Biqi Capsules were provided by Tianjin Darentang Pharmaceutical Jingwanhong Co. Ltd. (Tianjin, China) with the production batch number 31174066. The quality of Biqi Capsule is supervised by the National Medical Products Administration (NMPA) according to the Pharmacopoeia of the People’s Republic of China 2020 Edition, Part I, Page 1808. Papain (76,220), poloxamer (P4894), Lipopolysaccharide (LPS, L2880), and Dimethyl sulfoxide (DMSO, D8418) were purchased from Sigma-Aldrich (Shanghai, China). LY294002 was purchased from MedChemExpress (NJ, USA). Strychnine, Brucine, Liquiritin, Salvianolic Acid B, Glycyrrhizic Acid, Cryptotanshinone, and Tanshinone ⅡA were purchased from Sichuan Vicky Biotechnology Co., Ltd. (Chengdu, China). The primary antibodies against PI3K p85 (#4257), Akt (#9272), p-Akt (#9275), NF-κB p65 (#8242), p-NF-κB p65 (#3033), PCNA (#13110), the second antibodies against rabbit (#7074) and mouse (#7076) IgG were purchased from Cell Signaling Technology (MA, USA). The primary antibodies against p-PI3K p85 (ab182651) and β-actin (ab8226) were purchased from Abcam (Cambridge, UK). The primary antibodies against mTOR (A11355) and p-mTOR (AP0978) were purchased from ABclonal Technology (Wuhan, China). Rat PGE2 Elisa Kit (LCSJZF30697) was purchased from LunChangShuo Biotech. (Xiamen, China); Rat IL-6 Elisa Kit (E04640r) was purchased from CUSABIO (Wuhan, China); Rat TNF-α Elisa Kit (KRC3011), Rat IL-1β Elisa Kit (BMS630) and mouse IL-6 ELISA Kit (88-7064-88) were purchased from Invitrogen (CA, USA).

Healthy male SD rats (n = 60, body weight: 160g–180 g) were purchased from SPF Biotechnology Co., Ltd. (Beijing, China). All experimental procedures were approved by the Animal Care and Use Committee of the Tianjin University of Traditional Chinese Medicine (Authorization number: TCM-LAEC2022129). All animals were kept at 22°C–26 °C and 55% ± 5% humidity with a 12 h light/dark cycle and allowed free access to food and water during the experiments. After a week of adjustment, all of the rats were randomly divided into five groups as follows: Vehicle group (equal normal saline solution, i. g.), OA group (equal normal saline solution, i. g.), BQL group (Biqi Capsule low dose, 0.648 g/kg·d, i. g., double clinical dose), BQH group (Biqi Capsule high dose, 1.296 g/kg·d, i. g.), Celecoxib group (CB, as a positive control, 24 mg/kg·d, i. g.). The rat model of OA was established using the invention patent method of the author’s research group, which has been authorized by the China National Intellectual Property Administration (Approval number: CN102058877A). Previous studies have demonstrated that this model successfully simulates the pathophysiological processes of OA in humans and is suitable for pharmacological studies and efficacy evaluation of anti-osteoarthritic drugs (Zhanbiao et al., 2024). Briefly, the model of OA was induced by a sustained-release papain agent (4% papain and poloxamer). The sustained-release papain agent was injected into both knee joint compartments of rats anesthetized with pentobarbital and then monitored for 1.5 h at 50°C (Figure 1A).

For the in vivo study, rats had undergone a single intra-articular injection with 50μL/knee joint sustained-release papain agent. The day after the establishment of the OA model, BQL and BQH administration groups received the indicated dosage of Biqi Capsule content powder (dissolved in double distilled water) by oral gavage daily for seven consecutive days. Meanwhile, rats in the OA and CB groups were administered water and celecoxib, respectively. The dose was administered as described above.

