDockTools (version 1.5.6) was used for receptor and ligand preparation, including addition of polar hydrogen atoms and Gasteiger charge calculation. Molecular docking was performed using AutoDock Vina with a grid box encompassing the active site of each target protein. A binding free energy ≤−5.0 kcal/mol was considered indicative of favorable binding activity. Docking poses were visualized using PyMOL and Discovery Studio Visualizer.
2.8. Molecular dynamics simulation
The protein–ligand complex with the optimal molecular docking binding affinity was selected for a 100 ns all-atom MD simulation to validate the long-term binding stability between the compound and the target protein. Simulations were performed using the GROMACS software package (version 2021.4). The protein was parameterized using the AMBER99SB-ILDN all-atom force field. Ligand topology parameters were generated using Sobtop (version 1.0 dev5). The system was placed in a dodecahedral periodic boundary condition box with a minimum distance of 1.2 nm between solute atoms and box walls. The system was solvated using the TIP3P explicit water model and neutralized with Na + /Cl⁻ counter-ions. Energy minimization was performed using the steepest descent algorithm until the maximum force converged below 1000 kJ mol ⁻ ¹ nm ⁻ ¹. The system was subsequently equilibrated under NVT (300 K, 100 ps, leap-frog integrator) and NPT (1 bar, 100 ps, Berendsen pressure coupling) ensembles. Production MD simulation was conducted for 100 ns with an integration time step of 1 fs, and trajectory frames were saved every 10 ps. Long-range electrostatic interactions were treated using the Particle Mesh Ewald method. Trajectory post-processing was performed using gmx trjconv with the -pbc nojump flag to remove periodic boundary condition artifacts. Root-mean-square deviation (RMSD), root-mean-square fluctuation (RMSF), radius of gyration (Rg), hydrogen bond analysis, solvent-accessible surface area (SASA), and potential energy were analyzed using standard GROMACS analysis tools. Free energy landscape (FEL) analysis was performed using the gmx sham script with Rg and RMSD as reaction coordinates. Results were visualized using QtGrace (version 5.0.3).
3. Results
3.1. Bioactive compound identification and target prediction
Based on the screening criteria of OB ≥ 30% and DL ≥ 0.18, a total of 17 candidate bioactive compounds were identified through TCMSP and PubChem database searches (Table 1). Target information for each bioactive compound was retrieved from the SwissTargetPrediction server. After screening, combining, and removing duplicates, a total of 1,023 potential compound targets were obtained.
3.2. HBV-associated disease targets and intersection targets
A total of 1,676 HBV-related targets were retrieved and filtered from the GeneCards and OMIM databases. Venn diagram analysis using Venny 2.1.0 identified 142 intersection targets shared between 1,023 Isatis indigotica compound targets and 1,676 HBV disease targets (Fig 1; full list in S1 Table).
The intersection region contains 142 shared targets representing candidate therapeutic targets for Isatis indigotica against HBV-associated hepatic inflammation.
3.3. Compound–target network analysis
The 17 bioactive compounds of Isatis indigotica and their 142 intersection targets with HBV were imported into Cytoscape 3.10.3 to construct a visualized “compound–intersection target–HBV” network. This network contained 159 nodes (17 bioactive compounds and 142 core targets) and 526 interaction relationships (Fig 2). Green nodes represent bioactive compounds of Isatis indigotica, and blue nodes represent intersection targets; larger node size and deeper color indicate higher degree values and greater regulatory importance. Network topology analysis revealed that all bioactive compounds exhibit “one-to-many” target interaction characteristics, consistent with the polypharmacological synergistic regulatory mechanisms characteristic of TCM. Based on degree value ranking, acacetin, dinatin, sinensetin, eupatilin, and (IQ)6-(3-oxoindolin-2-ylidene)indolo[2,1-b]quinazolin-12-one were identified as the compounds most likely to exert major regulatory roles in HBV prevention and treatment (top 10 compounds listed in S2 Table). The complete compound-target network data are provided in S3 Table.
Green nodes represent bioactive compounds; blue nodes represent intersection targets. Node size and color depth are proportional to degree value.
3.4. Protein–protein interaction network analysis
The 142 intersection targets were imported into the STRING database for PPI analysis. The resulting network, optimized in Cytoscape 3.10.3 using the CytoNCA plug-in, comprised 142 nodes and 2,534 edges (Fig 3). Topology analysis demonstrated that nodes with larger degree values occupied more central positions in the network. The top 10 intersection targets ranked by degree value are presented in Table 2; the full ranking is available in S4 Table. The complete filtered PPI interaction dataset is provided in S5 Table. These hub nodes exhibit extremely high connectivity in the network and serve as key signaling nodes, suggesting that Isatis indigotica may modulate hepatic inflammatory processes by regulating these hub proteins.
