Animals.

Male Wistar rats, aged 7 to 8 weeks, were used in the experiments and housed in specific-pathogen-free facilities in pathogen-free conditions. The rats were fed ad libitum with sterile rat-specific diet feed (LAD 0100; Trophy Feed Technology, NanTong, China) and randomly distributed in cages. All experiments were conducted in strict conformance with guidelines adopted from the U.S. National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Ethics Committee of Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine.

Induction of CIA and Drug Administration.

Rats were immunized with an emulsion containing 100 μg bovine collagen II (Chondrex, Redmond, WA) with an 1-week interval. Collagen II emulsified 1:1 in incomplete Freund’s adjuvant (Chondrex) was injected intradermally around the base of the tail on day −7, and the same emulsion was injected by the same route on day 0 (Larsson et al., 2005; Brenner et al.,2007). Three experimental groups were injected intraperitoneally with 5, 25, and 50 mg DMY/kg rat body weight, respectively, and the same administrations were performed on the rats every other day for 5 weeks. Six rats in the normal group were injected with phosphate buffer (PBS) in the same way as those in the model group. Clinical monitoring of the rats was performed twice a week after the second immunization to determine body weight and test the volume of the hind paws. Simultaneously, rats were scored blindly on a scale from 0 to 3 per hind paw (0–12 per rat) as follows (Leavenworth et al., 2013): 0, normal; 1, mild swelling and/or erythema confined to the midfoot or ankle joint; 2, moderate edematous swelling extending from the ankle to the metatarsal joints; and 3, pronounced swelling encompassing the ankle, foot, and digits. Immunized rats were randomly divided into four groups: a vehicle control group injected with PBS and three DMY-treated groups (n = 7 in each group). DMY was of high-purity liquid chromatography grade, the molecular structure of which is (2R,3R)-3,5,7-trihydroxy-2-(3,4,5-trihydroxyphenyl)-2,3-dihydrochromen-4-one, and was ≥98% pure (Sigma-Aldrich, St. Louis, MO, US). DMY was administered intraperitoneally in a dosage of 5, 25, or 50 mg/kg in 1 ml volume every other day.

Enzyme-Linked Immunosorbent Assay of Inflammatory Cytokines.

Peripheral blood was collected in centrifuge tubes after the last administration of DMY and was centrifuged at 3500 rpm for 10 minutes at room temperature with a cooling centrifuge until coagulation. The clear plasma layer was obtained and stored at −80°C in a deep freezer before assaying. Levels of cytokines interleukin (IL)-1β, tumor necrosis factor (TNF)-α, and IL-17A in the plasma of experimental rats were detected by an enzyme-linked immunosorbent assay kit (Multisciences, Hangzhou, China) according to standard protocols. The optical density was measured using an automatic microplate reader (Infinite M200 PRO NanoQuant; Tecan, Männedorf, Switzerland) at a wavelength of 450 nm. Each sample was tested in duplicate, and concentrations were determined by reference to standard curves, generated by using the optical density values of standard solutions with 1:2 serial dilutions, for estimating cytokine content.

Histopathological Analysis of Joints.

On day 34 after the second immunization, the rats were anesthetized and rapidly sacrificed by cervical dislocation. Whole knee joints of the hind legs were carefully separated and fixed in 4% paraformaldehyde (PFA) for 24 hours. The joints were decalcified in 0.5 M EDTA, dehydrated through an alcohol gradient, embedded in paraffin wax, sectioned at 4-μm thickness in a sagittal plane, and subjected to routine hematoxylin and eosin staining. Sections were photographed with a cool CCD camera attached to a Zeiss Axioskop 2 plus microscope (Carl Zeiss Microimaging). The micrographs were evaluated by two blinded observers (B.-X.X. and Q.-S.W.). To determine the histopathology of the joint, inflammatory activity was observed in the samples with hematoxylin and eosin staining and scored from 0 to 3 according to the following criteria: 1, mild inflammatory infiltration with no synovial lining cell hyperplasia; 2, moderate infiltration with some synovial lining cell hyperplasia; and 3, severe infiltration with marked synovial lining cell hyperplasia (Buttgereit et al., 2009).

Synovial Cell Culture and Stimulation.

Synovial tissue specimens were isolated from rats with CIA and rinsed with PBS two to three times. Then the tissues were minced into small blocks (approximately 1-mm³ pieces), transferred to a small culture flask, and tiled uniformly with a spacing of approximately 1 mm. For the air-liquid interface culture, each flask contained a small amount of Dulbecco’s modified Eagle’s medium (DMEM; Thermo Fisher Scientific, Rockford, IL) with high glucose supplemented with 10% fetal bovine serum (FBS; Gibco Life Technologies, Carlsbad, CA) and 1% antibiotics (penicillin and streptomycin) (Thermo Fisher Scientific) at 37°C in a humidified thermostatic incubator (Thermo Fisher Scientific) under 5% carbon dioxide to ensure that pieces adhered to the flask (Liu et al., 2017). After 4–6 hours, an appropriate amount of fresh culture medium was added to the flask to continue the culture, and the medium was totally exchanged every 2 days with the same volume of fresh DMEM. Within 7 days, FLSs could migrate out from the tissue explant and achieve approximately 80% confluence (Shang et al., 2017); then the tissues were removed, and the liberated FLSs were trypsinized, collected, resuspended, planted for expansion, and used at passages 3–6 (Xiao et al., 2014; Angiolilli et al., 2017).

