4‐octyl Itaconate inhibits lipopolysaccharide (LPS)‐induced osteoarthritis via activating Nrf2 signalling pathway

Abstract Small molecule drug intervention for chondrocytes is a valuable method for the treatment of osteoarthritis (OA). The 4‐octyl itaconate (OI) is a cellular derivative of itaconate with sound cell permeability and transformation rate. We attempted to confirm the protective role of OI in chondrocytes and its regulatory mechanism. We used lipopolysaccharide (LPS) to induce chondrocyte inflammation injury. After the OI treatment, the secretion and mRNA expression of Il‐6, Il‐10, Mcp‐1 and Tnf‐α were detected by ELISA and qPCR. The protective effect of OI on articular cartilage was further verified in surgical destabilization of the medial meniscus model of OA. Cell death and apoptosis were evaluated based on CCK8, LDH, Typan blue staining, Annexin V and TUNEL analyses. The small interfering RNAs were used to knockout the Nrf2 gene of chondrocytes to verify the OI‐mediated Nrf2 signalling pathway. The results revealed that OI protects cells from LPS‐induced inflammatory injury and attenuates cell death and apoptosis induced by LPS. Similar protective effects were also observed on articular cartilage in mice. The OI activated Nrf2 signalling pathway and promoted the stable expression and translocation of Nrf2 into the nucleus. When the Nrf2 signalling pathway was blocked, the protective effect of OI was significantly counteracted in chondrocytes and a mouse arthritis model. Both itaconate and its derivative (i.e., OI) showed important medical effects in the treatment of OA.

the articular cartilage has limited ability to repair itself after acute or chronic injuries or joint degeneration. To date, there is no effective drug to improve the condition of OA, while the treatment is mainly to relieve the symptoms of patients with the surgical treatment as the only effective method to treat patients of OA at their late stages. [4][5][6] For both the post-traumatic arthritis and primary OA, the injury of chondrocytes caused by inflammatory reaction is closely related to the pathogenesis and development of OA. [7][8][9] Lipopolysaccharide (LPS) is one of the structural components in the outer wall of Gramnegative cells to provide a positive immune activation. 10 Local injection of LPS into the joint induces not only synovitis, but also inflammatory injury of articular cartilage. [11][12][13] The model of chondrocyte inflammation induced by LPS can effectively reflect the degree of the cartilage injury in OA.
In recent years, the understanding of metabolic physiology is advancing more rapidly than that of material and energy metabolisms.
The disorder of one kind of metabolite can affect many signalling pathways. [14][15][16][17] Metabolic inflammation also plays a crucial role in the pathogenesis of OA, opening a new direction to explore potential treatments of OA. 18 Furthermore, studies have shown that as an intermediate metabolite of tricarboxylic acid cycle (TCA cycle), itaconate plays important anti-inflammatory and anti-apoptotic roles. Specifically, itaconate weakens the succinate dehydrogenase (SDH) activity in cells, but not as effectively as traditional SDH inhibitors (e.g., sodium malonate), suggesting that itaconate may play an anti-inflammatory role through different mechanisms. 19 Itaconate and its derivatives have also been shown to regulate the IkBz-ATF3 inflammatory axis. 20 It has been confirmed that itaconate is a novel and effective activator of nuclear factor erythroid-2-related factor 2 (Nrf2), stabilizing and activating Nrf2 by alkylating cysteine residues on Keap1 and promoting the dissociation of Keap1-Nrf2 complex. Itaconate is an α, β unsaturated carboxylic acid with electrophilicity. Hence, it can interact with proteins containing a thiol group at the cellular level, in a process called the electrophilic stress response (ESR). 21 4-octyl itaconate has a ester group that allows it to infiltrate cells without the transporter protein. In addition, the electrophilic α, β unsaturated carboxylic acid of OI helps to bind to the Keap1 cysteine. 19 Based on these characteristics, a novel type of itaconate derivative 4-octyl itaconate (OI) was designed to effectively activate Nrf2 with anti-inflammatory and protective roles in macrophages and other cells. 19,22,23 Nuclear factor erythroid-2-related factor 2 is a crucial transcription factor mediating anti-inflammation and anti-apoptosis. [24][25][26] Under physiological conditions, Nrf2 and Keap1 join to form polymers, regulated by Keap1. 19,27,28 Once stimulated, the cysteine residue of Keap1 is modified, resulting in the release and accumulation of Nrf2 in the nucleus, binding with anti-oxidative response element (ARE), and ultimately initiating the transcription and expression of a variety of anti-inflammatory and anti-apoptotic genes, including Glutamate-Cysteine Ligase, Catalytic Subunit (Gclc), NADH Quinone Oxidoreductase 1 (Nqo1) and Haem Oxygenase-1 (Ho-1). 29,30 Furthermore, Nrf2 signalling pathway shows a strong antiinflammatory effect by limiting the progress of NF-κB. 29,31 Studies have demonstrated that activation of Nrf2 inhibits cartilage destruction in models of OA. 32,33 Meanwhile, the Nrf2 knockout mice showed more severe cartilage damage than the wild-type mice. 34,35 These results indicate that Nrf2 signalling pathway is involved in the formation and treatment of OA.

