The protective effect of sinapic acid in osteoarthritis: In vitro and in vivo studies

Abstract The anti‐inflammatory effect of sinapic acid (SA) has been reported in several studies. However, whether SA has the same effect on osteoarthritis (OA) has yet to be clearly elucidated. We designed a series of in vitro and in vivo procedures to verify the above conjecture. Compared with controls, SA‐pretreated human chondrocytes showed lower levels of interleukin (IL)‐1β‐induced IL‐6, prostaglandin E2 (PGE2), nitric oxide (NO) and tumour necrosis factor‐α (TNF‐α) in vitro. Meanwhile, SA could also reverse the degradation of type II collage and aggrecan, as well as the overproduction of matrix metalloproteinase‐9 (MMP‐9) and matrix metalloproteinase‐13 (MMP‐13), inducible nitric oxide synthase (iNOS), cyclooxygenase (COX)‐2 and a disintegrin and metalloproteinase thrombospondin motifs (ADAMTS)‐5. Furthermore, activation of nuclear factor κB (NF‐κB), which was induced by IL‐1β, was also inhibited by SA through the pathway of nuclear factor‐erythroid 2‐related factor‐2 (Nrf2)/heme oxygenase 1. In vivo, SA could delay the progress of mice OA models. We propose that SA may be applied as a potential therapeutic drug in OA treatment.

inflammation caused by IL-1β may be a potential therapeutic target for patients with OA. 7 It was well known that nuclear factor κB (NF-κB) participated in the production of several inflammatory factors which were involved in the pathogenesis of OA and induced by IL-1β. 8,9 As an important transcription factor in maintaining cell homeostasis, nuclear factor-erythroid 2-related factor-2 (Nrf2) could interact with a variety of signalling pathways to exert antioxidant and anti-inflammatory effects. An increasing amount of evidence demonstrates that there was an interaction between members of Nrf2 and NF-κB pathways. 10 Expression of antioxidant gene such as heme oxygenase-1 (HO-1), superoxide dismutase and NAD(P)Hquinone oxidoreductase-1 could be induced by the activation of Nrf2. 10,11 Among them, HO-1 plays a pivotal role in Nrf2-mediated NF-κB inhibition. 10 As reported, the production of inflammatory cytokines and NF-κB activation could be inhibited by the activation of the Nrf2 signalling pathway. [12][13][14] Furthermore, Nrf2 activation shows a chondroprotective potential in mice OA models. 15 Targeting Nrf2 may therefore be useful in treating cartilage degeneration.
Sinapic acid (SA) is widely found in spices, fruits, vegetables, cereals and oilseed crops. 16 It has been reported that SA has various biological functions such as antimicrobial, anti-inflammatory, antioxidant as well as anti-anxiety effects. 17,18 Previous studies discovered that SA could inhibit lipopolysaccharide-induced inflammation through NF-κB inactivation in vascular endothelial cells. 19 Meanwhile, SA reduced the production of PGE2, IL-1β, TNF-α, inducible nitric oxide synthase (iNOS) and cyclooxygenase (COX)-2 in RAW 264.7 macrophages. 20 Additionally, SA activated the Nrf2/HO-1 pathway, and led to NF-κB inactivation in cisplatin-induced nephrotoxicity in rats. 21 However, the antiinflammatory effect of SA in OA has never been investigated. In this study, we used in vivo and in vitro methods to find out whether SA has an anti-inflammatory effect of in OA and the possible underlying mechanism of SA.  Liquid supernatant was collected and then centrifuged at 100 g for 5 minutes. The inner cell mass was obtained and suspended in DMEM/ F12 with 10% FBS and 1% antibiotic mixture and incubated in an atmosphere of 95% air and 5% CO 2 at 37°C. The medium was changed every 2-3 days. Cells were passaged when up to 80% to 90% confluence using 0.25% trypsin-EDTA solution. To avoid the phenotype loss, only passage 1 and 2 were used in our study.

| Griess reaction and ELISAs
The interaction of NO in culture medium was determined by Griess reaction as previously described, while the concentration of PGE2, MMP-1, MMP-13 and ADAMTS-5 in the culture medium was detected using commercial ELISA kit (Minneapolis, MN) according to the manufacturer's instruction. All assays were performed in duplicate.   Table 1.

