Astilbin prevents osteoarthritis development through the TLR4/MD‐2 pathway

Abstract Osteoarthritis has become one of the main diseases affecting the life of many elderly people with high incidence of disability, and local chronic inflammation in the joint cavity is the most crucial pathological feature of osteoarthritis. Astilbin is the main active component in a variety of natural plants such as Hypericum perforatum and Sarcandra glabra, which possess antioxidant and anti‐inflammatory effects. At present, there is no study about the protective effect of Astilbin for osteoarthritis. The purpose of this study was to investigate the effect of Astilbin in human OA chondrocytes and mouse OA model, which was established by surgery‐mediated destabilization of the medial meniscus (DMM). In vitro, we found that Astilbin pre‐treatment inhibited lipopolysaccharide (LPS)‐induced overproduction of inflammation‐correlated cytokines such as nitric oxide (NO), prostaglandin E2 (PGE2), tumour necrosis factor α (TNF‐α) and interleukin 6 (IL‐6), and suppressed overexpression of inflammatory enzymes such as inducible nitric oxide synthase (iNOS) and cyclooxygenase 2 (COX‐2). Astilbin, on the other hand, prevented the LPS‐induced degradation of extracellular matrix (ECM) by down‐regulating MMP13 (matrix metalloproteinases 13) and ADAMTS5 (a disintegrin and metalloproteinase with thrombospondin motifs 5). Moreover, by inhibiting the formation of the TLR4/MD‐2/LPS complex, Astilbin blocked LPS‐induced activation of TLR4/NF‐κB signalling cascade. In vivo, Astilbin showed the chondro‐protective effect in the surgical‐induced OA mouse models. In conclusion, our findings provided evidence that develops Astilbin as a potential therapeutic drug for OA patients.


| INTRODUC TI ON
Osteoarthritis (OA) is a degenerative joint disease with substantial morbidity most common among older people. At present, there are 120 million OA patients in China. 1,2 The aetiologic agent of OA is related to age, obesity, gender and other factors. 3,4 There is currently no effective cure for OA. 1 Previous studies report that inflammation plays a crucial role in the pathogenesis of OA. [5][6][7] Kraus et al reported the presence of lipopolysaccharide (LPS) and lipopolysaccharide-binding protein (LBP) in serum and synovial fluid of OA patients. 8 LPS is known to cause low-grade inflammation and accelerate OA development. 9,10 Therefore, in this study, LPS was used to mimic OA in cell experiments.
Toll-like receptors (TLRs), are cell membrane receptors, participate in the progression of OA by activating a variety of downstream signalling. 11,12 Mild TLRs activation triggers a self-protective mechanism, but excessive activation of TLRs contributes to inflammatory responses due to the continuous secretion of pro-inflammatory cytokines and chemokines. Although there are many subtypes of TLRs, chondrocytes mainly express TLR4, which is reported to be involved in OA progression. 10,[13][14][15] As an effective promoter of OA-related disease-associated molecular patterns (DAMPs), LPS initiates a series of inflammatory reactions by activating TLR4 and subsequently leads to structural damage of joint cartilage. Myeloid differentiation protein-2 (MD-2) is the extracellular domain of TLR4, which recognizes LPS and assists it to bind TLR4. 16 After the LPS/TLR4/MD-2 complex is formed, downstream proteins such as myeloid differentiation factor 88 (MyD88), interleukin-1 (IL-1) receptor-related kinase (IRAKs) and tumour necrosis factor (TNF) receptor-related factor 6 (TRAF6) bind to the complex and further activate inflammation-related signalling pathways including the NF-κB pathway and eventually leading to the occurrence of inflammation and ECM catabolism. 12,17,18 Therefore, targeted inhibition of the TLR4/MD-2 pathway is a potentially promising therapy to prevent OA development.
Astilbin is a flavonoid widely found in daily diets and traditional medicines such as Dimorphandra Mollis, grape, Smilax corbularia, Hypericum perforatum, Sarcandra glabra and other natural plants. [19][20][21] Modern pharmacological research shows that Astilbin has anti-oxidation, 22 anti-inflammatory, 23,24 anti-tumour 25 and other pharmacological activities. It is reported that Astilbin can alleviate LPS-induced inflammatory response in macrophages 24 and suppress high glucose-induced inflammation in rat glomerular mesangial cells by intervening TLR4/MyD88/ NF-κB axis. 26 However, the effect of Astilbin on osteoarthritis is not clear. Therefore, in this study, we investigated the anti-inflammatory effect of Astilbin in LPS-induced human chondrocytes and its protective effect and underlying mechanism in the surgery-induced mouse OA model.