For the in vitro study, a solution from Biqi Capsule extract was prepared as follows (Wang et al., 2011): 10 g of Biqi Capsule content powder was mixed with 100 mL double-distilled water, extracted by ultrasonication twice at 37 °C for 1 h each. The extraction solution was centrifuged at 12,000rpm for 30 min and then freeze-dried to obtain the aqueous extract of Biqi Capsule. The extract was dissolved in a medium before use and filtered through a 0.22 µm pore size membrane filter (Millipore, USA).

Rats were euthanized on the eighth day to obtain the knee joints. The knee joints were fixed in 4% paraformaldehyde solution, decalcified with 1% nitric acid for 48h, dehydrated in gradient ethanol and xylene, and embedded in paraffin wax. Knee joint sections (4 μm) were stained with hematoxylin and eosin (HE) staining, and the cartilage abnormalities were evaluated by the modified Mankin score (Mankin and Lippiello, 1970; Mankin et al., 1971; Mankin et al., 1981; Takahashi et al., 2019; Sun et al., 2020).

Synovial tissues were rinsed with pre-chilled PBS, the blood was removed, and the tissues were dried with filter paper. The synovial tissues were weighed accurately and ground in a homogenizer (Servicebio, Wuhan, China) by adding 0.1 mg/μL PBS. The prepared tissue homogenate was centrifuged at 4 °C and 12,000 rpm for 15 min, and the supernatant was retained. The levels of PGE2, IL-6, IL-1β, and TNF-α were determined by ELISA according to the manufacturer’s instructions.

Total protein from synovial tissue/cells was collected with RIPA lysis containing Phosphatase Inhibitor Cocktail and Protease Inhibitor Cocktail (Yeasen), then ground in a homogenizer. Concentrations of protein were determined by the BCA Protein Assay Reagent Kit. 10 μg protein was separated by SDS-PAGE electrophoresis and transferred to PVDF membranes (Merck, NJ, USA). After blocking with 5% fresh nonfat milk in tris-buffered saline containing 0.05% Tween-20, the membranes were incubated with primary antibody overnight at 4 °C, followed by HRP-conjugated secondary antibody incubation. The membranes were incubated with Chemiluminescent HRP Substrate (Merck) for 1 min, then images were developed through Amersham Imager 600 (GE, MA, USA). The relative expression was quantified by ImageJ software.

RAW264.7 and SW1353 cells were purchased from Procell Life Science and Technology Co., Ltd (Wuhan, China). Passage numbers of all cells for the experiment were 5–10. The RAW264.7 cells were cultured in high-glucose Dulbecco’s Modified Eagle medium (DMEM; Gibco, MA, USA) containing 10% fetal bovine serum (FBS; Gibco), penicillin 100 U/mL (Gibco) and streptomycin 100 μg/mL (Gibco) at 37 °C in humidified conditions with 5% CO₂. SW1353 cells were cultured in Leibovitz’s L-15 medium (Procell) containing 10% FBS, penicillin 100 U/mL, and streptomycin 100 μg/mL at 37°C in humidified conditions with no CO₂.

The human chondrosarcoma-derived cell line SW1353 is a promising substitute for a chondrocyte experimental model. It produces sufficient proliferative activity and presents a consistent response to the phenotype of primary human chondrocytes. H₂O₂ stimulation induced apoptosis and necrosis, decreased cell viability, and increased eight-isoprostane F2-α expression in SW1353 cells. It is often used as a model for osteoarthritis studies (Park et al., 2018; Pang et al., 2021). For in vitro studies, H₂O₂-stimulated SW1353 cells were used as the model for chondrocyte damage, and the reversal of cell damage by drugs of interest was investigated using the CCK8 kit (Yeasen, Shanghai, China). Specifically, SW1353 cells were cultured in a complete medium containing 260 μmol/L H₂O₂. H₂O₂ stimulation and drug administration were performed simultaneously. A total of Biqi Capsule extract (10, 50, 100 μg/mL) as well as seven compounds including Strychnine, Brucine, Liquiritin, Salvianolic Acid B, Glycyrrhizic Acid, Cryptotanshinone, Tanshinone ⅡA (all concentrations are 0.2, 1, 5 μmol/L) were examined in this study. SW1353 cells were seeded with the density of 5×10³ cells per well in 96 well plates and incubated for 24 h before administration. The drugs were first dissolved in DMSO and then diluted with medium to the desired concentration while ensuring that the concentration of DMSO was less than 0.5‰. The DMSO was also supplemented in the Vehicle and H₂O₂ groups so that the DMSO concentration was the same in each group. The preparations were incubated for 12 h after administration and a CCK8 test was performed according to the manufacturer’s instructions.