| Western blotting
Total proteins were extracted from human chondrocytes using RIPA lysis buffer with 1 mmol/L Phenylmethanesulfonyl fluoride and on the ice for 10 minutes followed by 15 minutes centrifugation at 13 800 g and 4°C, and then protein concentration was measured using the BCA protein assay kit. Forty micrograms of protein were loaded onto an SDS-PAGE gel and transferred to a PVDF membrane (Bio-Rad). The membranes were blocked with 5% non-fat dry milk for 2 hours at room temperature and subsequently incubated sequentially with primary antibodies(all 1:1000 dilution) against p65, IκB-α, iNOs, COX-2, Nrf2, HO-1, MMP-9, MMP-13, Lamin-B and GADPH overnight at 4°C, and followed by subsequently incubation in HRP-conjugated secondary antibodies (1:3000) for 2 hours at room temperature. After washing three times with TBST for 5 minutes, the blots were visualized by electrochemiluminescence plus reagent (Invitrogen) and afterwards the intensity of these membranes was quantified with Image Lab 3.0 software (Bio-Rad).

| Immunofluorescence microscopy
Human chondrocytes were seeded on slices in glass plates in a sixwell plate at a density of 3 × 10 5 cells/mL and incubated for 24 hours. Afterwards, glass coverslips with chondrocyte monolayers were rinsed three times in PBS before fixation using 4% paraformaldehyde for 15 minutes and followed by permeation using the 0.5% Triton X-100 for 15 minutes at room temperature. Then, the sample was blocked with 5% goat serum for 1 hour at room temperature, rinsed with PBS and incubated with primary antibody against collagen-II and p65 (1:200) at 4°C overnight. On the next day, the glass plates were incubated with fluorescein-conjugated goat anti-rabbit IgG antibody (1:400) for 1 hour and labelled with DAPI (Invitrogen) for 1 minute. Finally, images were captured using a fluorescence microscope (Olympus Inc, Tokyo, Japan).

| Animal experiments
Ten-week-old C57BL/6 male wild-type mice were purchased from the Animal Center of Chinese Academy of Sciences, Shanghai, China.
The protocol for animal care and use conformed to the guidelines set forth by the Chinese National Institutes of Health and was approved by the Animal Care and Use Committee of Wenzhou Medical University. OA induction was performed as previously described.
Briefly, after anaesthesia with peritoneal injection of 2% pentobarbital, the cranial attachment of the medial meniscus to the tibial plateau (medial meniscotibial ligament) of the right knee was transected with a microsurgical knife. The lateral meniscotibial ligament was identified and protected during the surgery. A sham operation, consisting of an arthrotomy without the transaction of medial meniscotibial ligament, was also performed in the right knee joint of mice in sham group.
In this study, mice were randomly divided into three groups of 10 mice, including a sham control group (sham), an OA group and an OA treated with SA group (SA). Mice in SA group received a gavage of SA (20 mg/kg/d) for 14 days after surgery while mice in OA group received a gavage of 0.5% carboxymethylcellulose sodium. All animals were sacrificed at 8 weeks after surgery. Knee joint tissues were collected for further analysis.

| Effect of SA on ECM synthesis and degradation in IL-1β-induced human OA chondrocytes
We pre-treated chondrocytes with SA (40, 80, and 160 μmol/L) for 60 minutes and then chondrocytes were treated or untreated with IL-1β (10 ng/mL) for 24 hours. The results showed that IL-1β noticeably enhanced the expression of ADAMTS and MMP-13 at mRNA and protein level, MMP-9 at protein level in chondrocytes ( Figure 2C,D), which were all remarkably decreased after SA treatment ( Figure 3A).
Otherwise, as shown in Figure 3E and F, SA distinctly reduced mRNA downregulation of type II collagen (collagen-II) and aggrecan caused by IL-1β stimulation. Consistently, immunofluorescence results revealed that SA considerably inhibited the protein degradation of collagen-II and the expression of MMP-13 ( Figure 3C,D).

