Abstract Objective We investigated the underlying mechanism by which Rhododendron molle G. Don extraction (RME) mitigates rheumatoid arthritis (RA) in rats through network analysis and molecular docking. Methods The active metabolites of RME were analyzed by high performance liquid chromatography - evaporative light scattering detector (HPLC-ELSD). Subsequently, the potential targets were screened from various databases. Genes relative to RA and [42]GSE55457 data sets were obtained. Furthermore, the biological processes associated with the drug targets and [43]GSE55457 genes were analyzed. A protein - protein interaction (PPI) network was constructed and subjected to topological analysis. Key proteins within this network were identified and validated by in vivo experiments. Results: HPLC analysis showed that REM contained multiple anti-inflammatory active metabolites. Through network analysis, we found that the potential target of the drug and the differential genes of RA were involved in several signaling pathways, including Wnt, JAK-STAT and other signaling pathways which are known to play important roles in inflammatory responses. Dvl1 and GSK-3β were screened as the core targets by topological analysis of the potential target protein interaction network. The results of molecular docking showed that the core target had good affinity with the active metabolites of R. molle. It was confirmed that RME could remarkably downregulate the expression of Dvl1, Wnt1, p-GSK-3β, leading to a marked reduction in toe thickness and arthritis index (AI) of AIA rats, reducing the level of inflammatory factors, and alleviating the degree of tail ulceration in AIA rats in a dose-dependent manner, displaying therapeutic effects comparable to those of Tripterygium wilfordii polyglycoside tablets. Conclusion RME contains active metabolites with anti-inflammatory effects, which can significantly inhibit inflammation through multiple targets and pathways to effectively ameliorate RA. Supplementary Information The online version contains supplementary material available at 10.1186/s12906-025-04982-z. Keywords: Rhododendron molle G.Don, Rheumatoid arthritis, Network analysis, Dvl1/Wnt1/β-catenin/GSK-3β signaling pathway Introduction RA is an autoimmune disease typically characterized by synovitis and synovial hyperplasia [[44]1, [45]2]. The synovium, a connective tissue membrane covering the inner surface of the articular lumen, serves as the principal site of RA [[46]3]. Fibroblast-like synoviocyte (FLS) represent the major cell population within the synovial cells. Activation of autoimmune response and activation of RA-FLS produce inflammatory factors and chemokines to maintain synovial inflammation. FLS display “Tumor-like” cell behavior, such as uncontrolled proliferation, anti-apoptosis, and gradual migration or invasion to articular cartilage and bone tissue, which result in osteoclast proliferation, pannus formation, and synovial hyperplasia, ultimately causing bone destruction or osteonecrosis [[47]4, [48]5]. At present, the drugs for the treatment of RA mainly include non-steroidal anti-inflammatory drugs (NSAIDs), glucocorticoids (GCs), slow-acting anti-rheumatic drugs (SAARDs), etc [[49]6]., but long-term use of such drugs can cause adverse reactions such as liver injury and multiple infections [[50]7]. However, botanical drugs have the advantages of significant effects and small side effects in the treatment of RA. In recent years, studies on the active metabolites and pharmacological activities of R.molle have been conducted, revealing that this plant has strong anti-inflammatory and analgesic activities [[51]8–[52]10]. Previous research conducted by our research group has revealed that the serum containing R.molle extract can inhibit the proliferation of TNF-α treated RA-FLS by downregulating the EGFR/AKT signaling pathway, and reduce the levels of pro-inflammatory cytokines such as IL-1β, IL-17 A, and IL-6 [[53]11]. Additionally, it can also suppress the TNF-α levels in LPS-induced RAW264.7 cells [[54]8]. Previously, we identified Rhodojaponin-III, Rhodojaponin-II, Hyperoside and Quercetin as active metabolites in R.molle using HPLC fingerprinting combined with chemometric pattern recognition analysis [[55]12]. Guo et al. demonstrated that these four metabolites are rapidly absorbed into the bloodstream to exert pharmacological effects, including anti-inflammatory and analgesic [[56]13], immunomodulatory [[57]14], activities. They alleviate symptoms in CIA rats by regulating pro-inflammatory cytokines such as IL-1β and IL-6, playing a crucial regulatory role in the organism [[58]15]. Notably, rhodojaponin-III was found to suppress inflammatory responses and synovial hyperplasia in RA rats through modulation of the TLR4/MyD88/NF-κB signaling pathway [[59]16]. Hyperoside and Quercitrin exhibited potent anti-inflammatory effects by inhibiting TNF-α-induced MH7A cell proliferation, which was associated with reduced levels of pro-inflammatory cytokines including TNF-α and IL-1β. Serum metabolomics and spectrum-effect relationship analysis further demonstrated that Quercitrin and related compounds significantly correlated with RA efficacy, suggesting their potential as pharmacodynamic agents for RA treatment [[60]17, [61]18].Currently, the mechanism of R.molle against RA has not been studied in depth, and the active metabolites of R.molle are complex and have certain toxicity, which limits its further in-depth research, clinical development and application. This computer-based network analysis method, which combines bioinformatics, pharmacology and biology, can explore the potential relationship among botanical drugs, diseases, active metabolites and core targets [[62]19]. Therefore, this experiment explored the safe anti-inflammatory dosage of its extract through acute toxicity tests in vivo. By combining network analysis and molecular docking techniques, we aimed to uncover the potential mechanisms of action