Effects of Corydalis saxicola Bunting total alkaloids on lipopolysaccharide‑induced inflammatory responses in human oral epithelial cells

Introduction

Oral mucositis (OM) is an exaggerated inflammatory response characterized by pain, erythematous lesions and ulceration (1). Ionizing radiation and chemotherapy are the primary causes of OM, while oral bacterial colonization may further exacerbate inflammation (2). The oral cavity harbors a distinct and complex microbiota, the dynamic balance of which is regulated by numerous endogenous and exogenous factors. Microbial dysbiosis may facilitate the colonization and proliferation of pathogenic bacteria within oral epithelial tissues (3). These pathogens release antigens, endotoxins, such as lipopolysaccharide (LPS), and bacterial proteases that directly or indirectly damage epithelial and connective tissues. Concurrently, they stimulate epithelial cells to produce pro-inflammatory cytokines, including IL-1β, IL-6 and TNF-α, as well as chemokines, such as IL-8, thereby amplifying the inflammatory response (4,5).

Corydalis saxicola Bunting (CSB), a Traditional Zhuang Medicine from Guangxi, China, is widely used in traditional practice for the treatment of OM. To date, 81 chemical constituents have been identified in CSB, including alkaloids, steroids and flavonoids (6). The total alkaloids of Corydalis saxicola Bunting (CSBTA) are considered its principal active components. Although the anti-inflammatory effects of CSBTA have been demonstrated in various in vitro and in vivo disease models, their effects on OM remain largely unexplored (7). Our previous studies demonstrated that CSBTA suppressed macrophage pyroptosis and inflammatory responses induced by oral pathogens and their LPS (8-10).

To further investigate the therapeutic potential of CSBTA in OM, the present study employed network pharmacology to predict its bioactive components, core targets and signaling pathways. Molecular docking was subsequently performed to evaluate interactions between candidate bioactive components and their predicted targets. Finally, an LPS-induced human oral epithelial cell (HOEC) inflammatory model was established (11,12), to evaluate the effects of CSBTA and investigate its potential underlying mechanisms.

Materials and methods

Acquisition of CSBTA active ingredients and target prediction

The chemical constituents of CSBTA were identified through a literature review. The Simplified Molecular Input Line Entry System (SMILES) structures of these compounds were retrieved from the PubChem database (https://pubchem.ncbi.nlm.nih.gov) (13). Candidate active compounds were screened using SwissADME (http://www.swissadme.ch) (14) and their potential targets were predicted using the SwissTargetPrediction database (http://www.swisstargetprediction.ch) (15).

Collection of OM-associated genes

OM-associated targets were identified by searching the OMIM (https://omim.org) (16), DrugBank (https://go.drugbank.com) (17) and GeneCards (https://www.genecards.org) (18) databases using the keyword ‘oral mucositis’.

Construction of the protein-protein interaction (PPI) network and core target selection

Genes shared between the predicted CSBTA targets and OM-associated targets were identified and imported into the STRING database (version 12.5; https://string-db.org/) (19) to construct a PPI network. The resulting network was analyzed using Cytoscape (version 3.9.1; https://cytoscape.org/) with the NetworkAnalyzer plugin. Core targets were identified according to network topology parameters, with betweenness centrality and closeness centrality greater than their respective median values and degree >240.

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses

GO and KEGG enrichment analyses of the intersecting genes were performed using Metascape (https://metascape.org/) (20).

Molecular docking validation

The 2D structures of key CSBTA constituents were retrieved from the Traditional Chinese Medicine Systems Pharmacology (TCMSP; https://www.tcmsp-e.com/tcmsp.php/) database and the 3D structures of macromolecular targets were obtained from the Research Collaboratory for Structural Bioinformatics Protein Data Bank (RCSB PDB) database (https://www.rcsb.org/) (21). Molecular docking was performed using AutoDock Vina (version 1.1.2; https://vina.scripps.edu/) and the docking results were visualized using PyMOL (version 2.2.0; https://pymol.org/).

Preparation of CSBTA

CSBTA was extracted and characterized as previously described (9,10). CSBTA was dissolved in dimethyl sulfoxide before use.