Node size represents degree centrality; color intensity represents betweenness centrality. The top four hub genes (AKT1, IL6, TP53, TNF) are highlighted.
Among the top-ranked targets, AKT1, IL6, TP53, and TNF were the four core nodes with the highest degree connectivity in the PPI network. AKT1, a member of the serine/threonine kinase family, is a key signaling molecule regulating cell metabolism, survival, proliferation, and protein synthesis. TNF and IL6 are core cytokine genes driving inflammatory cascade reactions. TP53, the “guardian of the genome,” not only regulates cell cycle arrest and apoptosis but also participates in inflammatory response regulation by modulating inflammation-related gene expression. The high degree connectivity of these four genes in the PPI network suggests that Isatis indigotica may synergistically regulate these hub genes to intervene in downstream inflammatory gene expression programs.
3.5. GO and KEGG pathway enrichment analysis
GO enrichment analysis of the 142 intersection target genes yielded 5,267 biological process terms, 441 cellular component terms, and 864 molecular function terms. The top 20 most significant terms (sorted by ascending P-value) are shown in Fig 4.
The top 20 terms for biological processes, cellular components, and molecular functions are shown, sorted by ascending P-value. Bar length reflects the number of enriched genes.
Biological processes were predominantly enriched in protein phosphorylation, positive response to external stimuli, cellular response to nitrogen compounds, cell activation, and positive regulation of protein localization, suggesting that Isatis indigotica may modulate the activation status of inflammatory genes through post-translational modifications and cellular stress responses. Molecular functions were significantly enriched in protein kinase activity, protein tyrosine kinase activity, kinase binding, and protein domain-specific binding, indicating that Isatis indigotica compounds may regulate inflammatory gene transcription by intervening in kinase-mediated phosphorylation of downstream transcription factors.
KEGG pathway analysis identified 270 enriched pathways in total. With P < 0.01 as the significance threshold, the top 20 pathways are shown in Fig 5. Pathways closely related to the anti-inflammatory mechanism of Isatis indigotica include the NF-κB signaling pathway (the core pathway regulating pro-inflammatory gene transcription), the JAK-STAT signaling pathway (the primary cytokine receptor downstream gene activation pathway), the TNF signaling pathway (governing apoptosis and necroptosis), the lipid and atherosclerosis pathway, and the Th17 cell differentiation pathway. Notably, AKT1, IL6, TP53, and TNF occupy key node positions in both NF-κB and JAK-STAT pathways. Multi-component simultaneous action on these shared pathway nodes may block inflammatory signal network cascade amplification through a “multi-gene synergistic inhibition” strategy. Complete GO and KEGG enrichment results are provided in S6 Table.
The top 20 enriched pathways (P < 0.01) are displayed. Color gradient from dark to light represents P-value from low to high (i.e., significance from high to low).
3.6. Molecular docking results
Molecular docking was performed for the five core bioactive compounds against the four core target proteins (Fig 6). Binding energies of all compound–target pairs were below −5.0 kcal/mol (Table 3), indicating spontaneous formation of stable binding conformations. Among them, IQ showed the lowest binding energy with AKT1 (−9.35 kcal/mol), suggesting the strongest molecular recognition and binding affinity, which provides a structural basis for its regulatory effect on AKT1 gene function.
(A) Eupatilin–AKT1 complex showing five binding interactions at GLY-84, GLN-85, LYS-88, ALA-112, and TYR-113; (B) Dinatin–TNF complex showing four interactions at ASP-56, LEU-58, GLY-160, and ILE-247; (C) IQ–AKT1 complex showing three interactions at VAL-86, LYS-88, and ALA-112; (D) Sinensetin–AKT1 complex showing two interactions at GLN-85 and LYS-88; (E) Acacetin–TNF complex showing one interaction at ARG-62.
3.7. Molecular dynamics simulation analysis
Based on molecular docking results, the IQ–AKT1 complex with optimal binding affinity was selected for a 100 ns MD simulation to further validate the long-term stability of the protein–ligand complex.
RMSD analysis showed that ligand RMSD fluctuations remained below 0.04 nm throughout the entire 100 ns simulation, indicating overall conformational stability. The protein backbone RMSD converged at approximately 5 ns, and the system entered a stable single trajectory cluster after 40 ns, indicating that thermodynamic equilibrium was reached and that the IQ–AKT1 complex maintained a stable conformation (Fig 7A).