and molecular targets of the active metabolites of R.molle and RA. The key pathways Dvl1/Wnt1/β-catenin/GSK-3β were selected for experimental validation, which provided a theoretical basis for studying the treatment of RA with RME. Materials and methods Plant materials The medicinal material consisted of authenticated dried flowers of Rhododendron molle G. Don (Family: Ericaceae; Rhododendron; commonly referred to as “Yangzhizhu” in Traditional Chinese Medicine), which were collected during the typical flowering season (April 2023) from Henan Province, China (voucher specimen number: LXR-0410 was deposited in Hunan University of Traditional Chinese Medicine Herbarium). All samples were authenticated as dried flowers of Rhododendron molle G. Don (Ericaceae) by Professor Zhou Ribao from Hunan University of Chinese Medicine. The samples were cryopreserved at − 20 °C in the Key Research Laboratory of Germplasm Resources and Standardized Planting of Genuine Regional Medicinal Materials Produced of Hunan University of Traditional Chinese Medicine. A total of 100 g of powdered samples of R.molle was weighed and subjected to ultrasonic extraction with 75% ethanol (80 kHz) for 2 h. The combined filtrates were concentrated by rotary evaporation to recover the solvent, yielding an extract of R.molle. The extract was stored at 4 °C in a refrigerator, with an extraction yield of 28.25%. Experimental reagents M. Tuberculosis Des.H37RA was purchased from Becton, Dickinson and Company (Franklin Lakes, NJ, USA, 3026415). Mineral oil was purchased from Sigma Aldrich (St. Louis, MO, USA, MKCR3541). Tripterygium wilfordii polyglycoside tablets (2305105B) purchased from Zhejiang Dende Pharmaceutical Co., Ltd. (Shaoxing, China). Dexamethasone Acetate (LB22169) purchased from Zhejiang Juxian Pharmaceutical Co., Ltd. (Taizhou, China). The ELISA Kit TNF-α (JYM0635Ra), IL-1β (JYM0419Ra), IL-6 (JYM0646Ra), IL-17 A (JYM0480Ra) were purchased from Wuhan Genmei Biotechnology Co., Ltd. (Wuhan, China). The primary antibodies against β-actin (81115-1-RR), APC (19782-1-AP), Dvl1 (27384-1-AP), β-catenin (51067-2-AP), HRP-Goat Anti-Rabbit IgG (SA00001-2) were purchased from Proteintech Group, Inc (Rosemont, IL, USA). The primary antibodies against CKI-α (AF6660), Wnt1 (AF5315), P-GSK-3β (Ser9) (AF2016), GSK-3β (AF5016) were purchased from Affinity Bioscience (Liyang, China). Animals 70 SPF ICR mice, including male and female, weighed 18–22 g each, with a certificate number of 430,727,231,102,051,856. 60 male SPF ICR mice, weighing 18–22 g each, with a certificate number of 430,727,231,102,055,727, were obtained from Hunan Shrek Jingda Experimental Animal Co., Ltd (Changsha, China). 70 male SPF SD rats, weighing 70–90 g each, with a certificate number of 110,011,231,111,660,382, were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). Feeding Environment: sufficient food and water, temperature 18 ~ 26 °C, relative humidity 40%~70%, 8 ~ 12 h ventilation, 12 h light/dark cycle, adaptive feeding for 7 days. The methods utilized in this study were approved by the Ethics Committee of the Centre for Hunan University of Traditional Chinese Medicine laboratory animals (Ethical number: LL2023082402). Quality control of extract from R.molle HPLC-ELSD was used for detection (Agilent, 1200LC/Alltech, 3300), Agilent C[18] column (ZORBAX SB-C18, 4.6 mm×250 mm, 5 μm), column temperature 30 °C, sample injection volume 10 µL. Set to drift tube temperature 85 °C, gain 2, airflow 2.5 L/min. Based on network pharmacology and chemical recognition pattern analysis, we identified that Rhodojaponin-III, Rhodojaponin-II, Hyperin and Quercetin exhibited significant pharmacological activity, which could serve as active metabolites for RME. Consequently, this study quantitatively detected the four substances in RME, providing a solid material foundation for subsequent investigations into the RA pharmacodynamics of RME. Screening of targets for anti-rheumatoid arthritis of R.molle Using “rheumatoid arthritis” as the keyword, the disease gene data of RA were extracted from Genecards ([63]https://www.genecards.org), OMIM ([64]https://www.omim.org) and NCBI ([65]https://www.ncbi.nlm.nih.gov/gene) databases. Meanwhile, using “rheumatoid arthritis” and “homo sapiens” as keywords, the [66]GSE55457 [[67]20]dataset related to RA was downloaded from GEO database ([68]https://www.ncbi.nlm.nih.gov/geo/). All datasets were processed and analyzed using RStudio. The [69]GSE55457 dataset ([70]GPL96 platform) included transcriptomic analyses of 13 synovial samples from RA patients and 10 normal control synovial samples. Differential genes with folding changes greater than 1 and adjusted p-values less than 0.05 were screened. These genes were potential disease targets of RA, and KEGG enrichment analysis was carried out. The 2D structures and SMILES numbers of active metabolites of R.molle were obtained from PubChem ([71]https://pubchem.ncbi.nlm.nih.gov/). Key targets were obtained by Swiss Target Prediction ([72]http://www.swisstarg-etprediction.ch/) and PharmMapper ([73]http://www.lilabecust.cn/pharmmapper/) Prediction platforms. These main targets were uploaded to GeneMANIA ([74]https://genemania.org/) to predict the potential regulatory network associated with R.molle. The algorithm predicts potential targets that interact closely with the input gene [[75]21]. Uniprot ([76]https://www.uniprot.org/) queried for genes corresponding to the target, removed duplicate genes and non-human genes, and imported the screened target into a STRING ([77]https://www.stringdb.org/) to obtain the protein-protein interaction TSV information file. Then. we imported the file into Cytoscape 3.7.1 software to visualize the protein interaction network. KEGG pathway enrichment analysis was performed to identify signaling pathways significantly associated with R.molle drug targets. A network pharmacology diagram illustrating “active metabolites-targets-pathways-pharmacological effects-therapeutic