Sparstolonin B suppresses free fatty acid palmitate‐induced chondrocyte inflammation and mitigates post‐traumatic arthritis in obese mice

Abstract Abnormal lipid metabolism, such as systemic increased free fatty acid, results in overproduction of pro‐inflammatory enzymes and cytokines, which is crucial in the development of obesity‐related osteoarthritis (OA). However, there are only a few drugs that target the lipotoxicity of OA. Recent researches have documented that the traditional Chinese medicine, Sparstolonin B (Ssn B), exerted anti‐inflammatory effects in various diseases, but not yet in OA. On the basis of this evidence, our works purposed to evaluate the effect of Ssn B on free fatty acid (FFA) palmitate (PA)‐stimulated human osteoarthritic chondrocytes and obesity‐associated mouse OA model. We found that Ssn B suppressed PA‐triggered inflammatory response and extracellular matrix catabolism in a concentration‐dependent approach. In vivo, Ssn B treatment inhibited cartilage degeneration and subchondral bone calcification caused by joint mechanical imbalance and alleviated metabolic inflammation in obesity. Mechanistically, co‐immunoprecipitine and molecular docking analysis showed that the formation of tolllike receptor 4 (TLR4)/myeloid differentiation protein‐2 (MD‐2) complex caused by PA was blocked by Ssn B. Subsequently, it leads to inactivation of PA‐caused myeloid differentiation factor 88 (MyD88)‐dependent nuclear factor‐kappaB (NF‐κB) cascade. Together, these findings demonstrated that Ssn B is a potential treatment agent for joint degenerative diseases in obese individuals.


| INTRODUC TI ON
Osteoarthritis (OA), one of the commonest joint degenerative illness, is typified by cartilage loss, subchondral bone sclerosis as well as osteophyte formation. 1 A variety of underlying factors, including ageing, obesity, trauma, being female and congenital malformation, have obviously exhibited the potential to escalate the risk of OA. [2][3][4][5] Among these risk contributors, obesity seems to be great significance. 6 Elevated mechanical stress on the joint platform is recognized as one predisposing factor in obesity-related OA; however, it is likely not the only one. 7,8 Recent investigations have demonstrated that obesity-linked free fatty acids (FFAs), for example, palmitic acid (PA) and stearic acid (SA), were considered as novel pro-inflammatory factors and caused chondrocyte dysfunction and even death. 9,10 Moreover, in individuals with OA, the content of FFAs is elevated in the joint fluid, synovium and articular cartilage. [11][12][13] Toll-like receptors (TLRs) have been correlated with multiple diseases; nonetheless, accumulating evidences opine that TLR4 participates in the pathogenesis of osteoarthritis. [14][15][16] For obesity people, PA, as a kind of damage associated molecular patterns (DAMPs), was considered a novel agent for TLR4 activation. 16,17 However, the process of PA and TLR4 protein binding requires myeloid differentiation-2 (MD-2) protein as an auxiliary, which is involved in PA-mediated TLR4 complex formation through crosstalking with the TLR4 extracellular domain. 18 After the formation of the PA-TLR4-MD-2 complex, MyD88, IL 1 receptor associated kinases (IRAKs) as well as TNF receptor-associated factor 6 (TRAF6) are recruited to promote the transcription of p65 and activate nuclear factor kappa B (NF κB) pathway. 16 Finally, p65 in the nucleus binds to the promoter site of the target gene, triggering a cascade of catabolic and inflammatory reactions. 19 Therefore, specifical inhibition of the TLR4/ MD-2 cascade might be identified as a therapeutic schedule with great promise for OA, especially in obese individual.

Sparstolonin B (Ssn B) is an natural compound isolated form the
Chinese herb Sparganium stoloniferum (S.stoloniferum), whose tubers have long been applied in traditional Chinese medicine (TCM). 20 The Ssn B was identified as a selective TLR antagonist, showing significant antitumour, anti-obesity and anti-inflammatory effects in various diseases. [20][21][22][23][24][25][26] It could inhibit lipopolysaccharide (LPS)-mediated macrophages' inflammation by selective blockage of TLR4/NF-κB axis 20 and suppresses leptin-induced inflammation and lipid accumulation in hepatic Kupffer cells by decreasing TLR4 trafficking. 23 Moreover, Ssn B was reported to inhibit the LPS caused production of cytokines in 3T3-L1 adipocytes, as well as reduce high-fat diet-stimulated obesity in rats. 25 In addition, Ge et al. in recent study revealed that the Ssn B exerts a protective effect in the rat intervertebral disc degenerative model by targeting TLR4/MyD88/NF-κB-triggered inflammation. 27 Nonetheless, whether the Ssn B participates in the modulation of joint degenerative disease like osteoarthritis is not well elucidated. Based on the literature, the Ssn B was a potent anti-inflammatory agent and has a closed relation to the lipid metabolism. 25 Herein, we firstly verified that Ssn B alleviated the inflammation level and ECM degradation of human osteoarthritis chondrocytes exposed to palmitate and explained its potential molecular mechanism. Moreover, the Ssn B-induced protection in vivo was evaluated by mouse obesity-related OA model.