(A) RMSD of protein backbone (black) and ligand (red); (B) RMSF per residue; (C) number of hydrogen bonds between protein and ligand; (D) Rg of the complex; (E) potential energy contributions (Coulomb and Lennard-Jones); (F) SASA; (G) 2D free energy landscape; (H) 3D free energy landscape. Rg and RMSD were used as reaction coordinates for FEL analysis.
RMSF analysis showed that per-residue atomic displacement fluctuations were predominantly in the range of 0.05–0.2 nm, with overall low protein backbone flexibility and relatively stable side chain movements, favorable for maintaining a stable ligand conformation in the binding pocket (Fig 7B).
Hydrogen bond analysis revealed stable hydrogen bond interactions between the protein and small molecule throughout the simulation, providing key non-covalent binding forces for complex stabilization and supporting the persistence of IQ–AKT1 binding (Fig 7C).
Radius of gyration (Rg) analysis showed that the average Rg of the complex was 1.4 nm, reaching equilibrium rapidly after 5 ns, indicating that the complex maintained good structural compactness and stability with moderate conformational flexibility (Fig 7D).
Potential energy analysis showed that the short-range Coulomb interaction energy was −5.20 kJ/mol and the short-range Lennard-Jones potential energy was −78.27 kJ/mol. Both negative values confirm the presence of strong electrostatic attraction and hydrophobic interactions in the complex, supporting binding stability from an energetic perspective (Fig 7E).
SASA analysis showed that the average SASA of the complex was approximately 74 nm², with only small fluctuations (70–78 nm2) throughout the simulation, indicating that the protein–ligand interface remained stable without dissociation or significant conformational changes (Fig 7F).
Free energy landscape (FEL) analysis, using Rg and RMSD as reaction coordinates, showed that the lowest free energy basin was located at Rg = 1.40 nm and RMSD = 0.508 nm (ΔG < 2.5 kcal/mol), presenting a well-defined funnel-shaped potential energy surface. This indicates that the IQ–AKT1 complex preferentially occupies low-energy conformational states during simulation, with multiple interconvertible metastable states, further validating the binding stability of the complex from an energy landscape perspective (Fig 7G and 7H).
4. Discussion
This study systematically identified the core bioactive compounds of Isatis indigotica against HBV-associated hepatic inflammation using an integrated network pharmacology approach. The five hub compounds—acacetin, dinatin, sinensetin, eupatilin, and IQ—exhibited high degree values in the compound–target network, indicating broad target engagement and polypharmacological potential. Acacetin has been reported to inhibit oxidative stress through the Keap1/Nrf2 pathway and to promote HSP90-mediated COX-2 degradation [16,17], mechanisms that are highly consistent with the well-established ability of flavonoids to enhance hepatic antioxidant defense via the Nrf2 pathway [18–20]. Dinatin, a 6-methoxyflavone, exhibited a high degree value (48) in the network, suggesting broad target engagement. The polymethoxyflavone eupatilin possesses lipid metabolism-improving and hepatoprotective properties [21]. Our findings suggest that Isatis indigotica may exert systemic hepatoprotective effects that potentially involve metabolic pathways beyond the liver; however, as the present study is based solely on computational predictions and lacks metabolomics or systematic omics data, this hypothesis should be interpreted with caution and requires further experimental validation. IQ, an indoloquinazoline alkaloid, showed the strongest binding affinity in molecular docking, representing a novel finding not previously reported in the context of Isatis indigotica.
Four core hub targets were identified through PPI network analysis: AKT1, IL6, TP53, and TNF. These results align with the study by Deng et al., in which network pharmacology was employed to investigate the effects of Isatis indigotica on HBV-associated hepatitis, with common hub targets such as AKT1, IL-6, TP53, and TNF being identified [12]. AKT1 is an important member of the serine/threonine kinase family, playing key roles in regulating cell metabolism, growth, survival, proliferation, and protein synthesis [22]. In inflammatory responses, AKT signaling regulates downstream effectors including mTOR, eNOS, FOXO1, and GSK3β, thereby balancing pro-inflammatory and anti-inflammatory responses and influencing the development of inflammatory diseases [23–26]. Aberrant AKT1 activation can promote NF-κB nuclear translocation, driving transcriptional upregulation of pro-inflammatory genes including IL6 and TNF. TNF-α is an endogenous cytokine produced by activated monocytes and macrophages [27] that serves as a key mediator of immune and inflammatory responses; serum TNF-α levels are positively correlated with hepatitis disease activity [28,29]. IL-6 is a core regulatory factor in the hepatic inflammation–fibrosis axis, driving inflammatory amplification and fibrotic deposition primarily through the JAK/STAT3 and NF-κB pathways during chronic liver injury [30–32]. TP53, as an important tumor suppressor, not only mediates cell cycle arrest and apoptosis [33] but also regulates inflammation-related gene expression [34–37]. In the pathological continuum of “hepatitis–hepatic fibrosis–cirrhosis–hepatocellular carcinoma,” persistent inflammation is the core driver of hepatocyte malignant transformation; TP53 and AKT1 together serve as critical gatekeepers at the interface of inflammation and oncogenesis.