efficacy” relationships was constructed. The core pathways of R.molle were then cross-mapped with RA related pathways to determine overlapping pathways, thereby elucidating its anti-inflammatory and immunomodulatory mechanisms of action in the treatment of RA. The topological analysis of the gene of R.molle networks was performed, with the top 10 key proteins identified using the cytoHubba plug-in, applying default parameters based on the EcCentricity method. Molecular Docking Structures of the key target proteins GSK-3B (PDBID: 1o6l) and Dvl1(PDBID: 6LCB) were retrieved and downloaded from PDB data ([78]https://www.rcsb.org). Water molecules and ligands of the target proteins were removed using PyMOL software (2.5.2). Then, the active metabolites (Hyperoside, Quercitrin, Rhodojaponin-III, Rhodojaponin-II) were sourced from PubChem database, and the docking parameters were set by AutoDockTools (1.5.6) for molecular docking between effective chemical constituents and key target proteins. In this context, lower binding energy values indicate more stable binding conformations and an increased likelihood of interaction between receptor molecules and ligands. A binding energy threshold of less than − 5 kcal·mol^− 1 was set, indicating significant binding activity between the key active metabolites and the core targets of the disease. The resulting docking conformations were visualized and refined using PyMOL software. The safe dosage anti-inflammatory effect of RME in vivo Acute toxicity test of RME in vivo An acute toxicity test was performed according to OECD guideline No. 420 [[79]22]. The experiment was carried out to observe the toxic reaction of RME, calculate its 50% lethal dose(LD[50]), and determine the safe concentration range for the drug. Subsequently, the LD[50] of RME on ICR mice was calculated by Bliss method. ICR mice were randomly divided into 7 groups with 10 mice in each group, six groups were administered varying doses of RME, while one group served as a blank control. The blank control group was given normal saline according to body weight, RME was administrated orally by using gavage in single doses of 470 mg/kg, 630 mg/kg, 840 mg/kg, 1130 mg/kg, 1500 mg/kg, 2000 mg/kg. Before administration, the mice were fasted without access to water. On the next day, the mice in each group were given the corresponding dose of the drug. The volume of administration was 20 ml/kg, and the mice could be fed 4 h after administration. The toxic symptoms and survival status of mice in each group were observed at 10:00 am and 3:00 pm daily for 14 days, the time of toxic reaction, the time of death and the number of deaths were recorded. We employed humane endpoint procedures for mice that had not naturally succumbed to the experimental conditions, euthanizing them through intraperitoneal injection of an overdose of pentobarbital sodium (100 mg/kg) and subsequently confirming their death. The anti-inflammatory experiment of RME The purpose of this study was to evaluate the anti-inflammatory effect of RME on xylene-induced auricle swelling in mice. 60 male ICR mice, weighing 18–22 g, were randomly divided into 6 groups, with 10 mice in each group. The groups included a model group, a positive drug group (Dexamethasone Acetate group): 7.50 mg/kg, and three RME treatment groups at varying doses: 38.14 mg/kg, 72.28 mg/kg, 152.55 mg/kg. The model group was perfused with 0.3% CMC-Na solution at a volume of 20 mg/kg, once a day for 6 days. One hour after the last administration, 20 µL xylene was smeared evenly on both sides of the right ear to induce inflammation. After 30 min, animals were euthanized by injection of excessive sodium pentobarbital (100 mg/kg; i.p.), and we subsequently confirmed their death. Both the left and right ears were excised at the ear base, a circular section of 8 mm in diameter was perforated from the same region of both ears (along the ear margin). The degree of swelling and the swelling inhibition rate were subsequently calculated by weighing the excised ear samples. The evaluation method was as follows [[80]23]: swelling (mg) = right ear weight-left ear weight Swelling inhibition rate = (Model group- RME treatment groups)/Model group × 100%. Anti-RA mechanism of RME Establishment of 1AIA rat model, administration and general observation After one week of adaptive feeding, we established the AIA rat model by injecting 0.1 mL complete Freund’s adjuvant (CFA) containing 3 mg/ml of Mycobacterium tuberculosis into the tail root of SD rats. On the 14th day, rats with AIA symptoms were randomly assigned to the Model group, the TWG positive control group (9.45 mg/kg) and the RME treatment group (38.14 mg/kg, 76.28 mg/kg,152.55 mg/kg). Meanwhile, the rats in the CON group were given the same volume of normal saline, while rats in the Model group were given the same volume of 0.3% CMC-Na. The treatment lasted for 30 d. Following the establishment of the model, the rats were weighed every 3 d, and the activity, the mental state, and the morbidity were observed every day. Measurement of toe swelling, AI and immune organ index The thickness of the rat hind toes was measured every three days at the same location, with both left and right toes serving as controls for inter-group comparisons. On the 3 d after modeling, the AI was scored with a 5-grade scoring system: 0 points for normal condition without redness and swelling; 1 point for mild redness and swelling of local joint or toe; 2 points for moderate swelling of the toe or ankle; 3 points for complete swelling of the feet below the ankle or moderate swelling of the ankle; 4 points for total swelling and deformation of the feet. The sum of the accumulated scores of the limbs before and after treatment is the AI of each rat, up to a maximum of 16 points per rat [[81]24]. Following treatment, the rats were weighed and euthanized, and the spleen and thymus were excised. After rinsing with normal saline, excess surface fluid was removed