| Isolation of human primary osteoarthritic chondrocytes
Ethics Committee of the Second Affiliated Hospital of Wenzhou Medical University approved the tissue collection procedure and was as per Declaration of Helsinki guidelines. 28 Besides, the study subjects gave an informed consent. We acquired the knee cartilage samples from 6 participants consisting of 3 men and 3 women (aged 65-73 years) who received TKR (total knee replacement). The hyaline cartilage tissue was collected and chopped as much as possible and soaked in a DMEM/F12 solution of type II collagenase (2 mg/ mL) at 37°C for 4 hours. After washing and centrifuging with PBS, added DMEM/F12 medium containing 10%FBS and 1%antibiotic to resuspend the cells and seeded them in a six-well plate. To prevent loss of phenotype, we select chondrocytes within two generations for subsequent experiments.

| Cell experiment strategy
To explore protective influences of diverse concentrations of Ssn B, we exposed the cells to 500 μM PA, in combination with Ssn B pretreatment at varied levels (3, 10 and 30 μM), or alone. Moreover, our control group was not treated except medium replacement.
Chondrocytes were pretreated with twenty-four hours of Ssn B. As for the evaluation of NF-κB signal activity, the exposure time of PA was shortened to 2 h. While to examine the functional parameters, consisting of ECM marker, or inflammatory, we extended the duration to 24 h.

| Cell viability analysis
The Dojindo Co-CCK-8 kit (Japan) was employed to explore the Ssn B cytotoxicity on the chondrocytes, as described in the manufacturerprovided protocol. In brief, the 2nd passage chondrocyte was inoculated in 96-well plates (8000/well) for 24 hours, and then incubation in different concentration of Ssn B (0, 1, 3, 10, 30 and 100 μM) performed for 24 hours or 48 hours. After the specified time is over, we exhausted the medium in each well and introduced 100μl of DMEM/ F12 enriched with 10 μl of CCK-8 reagent to all the wells of the plate and then incubated at 37°C for an additional 2 hours. Finally, Thermofisher spectrophotometer was employed to determine the absorbance at 450 nm.

| Measurement of inflammatory factors and secreted proteins
Nitric oxide interaction in growth medium was assayed by the Griess reagent as documented previously. 29 The level of PGE2, TNFα, aggrecan, IL-6, ADAMTS-5, collagen II, as well as MMP-13 in the supernatants of the cell cultures was measured with the commercial ELISA kits (R&D Systems, Minneapolis, MN) as described by the manufacturer.

| Western blot analysis
After the treatment, the intracellular proteins were extracted by RIPA lysate, and then the protein concentration was detected by the BCA assay kit (Beyotime, China) to prepare the equivalent samples (40 μg).
Each sample was fractionated in SDS/PAGE gels and blotted onto the Bio-Rad PVDF membranes (USA). Subsequently, 5% dry milk was em- (1:500), as well as GADPH (1:5000) and incubated overnight at 4°C. Thereafter, we washed the membrane three times with TBST and incubated in the secondary antibody for 2 hours at RT (room temperature).
Thereafter, the films were rinsed thrice using TBST, and Invitrogenelectrochemiluminescence plus reagent employed to view the blots.
The Bio-Rad Image Lab software was finally employed to determine blot's intensity.

| Immunofluorescence
As for cellular fluorescence staining, the cells were seeded on glass slides and treated with the above-mentioned agents. Then, the glass slides were washed thrice with PBS, then fixed with paraformaldehyde and then permeabilized with 0.1% Triton X-100. Afterwards, 10% goat serum (dissolved in PBS) was employed to block non-specific antigens of each sample. Thereafter, PBS was employed to rinse the glass slides, which were inoculated by primary antibodies: collagen Ⅱ (1:200), p65 (1:200), as well as MMP-13 (1:200) at 4°C overnight.
Rinsing of glass plates was done the next day and then inoculated with Alexa Fluor ® 488-labelled secondary antibodies (diluted in PBS, 1:300) for 1h at RT, followed by labelling by using DAPI for 5 min.
Random selection of five fields for every slide was conducted for observation using the Olympus fluorescence microscope (Tokyo, Japan).

| Immunoprecipitation
We performed RIPA lysate to extract cellular protein and added enough TLR4 antibody to incubate for one hour. After incubation, added magnetic beads to enrich the immune complex in a 4°C overnight. Thereafter, ice-cold PBS was employed to rinse the precipitates four times and then boiled the sample buffer to release the proteins.
Finally, the content of MD-2 in the sample is detected by western blot.