The multi-component composition of Isatis indigotica may simultaneously act on these four shared hub nodes (AKT1, TNF, IL6, TP53), potentially synergistically suppressing the inflammatory gene network through the following mechanisms: (1) interrupting positive feedback amplification by inhibiting TNF-α and IL-6 protein activity; (2) suppressing transcriptional activation by stabilizing AKT1 structure or inhibiting its kinase activity, thereby reducing phosphorylation-dependent activation of NF-κB and STAT3 and decreasing transcription levels of downstream pro-inflammatory genes; and (3) restoring cell cycle regulation by facilitating TP53-mediated cell cycle arrest or orderly apoptosis of inflammatory-damaged hepatocytes to prevent malignant transformation. Consistently, an independent network pharmacology study on another medicine formula, XingQiChuShiYin (XQCSY), which is used for metabolic dysfunction-associated fatty liver disease (MAFLD) prevention, also identified AKT1, TNF, IL-6, and TP53 as key hub targets [38].
Beyond direct compound-target interactions, emerging evidence suggests that gut microbiota metabolic reprogramming acts as an early inducer of hepatic inflammatory lesions, and that microbiota-derived metabolites can directly activate hepatocellular inflammatory responses through TLR/NF-κB and JAK/STAT signaling pathways [39]. The hub genes identified in the present study—AKT1, TNF, and IL-6—are precisely the key effector nodes within this microbiota-metabolite-inflammatory signaling axis. This perspective extends the present work from a narrow “compound-target direct interaction” framework toward a broader “microbiota-host co-metabolism” paradigm, suggesting that Isatis indigotica may exert its hepatoprotective effects not solely through direct compound-target engagement on hepatocytes, but also through modulation of the gut microbiota-liver metabolic axis. Such a hypothesis provides a rational foundation for future investigations into the anti-inflammatory mechanisms of Isatis indigotica from a gut microecological perspective.
GO enrichment analysis further supported the above mechanisms at the gene function level. Significant enrichment of protein phosphorylation and protein kinase activity terms indicates that Isatis indigotica primarily intervenes in inflammatory signal transduction at the level of post-translational modifications. Enrichment of cell activation and cytokine-mediated signaling terms reflects regulatory effects on immune cell function-related genes. Significant KEGG enrichment of JAK-STAT and TNF pathways corroborates the pathological mechanism of abnormal STAT3 phosphorylation and upregulated pro-inflammatory factor expression following HBV infection [40,41]. Enrichment of the lipid and atherosclerosis pathway suggests crosstalk between hepatic inflammation and lipid metabolism dysregulation, as chronic HBV infection is frequently accompanied by abnormal lipid metabolism that further aggravates hepatic inflammation and fibrotic progression [42].
The IQ–AKT1 complex exhibited the strongest binding affinity in molecular docking (ΔG = −9.35 kcal/mol), and molecular dynamics simulation further validated its binding stability: the protein backbone RMSD converged rapidly, the binding pocket region maintained favorable conformational rigidity, and persistent hydrogen bonds were observed between the protein and ligand throughout the simulation. The free energy landscape displayed a characteristic funnel-shaped potential energy surface, indicating that the complex preferentially occupies low-energy conformational states and that significant dissociation is unlikely. Both the short-range Coulomb energy and Lennard-Jones potential energy were negative, suggesting that electrostatic attraction and hydrophobic embedding jointly constitute the physicochemical basis for the high-affinity IQ–AKT1 binding. IQ may thus stably occupy the AKT1 active pocket through a dual-mode mechanism involving hydrogen bond anchoring and hydrophobic residue embedding, thereby attenuating the upstream driving effect of AKT1 on NF-κB and JAK-STAT pathways.