| Enzyme-linked immunosorbent assay
The mouse ATDC5 cells and human C28I2 cells were inoculated in 24-well culture plates. After treated with LPS and OI, the superna-

| NF-κB p65 transcription factor assay
The NF-κB p65 transcription factor assay was conducted using the NF-κB p65 TF Assay Kit based on the manufacturer's instructions.
Nuclear lysate was extracted by Nuclear Protein Extraction Kit. The nuclear lysate of 10 μl was placed in a 96-well plate to react with the dsDNA. After multiple cleanings, 100 μl primary antibody of NF-κB (kept at room temperature for 2 h) and 100 μl goat anti-rabbit horseradish peroxidase (HRP) conjugate antibody (kept at room temperature for 2 h) were added into the reaction well in turn. After one-step colour reaction, the absorbance at 450 nm was observed and the OD value was calculated.

| Arthritis model in vivo
The male C57BL6 mice were purchased from Charles River. After anaesthesia, the medial meniscus ligament was cut-off with a microblade, dissociated, with the tissue and skin sutured layer by layer to establish a surgical destabilization of the medial meniscus model of osteoarthritis (DMM). LPS was used to enhance the development of OA in mice. Specifically, the mice were injected with LPS (2.5 mg/ kg) into the knee joint, and the control group was injected with the same volume of normal saline solution at the age of 8 weeks. Then, OI (50 mg/kg) was injected into the knee joint three times in a week on the 1st, 4th and 7th day. All the mice were treated with the left hindlimb to keep the mice free to move.

| Nrf2 knockout in a mouse arthritis model
To establish a mouse arthritis model. LPS (2.5 mg/kg) was used to enhance the development of OA in mice. Then, OI (50 mg/kg) was injected into the knee joint three times in a week on the 1st, 4th and 7th day, and ML385 was injected intraperitoneally 1 h before each injection. All the mice were treated with the left hindlimb to keep the mice free to move.

| Haematoxylin-eosin (HE) staining
The distal femur was dissected and fixed with 4% PFA and decalcified. Paraffin sections were quickly stained by HE with the cytoplasm stained red and the nucleus stained blue. TA B L E 1 Primer sequences used for qPCR. 'F' and 'R' indicate forward and reverse primers respectively

| Safranin O-Fast green staining
The distal femur was dissected and fixed with 4% PFA and decalcified. Paraffin sections were quickly stained by safranin O-fast green. The cartilage matrix was stained dark red, and the cytoplasm, muscle, collagen fibre and bone tissue were stained green to directly reflect the structure of articular cartilage and bone tissue.

| OARSI Scoring
The histology of articular cartilage in mice was graded based on the Osteoarthritis Research Society International (OARSI) score system. The score criteria included the absence/presence of safranin O staining, the fracture depth of articular cartilage and the cumulative degree of articular surface.

| Cell viability assay
The mouse ATDC5 cells and human C28I2 cells were seeded in 96well culture plates. Once the density of the cells treated with LPS and OI reached 70%, CCK-8 (10 μl) was added to stain for 30 min.
Then, the absorbance at 550 nm was observed, and the OD value was calculated.

| LDH cytotoxicity assay
The mouse ATDC5 cells and human C28I2 cells were inoculated in 96-well culture plates. When the density of cells treated with LPS and OI reached 70%, the supernatant (120 μl) was collected to react with 60 μl LDH working solution for 2 h. Then, the absorbance was observed at 490 nm, and OD value was calculated.

| Trypan Blue staining cell viability assay
The mouse ATDC5 cells and human C28I2 cells were cultured in 6well culture plates at a density of 1 × 10 6 cells/well. After the treatment of LPS and OI, cells were suspended by EDTA-free trypsin and added with equal volume of diluted 1× Trypan Blue stain. The blue 'dead' cells were counted using a blood cell count plate, and the ratio (100%) of 'dead' cells to the total cells was calculated.