using gauze, and the wet weight of the organs was measured on an electronic balance. The organ index was calculated. Organ index = organ mass (mg)/rat weight (g). Enzyme-linked immunosorbent assay (ELISA) After the last administration, rats were anesthetized with an intraperitoneal injection of pentobarbital solution (50 mg/kg; i.p.). Blood samples were collected from abdominal aorta and centrifuged at 4 °C for 15 min at 3000 rpm to separate the upper layer serum. An appropriate amount of serum was taken and the experiment was carried out according to the instructions of the ELISA kit. Following the termination of the reaction, the absorbance was measured at a wavelength of 450 nm. The levels of pro-inflammatory factors TNF-α, IL-1β, IL-6 and IL-17 A were quantified using a standard curve. The pathological changes of rat ankle synovium were observed by HE staining We implemented humane endpoint measures on rats. We anesthetized them with an intraperitoneal injection of pentobarbital solution (100 mg/kg; i.p.) and confirmed their death. The ankle joints of rats were harvested. After removal of the excess tissue, the rat ankles were fixed with paraformaldehyde, decalcified, embedded in paraffin, and then sectioned. The sections were stained with hematoxylin-eosin (HE), and then sealed with a coverslip. Histopathological changes in the synovial tissue of the ankle joints from each group were observed under an electron microscope. Pathological scores (synovial hyperplasia, inflammatory cell infiltration, pannus formation) were made, and these scores ranged from 0 to 4. A score of 0 indicates the absence of synovial tissue hyperplasia, inflammatory cell infiltration, pannus formation, and normal joint space [[82]25]. Western blot was used to detect the expression of key proteins in synovium RIPA lysate, phosphatase inhibitor, protease inhibitor and PMSF were prepared based on the weight of the synovium. The mixture was added to homogenize the synovium, which was then lysed on ice for 30 min, and the supernatant was separated by centrifugation at 4 ℃, 12,000 rpm for 25 min. The protein concentration in the supernatant was determined by BCA. Proteins were denatured at 98 ℃ for 10 min and stored at -20 ℃. Separation of proteins was performed by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), the protein was transferred to PVDF membrane after 1 h of wet rotation at 250 mA. The membrane was blocked, washed with TBST, and incubated with primary antibodies (β-actin, 1:10000; Dvl1, 1:2000; Wnt1, 1:1000; β-catenin, 1:10000; CKI-α, 1:1000; APC, 1:1000; GSK-3β, 1:1000; P-GSK-3β(Ser9), 1:1000) overnight at 4 ℃. The membrane was then washed with TBST, and diluted secondary antibodies(HRP-goat anti-rabbit IgG, 1:10000) were added. After being washed with TBST, the mixture of ECL luminescent developer A and B was added and incubated for 2–3 min. The Tanon imaging system was used to develop the film, and the band gray values were analyzed by ImageJ software. Statistical analysis All data are entered into SPSS 25 software for processing. One-way analysis of variance (ANOVA) was employed when the measurement data obeyed the normal distribution with homogeneous variance, utilizing the LSD method for post-hoc comparisons. For data that did not meet the normality or homogeneity assumptions, a non-parametric test was performed, specifically the Kruskal-Wallis H test for group comparisons. P < 0.05 was considered statistically significant. Result Quality control of extract from R.molle Conducted quality control on RME to ensure stable content of active metabolites in R.moll, thereby guaranteeing its safety and effectiveness. In the early stages of our research, we conducted quantitative analysis of 4 active metabolites in R.moll samples collected from different flowering stages and origins, and established a fingerprint profile of R.moll for the first time [[83]12]. In this study, the linear relationships of Rhodojaponin-III, Rhodojaponin-II, Hyperin and Quercetin were determined by HPLC-ELSD, as shown in Table [84]1, while the liquid chromatography profile of RME was illustrated in Fig. [85]1. According to the linear relationship, the contents of four active metabolites in RME were 0.866%, 1.038%, 0.996% and 1.661% respectively. These active metabolites exhibit significant anti-inflammatory and analgesic activities, and play important regulatory roles in the body. Table 1. Regression equation, correlation coefficient and linear range of 4 active metabolites in RME Active metabolites Regression equation r ^2 Range of linearity / mg/mL Rhodojaponin-III y = 1.792x + 1.680 0.9993 0.2963 ~ 1.1000 Rhodojaponin-II y = 1.465x + 1.734 0.9995 0.4882 ~ 1.0000 Hyperin y = 2.890x + 0.775 0.9994 0.4882 ~ 1.0000 Quercitrin y = 1.237x + 0.739 0.9998 0.2634 ~ 1.3333 [86]Open in a new tab Fig. 1. [87]Fig. 1 [88]Open in a new tab RME HPLC-ELSD Chromatogram. A) Mixed standards. B) RME liquid chromatogram figure. (1) Rhodojaponin-III; (2) Hyperin; (3) Quercetin; (4) Rhodojaponin-II Screening of RA genes We obtained 6786, 1535 and 198 RA-related genes from Genecards, OMIM and NCBI databases. By combining the genes obtained from the three databases, we determined thatthe final number of RA disease-related genes was 6842 (Fig. [89]2A). A human gene dataset related to RA, [90]GSE55457, was downloaded from the GEO database. Synovial samples of 13 RA patients were taken as the “Model group”, while 10 normal human synovial samples were taken as the “Normal group” for transcriptome analysis. A total of 1108 differentially expressed genes were obtained, with 585 genes up-regulated and 523 genes down-regulated. The up-regulated genes included UBD, CXCL9, CXCL10, CCDC88C, TNFSF10, etc., and ZBTB7C, ADAM15, TMEM259, KRTAP5-8 were down-regulated genes (Fig. [91]2B-C). By integrating the database and differential genes, 378 RA-related genes were finally obtained (Fig. [92]2D), and KEGG enrichment analysis was performed on 378 genes (Fig. [93]2E), and