Notably, IQ is a naturally occurring indoloquinazoline alkaloid (tryptanthrin) found in Isatis indigotica and Indigo naturalis. Previous proteomics-based studies have demonstrated that tryptanthrin can concurrently regulate TLR4/NF-κB and JAK/STAT3 signaling pathways [43]. Furthermore, Niu et al. observed significant tryptanthrin-mediated suppression of NF-κB/MAPK activation and reduction of TNF-α and IL-6 levels in both in vitro and in vivo pulmonary fibrosis models, findings that align remarkably well with the hub genes and core pathways identified in the present network pharmacology analysis [44]. The present study provides the first structural dynamics evidence at the atomic level for IQ targeting the upstream kinase AKT1 through molecular docking and MD simulation.
Food and medicine homology components have been shown to act through dual pathways, direct anti-inflammatory activity following gastrointestinal absorption and indirect metabolic regulation via gut microbiota remodeling, which are highly compatible with the oral pharmacokinetic properties of IQ and its multi-target regulatory roles on NF-κB and JAK-STAT cascades [45]. Therefore, future research should not only verify the direct IQ–AKT1 binding and its kinase regulatory mechanism, but also investigate whether the active components of Isatis indigotica may indirectly potentiate their systemic anti-inflammatory effects through modulation of the gut microbiota–liver metabolic axis.
Several limitations of this study should be acknowledged. First, the findings are based entirely on computational predictions and require experimental validation, including in vitro cell-based assays and in vivo animal studies, to confirm the proposed mechanisms. Second, the TCMSP screening thresholds (OB ≥ 30%, DL ≥ 0.18), while widely used, may not capture all pharmacologically relevant compounds of Isatis indigotica. Third, network pharmacology analyses rely on the completeness and accuracy of publicly available databases, which are subject to ongoing updates. Fourth, the absence of positive reference inhibitors and negative control ligands in the molecular docking experiments limits the quantitative benchmarking of binding specificity; future studies should include appropriate controls to validate the relative binding affinity of IQ to AKT1. Future studies should prioritize experimental validation of the predicted IQ–AKT1 interaction and the multi-target inhibitory mechanism of Isatis indigotica active compounds in relevant cell and animal models of HBV-associated hepatic inflammation.
5. Conclusion
This study systematically elucidated the bioactive compounds and target gene regulatory mechanisms of Isatis indigotica against hepatic inflammation using an integrated network pharmacology, molecular docking, and molecular dynamics simulation approach. The results suggest that Isatis indigotica may regulate hub hepatic inflammation genes—including AKT1, TNF, IL6, and TP53—through core active components such as acacetin, dinatin, sinensetin, eupatilin, and IQ, by intervening in the NF-κB, JAK-STAT, and TNF inflammatory signaling pathways. Molecular simulation supported stable binding between IQ and AKT1 from a structural dynamics perspective, providing a molecular basis for the gene regulatory functions of these compounds. Given the medicinal and edible homology characteristics of Isatis indigotica, future studies can further explore its application as a functional food ingredient for primary prevention in high-risk hepatic inflammation populations. Particularly, IQ, as a candidate active molecule targeting AKT1, represents a promising lead compound for the development of anti-hepatic inflammation agents derived from Isatis indigotica.
Supporting information
S1 Table. Complete list of 142 intersection targets of Isatis indigotica against HBV-associated hepatic inflammation.
https://doi.org/10.1371/journal.pone.0355930.s001
(XLSX)
S2 Table. Top 10 bioactive compounds ranked by degree in the compound-target network.
Ranking includes MOL ID, compound name, degree, betweenness centrality, and closeness centrality.
https://doi.org/10.1371/journal.pone.0355930.s002
(XLSX)
S3 Table. Compound-target network data.
Full compound-target interaction dataset used to construct the compound-target network (Fig 2).
https://doi.org/10.1371/journal.pone.0355930.s003
(XLSX)
S4 Table. Top 10 hub targets ranked by degree centrality in the PPI network.
Ranking includes gene symbol, degree, betweenness centrality, and closeness centrality.
https://doi.org/10.1371/journal.pone.0355930.s004
(XLSX)
S5 Table. Filtered protein-protein interaction data.