For in vitro studies, LPS-stimulated RAW264.7 cells were used as an inflammatory model for the screening of effective anti-inflammatory drugs. LPS stimulation and drug administration were performed simultaneously. RAW264.7 cells were seeded with the density of 1×10⁴ cells per well in 96 well plates; after 24 h incubation, the culture medium was replaced with serum-free medium for another 12 h. According to the preliminary results from our pre-experiments, the cells were treated with LPS (500 ng/mL) and Biqi Capsule extract (10, 50, 100 μg/mL), Strychnine (0.1, 1, 10 μmol/L), Brucine (0.1, 1, 10 μmol/L), Liquiritin (0.1, 1, 10 μmol/L), Salvianolic Acid B (0.1, 1, 10 μmol/L), Glycyrrhizic Acid (0.1, 1, 10 μmol/L), Cryptotanshinone (0.1, 1, 10 μmol/L), Tanshinone ⅡA (0.1, 1, 10 μmol/L) for 12 h. Next, the levels of IL-6 in the supernatant were determined by ELISA according to the manufacturer’s instructions.

The receptor is the crystal structure of RAC1, which is one of the critical targets according to the PPI network. The structure was obtained from the RCSB protein data bank, with the corresponding PDB code 1RYF. The ligands are Glycyrrhizic acid and Liquirtin, the structure files of which were obtained from PubChem.

Discovery Studio is a highly visual commercial software integrated by BIOVIA for life science research and used to perform molecular docking to investigate the interaction patterns between active compounds and critical targets. The receptor and ligands were processed by the prepare protein and prepare ligands modules, respectively, and then applied CHARMm force field on all ligands. The coordinates of the receptor active site were set to −1.23843, 73.5704, 38.9122, and a radius of 11. Parameters such as CDocker energy and CDocker interaction energy were used to evaluate the molecular docking results. The conformation with the highest CDocker interaction energy score value was selected as the most reliable binding conformation for further analysis.

Studies were designed to generate groups of equal size using randomization and blinded analysis. The data was presented as mean ± SEM of n measurements. N identifies the number of independent samples. Data were tested for normality using the Shapiro–Wilk test. Significant differences were assessed by either Student’s t-test or one-way ANOVA, according to the number of groups compared. When significant variations were found by one-way ANOVA, the Tukey-Kramer post-hoc test for multiple comparisons was performed only if F achieved a p-value <0.05. All the statistical analyses were done using GraphPad Prism 9.5. Differences were considered significant at p < 0.05.

We assessed the efficacy of Biqi Capsule on OA using a papain-induced OA model. Firstly, no changes were observed in the rate of body weight change in Biqi Capsule treated or OA rats when compared with the Vehicle group during the 7 days (Figure 1B), which partly indicated that the Biqi Capsule treatment exerted no apparent toxic effects on OA rats. HE staining of the tissues from the knee joint of papain-treated rats showed apparent cartilage damage, mainly manifesting as localized greyish-white cartilage surface with no luster; the cartilage migration layer matrix was edematous with light red and uneven staining. Some chondrocytes showed degeneration of different sizes, clustering, and cellular disorganization; the chondrocytes in the mid and deep zones were dissolved and necrotic, with homogeneous red staining and loss of original structure, and in a few cases, there was mild hyperplasia and inflammatory cell infiltration in the synovium (Figure 1C, D). Administration of high-dose Biqi Capsule significantly ameliorated cartilage destruction and chondrocyte damage, showing similar results as celecoxib as a positive control.