the results showed that the enrichment pathways mainly involved JAK-STAT, Wnt, Th1 and Th2 cell differentiation, TLR signal pathways. Fig. 2. [94]Fig. 2 [95]Open in a new tab Analysis of disease targets and differential genes in RA. A) RA database target intersection Venn plot. B) Differential gene volcano plot, with red indicating up-regulated genes, green indicates down-regulated genes. C) Differential gene expression heat map. D) Intersection Venn map of RA related genes and [96]GSE55457 differential genes obtained from the database. E) KEGG enrichment analysis of 378 intersecting genes Target selection, PPI network, “active metabolites-target-pathway-pharmacological effect-efficacy” network construction of R.molle By searching the SWISS TargetPrediction and PharmMapper databases, we obtained 84 chemical targets of the active metabolites of R.molle. We then uploaded these targets to the GeneMANIA database to predict their indirect regulatory proteins, and as a result, 20 additional indirect targets were identified, and the PPI network diagram of R.molle was constructed by Cytoscape software (Fig. [97]3A). To further analyze the function of the network, KEGG enrichment analysis was performed on these 104 genes (Fig. [98]3B), and the results show that enrichment pathways include VEGF, PI3K-Akt, HIF-1 and Wnt signaling pathway, all of which have been shown to be closely associated with the development of RA [[99]26]. The network of “active metabolites-target-pathway” and “pathway-pharmacological effect-efficacy” was constructed (Fig. [100]3C-D), which showed that R.molle had the characteristics of multi-target and multi-pathway in the treatment of diseases. R.molle played the role of expelling wind and removing dampness, dispersing blood stasis and relieving pain through anti-inflammatory, immune regulation and other pharmacological effects. Fig. 3. [101]Fig. 3 [102]Open in a new tab PPI network and enrichment analysis of R.molle related targets. A) PPI network of R.molle active metabolites related target proteins. B) KEGG enrichment analysis of 104 target genes. C) “Active metabolites-target-pathway” network diagram of R.molle. D) “Pathway-pharmacological effect-efficacy” network diagram of R.molle Key pathway acquisition and key target topology analysis of R.molle in the treatment of RA The main pathways related to R.molle were mapped with RA-related pathways, and it can be concluded that JAK-STAT, Wnt, and TNF signaling pathways were intersecting pathways (Fig. [103]4A), which were potential pathways for R.molle to treat RA. Through topological analysis of the network of 104 genes, the top 10 primary genes (Fig. [104]4B) were screened, including GSK-3β, Dvl1, CSNK2A1, etc. GSK-3β and Dvl1 are key proteins in the Wnt pathway, and they play important regulatory roles in the signal transduction process of the pathway. Fig. 4. [105]Fig. 4 [106]Open in a new tab Analysis of key pathways and core targets of R.molle in the treatment of RA. A) Key pathways of R.molle in the treatment of RA. B) Topological analysis of core targets of R.molle Molecular Docking of core targets To verify the binding affinities of the active metabolites of R.molle, namely Rhodojaponin-III, Rhodojaponin-II, Hyperoside and Quercitrin, with the core targets GSK-3β and Dvl1 in the Wnt pathway, molecular docking was performed between these metabolites and the core targets. The docking scoring values can be seen in Table [107]2, and the docking results are shown in Fig. [108]5. Notably, the docking binding energies of all 4 active metabolites with each of the core targets were observed to be below − 5 kcal/mol, indicating a favorable docking interaction. Table 2. Molecular Docking scoring values Protein PDB ID Active metabolites Docking energy Dvl1 6lcb Rhodojaponin-III -7.6 Rhodojaponin⁃II -7.8 Hyperoside -6.8 Quercitrin -7.3 GSK-3β 1o6l Rhodojaponin-III -8.0 Rhodojaponin⁃II -8.3 Hyperoside -9.3 Quercitrin -9.5 [109]Open in a new tab Fig. 5. [110]Fig. 5 [111]Open in a new tab Molecular docking results of 4 active metabolites of R.molle with core targets. A) Dvl1 (PDBID:6lcb). B) GSK-3β (PDBID:1o6l) Study on anti-inflammatory properties of R.molle extract in vivo Acute toxicity test in mice To assess the potential of R.molle as a treatment for RA, it is essential to explore its safe dosage to ensure the safety in the treatment of RA. The results of acute toxicity test for RME are presented in Table [112]3. The LD[50] value of RME in mice was determined to be 1340 g/kg using the Bliss method. Mice that exhibited poisoning showed the disappearance of righting reflex, spasm and decreased spontaneous activity, with no significant difference in poisoning symptoms between male and female mice. Those that survived were able to regain normal activity after 24 h. Based on the LD[50] value for RME, the estimated 50% lethal dose for a 60 kg human would be approximately 8.83 g, which corresponds to about 31.27 g of crude R.molle, approximately 20 times the clinically safe dose of 1.5 g of crude drug for a 60 kg human. After normal feeding for 14 d, the non-deceased mice were euthanized. Autopsy results indicated that there were no significant changes in the organs unrelated to the therapeutic effects. The dosages of RME utilized in the subsequent studies fell within the safe range. Table 3. Acute toxicity test results of R.molle extract (n = 10) Dose administered (mg/kg) The number of deaths State of death 2000 10 Died 5 min after administration 1500 4 Died 7 min after administration 1130 3 Died 9 min after administration 840 1 Died 12 min after administration 630 1 Died 15 min after administration 470 0 The mice were depressed and unresponsive within 3 h after administration, and recovered gradually within 5 h, and returned to their pre-administration state after 24 h 0 0 Normal [113]Open in a new tab The anti-inflammatory effect of RME The results of RME anti-acute inflammation experiment are shown in Table [114]4; Fig. [115]6. Compared