PPI interactions from the STRING database filtered at combined score ≥ 0.9, used to construct the PPI network (Fig 3).
https://doi.org/10.1371/journal.pone.0355930.s005
(TSV)
S6 Table. Full GO and KEGG pathway enrichment analysis results for the 142 intersection targets.
https://doi.org/10.1371/journal.pone.0355930.s006
(XLSX)
Acknowledgments
The authors thank the developers of the TCMSP, SwissTargetPrediction, GeneCards, OMIM, STRING, and Metascape databases for providing open-access data resources. We also acknowledge the GROMACS and AutoDock Vina development teams for their freely available software tools.
References
- 1. Liu Y, Xu Q, Deng F, Zheng Z, Luo J, Wang P, et al. HERC2 promotes inflammation-driven cancer stemness and immune evasion in hepatocellular carcinoma by activating STAT3 pathway. J Exp Clin Cancer Res. 2023;42(1):38. pmid:36721234
- 2. Subramaniam NK, Mann KK. Mechanisms of metal-induced hepatic inflammation. Curr Environ Health Rep. 2024;11(4):547–56. pmid:39499483
- 3. Quintavalle C, Meyer-Schaller N, Roessler S, Calabrese D, Marone R, Riedl T, et al. miR-579-3p controls hepatocellular carcinoma formation by regulating the phosphoinositide 3-kinase-protein kinase b pathway in chronically inflamed liver. Hepatol Commun. 2022;6(6):1467–81. pmid:35132819
- 4. Luxenburger H, Neumann-Haefelin C. Liver-resident CD8+ T cells in viral hepatitis: not always good guys. J Clin Invest. 2023;133(1):e165033. pmid:36594469
- 5. Hao B, Liu Y, Wang B, Wu H, Chen Y, Zhang L. Hepatitis B surface antigen: carcinogenesis mechanisms and clinical implications in hepatocellular carcinoma. Exp Hematol Oncol. 2025;14(1):44. pmid:40141002
- 6. Ligat G, Verrier ER, Nassal M, Baumert TF. Hepatitis B virus-host interactions and novel targets for viral cure. Curr Opin Virol. 2021;49:41–51. pmid:34029994
- 7. Zhang J, Zhang Y, Khanal S, Cao D, Zhao J, Dang X, et al. Synthetic gRNA/Cas9 ribonucleoprotein targeting HBV DNA inhibits viral replication. J Med Virol. 2023;95(7):e28952. pmid:37455550
- 8. Pak U, Yu Y, Ning X, Ho C, Ji L, Mayo KH, et al. Comparative study of water-soluble polysaccharides isolated from leaves and roots of Isatis indigotica Fort. Int J Biol Macromol. 2022;206:642–52. pmid:35247423
- 9. Tan Y, Yang J, Sun S, Zhu H, Jiang Y, Wei X. Functional analysis of tandem glycosyltransferases catalyzing the O-glycosylation of flavonoid 4’-O-glycosyl-6-C-glycosides biosynthesis from Isatis indigotica. Plant Physiol Biochem. 2025;224:109920.
- 10. Tong H, Zhang J, Jiang L, Lu Y, Lu T. Mechanistic study on HuaganJiedu decoction in treating hepatitis B based on multi-omics analysis of network pharmacology, metabolomics, and transcriptomics. Phytomedicine. 2025;149:157568. pmid:41297320
- 11. Liu Y-F, Ling N, Zhang B, Chen C, Mo X-N, Cai J-Y, et al. Flavonoid-Rich mulberry leaf extract modulate lipid metabolism, antioxidant capacity, and gut microbiota in high-fat diet-induced obesity: potential roles of FGF21 and SOCS2. Food & Medicine Homology. 2024;1(2):9420016.
- 12. Deng J, Ma Y, He Y, Yang H, Chen Y, Wang L, et al. A Network Pharmacology-Based Investigation to the Pharmacodynamic Material Basis and Mechanisms of the Anti-Inflammatory and Anti-Viral Effect of Isatis indigotica. Drug Des Devel Ther. 2021;15:3193–206. pmid:34321868
- 13. Thombre KR, Gabhane V, Gupta KR, Umekar MJ. Network pharmacology in the multi-omics era: uncovering novel therapeutic strategies. Discov Chem. 2026;3:111.
- 14. Gu M, Liu H, Bi C, Situ W, Du H, Lin A, et al. Integrating network pharmacology, molecular docking and dynamics simulation to decipher the antipyretic mechanisms of Xiaochaihu granules. Front Med (Lausanne). 2026;13:1772991. pmid:41797801
- 15. Terkimbi SD, Mujinya R, Kayanja KL, Sunday BY, Dangana RS, Paul-Chima UO. Integrative network pharmacology and molecular docking approaches in herbal medicine research. A systematic review of applications, advances, and translational potential. F1000Res. 2025;14:1384.