| Flow cytometry
The mouse ATDC5 cells and human C28I2 cells were cultured in 6-

| TUNEL assay
The mouse ATDC5 cells and human C28I2 cells seeded in 96-well culture plates at 0.5 × 10 5 cells/well were permeabilized with proteinase K (20 μg/ml) at room temperature for 5 min. Each sample was stained with 50 μl TUNEL TdT Enzyme working solution for 30 min.
The fluorescence intensity of the sample was detected with EX at 546 nm and EM at 570 nm.

| Western blotting assay
The mouse ATDC5 cells and human C28I2 cells were cultured in 100 mm Cell Culture Dishes with a density of 2 × 10 6 cells/dish.
After the treatment of OI, the cells were collected by centrifugation.

| Immunofluorescence
Based on a previous study, the cells were immobilized with 4% PFA for 30 min and sealed with 5% BSA containing 0.1% Triton-X for 1 h. 44 The cells were incubated with primary antibody at 4°C overnight and then incubated with secondary antibody at 25°C for 1 h.
The protein expression and localization performed in dark were observed under the inverted laser scanning confocal microscope.

| Nqo1 activity assay
As previously described, a total of 100 μl menadione was used as substrate to react with the lysate for 48 h. 19 The activity of Nqo1 was assessed in cell lysates. The OD value of the mixture was calculated at 565 nm.

| Nrf2 siRNA
Based on the Nrf2 sequence, the Nrf2 siRNAs were designed and synthesized by Genepharma (

| Statistical analysis
Each experiment with 3 or 4 samples included was repeated thrice.
Data presented as the mean ± standard deviation were subjected to one-way ANOVA (nonparametric or mixed) and two-tailed

| OI inhibits the inflammation induced by LPS in mouse ATDC5 cells and human C28I2 cells
To investigate the potential effects of OI on pro-inflammatory cytokines, the mouse ATDC5 cells were first induced with LPS (10 μg/ ml) for 12 h. Compared with the control group, the levels of Il-6 and Mcp-1 were significantly increased with the LPS induction ( Figure 1A,B), whereas no significant change was revealed in the expression levels of Il-10 and Tnfα ( Figure 1C,D). With the treatment of OI, the levels of Il-6 and Mcp-1 in the supernatant were notably inhibited ( Figure 1A,B). The levels of IL-10 and TNFα exhibited a low sensitivity to OI ( Figure 1C,D). LPS-induced production of Il-6  in vivo of OA induced by LPS ( Figure 2E). These results demonstrated that OI alleviated the inflammatory damage caused by LPS in DMM mice.

| OI attenuates cell death and apoptosis in mouse ATDC5 cells and human C28I2 cells induced by LPS
Results showed that the mouse ATDC5 cells treated with LPS (10 μg/ ml, 12 h) showed decreased cell viability ( Figure 3A), increased LDH release level ( Figure 3B) and induced cell death ( Figure 3C).

Treatment of OI significantly inhibited LPS-induced cytotoxicity in
the ATDC5 cells in a dose-dependent manner ( Figure 3A,B). A single treatment of OI at 25, 50, 100 and 200 μM revealed no significant difference in cytotoxicity ( Figure 3A,B). OI was also demonstrated to show the ability to reduce the damage to LPS-induced cell viability ( Figure 3D), increase LDH release levels ( Figure 3E) and induce cell death ( Figure 3F) in the human C28I2 cells. The results showed that OI (100 μM) efficiently protected the ATDC5 and C28I2 cells from the injury induced by LPS ( Figure 3A,B,D,E). Therefore, the 100 μM was chosen as the optimal concentration of OI for further experiments.
The potential effect of OI on apoptosis was further verified by flow cytometry analysis. Results showed that the ratio of apoptosis and death rate induced by LPS was significantly increased, while the ratio of Annexin V was decreased in the ATDC5 cells treated with OI ( Figure 3G,H). The TUNEL staining ratio was also decreased in the ATDC5 cells treated with OI ( Figure 3I). The apoptosis activa-

tion was evidently induced by LPS as demonstrated by increased
Annexin V ratio ( Figure 3J,K) and TUNEL staining ( Figure 3L). These results showed that OI played an important role in the protection of chondrocytes.