F I G U R E 4
Effect of SA on IL-1β-induced NF-κB activation in human OA chondrocytes. Chondrocytes were pre-treated for 1 hour with various concentrations of SA (40, 80, and 160 μmol/L), followed by stimulation with or without IL-1β (10 ng/mL) for 2 hours. NF-κB p65 activation (in nucleus) and IκB-α degradation (in cytoplasm) were determined by Western blot (A) and quantification analysis (B). The nuclei translocation of p65 was detected by immunofluorescence combined with DAPI staining for nuclei (C). The values are mean ± SD of three independent experiments. ## P < 0.01 compared with control group. *P < 0.05, **P < 0.01 compared with IL-1β group. IL, interleukin; IκB, inhibitor of kappa B; NF-κB, nuclear factor κB; OA, osteoarthritis; SA, sinapic acid F I G U R E 5 Effect of SA on Nrf2/HO-1 pathway. Chondrocytes were pre-treated for 1 hour with various concentrations of SA (40, 80, and 160 μmol/L), followed by stimulation with or without IL-1β (10 ng/mL) for 2 hours. Nrf2 activation (in nucleus) and HO-1 (in cytoplasm) were determined by Western blot (A) and quantification analysis (B). After Nrf2 knock down, the protein expressions of Nrf2 and p65 in nuclear and HO-1 in cytoplasm in chondrocytes treated as above were visualized by Western blot (C), and quantified in (D). The production of PGE2, MMP-13 and collagen-ll was assessed by ELISA, and the expression of NO was assessed using Griess reagent (E). The values are mean ± SD of three independent experiments. ## P < 0.01 compared with Control group. **P < 0.01 compared with IL-1β group. HO-1, heme oxygenase-1; IL, interleukin; Nrf2, nuclear factor-erythroid 2-related factor-2; SA, sinapic acid; MMP, matrix metalloproteinase; NO, nitric oxide; PGE2, prostaglandin E2

| Effect of SA on NF-κB activation in IL-1βinduced human OA chondrocytes
IL-1β stimulation led to the degradation of IκBα, and increasing phosphorylation of NF-κB p65 and the IκB-α in human OA chondrocytes, which could be dramatically inhibited by SA ( Figure 4A).
In order to determine whether SA could inhibit NF-κB p65 nuclear translocation induced by IL-1β, chondrocytes were pre-treated with or without SA (160 μmol/L) for 60 minutes and followed by a stimulation with IL-1β (10 ng/mL) for 120 minutes, after which immunofluorescence staining was performed. As shown in Fig

| DISCUSSION
The incidence of OA has increased alongside an increasingly ageing population. 22 There is an increasing evidences that inflammation plays a critical role in the initiation and progression of OA. 23  MMPs, the most important proteolytic system in degrading the ECM, comprise a class of proteolytic enzymes. As reported in several studies, MMP-9 and MMP-13 could irreversibly and efficiently digest collagen-II. 30 ADAMTS-5 plays the key role in the digestion of aggrecan. 31 Growing evidence has demonstrated that in clinical patients, the progression of OA could be delayed by ADAMTS inhibitors, 32 especially ADAMTS-5 inhibitors, which lessened aggrecan loss in human OA cartilage explants through siRNA. 33 In our research, we discovered that SA clearly reduced IL-1β-induced MMP-13 and ADAMTS-5 production in human OA chondrocytes at mRNA as well as protein levels.
It is well known that NF-κB signalling pathway could regulator inflammatory mediators associated with OA development. 9 As reported previously, phosphorylation of IκBα and p65 release could be triggered by IL-1β stimulation, consequently translocated from the cytoplasm to the nucleus and lead to the inflammatory mediators production. 34 In addition, NF-κB p65-specific siRNA activation could lead to the inhibition of NF-κB p65 and COX-2, iNOS, MMP-9 and MMP-13 production induced by IL-1β in human chondrocytes. Furthermore, the Nrf2/HO-1 pathway is deemed a target for inflammation induced by anti-NF-κB. 10 To maintain its stability, Nrf2 normally binds to Kelch-like ECH-associated protein-1 (Keap1) under normal physiological conditions. Nrf2, dissociating from Keap1, enters the nucleus and later unite with antioxidant-responsive elements, resulting in the downstream genes upregulation, such as HO-1. It has been reported that Nrf2 and HO-1 could lighten inflammation through the inhibition of p65 translocation. 10 It has also been reported that Nrf2 activation could prevent OA progress via inflammatory response inhibition. 15 Moreover, SA led to the inhibition of NF-κB in cisplatininduced nephrotoxicity in rats via activation of the Nrf2/HO1 pathway. 21 In our results, SA inhibited the IL-1β-induced inflammatory response via the activation of Nrf2.
In conclusion, our findings demonstrated that SA could suppress

CONFLI CT OF INTEREST
The authors declare that they have no conflict of interest.