to the Model group, the middle dose RME and high dose RME treatment significantly reduced xylene-induced auricle swelling in mice (P < 0.01 or P < 0.001). These findings suggest that RME may inhibit exudation and edema during the early stage of inflammation. Table 4. Effect of RME on xylene-induced auricle swelling in mice (n ± s, n = 10) Group Dose/(mg/kg) Swelling/mg Swelling inhibition rate /% Model group — 16.30 ± 7.25 — DXMS group 7.50 5.70 ± 2.32^*** 63.81 RME group 55.09 12.20 ± 3.78 25.15 110.18 11.40 ± 3.53^** 27.62 220.35 6.70 ± 4.71^*** 58.90 [116]Open in a new tab Note: ^**P < 0.01, ^***P < 0.001 compared with Model group Fig. 6. Fig. 6 [117]Open in a new tab The effect of RME on acute ear swelling in mice. All data presented as mean ± SEM (n = 10). ^**P < 0.01, ^***P < 0.001 compared with the Model group Therapeutic effect of RME on AIA rats General observation To evaluate the therapeutic effects of RME on AIA rats, the experimental design is depicted in Fig. [118]7A. The changes of body weight of rats before and after treatment are presented in Fig. [119]7B. Compared to the CON group, the body weight of the rats in the Model group was significantly decreased (P < 0.001).In contrast, the body weights of the rats in the TWG group, the RME-M and the RME-H group were significantly increased compared to the Model group (P < 0.001).However, no significant difference was observed between the RME-L group and Model group (P > 0.05). In the Model and RME-L groups, the rats exhibited yellow, dull fur, stiff tails with severe ulceration, and reduced activity, while the RME-M and RME-H groups showed varying degrees of alleviation of these symptoms, demonstrating a dose-dependent response. Fig. 7. [120]Fig. 7 [121]Open in a new tab Effects of RME on body weight changes in AIA rats. A) Overview of the experimental timeline. B) Changes in body weight of rats. All data are expressed as mean ± SEM (n = 10). Compared with the CON group ^***P < 0.001, compared with the Model group ^###P < 0.001 RME can reduce the severity of AIA rats 9 d after the induction of AIA, the rats began to exhibit erythema or redness in their feet, with peak symptoms observed at 14 d after induction. The Model group displayed significantly swollen and deformed hind feet, stiff tail, and ulcerated skin (Fig. [122]8A). Compared with the CON group, the Model group showed a significant increase in the AI score (Fig. [123]8B), toe thickness (Fig. [124]8C), thymus index (Fig. [125]8D) and spleen index (Fig. [126]8E) (P < 0.001). Following treatment, the deformity of the hind paws was reduced, and the AI score, toe thickness, thymus index, and spleen index all significantly decreased (P < 0.001). Notably, different doses of RME demonstrated a dose-dependent therapeutic effect, with the RME-H group exhibiting results comparable to those of the TWG Group. Fig. 8. [127]Fig. 8 [128]Open in a new tab RME can obviously relieve AIA rat symptoms. A) Typical figure of the rat hind paw. B-C) The arthritis index of hind paw and the thickness of toe at different time points. D) Spleen index. E) Thymus index. All data are expressed as mean ± SEM (n = 10). ^***P < 0.001 compared with the CON group, ^##P < 0.01, ^###P < 0.001 compared with the Model group The effect of RME on the pathological changes of the ankle joint of AIA rats The HE staining results of the ankle joints of the rats in each group were shown in Fig. [129]9A. In the CON group, the ankle joint structure was intact; the articular cartilage surface was smooth and even, with synovial cells present in 1–2 layers. In contrast, the Model group exhibited typical pathological features associated with RA, including synovial cell proliferation, infiltration of inflammatory cells, pannus formation, severe destruction of bone and joint. The severity of these changes was significantly different from that in the CON group (P < 0.001), as shown in Fig. [130]9B. Pathological changes in the ankle joint of the TWG and RME group were significantly improved compared to the Model group, with a dose-dependent response observed. The number of synovial cell layers in the TWG and RME-H group was comparable to the normal level, and the histopathological scores in these treatment groups were significantly different from those in the Model group (P < 0.001). Based on these results, it is suggested that R.molle has a pronounced protective effect on bone and joint. Fig. 9. [131]Fig. 9 [132]Open in a new tab RME can alleviate ankle injury in AIA rats. A) HE-stained micrographs of typical ankle sections, with red arrows indicating joint lumen, black arrows indicating synovial hyperplasia, blue arrows indicating inflammatory response, and yellow arrows indicating pannus, scale bar = 100 μm. B) Evaluation of pathological changes (Synovial hyperplasia, Inflammatory cell infiltration, Pannus Formation). All data are expressed as mean ± SEM (n = 3). ^##P < 0.01, ^###P < 0.001 compared with the Model group The effect of RME on inflammatory cytokines in AIA rats The effect of RME on serum cytokine levels in rats from each group was presented in Fig. [133]10. Compared with the CON group, the serum levels of TNF-α, IL-1β, IL-6 and IL-17 A were significantly elevated in the Model group (P < 0.01). Treatment with RME (38.14, 76.28, 152.55 mg/kg) could significantly reduce the levels of pro-inflammatory cytokines in a dose-dependent manner (P < 0.01). These results indicate that RME can inhibit the abnormal proliferation of synovial cells, prevent bone destruction, and protect bone and joint by decreasing the levels of inflammatory factors, including TNF-α, IL-1β, IL-6 and IL-17 A. Fig. 10. [134]Fig. 10 [135]Open in a new tab Effect of RME on inflammatory cytokines in serum of rats. All data are expressed as mean ± SEM (n = 5). ^**P < 0.01 compared with the CON group, ^##P < 0.01 compared with the Model group The effect of RME on Dvl1/Wnt1/β-catenin/GSK-3β related pathway protein in AIA rats The effect of RME on