| Primary human chondrocyte culture
The human articular cartilage sample collection was approved by the medical ethics committee of the Second Affiliated Hospital of Wenzhou Medical University (ethics standard: LCKY-2017- 25) and followed the guidelines of Helsinki Declaration. 27 The cartilage tissues were obtained from six patients (58-69 years old, 3 males and 3 females) who underwent total knee arthroplasty, and informed consent was obtained from all patients before the start of the study. To ensure that full-thickness cartilage was obtained, the cartilage tissue was cut at the posterior femoral condyles and avoided selecting cartilage tissue with a damaged surface. The cartilage tissues were cut into 1.0 × 1.0 × 1.0 mm 3 pieces and PBS was used to blow and wash the tissues repeatedly. The cartilage slices were treated with 2mg/ml type II collagenase and digested for 4h in Dulbecco's modified Eagle's medium (DMEM)/F12 at 37°C. Following the basic digestion of the tissue, centrifugation at 1000 rmp for 5 minutes was done, and the supernatant discarded.
The chondrocytes were cultured at a density of 2 × 10 5 cells/ml in a 6-well plate in DMEM/F12 supplemented with 10% FBS and 1% antibiotics and incubated at 37°C with 5% CO 2 . The morphology and the adherence of the cells were observed after one day. The media was changed every 2-3 days, and the cells passaged after attaining 70%-90% confluence. To avoid changes in cell phenotype, only the first to third generations of chondrocytes were considered for these experiments.

| Animal model
A total of forty-five 8-week-old C57BL/6 male wild-type (WT) mice were purchased from the Shanghai animal research centre of the Chinese Academy of Sciences and were randomly divided into three groups (n = 15). The animal protection and use regulations were performed in accordance with the guidelines for the protection and use of experimental animals issued by the National Institutes of Health and approved by the animal protection and use Committee of Wenzhou Medical University (code of Ethics: wydw2014-0127).
The mice were raised in a relatively suitable environment (temperature 20-28 ℃, humidity 45%-70%, 24-hour cycle light, free food, and drinking water). Subsequently, the experimental mice underwent surgical destabilization of the medial meniscus (DMM) to induce OA. 28 Briefly, after intraperitoneal anaesthesia with 2% (w/V) pentobarbital (40 mg/kg), the skin and joint capsule of the right knee were cut along the medial side of patellar ligament using a mini blade, and the muscle and fat pad separated to expose the joint cavity and medial meniscus. The ligament of the medial meniscus was cut off using a knife or scissors. Finally, the tissue and skin were sutured layer by layer. Following intraperitoneal anaesthesia, the sham operation was performed on the right knee joint of the sham group mice, the skin and joint capsule were cut along the medial patellar ligament, and then, the wound was sutured. After the operation, penicillin was injected to prevent infection.

| Griess reaction and ELISA
The light absorption value of the sample was detected under the ultraviolet spectrophotometer, and the NO content of the corresponding sample was calculated according to the manufacturer's instructions. The blood samples collected from the abdominal aorta were centrifuged for 15 minutes at 1000 g. The concentration of TNF-α, PGE2, IL-6, Collagen II, aggrecan, ADAMTS-5 and MMP13 in each sample was detected using commercial ELISA kits (R&D Systems, Minneapolis, MN) according to the manufacturer's instructions. All tests were conducted five times.

| Protein extraction
Total proteins were extracted from chondrocytes using RIPA lysis buffer. Cytoplasmic and nuclear proteins were extracted using the corresponding protein extraction kit (Beyotime, Shanghai, China).
The sample solution was lysed on ice by ultrasound and centrifuged at 15000 xg at 4°C for 30 minutes. The concentration of the protein was determined using the BCA protein assay kit (Beyotime, Shanghai, China). After washing with TBST 3 times, antibody binding was visualized by Electrochemiluminescence Plus reagent (Invitrogen), and the Image Lab 3.0 software (Bio-Rad) used to quantitatively analyse the intensity of the bands.