Establishment of the LPS-induced HOEC model

HOECs (cat. no. HTX1862; Shenzhen OTWO Biotechnology Co., Ltd.) were commercially obtained as an immortalized human oral epithelial cell line established by the supplier for research use. No Cellosaurus CVCL accession number is available. Supplier information is available at http://www.otwobiotech.com/product_detail3/331.html, which includes information on the use of this cell line in previously published studies. HOECs were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, penicillin and streptomycin at 37˚C in a humidified atmosphere containing 5% CO2. The morphological characteristics of cells were observed under an inverted phase-contrast microscope at x100 and x400 magnification. Cells were seeded in 6-well plates at a density of 5x105 cells/ml. LPS (Escherichia coli serotype O55:B5; MilliporeSigma) was used to prepare a 5 mg/ml stock solution. For experimental stimulation, 2 µl of the stock solution was diluted in 2 ml of complete medium to achieve a final LPS concentration of 1 µg/ml (22). HOECs were treated with 1 µg/ml LPS for 1, 3, 6, 12 or 24 h, 37˚C, after which the mRNA expression levels of IL-1β, IL-6, IL-8 and TNF-α were determined using reverse transcription-quantitative (RT-q)PCR. Treatment with 1 µg/ml LPS for 3 h markedly increased the relative mRNA expression levels of IL-1β, IL-6 and TNF-α. Therefore, treatment with 1 µg/ml LPS for 3 h was selected as the modeling condition for subsequent experiments (Fig. S1A). As commercially available established human cells were used, no additional ethical approval was required.

Cell viability assay

HOECs were treated with CSBTA (0, 20, 40, 80, 160 or 320 µg/ml) or SR11302 (0, 1, 5 or 10 µM) for 24 h, 37˚C. Cell viability was assessed using the Cell Counting Kit-8 (CCK-8; Biosharp Life Sciences). Absorbance was measured at 450 nm using a microplate reader (Bio-Rad Laboratories, Inc.).

Drug treatment

HOECs were divided into four groups: i) Blank control group, untreated HOECs; ii) model group, LPS-induced HOECs; iii) SR11302 group, LPS-induced HOECs treated with 1 µM SR11302 [activator protein-1 (AP-1) inhibitor; MedChemExpress] for 24 h (23); and iv) CSBTA group, LPS-induced HOECs treated with 20 µg/ml CSBTA for 24 h (10).

RT-qPCR analysis

Total RNA was extracted from HOECs seeded in 6-well plates at a density of 5x105 cells/ml (2 ml per well) using TRIzol® (Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions and reverse-transcribed into cDNA using the SuperScript III First-Strand Synthesis System (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. RT-qPCR was performed using a 2XReal Star Green Fast Mixture with ROXII (GenStar) on a Thermo Fisher Scientific, Inc. system under the following conditions: Initial denaturation at 95˚C for 2 min; 40 cycles of 95˚C for 15 sec and 60˚C for 30 sec; followed by 95˚C for 15 sec, 60˚C for 1 min and 95˚C for 15 sec. The fluorophore was SYBR Green, and the supplier was GenStar. GAPDH was used as the reference gene. Primers were synthesized by Sangon Biotech Co., Ltd. and are listed in Table I. Relative mRNA expression levels were calculated using the 2-ΔΔCq method (24).

Table I Gene primer base sequence.

Table I

Gene primer base sequence.

Gene	Sequences
IL-1β	Forward GCCAGTGAAATGATGGCTTATT
	Reverse AGGAGCACTTCATCTGTTTAGG
IL-6	Forward CACTGGTCTTTTGGAGTTTGAG
	Reverse GGACTTTTGTACTCATCTGCAC
IL-8	Forward AACTGAGAGTGATTGAGAGTGG
	Reverse ATGAATTCTCAGCCCTCTTCAA
TNF-α	Forward AGCCCTGGTATGAGCCCATCTATC
	Reverse TCCCAAAGTAGACCTGCCCAGAC
JUN	Forward AAGAACTCGGACCTCCTCACCTC
	Reverse GCCCGTTGCTGGACTGGATTATC
GAPDH	Forward CAGGAGGCATTGCTGATGAT
	Reverse GAAGGCTGGGGCTCATTT

Enzyme-linked immunosorbent assay (ELISA)

Cytokine levels (IL-1β, IL-6, IL-8 and TNF-α) in the cell culture supernatants were measured using commercial ELISA kits (Meimian; cat. nos: MM-0181H1 IL-1β, MM-0049H2 IL-6, MM-1558H2 IL-8 and MM-0122H1 TNF-α).