| OI activates Nrf2 signalling pathway and downstream genes in mouse ATDC5 cells and human C28I2 cells
The potential effects of OI on Nrf2 signalling pathway in ATDC5 cells were investigated. The expression of Nrf2 protein in cytoplasm and nucleus was detected after the stimulation of OI (100 μM, 12 h). The results revealed that the expression of Nrf2 protein was slightly increased in cytoplasm but significantly increased in nucleus ( Figure 4A,B). The results of immunofluorescense analysis confirmed that Nrf2 was strongly expressed and accumulated in the nucleus under the OI treatment ( Figure 4C). The results of the expression and localization analyses of NRF2 protein in C28I2 cells further showed that NRF2 was stabilized with enhanced nuclear translocation under the treatment of OI ( Figure 4D-F), which was the crucial step of NRF2 activation.
The effects of OI on Nrf2 downstream gene expression were further verified by IF and activity analyses. Results showed that the expression of Gclc protein was significantly increased in ATDC5 cells treated with OI ( Figure 4G) and the activity of Nqo1 was significantly increased ( Figure 4H). The transcription levels of the Nrf2-dependent genes Gclc and Nqo1 were significantly increased in ATDC5 cells ( Figure 4I). However, the overexpression of Nrf2 mRNA was not achieved by the treatment of OI ( Figure 4I). The Nrf2-dependent molecule GCLC expression F I G U R E 4 4-octyl itaconate activates Nrf2 signalling pathway and downstream genes in mouse ATDC5 cells and human C28I2 cells. The ATDC5 cells and the C28I2 cells are treated with OI (100 μM) for 12 h. The expression of Nrf2 protein in cytoplasm and nucleus is analysed by Western blot (A, B, D and E). The protein expression and localization are detected by IF (C and F). The protein expression and localization of Gclc are detected by IF (G and J). Using menadione as substrate, the Nqo1 activity of lysates is determined by ELISA (H and K). The mRNA levels of Nrf2, Gclc and Nqo1 in cells are detected by qPCR (I and L). Data are presented as mean ± standard deviation (n = 3). Symbols '*' indicates the significant difference set at p < 0.05 in comparison with control ('Ctrl') and NQO1 activity were significantly increased in C28I2 cells treated with OI ( Figure 4J,K). The transcriptions of genes GCLC and NQO1 were also enhanced in C28I2 cells ( Figure 4L). Our results showed that OI stimulated Nrf2 signalling pathway in both the ATDC5 and C28I2 cells and promoted downstream genes expression.

| Nrf2 activation involved in inhibiting LPSinduced pro-inflammatory reactions in mouse ATDC5 cells and human C28I2 cells treated with OI
The interference of OI on LPS-induced ATDC5 and C28I2 cells was Furthermore, the changes in serum Il-6 levels are consistent with the above performance ( Figure 6E). These results revealed that OI relieved cartilage injury by activating the Nrf2 signalling pathway in a mouse arthritis model.

| DISCUSS ION
Osteoarthritis affects 4%~7% of the population worldwide, causing severe clinical challenges and financial burdens for patients. 45  Our results showed that OI promoted the expression of Nrf2 and downstream genes GCLC and NQO1 not by inducing Nrf2 gene transcription but by activating inactivated Nrf2 in the cytoplasm. These results provide strong evidence to support the application of itaconate in the treatment of inflammation. 19 Previous studies have shown that itaconate impairs SDH activity in cells but not as effectively as traditional SDH inhibitors (e.g., sodium malonate), suggesting that itaconate may play an antiinflammatory role through different mechanisms. 19 Besides activating Nrf2, itaconate and its derivatives have also been shown to regulate the IkBz-ATF3 inflammatory axis. 20 Our results suggest that OI activates Nrf2 signalling pathway in chondrocytes.
Specifically, OI induces stabilization of Nrf2 protein and nuclear transposition, leading to the expression of Nrf2 target genes Gclc and Nqo1. More importantly, once the Nrf2 signalling pathway is blocked, the protective effect of OI on chondrocytes is not observed. These results confirm that Nrf2 signalling activation mediates OI to inhibit LPS-induced chondrocyte inflammation. We note that it is important to further explore the potential functional mechanism of Nrf2 in the inhibition of LPS-induced inflammatory injury of articular cartilage and related target genes activated by OI as well as the possible anti-inflammatory and protective functions of Nrf2.
In conclusion, we report for the first time the effect and functional mechanism of itaconate in the treatment of cartilage inflammation both in vitro and in vivo. OI can effectively activate Nrf2 signalling pathway and downstream related anti-inflammatory and anti-apoptotic genes, ultimately inhibiting the production of LPSinduced inflammatory factors, the NF-κB activation, cell necrosis and apoptosis (Figure 7). Therefore, we conclude that itaconate contains significant importance for investigating the treatment of OA, opening a novel direction to explore other metabolites in the treatment of inflammation. More clinical trials are needed to verify in vivo the application of itaconate in the treatment of OA. We speculate that more in-depth studies of metabolites in OA signalling pathway are necessary to make a breakthrough in the treatment of OA.