Written informed consent was obtained from each patient for inclusion in the study and the use of their synovial tissues for research. Synovial tissues were collected from patients who underwent total knee replacement due to knee OA (OA group, n=5; one woman, four men; age range, 64–74 years; median age, 69 years) and those who underwent knee arthroscopic surgery due to anterior cruciate ligament injury (Control group, n=5; one woman, four men; age range, 63–71 years; median age, 66 years). This study was performed in accordance with the Ethical Standards of the Declaration of Helsinki and was approved by the Ethics Committee of Suzhou Municipal Hospital (approval no. K-2021-066-K01).

Male C57BL/6 (C57) mice weighing 25–35 g were obtained from the Sino-British SIPPR/BK Lab (Shanghai, China). All mice were provided ad libitum access to food and water, housed at a temperature of 23–25°C and a humidity of 45–55%, with a 12/12-h light/dark cycle for 1 week prior to experimentation. The mice were randomly divided into three groups (n=5 per group): Control group, an MIA-OA group (MIA), and an MJN110 group (MIA + MJN110).

The MIA model of OA pain in the mouse was established as previously described (22). Briefly, the mice in the MIA-OA and MJN110 groups were anesthetized with 50 mg/kg pentobarbital sodium (intraperitoneal injection). The ipsilateral knee joint was trimmed and wiped with alcohol, the knee was kept in a bent position to identify the precise site for injection, and the 26 G needle was inserted to inject MIA (0.1 mg/20 µl) (MilliporeSigma) through the patellar tendon to the gap beneath the patella. A total of 3 weeks after the MIA injection, the mice in the MJN110 group were intraperitoneally injected with 1 mg/kg MJN110 (MilliporeSigma) once a day for a week. On day 28, the mice were sacrificed (CO₂ euthanasia, the container was gradually filled with carbon dioxide at a rate of 49.5% vol/min) for synovial tissue collection.

All animal experiments were approved by the Institutional Animal Care and Use Committee of Nanjing Medical University (approval no. Y20210267).

In the logarithmic growth phase, macrophages (RAW264.7, Beyotime Institute of Biotechnology, cat. no. C7505) were selected and seeded into 6-well plates. After cell adhesion, the cell culture medium (DMEM/F12, 10% FBS, 1% Penicillin-Streptomycin) (Gibco; Thermo Fisher Scientific, Inc.) was discarded and washed once with PBS. Then, the M1 polarization induction medium containing 40 ng/ml LPS (MilliporeSigma) or M2 polarization induction medium containing 40 ng/ml IL-4 (MedChemExpress) was added as appropriate (23–25). Finally, the macrophages were incubated with 5% CO₂ at 37°C for 36 h.

MAGL-siRNA was designed and synthesized by Gima Gene Company and divided into a Control group (untransfected), a siRNA Negative Control group (NC siRNA, sense 5′-UUCUCCGAACGUGUCACGUTT-3′ and antisense, 5′-ACGUGACACGUUCGGAGAATT-3′), and three target gene transfection groups (MAGL siRNA 1 sense, 5′-GCUGGACAUGCUGGUAUUUTT-3′ and antisense, 5′-AAAUACCAGCAUGUCCAGCTT-3′; MAGL siRNA 2, sense 5′-CCAUGACCAUGUUGGCCAUTT-3′ and antisense, 5′-AUGGCCAACAUGGUCAUGGTT-3′; and MAGL siRNA 3 sense, 5′-GCCUACCUGCUCAUGGAAUTT-3′ antisense, 5′-AUUCCAUGAGCAGGUAGGCTT-3′). Macrophages in each group were seeded in 6-well plates. The following experiments were performed in accordance with the instructions of the Gima gene product. The siRNA was diluted with buffer solution and gently mixed to prepare the siRNA transfection diluent. Opti-MEM (200 µl) was diluted with Lipofectamine® 3000 (5 µl) (Invitrogen; Thermo Fisher Scientific, Inc.) and mixed and incubated for 5 min at room temperature. The diluted Lipofectamine® with 100 pmol siRNA was gently mixed and incubated for 20 min. The mixed solution of the complex was added to the cell culture plate, and the cells were incubated for 48 h.