Western blotting

Proteins were extracted from HOECs using radioimmunoprecipitation assay (RIPA) lysis buffer (Fudebio-tech) and quantified using a bicinchoninic acid (BCA) assay (Fudebio-tech). Equal amounts (30 µg) of protein were separated on a 10% SDS-PAGE gel with a 5% stacking gel and transferred onto 0.22-µm PVDF membranes. After blocking with 2% skimmed milk in Tris-buffered saline (TBS) containing 0.1% Tween-20 (TBST) at room temperature for 1 h, membranes were incubated overnight at 4˚C with primary antibodies against GAPDH (Wuhan Servicebio Technology Co., Ltd.; cat. no. GB15004; 1:2,000), phosphorylated (p)c-Jun was detected with a rabbit anti-p-c-Jun (Ser73) monoclonal antibody (clone D47G9; Cell Signaling Technology, Inc.; cat. no. 3270; 1:1,000) and total c-Jun with a rabbit anti-c-Jun monoclonal antibody (clone 60A8; Cell Signaling Technology, Inc.; cat. no. 9165; 1:1,000). After washing, membranes were incubated with secondary antibodies (Wuhan Servicebio Technology Co., Ltd.; cat. no. G1213; 1:5,000) for 1 h at room temperature. Protein bands were visualized using a chemiluminescence imaging system (ChemiDocM; Bio-Rad Laboratories, Inc.) and quantified using ImageJ (version 6.0; National Institutes of Health). Relative protein expression levels were normalized to GAPDH.

Statistical analysis

All quantitative experiments were performed using three independent biological replicates. Data distribution and variance assumptions were assessed prior to statistical analysis. Data are presented as the mean ± standard deviation (SD). Group differences were assessed by one-way ANOVA, followed by Tukey's honestly significant difference (HSD) test for pairwise comparisons. Statistical analyses were performed using SPSS (version, 22.0; IBM Corp.). P<0.05 was considered to indicate a statistically significant difference.

Results

CSBTA active ingredients and target prediction

A total of 49 chemical constituents of CSBTA were identified through a literature review and their SMILES structures were retrieved from the PubChem database. After screening using SwissADME (high gastrointestinal absorption and ≥2 ‘YES’ drug-likeness properties), 33 active compounds were retained. Potential targets of these compounds were predicted using SwissTargetPrediction, yielding 538 targets after excluding those with a probability score of zero (Fig. 1A).

Figure 1 Network pharmacology and molecular docking. (A) CSBTA effective chemical components and potential targets; 33 effective chemical components of CSBTA with a total of 538 potential targets. (B) The intersection of CSBTA chemical components targets with disease gene targets. 244 intersection genes between the component targets and the disease gene targets were identified. (C) Core Targets. The size and darkness of the circles representing the targets indicate their degree values, with larger and darker circles indicating higher degrees. Thicker and darker lines connecting the targets indicate closer relationships between them. (D) GO enrichment analysis. (E) KEGG enrichment analysis. The bubble plots of the 20 most significant signaling pathways based on KEGG enrichment analysis. The X-axis and Y-axis stand for the gene ratios and full names of the processes, respectively. The color and size of each bubble represent the P-value and gene count, respectively. (F) Heat maps of lowest binding energy between core targets and key components. The darker the color, the lower the binding energy, indicating a tighter binding. Binding energies below-5 kcal/mol indicate binding capability, while those below-8 kcal/mol indicate tight binding. (G) Molecular docking results. The purple molecules represent the target JUN, the green ones indicate the sites on the target where the compounds bind, labeled with their names. The orange molecules represent the key chemical components, while the yellow dashed lines denote the hydrogen bonds directly formed between the target and the compounds, labeled with their lengths. CSBTA, Corydalis saxicola Bunting total alkaloids; JUN, Jun proto-oncogene; AP-1, activator protein 1; EGFR, epidermal growth factor receptor; CASP3, caspase 3; AKT1, AKT serine/threonine kinase 1; STAT3, signal transducer and activator of transcription 3; GO, Gene Ontology; BP, biological processes; CC, cellular components; MF, molecular functions; KEGG, Kyoto Encyclopedia of Genes and Genomes.

OM-associated target genes

A total of 1,868 OM-associated targets were obtained from the OMIM, DrugBank and GeneCards databases following duplicate removal.