| MD-2 and TLR4 competitive ELISA assay
Human MD-2 or TLR4 antibody was coated on 96-well plates using 10 mmol/L Tris-HCl buffer (pH 7.5). The plates were incubated at 4 ℃ overnight, washed thrice with poly (butylene succinate-cobutylene terephthalate) PBST, followed by blocking with 3% bovine serum albumin at room temperature for 1.5 hours. After blocking, 4 μg/mL of recombinant human MD-2 (rhMD-2) or recombinant Toll-like receptor 4 (rhTLR4) was added to 10 mmol/L Tris-HCl buffer (pH 7.5), and the plates were further incubated at room temperature for 1.5 hours. After washing with PBST, biotin-LPS and different concentrations of Astilbin (0, 10, 20 or 40 μmol/L) were added at room temperature for 1.5 hours. The horseradish peroxidase combined with streptavidin (Beyotime, Shanghai, China) was added at room temperature and incubated for 1 hour. After the addition of tetramethylbenzidine substrate (eBioscience, San Diego, CA, USA), horseradish peroxidase activity was then measured at 450 nm using a SpectraMax M5 plate reader (Molecular Devices, Sunnyvale, CA, USA).

| Immunoprecipitation
Cell lysates were obtained as mentioned above, incubated with sufficient anti-TLR4 antibody for 1h, and protein G-Sepharose beads added to retrieve the immune complexes overnight at 4°C in a shaking bed. The precipitates were washed with cold PBS for 4 times and boiled in sample buffer to release the proteins. Similarly, the level of MD-2 was further detected by immunoblotting with the anti-MD-2 antibody (IB).

| Molecular modelling
ChemBioDraw software was used to draw the molecular structure of Astilbin, and the energy of Astilbin minimized using ChemBio3D.
Access to the Protein Data Bank (https://www.rcsb.org/) to obtain the crystal structure of the human MD2/lipid IVa complex (PDB code 2E59) was done. After minimizing the energy with PyMOL (version 1.7.6), the lowest energy conformation for docking was determined using default parameters. To complete the proteinligand docking analysis, AutoDock Tools (version 1.5.6) was used.
Finally, UCSF PyMOL was used to create the image of a 3D view.

| X-ray imaging method
After 8 weeks of operation, all animals within or without treatment were received X-ray examination. A digital X-ray machine (Kubtec Model XPERT.8; KUB Technologies Inc) was used to assess changes in knee joint space, osteophyte formation and cartilage surface calcification. Appropriate images are obtained using the following settings: 50kV and 160 μA.

| Histological analysis
The knee joint specimens were fixed with 4% paraformaldehyde Then, the extent of articular cartilage destruction was graded by the summed OARSI scores (0-12) from the medial femoral condyle and medial tibial plateau.

| Statistical analysis
Each experiment was repeated at least 5 times. The results were expressed as the mean ± standard deviation, and SPSS 20.0 statistical software was used to process all data of the experiment.
Comparisons between control and treatment groups were analysed using Mann-Whitney U test. The difference was statistically significant when P < 0.05.

| Effect of Astilbin on human chondrocyte viability
The molecular structure of Astilbin is shown in Figure 1A. We em-

| Effect of Astilbin on the expression of iNOS, COX-2, PGE2, NO, TNF-α and IL-6 in human osteoarthritis chondrocytes induced by LPS
To determine the effect of Astilbin on LPS-induced inflammatory response in human OA chondrocytes, the expression levels of several inflammatory markers were determined. As shown in Figure 2A

| Effect of Astilbin on LPS-induced extracellular matrix (ECM) degradation in human OA chondrocytes
The effect of Astilbin on the degradation of collagen II and aggregates induced by LPS in human OA chondrocytes was investigated.
Compared with LPS alone group, Astilbin treatment increased the expression of aggrecan and collagen II and suppressed the production of matrix metalloproteinases (MMP)-13 and ADAMTS-5 in a dose-dependent manner ( Figure 3A). Additionally, the results of immunofluorescence staining of collagen II and MMP13 were consistent with ELISA results ( Figure 3B). Taken together, these results indicated that Astilbin alleviates LPS-mediated ECM degradation in human OA chondrocytes.