expression of vital proteins in Dvl1/Wnt1/β-catenin/GSK-3β pathway in synovial tissues of rats from each group is illustrated in Fig. [136]11. Compared with the CON group, the Dvl1 mediated Wnt1/β-catenin/GSK-3β signaling pathway was abnormally activated in the synovium of the Model group. The expression levels of Dvl1, Wnt1, β-catenin, CK1-α, APC, p-GSK-3β(SER9) proteins were significantly elevated in the Model group (P < 0.001). RME treatment inhibited the expression of key proteins in the Dvl1/Wnt1/β-catenin/GSK-3β pathway in a dose-dependent manner, with significantly reduced expression levels observed in the RME-M and RME-H groups compared to the Model group (P < 0.01, P < 0.001). The results indicate that RME can inhibit the abnormal activation of the Dvl1/Wnt1/β-catenin/GSK-3β signaling pathway in synovial tissue of rats, thereby offering protective effects against rheumatoid arthritis. Fig. 11. [137]Fig. 11 [138]Open in a new tab RME inhibited the expression of key proteins of Dvl1/Wnt1/β-catenin/GSK-3β signaling pathway. All data are expressed as mean ± SEM (n = 3). ^***P < 0.001compared with the CON group, ^#P < 0.05, ^##P < 0.01, ^###P < 0.001 compared with the Model group Discussion Rheumatoid arthritis is a chronic systemic autoimmune disease characterized by high morbidity and disability [[139]27]. The etiology and pathogenesis of RA are extremely complex, and effective treatment remains elusive [[140]26]. In recent years, an increasing number of botanical drugs have demonstrated superior anti-RA activity and are now widely used, both as primary treatments and as complementary therapies, notable examples include tripterygium glycosides tablets[[141]28]. The botanical drug R.molle shows significant promise due to its comprehensive benefits and potential clinical applications. Previous studies have indicated that the root extract of R.molle can appreciably restrain histamine-induced joint swelling and subcutaneous inflammation in mice [[142]29], and possesses strong anti-inflammatory and analgesic activities [[143]30]. However, R.molle has certain toxicity, which can be viewed as a reflection of its high efficacy. The concept of “toxicity-effect duality” is the characteristic feature of botanical drugs. Within an appropriate dosage, they exhibit “efficacy”, whereas excessive use leads to “toxicity”. By using processing techniques and combining it with other botanical drugs, the excessive corrective and regulatory effects of R.molle can be mitigated, thereby maximizing its therapeutic efficacy while minimizing its toxic potential [[144]11]. Therefore, we first explored the safe and effective dose of RME through in vivo experiments, revealing that high-dose RME had certain toxicity and the oral LD[50] of RME was 1340 g/kg BW, which is classified as “low toxicity” in the Globally Harmonized System [[145]22]. Notably, the blood toxicity and reproductive toxicity and other adverse reactions are lower than tripterygium glycosides tablets [[146]29], indicating promising application potential. Based on the determination of the safe dosage of R.molle, it is also of great significance to investigate the mechanisms underlying its anti-RA effects. Based on the research and screening conducted by our research group, it was found that Rhodojaponin-III, Rhodojaponin-II, Hyperin and Quercetin are the primary active metabolites responsible for the therapeutic effects of R.molle [[147]12]. In this study, HPLC-ELSD was employed for quality control of these 4 metabolites in RME to ensure its safety and effectiveness. Due to the complexity of components and the diversity of action pathways in botanical drugs, this study employed network analysis to mine the targets of R.molle against RA from public databases. The results revealed that its active metabolites are closely related to RA treatment. Previous studies have stated that Rhodojaponin-III and Rhodojaponin-II are the essential diterpenoids of R.molle, and their bioavailability and biotransformation are much higher than those of other diterpenoids [[148]31]. They exert anti-inflammatory role by regulating Wnt1 [[149]32]and AKT/NF-κB [[150]33]signaling pathways to diminish the secretion of inflammatory cytokines. The flavonoids Hyperoside and Quercetin can downregulate the AKT/NF-κB [[151]34]and PI3K-AKT [[152]35]signaling pathways to restrain the release of pro-inflammatory cytokines, and have remarkable anti-inflammatory effects. Through network analysis, a total of 108 potential targets and 378 RA-related genes were obtained. KEGG enrichment analysis indicated that the Wnt, JAK-STAT, TNF signaling pathways may serve as critical pathways for the therapeutic effects of R. molle on RA. Some research has confirmed that R.molle can indeed ameliorate RA by regulating JAK-STAT, TNF, TLR4/NF-κB and AKT signaling pathways [[153]16, [154]36]. KEGG enrichment analysis revealed that a large number of genes were enriched in the Wnt signaling pathway, and Wnt1 was the main subtype of expression in synovial tissue [[155]37]. The Wnt1/β-catenin signaling pathway is a highly conserved signaling pathway in organisms, which is closely related to cell differentiation, proliferation and programmed cell death [[156]38]. Dysregulation of Wnt1/β-catenin signaling pathway will accelerate the development of cancer, bone metabolism and other diseases [[157]39]. However, the Wnt1/β-catenin signaling pathway has been poorly studied in the field of RA. Notably, this pathway may influence FLS proliferation or OB/OC balance in RA pathology in a variety of ways [[158]40–[159]42]. Synovitis is a major pathological feature of RA and the primary cause of irreversible damage to bone and articular cartilage [[160]43]. To further explore the effect of RME on RA and its potential molecular mechanism, we preliminarily evaluated its anti-inflammatory effects using the xylene-induced mouse ear swelling model. The results demonstrated that RME inhibited mouse ear swelling in a dose-dependent