Next, we measured inflammatory factors, including IL-1β, TNF-α, IL-6, and PGE2, in the synovium of rats. It was realized that Biqi Capsule significantly reduced IL-1β, IL-6, and TNF-α levels in the synovium of the OA rats (Figure 1E-G). Biqi Capsule also reduced PGE2 in the synovium, which led to hyperalgesia and allodynia (Figure 1H). Altogether, Biqi Capsule inhibited the expression and secretion of multi-cytokines, consequently inhibiting OA development.

NF-κB is a key signaling mediator of inflammation. Immunoblotting of synovial tissue showed a significant increase in the phosphorylation level of NF-κB in the OA group. These increased levels of phosphorylation were inhibited in a dose-dependent manner by Biqi Capsule (Figure 1I, J).

Network pharmacology analysis was performed to clarify the mechanisms of Biqi Capsule. Uploading seven components of Biqi Capsule to the PharmMapper server resulted in 275 candidate compound targets. By searching the OMIM, TTD, and GeneCards databases, 3221 OA-associated targets were retrieved. The 68 potential therapeutic targets of Biqi Capsule for OA were identified using Venn diagrams (Figure 2B). Furthermore, we conducted a KEGG and GO enrichment analysis on the 68 targets (Figures 2C, D). The significant and highest-scoring pathway was the PI3K/AKT pathway. Biological processes are mostly associated with inflammatory responses and oxidative stress. This suggests that the pharmacological mechanism of Biqi Capsule for the treatment of OA may be related to the inhibition of inflammatory response and oxidative stress, via PI3K/AKT signaling pathway.

Network Pharmacology showed that the PI3K/AKT pathway is a critical pathway for Biqi Capsule in the treatment of OA. Activated PI3K/AKT/mTOR signaling has been found in many studies to promote chondrocyte proliferation and reduce cartilage damage (Sun et al., 2020). To confirm the anti-chondrocyte damage effect of Biqi Capsule, we examined the effect of Biqi Capsule extract on chondrosarcoma cell line SW1353 cell damage induced by H₂O₂ (Figure 2E). After H₂O₂ stimulation, the cell viability was significantly reduced, in association with the reduction of phosphorylation of PI3K, AKT, and mTOR. Biqi Capsule extract (100 μg/mL) increased the viability in H₂O₂-stimulated SW1353 cells. Mechanically, Biqi Capsule extract (100 μg/mL) upregulated the phosphorylation of PI3K, AKT, and mTOR in H₂O₂-stimulated SW1353 cells (Figure 2E∼G). According to one study, PI3K inhibitor LY294002 could strongly inhibit the activation of PI3K in SW1353 cells both in vivo and in vitro (Huang et al., 2017; Xu et al., 2019; Zhu et al., 2019; Xu et al., 2020). In line with our current study, it was revealed that Biqi Capsule extract could not reverse the H₂O₂-stimulated damage in SW1353 cells that were treated with LY294002 (Figure 2H). Given these findings, we assumed that Biqi Capsule could ameliorate OA by stimulating, as well as enhancing the activity of the PI3K/AKT/mTOR pathway in knee joint cartilage.

To explore the active components of Biqi Capsule against chondrosarcoma cell damage, we screened for the effects of the seven components of Biqi Capsule, including Strychnine, Brucine, Liquiritin, Salvianolic Acid B, Glycyrrhizic Acid, Cryptotanshinone, Tanshinone ⅡA on the viability of H₂O₂-stimulated SW1353 cells. Among these components, Glycyrrhizic Acid and Liquiritin were found to increase cell viability. Moreover, the effects of Glycyrrhizic Acid and Liquiritin were also found to be dose-dependent (Figure 3).