| Effect of Astilbin on LPS-induced NF-κB activation in human OA chondrocytes
To determine the anti-inflammatory mechanism of Astilbin, we assessed the change of NF-κB signalling activity. We found that LPS stimulated IκBα degradation in the cytoplasm and increased p65 content in the nucleus. However, Astilbin pre-treatment (40 μmol/L) inhibited this phenomenon, and the NF-κB activity in the Astilbinalone group did not change ( Figure 4A,B, and C). Subsequently, we examined the p65 translocation in human OA chondrocytes induced by LPS. Immunofluorescence staining in. 4D shows that the p65 positive spot in the LPS-stimulated group was translocated from the cytoplasm to the nucleus compared with the control group. However, pre-treatment with Astilbin (40 μmol/L) sharply reduced the translocation of p65. These results indicated that Astilbin could inhibit NF-κB signalling activity in LPS-exposed human OA chondrocytes.

| Effects of Astilbin on the interaction of TLR4 and MD-2 in LPS-induced human OA chondrocytes
Accumulating evidence demonstrates that LPS binds to TLR4 by recognizing the adaptor protein MD-2. We first used competitive ELISA to explore the effect of Astilbin in the TLR4/MD-2 axis, which is an upstream molecule of NF-κB signalling. As shown in Figure 5A, Astilbin reduced the binding of LPS to rhMD-2, but not to rhTLR4, which indicated that Astilbin acts on the TLR4/  Figure 5C shows that Astilbin has a high affinity for the hydrophobic position of MD-2 (−6.5 kcal/mol), and a hydrogen bond is formed be-  As shown in Figure 5D, LPS promoted the expression of MyD88, F I G U R E 2 Influence of Astilbin in LPS-induced inflammatory reaction in human chondrocytes. (A,B) iNOS and COX-2 protein level in chondrocytes measured by Western blot. C, Effect of Astilbin on LPS-exposed IL-6, TNF-α, PGE2 and NO production in human chondrocytes. The data in the figures represent the averages ± SD. Significant differences among different groups are indicated as ## P < 0.01, vs control group; **P < 0.01 vs LPS alone treatment group, n = 5 IRAK-1 and TRAF-6, whereas this phenomenon was reversed by Astilbin pre-treatment.

| Astilbin ameliorates OA development in the mouse model of DMM
To study the protective ability of Astilbin in the OA process in vivo, the DMM model was established by surgery. Mice in the Astilbin group received 40 mg/kg of Astilbin orally once a day for 8 weeks and were then sacrificed for imaging and histological analysis. As shown on the radiographs in Figure 6A, the surface density of cartilage in the DMM group was higher compared with the sham group, and the joint space was severely narrowed. However, in the Astilbin treatment group, although the joint space was narrowed, the surface calcification of cartilage was lighter. Histological analysis of OA was performed by Safranin O staining ( Figure 6B) and quantified using the Osteoarthritis Research Society International (OARSI) scoring F I G U R E 3 Influence of Astilbin in LPS-mediated ECM degeneration in human chondrocytes. A, The level of Collagen II, aggrecan, MMP13 and ADAMTS-5 in chondrocytes treated as above was visualized by ELISA. B, The representative collagen II and MMP13 were detected by the immunofluorescence combined with DAPI staining for nuclei (scale bar: 50 µm or 30 µm). The data in the figures represent the averages ± SD. Significant differences among different groups are indicated as ## P < 0.01, vs control group; *P < 0.05, **P < 0.01 vs LPS alone treatment group, n = 5 system. We found that the cartilage surface in the sham group was smooth and stained red. In contrast, cartilage wear and erosion were observed in the DMM group. The degree of cartilage erosion in the Astilbin treatment group was lower compared with the DMM group.
The OARSI score ( Figure 6C) of the DMM group was significantly higher than that of the sham group, which was consistent with the Safranin O staining. However, the OARSI score in the Astilbin group was lower compared with the DMM group. In conclusion, these results suggested that Astilbin delays the development of OA in vivo.