- 16. Wang B-B, Wei F-H, Xiao Y, Guo H-X, Ji Z-H, Hu J-P, et al. Acacetin Alleviates DSS-Induced Ulcerative Colitis by Targeting Keap1/Nrf2 Pathway to Inhibit Endoplasmic Reticulum Stress-Mediated Apoptosis. J Agric Food Chem. 2026;74(5):4420–34. pmid:41612915
- 17. Wang W, Zhai S, Yang W, Gao H, Chang N, Zhang M, et al. Acacetin alleviates rheumatoid arthritis by targeting HSP90 ATPase domain to promote COX-2 degradation. Phytomedicine. 2024;135:156171. pmid:39489991
- 18. Pei X, Mu R, Liu S, Xiong L, Wang X, Yao Q, et al. 7,8-Dihydroxyflavone protects acetaminophen induced liver injury through activating PI3K/Akt/NRF2/GPX4 mediated ferroptosis suppression. Free Radic Biol Med. 2026;242:275–87. pmid:41173313
- 19. Zheng Y, Wang C, Liu W, Chen J, Sun Y, Chang D, et al. Upregulation of Nrf2 signaling: A key molecular mechanism of Baicalin’s neuroprotective action against diabetes-induced cognitive impairment. Biomed Pharmacother. 2024;174:116579. pmid:38631145
- 20. Shi M, Zhang J, Li M, Zhao Y, Guo Y, Xu J, et al. Liquiritigenin Confers Liver Protection by Enhancing NRF2 Signaling through Both Canonical and Non-canonical Signaling Pathways. J Med Chem. 2023;66(16):11324–34. pmid:37534604
- 21. Lee HY, Nam Y, Choi WS, Kim TW, Lee J, Sohn UD. The hepato-protective effect of eupatilin on an alcoholic liver disease model of rats. Korean J Physiol Pharmacol. 2020;24(5):385–94. pmid:32830145
- 22. Wang Z, Dou X, Shan Q, Ning Y, Wang J, Wang T, et al. Targeting AKT to treat liver disease: Opportunities and challenges. Biochem Pharmacol. 2025;242(Pt 3):117208. pmid:40752547
- 23. Cui G, Zhou Y, Liao W, Cossu E, Evert M, Zhang S, et al. Inhibition of GSK3 and TSC2 Mediates the Oncogenic Activity of AKT in Hepatocellular Carcinoma. Cancer Res. 2025;85(24):5049–65. pmid:40939191
- 24. Zhou Y, Zhang S, Qiu G, Wang X, Yonemura A, Xu H, et al. TSC/mTORC1 mediates mTORC2/AKT1 signaling in c-MYC-induced murine hepatocarcinogenesis via centromere protein M. J Clin Invest. 2024;134(22):e174415.
- 25. Huang Y, Luo W, Chen S, Su H, Zhu W, Wei Y, et al. Isovitexin alleviates hepatic fibrosis by regulating miR-21-mediated PI3K/Akt signaling and glutathione metabolic pathway: based on transcriptomics and metabolomics. Phytomedicine. 2023;121:155117. pmid:37820467
- 26. Manning BD, Toker A. AKT/PKB Signaling: Navigating the Network. Cell. 2017;169(3):381–405. pmid:28431241
- 27. García-Domínguez M. Targeting TNF-α: An immunotherapeutic approach for alleviating neuropathic pain. Int Rev Cell Mol Biol. 2025;396:81–94. pmid:40701698
- 28. Qu X, Yang T, Wang X, Xu D, Yu Y, Li J, et al. Macrophage RIPK3 triggers inflammation and cell death via the XBP1-Foxo1 axis in liver ischaemia-reperfusion injury. JHEP Rep. 2023;5(11):100879. pmid:37841640
- 29. Meng W-S, Sun J, Lu Y, Cao T-T, Chi M-Y, Gong Z-P, et al. Biancaea decapetala (Roth) O.Deg. extract exerts an anti-inflammatory effect by regulating the TNF/Akt/NF-κB pathway. Phytomedicine. 2023;119:154983. pmid:37586161
- 30. Zhang Z, Yao T, Zhao N, Liu H, Cheng H, Gonzalez FJ, et al. Disruption of peroxisome proliferator-activated receptor α in hepatocytes protects against acetaminophen-induced liver injury by activating the IL-6/STAT3 pathway. Int J Biol Sci. 2022;18(6):2317–28. pmid:35414769