PPI network construction and core target identification

Intersection analysis identified 244 common targets between CSBTA and OM (Fig. 1B). PPI network construction using STRING and Cytoscape (version, 3.9.1) identified AKT1, STAT3, EGFR, CASP3 and JUN as core targets (Fig. 1C). Moreover, five key CSBTA constituents, namely, dihydrochelerythrine, oxyacanthine, magnoflorine, cheilanthifoline and epiberberine, were identified as critical compounds.

GO and KEGG enrichment analyses

KEGG enrichment analysis indicated that the CSBTA core targets were markedly associated with 208 pathways, while GO enrichment analysis identified 2,430 biological processes, 137 cellular components and 272 molecular functions (P<0.05; Fig. 1D and E).

Molecular docking validation

Molecular docking demonstrated favorable binding between JUN and key CSBTA constituents, with binding energies of -9.0, -9.7, -8.2, -8.0 and -8.2 kcal/mol for dihydrochelerythrine, oxyacanthine, magnoflorine, cheilanthifoline and epiberberine, respectively (Fig. 1F and G).

Morphology of HOECs

HOECs cultured under standard conditions exhibited a typical cobblestone-like morphology, characterized by clear cell boundaries, a uniform polygonal shape and colony formation (Fig. 2A).

Figure 2 Effects of CSBTA on LPS-induced inflammatory responses and JUN/AP-1-related signaling in HOECs. (A) HOEC cell morphology. The cells proliferate and adhere to the substrate, exhibiting irregular polygons that are uniformly sized and tightly arranged, resembling ‘cobblestones’, similar to squamous epithelial cells. (B) CSBTA reduces the expression of inflammatory cytokines in LPS-induced HOECs. (C) CSBTA reduces the expression of AP-1 in LPS-induced HOECs. *P<0.05, **P<0.01, ***P<0.005, ****P<0.001. CSBTA, Corydalis saxicola Bunting total alkaloids; HOECs, human oral epithelial cells; LPS, lipopolysaccharide; JUN, Jun proto-oncogene; AP-1, activator protein 1; NC, negative control.

Cytotoxicity of CSBTA and SR11302

CSBTA exhibited no cytotoxicity at concentrations up to 40 µg/ml (Table II). The half-maximal inhibitory concentration (IC50) of CSBTA in HOECs was 174.2 µg/ml (Fig. S1B). Similarly, SR11302 treatment did not markedly affect cell viability at any of the tested concentrations (Table III).

Table II Relative cell viability after stimulation with CSBTA.

Table II

Relative cell viability after stimulation with CSBTA.

Group, CSBTA, µg/ml	Relative cell viability (%)
0	100
20	103.46±3.08
40	99.89±2.61
80	87.32±7.69a
160	62.11±10.35b
320	0b

[i] ^avs. Control P<0.01,

[ii] ^bvs. Control P<0.001. CSBTA, Corydalis saxicola Bunting total alkaloids.

Table III Relative cell viability after stimulation with SR11302.

Table III

Relative cell viability after stimulation with SR11302.

Group, SR11302, µM	Relative cell viability (%)
0	100
1	104.82±9.35
5	106.14±8.80
10	102.29±4.50

Effect of CSBTA on LPS-induced cytokine expression

Stimulation of HOECs with 1 µg/ml LPS for 3 h markedly increased IL-1β, IL-6, TNF-α and IL-8 expression compared with the control group (P<0.05). Following establishment of the LPS-induced inflammatory response, subsequent treatment with either 20 µg/ml CSBTA or 1 µM SR11302 for 24 h markedly reduced the expression of these cytokines (P<0.05), with no significant difference between the two treatment groups (P>0.05; Fig. 2B).

Effect of CSBTA on AP-1 expression in HOECs

LPS stimulation markedly increased AP-1 mRNA expression and c-Jun and p-c-Jun protein expression in HOECs (P<0.05). Subsequent treatment with CSBTA (20 µg/ml) or SR11302 (1 µM) for 24 h markedly reduced AP-1 mRNA expression and c-Jun and p-c-Jun protein expression (P<0.05). No significant difference was observed between the CSBTA and SR11302 treatment groups (P>0.05; Fig. 2C).

Discussion

OM is characterized by inflammatory injury of the oral epithelium and is frequently associated with increased production of pro-inflammatory cytokines. In the present study, an LPS-induced inflammatory model was established in HOECs to evaluate the therapeutic effects of CSBTA. The results demonstrated that CSBTA markedly reduced the expression of key inflammatory mediators, including IL-1β, IL-6, IL-8 and TNF-α, indicating its potential anti-inflammatory effects in oral epithelial cells.