| Molecular modelling
We performed ChemBioDraw to form the Ssn B's structure.
ChemBio3D was employed in minimizing its energy. We down- in the other group were fed an STD or HFD and treated with an equal volume of saline by intraperitoneal injection. Notably, based on the previous research, we selected the concentration of 5 mg/ kg/day to explore the protective effect of Ssn B in vivo. 23,25,[30][31][32] As for DMM surgery, in brief it entails cutting the tendon between tibial plateau and the medial meniscus to achieve meniscal instability. 33 The mice were allowed unlimited activity along with ad libitum access to food and water for 12 weeks.

| X-ray assay
After 12 weeks of surgery, animals received an X-ray. Knee joint X-ray was produced by the digital X-ray machine (Kubtec Model XPERT.8; KUB Technologies Inc) at 160 μA and 50 kV.

| Safranin O staining
Knee tissue is collected and fixed with 4% paraformaldehyde for one day. Afterwards, the samples were decalcified by 10% EDTA solution for one month. After dehydration with gradient alcohol, the tissue was embedded in paraffin and then cut into 5 μm slices. For each joint sample, we selected a slice every 50 μm, a total of 10 slices in total, and performed SO staining. Finally, our histological researchers accessed the degree of cartilage degeneration, according to the Osteoarthritis Research Society International (OARSI) scoring system. 33,34 Briefly,

| ELISA test of mouse serum
To explore the contents of IL-1β, TNFα, as well as IL-6 in serum of mice, the blood specimens collected. After centrifugation to remove cells and cell debris, the ELISA kits were used to detect the levels of the above-mentioned inflammatory factors in the serum.

| Statistical analyses
Our experiments were conducted at least 5 times. These data are indicated by mean ± SD. SPSS software program 20.0 was employed to analyse the data. Data were assessed with one-way analysis of variance (ANOVA) and the n by Tukey's test for comparison between control and treatment groups. p < 0.05 were considered significant.

| Influences of Ssn B on human chondrocytic viability
Chemical structure of Ssn B is shown in Figure 1A. The cytotoxic effect of Ssn B on chondrocytes was assayed at an ascending concentration of Ssn B (0, 1, 3, 10, 30 and 100 μM) for 24 and 48 hours with the CCK-8 kit. Chondrocytic viability increased after Ssn B treatment at 3, 10 and30 μM, but decreased when the Ssn B concentration increased to 100 μM ( Figure 1B). Therefore, 3, 10 or 30 µM Ssn B was employed in the following experiments.

| Influences of Ssn B on the expressions of inflammatory factors in PA-mediated human osteoarthritic chondrocytes
In order to explore whether Ssn B represses the inflammation in PAtreated human osteoarthritic chondrocytes, diverse-related biomarkers were assayed. Western blot was used to assess the production of iNOS and COX-2. In Figure 1C

| Influences of Ssn B on ECM degradation in PA-stimulated human osteoarthritic chondrocytes
Then, we evaluated the role of Ssn B on PA-caused chondrocytic ECM degradation. In Figure 2A, Ssn B elevated the expression of collagen II and aggrecan, but repressed MMP-13 and ADAMTS-5 production in a concentration-dependent manner, compared with PA-stimulated group. Besides, collagen II and MMP-13 immunofluorescence results of chondrocytes were consistent with the ELISA results ( Figure 2B). Altogether, these results reveal that Ssn B alleviates ECM degradation in PA-stimulated chondrocytes.

| Influences of Ssn B on PA-mediated NF-κB activation in PA-stimulated human osteoarthritic chondrocytes
To evaluate the mechanism of Ssn B, NF-κB signalling, the classic inflammation-associated pathway, was examined by western blot analysis. PA remarkably caused the upregulation of p-IκBα and p-p65 and promoted the degradation of IκBα. Nonetheless, these effects were remarkably suppressed by Ssn B pretreatment at the concentration of 10 μM (Figure 3A and B). We performed p65 immunofluorescence staining to assess NF-κB signalling activity in PA-treated chondrocytes. The chondrocytes were clustered into 3 groups: control, PA and PA + Ssn B. In control group, p65-positive spots were mainly located in the cytoplasm. Nevertheless, after PA treatment, the p65 was remarkably translocated into the nucleus. Ssn B pretreatment mitigated p65 translocation ( Figure 3C). Altogether, these data illustrated a suppressive effect of Ssn B on the activity of NF-κB signalling in PA-treated chondrocytes.  Figure 4A