The cells were incubated in blocking buffer (0.5% BSA in 1× PBS) for 30 min at room temperature and stained with FITC-conjugated anti-inducible nitric oxide synthase (iNOS; 1:20, BD Biosciences, cat. no. 610330) and PE-conjugated anti-arginase 1 (Arg1; 1:20, R&D Systems, cat. no. IC5868P) at 4°C in the dark for 30 min. The cells were analyzed using a BD FACSAria^™ II flow cytometer (BD Biosciences) and the BD FACSDiva^™ software version 8.0 (BD Biosciences).

Immunohistochemistry was performed as previously described (26). Briefly, the synovial tissues were cut into 7 µm thick sections. After blocking in 5% goat serum in PBS for 1 h at room temperature, the tissue sections were incubated with primary antibodies against anti-MAGL (1:100, Abcam, cat. no. ab246902), anti-iNOS (1:200, ProteinTech Group, Inc., cat. no. 22226-1), anti-Arg1 (1:200, ProteinTech Group, Inc., cat. no. 16001-1), anti-CD80 (1:300, ProteinTech Group, Inc., cat. no. 14292-1), or anti-CD206 (1:300, ProteinTech Group, Inc., cat. no. 18704-1) overnight at 4°C followed by the secondary antibody (1:1,000, Abcam, cat. no. ab6721) for 2 h at room temperature. An Olympus CH30 (Olympus Corporation) microscope was used to capture images (×200 or ×400).

H&E staining was performed and evaluated as previously described (27). The synovial tissues of the OA patients or OA mice were fixed with 4% paraformaldehyde for 24 h at 4°C and decalcified in 10% EDTA for 2 weeks. The synovial tissues were embedded in paraffin before cutting into 7 µm thick sections and then stained with H&E. Evaluations were performed using an Olympus CH30 microscope (×200 or ×400). The inflammation score was evaluated by scoring the severity of the infiltration of inflammatory cells as follows: 0, no; 1, mild; 2, moderate; 3, severe.

Western blotting was performed as previously described (28). The protein concentrations of the synovial tissues or cells were determined using a BCA Protein Assay kit (Thermo Fisher Scientific, Inc.). Proteins (30 µg) were separated by SDS-PAGE on a 10% gel and transferred to PVDF membranes (MilliporeSigma). The PVDF membranes were blocked with 5% non-fat dry milk for 2 h, followed by incubation with one of the following primary antibodies at 4°C overnight: Anti-MAGL (1:1,000, Abcam, cat. no. ab246902), anti-iNOS (1:1,000, ProteinTech Group, Inc., cat. no. 22226-1), anti-Arg1 (1:5,000, ProteinTech Group, Inc., cat. no. 16001-1), anti-TNF-α (1:1,000, ABclonal, cat. no. A20851), anti-IL-1β (1:1,000, ABclonal, cat. no. A16288), anti-IL-6 (1:1,000, ABclonal, cat. no. A11115), anti-PTEN-induced kinase 1 (PINK1) (1:1,000, ProteinTech Group, Inc., cat. no. 23274-1), anti-Parkin (1:1,000, ProteinTech Group, Inc., cat. no. 14060-1), or anti-β-actin (1:5,000, ProteinTech Group, Inc., cat. no. 81115-1). The following day, the membranes were washed and incubated with the secondary antibody (1:5,000, Abcam, cat. no. ab205718) for 2 h at room temperature. The proteins were detected using Pierce ECL Western Blotting Substrate (Thermo Fisher).

The synovial tissues were permeabilized in 0.1% Triton X-100 for 20 min and then blocked with 5% bovine serum albumin for 1 h at room temperature. Then the synovial tissues were incubated with anti-MAGL (1:200, Abcam, cat. no. ab246902) and anti-iNOS (1:200; Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. MA5-17139) antibodies overnight at 4°C, followed by incubation with the secondary antibody (1:1,000, Abcam, cat. nos. ab150084 and ab150113) for 2 h at room temperature in the dark. An Olympus Fluoview FV3000 (Olympus Corporation) confocal microscope was used to capture images.

The synovial tissues were cut into 1 mm³ segments, preserved in 3% glutaraldehyde for 2 h at 4°C, treated with 1% osmic acid for 2 h at 4°C, and dried with various acetone concentrations before encasing in resin. The samples were sliced into extremely thin sections (90 nm) using an ultramicrotome. A Hitachi H-7560 (Hitachi) transmission electron microscope was used to capture images.