| D ISCUSS I ON
Osteoarthritis is a common clinical degenerative joint disease with increasing morbidity rate. [1][2][3] Unfortunately, the aetiology of OA is complex and remains unclear. 3,30 Although non-steroidal anti-inflammatory drugs (NSAIDs) can effectively alleviate the pain in patients, they also have obvious limitations, such as intestinal adverse reactions, and do not delay the progression of OA. [31][32][33] Therefore, it is urgent to explore drugs, which can prevent and delay the progression of OA. 3,34 Accumulating evidence demonstrates that inflammation is closely related to the development of OA. 35,36 LPS has been widely used as an agent to stimulate the inflammation and apoptosis in chondrocytes in vitro. 9,37 In previous studies, the dosage of LPS varying from 0.1 to 50 μg/mL was applied to chondrocytes, a high concentration of LPS was used to study the mechanism of cell death, whereas the low dosage of LPS was considered more suitable for mimicking the inflammatory environment. [37][38][39] In this study, we used LPS at 1 μg/mL for cellular experiments.
Astilbin is a flavonoid extracted, separated and identified from Dimorphandra Mollis, grape, Smilax corbularia, Hypericum perforatum, Sarcandra glabra and other natural plants. [19][20][21] Modern pharmacology research shows that it possesses anti-oxidation, 22 anti-inflammatory, 23,24 anti-tumour 25 and other pharmacological activities. However, the mechanism of protection in OA remains unclear. Therefore, in this study, we studied the anti-inflammatory effect of Astilbin in OA.
Previous studies have reported that the progression of OA is mainly caused by several pro-inflammatory mediators, iNOS catalyses NO or COX-2 to induce PEG2 production, then increases the content of MMPs and ADAMTs, finally promoting the degradation of collagen II and proteoglycan. 40,41 Moreover, TNF-α and IL-6 also participate in the development of OA. 42 In this study, we found that Astilbin treatment inhibited the LPS-induced increase in PGE2 and NO in human OA chondrocytes and up-regulation of COX-2 and iNOS, as well as TNF-α and IL-6. Besides its anti-inflammatory effects, Astilbin also inhibits ECM degradation. Following LPS stimulation, the synthesis and secretion of MMP-13 and ADAMTS-5 increased in human OA chondrocytes, and the expression of collagen II and aggrecan decreased, and these phenomena were reversed after Astilbin pre-treatment. These results suggest that Astilbin has a potential therapeutic role in preventing OA progression.
To explore the protective mechanism of Astilbin during OA, the effect of Astilbin on the NF-κB pathway was studied, which is a crucial inflammatory signalling pathway for OA. 43,44 The results showed that Astilbin inhibited LPS-induced NF-κB activation in human chondrocytes. However, even though we have confirmed that Astilbin can exert anti-inflammatory effects by targeting NF-κB, its potential upstream cascade effect also requires to be studied further. Previous studies reveal that TLR4 signalling is currently one of the most widely studied upstream cascades, and the concentration of TLR4 in NF-κB activation and OA development has received increasing attention. 16,45 TLR4-targeted antagonist is emerging as a new drug to alleviate the progression of OA. Therefore, in this study, we investigated whether Astilbin affects the TLR4/MD-2 axis using competitive ELISA. Astilbin was found to prevent the binding between LPS and MD-2 but F I G U R E 4 Influence of Astilbin in LPSmediated activation of NF-κB pathway in human chondrocytes. (A,C,D) The protein expressions of IκBα in cytoplasm and p65 in nuclear in chondrocytes treated as above were detected by Western blot. B, The nuclei translocation of p65 was detected by the immunofluorescence combined with DAPI staining for nuclei (scale bar: 20 µm). The data in the figures represent the averages ± SD. Significant differences among different groups are indicated as ## P < 0.01, vs control group; **P < 0.01 vs LPS alone treatment group, n = 5 showed no significant effect on TLR4. Co-immunoprecipitation results showed that the interaction between TLR4 and MD- In conclusion, this study demonstrated that Astilbin alleviates LPS-induced inflammation and ECM catabolism by inhibiting the TLR4/MD-2 axis and NF-κB signalling pathway in human OA chondrocytes. In vivo, Astilbin protects the cartilage and delays the progression of OA. This study demonstrates that Astilbin can be considered as a potential therapeutic drug for OA.

ACK N OWLED G EM ENTS
This work was funded by National Natural Science Funding of China (81601983) and Wenzhou Science and Technology Bureau Foundation (Y20190270).

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are available from the corresponding author upon request.

Ping Shang
https://orcid.org/0000-0002-0743-5264 F I G U R E 6 Astilbin attenuates OA development in vivo. A, Digital X-ray image of mouse knee joints from different experimental groups. Narrowing of joint space was found in both OA and treatment group (white triangles), the calcification of cartilage surface was obviously shown in OA group (white arrows). B, Representative S-O staining of cartilage from different experimental groups at 8 wk post-surgery (scale bar: 200 µm). C, Diagrams showed the OARIS socres of cartilage. The data in the figures represent the averages ± SD. Significant differences among different groups are indicated as ## P < 0.01, **P < 0.01, n = 15