manner, suggesting substantial anti-inflammatory properties. Then, HE staining revealed that the synovium of rats induced by AIA was severely destroyed, inflammatory cells infiltrated into the synovium with pannus formation. After RME treatment, the number of cell layer in the synovium returned to normal level, alongside reduced angiogenesis and alleviated bone erosion, indicating the anti-inflammatory effects and demonstrating a promising potential against RA. Topological analysis revealed that GSK-3β, Dvl1, and other proteins play crucial roles in the target network. Dvl1, an essential component of the Wnt1 receptor complex-related membrane protein, acts as a key regulatory point for Wnt1-induced downstream signaling pathways [[161]37]. Interfering with the interaction between Dvl1 and β-catenin can inhibit the Wnt1 pathway. GSK-3β is a key downstream target of the Wnt signaling pathway, and silencing Dvl1 can down-regulate GSK-3β/β-catenin signal transduction [[162]44]. Molecular docking studies demonstrated that Rhodojaponin-III, Rhodojaponin-II, Hyperin and Quercetin exhibited strong docking affinity with Dvl1 and GSK-3β, identifying them as effective targets of R.molle for RA treatment. Our study confirmed that RME could down-regulate the expression of Dvl1 and p-GSK-3β(Ser9) in the synovium of AIA rats. Additionally, compared to the Model group, the expression of Wnt1 and β-catenin in the synovium of AIA rats treated with RME was significantly decreased. RME administration also led to a downregulation of key proteins such as CK1-α and APC. CK1-α, which plays an active role in modulating the Wnt1 signaling pathway as a molecular switch within this pathway, was among the down - regulated proteins.The functional deficiency of CK1-α impacts the activity of β-catenin in the cytoplasm [[163]45]. APC is a multifunctional protein that can specifically bind to β-catenin to regulate the Wnt1 signaling pathway [[164]46]. The above results indicate that the Wnt1/β-catenin signaling pathway may be a potential new target for the treatment of RA in R.molle. Upon Down- regulation of the Wnt1 pathway, β-catenin is phosphorylated and degraded by the protein complex formed by GSK-3β, APC and Axin [[165]47]. The inactivation of β-catenin following degradation affects the binding of T-cell factor (TCF)/lymphoid enhancer binding factor (LEF), thereby impacting the transcription and translation processes of downstream target genes such as TNF-α and IL-1β [[166]41, [167]48, [168]49]. This, in turn, exacerbates swelling and cartilage and bone damage in RA. The results demonstrated that varying doses of RME reduced serum levels of TNF-α, IL-1β, IL-6 and IL-17 A in AIA rats, significantly alleviating toe swelling and lowering the AI. In addition, RME administration was associated with a notable reduction in immune organ indices, such as thymus and spleen, suggesting that RME can effectively alleviate immune dysfunction triggered by RA and inhibit abnormal growth or hyperplasia of immune organs [[169]50]. It is suggested that RME can inhibit the secretion of pro-inflammatory factors by disrupting the activation of the Dvl1/Wnt1/β-catenin/GSK-3β signaling pathway, thereby protecting bone and joints and potentially improving RA outcomes. (Fig. [170]12). Fig. 12. [171]Fig. 12 [172]Open in a new tab The mechanism of RME inhibiting the secretion of inflammatory factors by regulating the Dvl1/Wnt1/β-catenin/GSK-3β signaling pathway in the treatment of RA. (FZD: frizzled protein, LRP: low density lipoprotein-related receptor, TCF: T cytokine, LEF: lymphocyte enhancing factor) This project demonstrates that RME can attenuate the inflammatory response in RA by modulating the Dvl1/Wnt1/β-catenin/GSK-3β signaling pathway. Due to potential data noise or incomplete data in network analysis, coupled with the complexity of the metabolic network of botanical drugs, this study focuses on the more established possibilities within network analysis and discusses the most representative core targets and pathways, thereby laying a foundation for subsequent in-depth research. This study preliminarily elucidated the active metabolites cluster of R.molle and its partial mechanisms of action in RA treatment. However, potential additional active metabolites and their synergistic regulatory networks require further investigation. Subsequent research will integrate untargeted metabolomics and LC-MS/MS to analyze the dynamic distribution patterns of components absorbed into the blood and their metabolites. Through in vitro (e.g., LPS induced RAW264.7 cell inflammation model) and in vivo (CIA rat model) experiments, we will systematically evaluate the therapeutic differences between single active components (e.g., Rhodojaponin-III) and multi-component combinations. By incorporating molecular docking and gene knockout technologies, we will specifically validate the regulatory effects on core targets within key pathways such as Dvl1/Wnt1/β-catenin/GSK-3β, thereby comprehensively clarifying the multi-component, multi-target, and multi-pathway mechanism of R. molle in RA treatment. Conclusion In conclusion, through network analysis and experimental validation, this study predicted and analyzed the potential targets of R.molle for the treatment of RA. The results indicated that the active metabolites in R.molle can alleviate RA through multiple targets and pathways. In vivo studies have shown that R.molle can inhibit synovial hyperplasia, inflammatory cell infiltration and pannus formation in rats through the Dvl1/Wnt1/β-catenin/GSK-3β signaling pathway, thereby alleviating the inflammatory response in rats and maintaining the balance between osteocytes and articular cartilage to achieve the therapeutic effect on RA. Electronic supplementary material Below is the link to the electronic supplementary material. [173]Supplementary Material 1^ (324.1KB, pdf) [174]Supplementary Material 2^ (13MB, zip) [175]Supplementary Material 3^ (87KB, pdf) Acknowledgements