The mechanical pain threshold of mice was measured using an electronic von Frey (Ugo Basile S.R.L, cat. no. 38450) (29). Briefly, the mice were individually placed in a cage with a grid floor for 1 h to adapt to the new environment. An increasing force (g) was applied on the plantar surface of the hind using rigid 0.5 mm diameter polypropylene tips. The mechanical threshold was based on the pressure at which paw withdrawal occurred. This test was done before, and on days 3, 7, 10, 14, 17, 21, and 28 after the injection of MIA. The mechanical threshold was tested thrice, and the mean value was used each time.

The thermal pain threshold of mice was measured using a hot plate (Ugo Basile S.R.L., cat. no. 35250) as described previously (30). The mice were placed on a metal surface of the hot plate equipment maintained at 55±0.1°C. The hot plate was surrounded by a transparent plastic barrier. The latency to jumping off the plate or licking a hind paw was recorded; 30 seconds was used as a cut-off time to protect the paws against injury.

All experiments were performed at least three times. Data are presented as the mean ± SD. Statistical analysis was performed using GraphPad Prism version 8.0.1 (GraphPad Software, Inc.). Normality and homogeneity were evaluated using Shapiro-Wilk and Levene tests, respectively. Differences between groups were compared using an unpaired Student's t-test (two groups), a Mann-Whitney U test (two groups), a Kruskal-Wallis followed by Dunn's test, or a one-way or two-way ANOVA followed by a Tukey's post hoc test for multiple comparisons. P<0.05 was considered to indicate a statistically significant difference.

To investigate the potential role of MAGL in OA, we first collected synovial tissues from patients who underwent total knee replacement due to knee OA (OA group, n=5) and from patients who underwent arthroscopic knee surgery due to anterior cruciate ligament injury (Control group, n=5) to perform H&E staining. As shown in Fig. 1A and B, the synovial tissues of the OA group were thickened, and there was a greater degree of infiltration of inflammatory cells compared with the Control group. Additionally, the synovial tissue score in the OA group was significantly increased compared with that of the Control group (P<0.01). Immunohistochemical staining and western blotting were performed to identify the MAGL levels in the synovial tissue. As shown in Fig. 1C and D, compared with the Control group, the average optical density of MAGL in the OA group was significantly increased (P<0.01). The protein expression levels of MAGL in the synovial tissues of the Control group were low, while that of the OA group was significantly higher (P<0.01) (Fig. 1E and F).

To further determine the polarization phenotype of synovial macrophages from knee OA patients, the expression levels of the M1 polarization markers, iNOS and CD80, and the M2 polarization markers, Arg1 and CD206, were determined by immunohistochemical staining. As shown in Fig. 1G and H, compared with the Control group, the mean optical density of iNOS and CD80 in the OA group was significantly higher (P<0.01). The expression levels of Arg1 and CD206 were low in the synovial tissues of both the Control group and the OA group, and there was no significant difference in the average optical density between the Control group and OA group (P>0.05) (Fig. 1I and J). Furthermore, it was found that the colocalization of MAGL with the M1 macrophage marker iNOS in the synovial tissues of the OA group increased when compared with that of the Control group (P<0.05) (Fig. 1K and L). Hence, these results indicated that MAGL was upregulated in patients with knee OA, with synovial macrophages exhibiting polarization towards an M1 phenotype.

To further investigate the role of MAGL in OA, MJN110, a potent pharmacological inhibitor of MAGL, was intraperitoneally injected in MIA-induced OA mice once a day for a week. First, the expression levels of MAGL in the synovial tissue were detected. As shown in Fig. 2A and B, compared with the Control group, the average optical density of MAGL in the MIA-OA group was significantly increased (P<0.01), and the average optical density of MAGL in the MJN110 group was statistically significantly lower than that in MIA-OA group (P<0.01). Next, the effect of MAGL inhibition on the polarization of synovial macrophages in MIA-induced OA mice was investigated by detecting the expression levels of iNOS and Arg1 in the synovial tissues. As shown in Fig. 2C and D, the mean optical density of iNOS in the MIA-OA group was significantly increased compared with the Control and MJN110 groups (P<0.01). There was no significant difference in the expression levels of Arg1 in the Control group and the MIA-OA group (P>0.05). Compared with the MIA-OA group, the average optical density of Arg1 in the MJN110 group was significantly increased (P<0.01) (Fig. 2E and F).