| Influences of Ssn B on obesity -related OA in mouse
We developed the mouse obesity-related OA model by feeding mice with HFD and surgically removing medial meniscus ligament to explore the role of Ssn in vivo. The timeline for animal processing is indicated in Figure 5A. By weighing regularly, we found that HFDfed mice were significantly heavier than the STD-fed mice, within 0.5 to 3 months ( Figure 5B). From the 3rd to 6th month, the weight of the mice fed by HFD continued to increase, while the mice in the STD group maintained a weight of about 25g ( Figure 5C). Notably, regardless of the STD-fed or HFD-fed mice, Ssn B treatment has not changed the weight of mice. All animals were sacrificed 6 months later to collect knee joint tissue for further evaluation. As for the radiograph, DMM group showed excessive narrowing of the joint space and osteosclerosis occurs in the load-bearing area of the tibial plateau, which was more obvious in the HFD + DMM group ( Figure 5D).
But these phenomena were alleviated after Ssn B administration.
In SO staining, we found that HFD feeding further aggravated the destruction of cartilage structure with rougher surface, fewer red areas and a higher OARSI score, suggesting that HFD could accelerate DMM-induced cartilage degeneration in mice. Nevertheless, these pathological changes were ameliorated to varying degrees in Ssn B-treated mice ( Figure 5E and F). Based on immunohistochemical staining, there are more MD-2-positive chondrocytes in cartilage tissue of HFD-fed mice, but Ssn B treatment could alleviate this phenomenon ( Figure 5G and H). Detecting serum IL-6, TNFα, as well as IL-1β levels by ELISA, we found that Ssn B administration reduced the serum inflammatory factor levels in OA mice and obese OA mice ( Figure 5I-K).

| DISCUSS ION
Obesity is causally linked to joint disease. 6,35 The abnormal lipid accumulation not only increased body mass, leading to a heavy mechanical stress to the joint, but also is responsible for the pathological development of OA. 36 Firstly, for obese patients, the risk of developing arthritis in nonweight-loading joints was also increased. 7,37,38 Secondly, the high-fat diet-mediated obese mice present a significant joint degeneration and systemic inflammation, while it is reported that escalated mechanical loading of joint via exercise did not enhance OA, but rather ameliorate OA. 39 Therefore, the lipotoxicity resulting from dysregulated lipid metabolism in OA development should be seriously con-  The data present as averages ± SD. ##p < 0.01, **p < 0.01, n = 6 It has been documented that traditional Chinese medicine exerts chondroprotective effects by suppressing inflammatory response and possesses the potential to treat OA. 41 However, no such agents have been reported to be applied in targeting obesity-related OA. Herein, we demonstrated that Ssn B, a core bioactive component of the traditional Chinese herb S.stoloniferum, as a novel TLR's inhibitor, significantly inhibits PA-mediated inflammation in chondrocytes. In addition, we explored the specific protection mechanism of Ssn B.
Free fatty acids, particularly PA, were reported to be associated with chronic inflammation in chondrocytes. 10,40 Specifically, PA could promote pro-inflammatory cytokine production via activating NF-κB signal. 42 With the stimulation of PA, the IκBα is phosphorylated and then frees and initiates the p65 translocated from the cytoplasm to the nucleus; IκBα is consequently degraded in the cytoplasm. p65 in the nucleus promotes the transcription of related gene of catabolic enzymes and inflammatory mediators. 19 Nitric oxide and PGE2 are two major inflammatory productions, catalysed by iNOS and COX-2, respectively, which not only improve the release of MMPs and ADAMTS, but also disturb collagen II and proteoglycan synthesis, leading to the ECM degradation and ultimate cartilage corrosion. 29 All of these factors mentioned above, together with TNFα and IL-6, Despite numerous cascades are regarded as upstream targets for NF-κB-involved inflammation, the TLR4 signalling cascade is one of the most widely studied pathways being closely linked to FFA-correlated inflammation. 16 The classic TLR4 pathway involves MyD88-dependent and MyD88-independent cascades that stimulate signalling through TRIF (TIR domain-containing adaptorinducing interferonβ). Of note, PA-induced inflammation was reported to be MyD88-dependent. 17 In brief, upon PA-triggered TLR4 dimerization, the bridging adaptor MyD88 adaptor like (MAL; also referred to as TIR domain-containing adaptor protein, or TIRAP) mobilizes MyD88 resulting in its polymerization and interaction with several IRAKs and then autophosphorylate and activate TRAF6, which followed by the stimulation of the central pro-inflammatory transcription factor NF κB to trigger the subsequent inflammatory responses. 16 Actually, MD-2 is the key participant to construct the TLR4/MD-2 complex to mediate the deleterious response during several DAMPs like LPS and PA-induced TLR4 activation. 18,43 In the current study, we firstly conduct Co-IP to find that Ssn B decreased the interaction be-

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

DATA AVA I L A B I L I T Y S TAT E M E N T
Data available on request from the authors.

Enxing Xue
https://orcid.org/0000-0003-3701-2120 F I G U R E 6 Schematic illustration of the potential protective effects of Ssn B in osteoarthritis development