LPS, a major component of the outer membrane of Gram-negative bacteria, exerts pathogenic effects primarily through the activation of immune and epithelial cells, induction of cytokine release and amplification of inflammatory cascades, ultimately leading to tissue damage (25). As a classical pathogen-associated molecular pattern, LPS plays a central role in innate immunity and is widely used to establish in vitro inflammatory models (26,27). Previous studies have demonstrated that LPS induces time-dependent upregulation of TNF-α and IL-6 in taste organoids and primary lingual epithelium, confirming that oral epithelial cells are responsive to LPS stimulation (22). Similarly, Gasiorek et al (28) showed that LPS activates pro-inflammatory signaling pathways in oral keratinocytes, leading to increased expression of IL-1β, IL-6, IL-8 and TNF-α through NF-κB and STAT1 activation (28). Consistent with these findings, LPS stimulation of HOECs in the present study markedly increased the expression of IL-1β, IL-6, IL-8 and TNF-α. These cytokines are closely associated with various oral mucosal inflammatory conditions, including peri-implant mucositis, recurrent aphthous stomatitis, oral lichen planus and oral candidiasis (29-32).

A total of 49 chemical constituents of CSBTA were identified through a literature review and database searches. Following analysis using SwissTargetPrediction, 538 potential targets were identified. Network pharmacology analysis identified several active constituents of CSBTA, including dihydrochelerythrine, oxyacanthine, magnoflorine, cheilanthifoline and epiberberine, as potential modulators of OM. These compounds were predicted to act on core targets including AKT1, STAT3, EGFR, CASP3 and JUN. Moreover, molecular docking analysis demonstrated favorable binding between JUN and the key CSBTA constituents, supporting JUN as a potential target of these compounds.

AP-1, a dimeric transcription factor composed of Jun and Fos subunits, plays a critical role in inflammatory signaling and early mucosal injury responses (12,33). It functions as a key integrator of multiple signaling pathways involved in mucosal inflammation. Upon LPS stimulation, activation of the PI3K/AKT pathway enhances AP-1 activity and promotes the transcription of pro-inflammatory genes, partly through modulation of c-Jun activation and stability (23). Concurrently, EGFR signaling induces AP-1 activation through MAPK cascades, particularly through JNK-mediated phosphorylation of c-Jun (34). As a major downstream transcription factor of the MAPK/JNK pathway, AP-1 regulates the expression of multiple pro-inflammatory cytokines, including IL-1β, IL-6 and TNF-α (35). In addition, STAT3 cooperates with AP-1 in the transcriptional regulation of inflammation-associated genes, including matrix metalloproteinases, through coordinated promoter interactions (36). Beyond its role in inflammatory gene expression, AP-1 is also involved in the regulation of apoptosis. In particular, AP-1 components, such as c-Jun, modulate the expression of apoptosis-associated genes, including CASP3, thereby associating stress-activated signaling pathways to cell fate determination. The interplay between AP-1 and CASP3 may therefore contribute to the regulation of cell fate in the context of mucosal inflammation (37). Given its central role in integrating inflammatory signaling networks and coordinating inflammatory responses and cell fate, AP-1 was selected as a key target for further investigation in the present study.

Consistent with the central role of AP-1 in inflammatory signaling, accumulating evidence supports the therapeutic relevance of targeting this transcription factor (38,39). Previous studies have demonstrated that anomalin attenuates LPS-induced acute lung injury by suppressing AP-1 activity (40), while extracts from Xanthium strumarium and Phyllanthus acidus inhibit AP-1-mediated cytokine release in macrophages (41,42). Similarly, curcumin has been reported to downregulate pro-inflammatory cytokines in lymphoma models through modulation of AP-1 signaling (43). In the present study, five compounds, namely, dihydrochelerythrine, oxyacanthine, magnoflorine, cheilanthifoline and epiberberine, were identified through network pharmacology analysis and have previously been reported as important constituents of CSBTA (6,44). Among these compounds, magnoflorine has the most substantial evidence supporting its anti-inflammatory activity in LPS-induced models. Previous studies demonstrated that magnoflorine attenuates inflammatory responses both in vitro and in vivo, through the suppression of TLR4-mediated NF-κB and MAPK signaling pathways, thereby reducing the production of pro-inflammatory cytokines, such as TNF-α, IL-1β and IL-6(45). Given that MAPK cascades, including JNK, ERK and p38, are key upstream regulators of AP-1 activation, these findings suggest that magnoflorine may also modulate AP-1-associated inflammatory signaling. By contrast, evidence regarding dihydrochelerythrine, oxyacanthine, cheilanthifoline and epiberberine remains relatively limited. Although some studies have reported their anti-inflammatory potential, including inhibitory effects on pro-inflammatory mediators, such as TNF-α (46,47), direct mechanistic investigations, particularly those examining AP-1 signaling in LPS-induced inflammation, are lacking. Notably, the present results demonstrated that CSBTA markedly reduced AP-1 expression and the levels of IL-1β, IL-6, IL-8 and TNF-α in LPS-stimulated HOECs, with effects comparable to those of the AP-1 inhibitor, SR11302. These findings support the possibility that the anti-inflammatory effects of CSBTA are mediated, at least in part, through inhibition of AP-1 signaling and subsequent attenuation of downstream pro-inflammatory responses associated with OM.