To further identify the role of MAGL inhibition in OA pain treatment, H&E staining, and mechanical and thermal threshold tests were performed. Based on the staining results (Fig. 2G and H), it was found that the synovial tissue score of the MJN110 group was significantly lower when compared with the MIA-OA group (P<0.01). The results of the mechanical pain and thermal pain tests are shown in Fig. 2I and J. Compared with the Control group, the thresholds of mechanical and thermal pain in the MIA-OA group were significantly reduced after 7 days of MIA injection, which lasted until day 28 (P<0.05). After 7 days of continuous administration of MJN110, the thresholds of mechanical and thermal pain were significantly higher than those of the MIA-OA group (P<0.05). These results suggest that the inhibition of MAGL promotes the polarization of synovial macrophages towards an M2 phenotype and alleviates pain behaviors in OA mice.

Next, the function of MAGL in vitro was explored. Mice macrophages were divided into a Control group, the M1 group (LPS polarization inducing), and the M2 (IL-4 polarization inducing) group. The representative pictures of the cellular morphologies of the three groups are shown in Fig. 3A. Initially, flow cytometry experiments were used to verify the effectiveness of the induction medium. As shown in Fig. 3B, compared with the M2 group, the proportion of iNOS-positive cells in the M1 group was significantly higher. In comparison, the proportion of Arg1-positive cells in the M1 group was significantly lower than that in the M2 group (P<0.01). Meanwhile, western blotting was used to determine the trends of protein levels of iNOS and Arg1 in macrophages in the three groups (Fig. 3C-E).

Subsequently, western blotting was used to examine the expression of MAGL in the macrophages in the three groups. As shown in Fig. 3F-G, compared with the Control group, the protein expression levels of MAGL in the macrophages of the M1 group significantly increased. In contrast, the protein expression levels of MAGL of the M2 group significantly decreased compared with that of the M1 group (P<0.01).

Mice macrophages (LPS polarization inducing) were transfected with MAGL siRNA or NC siRNA (Fig. 4A). Western blotting and flow cytometry experiments confirmed the effect on the polarization of macrophages. As shown in Fig. 4B-E, the protein expression levels of MAGL and iNOS in the LPS group significantly increased compared with the Control group (P<0.01), and the protein expression levels of iNOS significantly decreased following MAGL knockdown when compared with the LPS group. In contrast, Arg1 expression significantly increased (P<0.01). As shown in Fig. 4F, compared with the Control group, the proportion of iNOS-positive cells in the LPS group was significantly higher. In contrast, the proportion of Arg1 positive cells was significantly higher than that in the LPS group after MAGL knockdown (P<0.01). The expressions of inflammatory factors IL-1β, TNF-α, and IL-6 were also examined. As shown in Fig. 4G-J, compared with the control group, the expressions of IL-1β, TNF-α, and IL-6 significantly increased after M1 polarization (P<0.01), while MAGL knockdown significantly reduced the expression of IL-1β, TNF-α, and IL-6 compared with the LPS group (P<0.01). Based on the above results, MAGL knockdown suppressed polarization towards an M1 phenotype and promoted polarization towards the M2 phenotype.

As shown in the electron microscopy images in Fig. 5A, mitochondria with abnormal morphology with mitophagy were observed in the synovial tissues of patients in the Control and OA groups. Fluorescence imaging and mitophagy protein markers were evaluated further to verify the possible role of MAGL on mitophagy. The mitophagosome is a red fluorescent probe, which is bound to mitochondria in the cell by chemical bonds, while the lysosome is a green fluorescent probe; thus, yellow fluorescence is observed when they are colocalized. The amount of yellow fluorescence represented the levels of mitophagy. As shown in Fig. 5B, there was almost no yellow fluorescence in the Control group, only a little yellow fluorescence in the LPS group, and a considerable amount of yellow fluorescence in the macrophages in the MAGL siRNA group. The PINK1 and Parkin expression levels in macrophages were examined by western blotting. As shown in Fig. 5C-E, compared with the LPS group, the protein expression levels of PINK1 and Parkin in the MAGL siRNA group were significantly increased (P<0.01). The above results indicated that MAGL knockdown by siRNA increased mitophagy levels in M1 macrophages.