The treatment with Glycyrrhizic Acid (5 μmol/L) and Liquiritin (5 μmol/L) resulted in a significant increase in cell viability (Figure 3D–F), so we investigated whether Glycyrrhizic Acid (5 μmol/L) and Liquiritin (5 μmol/L) could regulate the PI3K/AKT/mTOR pathway. Glycyrrhizic Acid and Liquiritin upregulated the phosphorylation of PI3K, AKT, and mTOR in H₂O₂-stimulated SW1353 cells (Figure 4A–D). Moreover, PI3K inhibitor blocks the protective effect of Glycyrrhizic Acid and Liquiritin on H₂O₂-induced SW1353 cells (Figure 4E).

In the PPI network, we obtained one key core target, RAC1, which is an upstream target of the PI3K/AKT/mTOR pathway. Glycyrrhizic Acid and Liquiritin treatment modulated the PI3K/AKT/mTOR pathway. Thus, we speculated that Glycyrrhizic Acid and Liquiritin actively target RAC1 to reduce chondrocyte damage via the RAC1/PI3K/AKT/mTOR pathway. To confirm this, we performed molecular docking studies to gain insight into the binding modes of Glycyrrhizic Acid and Liquiritin upon binding to vital regulatory sites on RAC1, respectively. A negative value of CDocker Energy indicates that the ligand-receptor has decent binding activity. We found that the CDocker Energy of Glycyrrhizic Acid and Liquiritin with RAC1 were −91.3 and −53.9 respectively, which are all much less than zero, indicating that the docking structures of these compounds with RAC1 are considerably stable. We selected the docking structure with the lowest CDocker interaction energy after docking with RAC1 for further analysis. Figure 4F shows the chemical bonds and distances between the different atomic groups of each compound interacting separately with the different amino acids of the RAC1 protein. Glycyrrhizic Acid forms six hydrogen bonds with A13, A15, A137 and A178, Liquiritin forms six hydrogen bonds with A16, A33, A135 and A137. This finding indicates that the Glycyrrhizic Acid and Liquiritin may induce a conformational change in RAC1 that promotes binding to other downstream effectors.

In vivo experiments and network pharmacology analysis showed the anti-inflammatory effect of Biqi Capsule. Thus, we examined the effects of Biqi Capsule extract and the seven compounds on the expression of IL-6 in LPS-stimulated RAW264.7 cells. The CCK8 test showed that none of the experimental concentrations of all drugs used were toxic to RAW264.7 cells (Figure 5A). After LPS stimulation, the IL-6 concentrations were significantly increased. Treatment with Biqi Capsule extract, Brucine, Liquiritin, Salvianolic acid B, Glycyrrhizic Acid, Cryptotanshinone, and Tanshinone ⅡA decreased IL-6 production (Figure 5B, D–I). In addition, the effect was found to be dose-dependent. Among these compounds, Glycyrrhizic Acid showed the most potent anti-inflammatory activity with EC₅₀: 2.25 μmol/L.

NF-κB signaling plays a central role in pro-inflammatory stress-related responses, which also mediates the expression of IL-6. To determine whether the anti-IL-6 effect of Biqi Capsule extract was related to NF-κB, we first detected the activation of NF-κB in RAW264.7 cells. Nuclear translocation and phosphorylation of NF-κB lead to the transcription of genes encoding proinflammatory mediators. Immunoblotting showed that the nuclear translocation and phosphorylation (Figure 5J–K) of NF-κB (p65) were all significantly increased after LPS stimulation. Notably, nuclear translocation and phosphorylation of NF-κB were inhibited by Biqi Capsule extract. In addition, the Biqi Capsule also inhibited the phosphorylation of NF-κB and the expression of IL-6 in OA rats. Taking these findings into consideration, we suggest that the therapeutic effect of Biqi Capsule on the inflammatory response in OA is achieved via the NF-κB signaling pathway, and the bioactive components responsible for this observable effect are Brucine, Liquiritin, Salvianolic Acid B, Glycyrrhizic Acid, Cryptotanshinone, and Tanshinone ⅡA.