- 31. Mohamed MR, Zhao G, Volkert I, Molinaro A, Schneider CV, Strnad P, et al. Differential Protective Roles of c-Jun N-terminal Kinase-2 in Nonparenchymal Liver Cells and Hepatocytes During Cholestasis. Cell Mol Gastroenterol Hepatol. 2025;19(11):101588. pmid:40681041
- 32. Li Y, Zhao J, Yin Y, Li K, Zhang C, Zheng Y. The Role of IL-6 in Fibrotic Diseases: Molecular and Cellular Mechanisms. Int J Biol Sci. 2022;18(14):5405–14. pmid:36147459
- 33. Zarneshan SN, Fakhri S, Bachtel G, Bishayee A. Exploiting pivotal mechanisms behind the senescence-like cell cycle arrest in cancer. Adv Protein Chem Struct Biol. 2023;135:1–19. pmid:37061329
- 34. Baliakas P, Soussi T. The TP53 tumor suppressor gene: From molecular biology to clinical investigations. J Intern Med. 2025;298(2):78–96. pmid:40524430
- 35. Rajagopalan A, Feng Y, Gayatri MB, Ranheim EA, Klungness T, Matson DR, et al. A gain-of-function p53 mutant synergizes with oncogenic NRAS to promote acute myeloid leukemia in mice. J Clin Invest. 2023;133(24):e173116. pmid:37847561
- 36. Busbee PB, Bam M, Yang X, Abdulla OA, Zhou J, Ginsberg JPJ, et al. Dysregulated TP53 Among PTSD Patients Leads to Downregulation of miRNA let-7a and Promotes an Inflammatory Th17 Phenotype. Front Immunol. 2022;12:815840. pmid:35058939
- 37. Marruecos L, Manils J, Moreta C, Gómez D, Filgaira I, Serafin A, et al. Single loss of a Trp53 allele triggers an increased oxidative, DNA damage and cytokine inflammatory responses through deregulation of IκBα expression. Cell Death Dis. 2021;12(4):359. pmid:33824284
- 38. Ren C-X, Gao M-Y, Li N, Tang C, Chu G-H, Yusuf A, et al. Identification and mechanism elucidation of medicative diet for food therapy XQCSY in NAFLD prevention: an integrative in silico study. Food & Medicine Homology. 2024;1(2):9420015.
- 39. Wang Y, Huang B, Wei X, Guan Y, Li L, Zheng Y, et al. Gut microbiota metabolic reprogramming drives the development of metabolic diseases in the host. Gut Microbes. 2026;18(1):2644681. pmid:41830551
- 40. Ringelhan M, Schuehle S, van de Klundert M, Kotsiliti E, Plissonnier M-L, Faure-Dupuy S, et al. HBV-related HCC development in mice is STAT3 dependent and indicates an oncogenic effect of HBx. JHEP Rep. 2024;6(10):101128. pmid:39290403
- 41. Wuerger LTD, Kudiabor F, Alarcan J, Templin M, Poetz O, Sieg H, et al. Okadaic Acid Activates JAK/STAT Signaling to Affect Xenobiotic Metabolism in HepaRG Cells. Cells. 2023;12(5):770. pmid:36899906
- 42. Li Y, Zhu Y, Feng S, Ishida Y, Chiu T-P, Saito T, et al. Macrophages activated by hepatitis B virus have distinct metabolic profiles and suppress the virus via IL-1β to downregulate PPARα and FOXO3. Cell Rep. 2022;38(4):110284. pmid:35081341
- 43. Zhu J, Cheng W, He T-T, Hou B-L, Lei L-Y, Wang Z, et al. Exploring the Anti-Inflammatory Effect of Tryptanthrin by Regulating TLR4/MyD88/ROS/NF-κB, JAK/STAT3, and Keap1/Nrf2 Signaling Pathways. ACS Omega. 2024;9(28):30904–18. pmid:39035974
- 44. Niu M, Wang Y-Z, Deng X-M, Wu X, Hua Z-Y, Lv T-T. Tryptanthrin alleviate lung fibrosis via suppression of MAPK/NF-κB and TGF-β1/SMAD signaling pathways in vitro and in vivo. Toxicol Appl Pharmacol. 2025;498:117285. pmid:40089192
- 45. Zhai SY, Li LR, Jia JH, Cao YX, Li MY, Zheng YF. Food and medicine homology components ameliorate metabolic diseases via dual pathways: blood absorption and nonabsorptive pathways. Food Med Homol. 2026;3(2):9420080.