Several limitations of the present study should be acknowledged. First, the experiments were limited to an in vitro cell model and the effects of CSBTA were not evaluated in vivo. In addition, as CSBTA was administered after LPS stimulation, the present design evaluates attenuation of an established inflammatory response rather than prevention; future studies using pretreatment or co-treatment are required to assess potential preventive effects. Although network pharmacology and molecular docking provided supportive evidence, these approaches are predictive and do not establish direct molecular interactions. The present study also primarily focused on AP-1 signaling and provided only preliminary evidence that CSBTA suppresses oral epithelial inflammation through modulation of this pathway. Other predicted targets and signaling pathways, including AKT1, STAT3, NF-κB signaling, oxidative stress and apoptosis-associated mechanisms, were not systematically investigated. Further studies using animal models and additional mechanistic approaches are therefore required to validate these findings and clarify the precise molecular mechanisms underlying the effects of CSBTA.

In conclusion, the present study demonstrated that CSBTA may attenuate LPS-induced inflammatory responses in HOECs, potentially through modulation of AP-1 signaling. These findings provide preliminary experimental evidence supporting the anti-inflammatory properties of CSBTA and warrant further investigation of its potential application in OM and other oral mucosal inflammatory conditions.

Supplementary Material

LPS-induced inflammatory cytokine expression in HOECs and cytotoxicity of CSBTA. (A) Time-dependent mRNA expression of inflammatory cytokines in HOECs following LPS stimulation. HOECs were stimulated with 1 μg/ml LPS for indicated time points (1-24 h). RT-qPCR analysis showed significant upregulation of IL-6 and TNF-α at 1 h, IL-1β/IL-6/TNF-α at 3 h and IL-1β at 6 h. IL-8 showed no significant response at any time point. Maximum inflammatory response was observed at 3 h; thus, 1 μg/ml LPS for 3 h was selected as the modeling condition. *P<0.05, **P<0.01, ****P<0.001. Cell viability was evaluated using a CCK-8 assay and nonlinear regression analysis revealed that the IC50 of CSBTA in HOECs was 174. 2 μg/ml. (B) IC50 curve of CSBTA on HOECs. HOECs, human oral epithelial cells; LPS, lipopolysaccharide; RT-qPCR, reverse transcription-quantitative PCR; IC50, half-maximal inhibitory concentration; CSBTA, Corydalis saxicola Bunting total alkaloids; NC, negative control.

Acknowledgements

Not applicable.

Funding

Funding: The present study was supported by grants from General Project of Guangxi Natural Science Foundation (grant no. 2025GXNSFAA069899), the National Natural Science Foundation of China (grant no. 82460191), Guangxi Young Elite Scientist Sponsorship Program (grant no. GXYESS2025217) and First-class discipline innovation-driven talent program of Guangxi Medical University.

Availability of data and materials

The data generated in the present study may be requested from the corresponding author.

Authors' contributions

YQ was responsible for methodology, experiments, statistical analysis and preparation of the original draft. YL was responsible for methodology, experiments, statistical analysis, network pharmacology and preparation of the original draft. JW, QJ and ZL were responsible for cell experiments. ZS was responsible for extraction and quality control of the drugs used throughout the study. RT was responsible for supervision, review and editing. XY was responsible for supervision, review and editing, study conception and coordination and project administration. YL and XY confirm the authenticity of all the raw data.

Ethics approval and consent to participate

The overall study was reviewed and approved by the Ethics Committee of Guangxi Medical University (approval no. 2024KY0078).

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References