Identification of FLSs.

FLSs that migrated from synovial tissues on days 0, 4, 5, and 6 were observed. FLSs were seeded into a six-well plate with 8 × 10⁴ cells/well and cultured for 24 hours. The cells were fixed with 100% methanol for 5 minutes, permeabilized with 0.1% Triton X-100 for 5 minutes, and then blocked with 1% bovine serum albumin (BSA) in 0.1% PBS-Tween for 1 hour. Immunofluorescence staining of FLSs was performed with purified anti-Vimentin antibody (Abcam, Cambridge, UK) at a working dilution of 1:250 overnight at 4°C and was followed by a further incubation at room temperature for 1 hour with goat anti-rabbit IgG secondary antibody (1:300 dilution; Santa Cruz Biotechnology, Dallas, TX). Nuclear DNA was labeled with 4′,6-diamidino-2-phenylindole (Sigma-Aldrich) (Fan et al., 2017; Yang et al., 2017). In addition, phenotype identification and purity analysis of FLSs was performed by fluorescence-activated cell sorting analysis (Doss et al., 2016). FLSs were seeded into a six-well plate with 3 × 10⁵ cells/well and collected after cultivation for 24 hours. Cells were then resuspended with 100 μl fluorescence-activated cell sorting buffer (1% BSA in 1× PBS) containing 0.5 mg/ml fluorescein isothiocyanate–conjugated monoclonal CD90 (also called Thy-1 or Thy1) antibody (Biolegend, San Diego, CA) and incubated at 4°C for 30 minutes under dark conditions. Analysis was conducted by flow cytometry (BD LSRFortessa X-20 cell analyzer; BD Biosciences, San Diego, CA).

Cell Proliferation Assay.

Cytotoxicity experiments were performed to measure cell proliferation after administration of DMY to synovial cells alone. In general, FLSs were seeded in 96-well plates (Corning Glassworks, Corning, NY) with 2500 cells/well; 24 hours later, they were randomized into four groups and treated with vehicle or a test compound DMY (3.125–800 μM) for 24 and 48 hours. All tests were performed in triplicate, and cell proliferation rates were assessed by the Cell Counting Kit-8 (CCK-8) assay (Dojindo, Kumamoto, Japan) following the manufacturer’s recommendation. At each time point, absorbance was measured at 450 nm using a microplate reader (Shang et al., 2017). To detect the inhibitory effect of DMY on FLSs, cells were stimulated with 10 ng/ml IL-1β (R&D Systems, Minneapolis, MN) to maintain the inflammation condition and then incubated in the presence of DMY at different concentrations.

In Vitro Migration Assay.

FLS migration ability was evaluated in a 24-well microchemotaxis chamber (BioCoat; Corning, Shanghai, China) with an 8-μm pore size. DMY and cells in DMEM without FBS were loaded into the upper chambers, and 10% FBS medium was added to the lower chambers to evaluate the effect of DMY on migration. After 16 hours, a cotton swab was used to remove the remaining cells from the upper surface of the filter, and the cells adhering beneath were fixed in 4% PFA for 20 minutes and dyed with 0.1% crystal violet for 30 minutes. Migrated cells were counted in five random fields for each assay at 20× magnification using light microscopy (Shang et al., 2017). In addition, a wound scratch assay in vitro was adopted to evaluate FLS migration. FLSs were seeded in 60-mm culture dishes (Corning Glassworks) at a density of 5 × 10⁵ cells/ml; when they reached 90% confluence, one parallel wound was created using a sterile 200-μl pipette tip. The detached cells were washed away with serum-free media, and the cultured FLSs were then treated with DMY (at concentrations of 6.25, 12.5, and 25 μM) without FBS. Photographs were taken at the indicated time points (0, 16, 24, and 48 hours), and migration was quantified by calculating the percentage of the area of FLSs migrating to the scratch with ImageJ software (National Institutes of Health, Bethesda, MD) (Wang et al., 2016).

Flow Cytometric Analysis of Apoptosis.

For analysis of apoptosis, FLSs were exposed to several concentrations (6.25, 12.5, and 25 μM) of DMY in DMEM for 24 hours. Approximately 5 × 10⁵ cells were collected and stained with an Annexin V-APC (allophycocyanin)/propidium iodide (PI) apoptosis kit (eBioscience, San Diego, CA) following the manufacturer’s protocol and analyzed by flow cytometry (with a BD LSRFortessa X-20 cell analyzer). Early apoptotic cells were defined as Annexin V positive and PI negative in the lower-right quadrant of the graph, whereas late apoptotic cells were positive for both Annexin V and PI in the upper-right quadrant. The percentages in these two quadrants were calculated to represent the total apoptotic rate (Shang et al., 2017; Liu et al., 2017).

Real-Time Polymerase Chain Reaction.

After the same treatment as above, FLSs were homogenized in TRIzol reagent (Ambion, Shanghai, China) to extract total cellular RNA, which was subsequently reverse-transcribed to cDNA using the cDNA reverse transcription kit (Takara Bio, Tokyo, Japan) following the standard protocol. Real-time polymerase chain reaction (PCR) primers for TNF-α, IL-1β, IL-6, and β-actin were synthesized by Shanghai Sangon Biological Engineering Technology and Services Co., Ltd. (Shanghai, China). The primers were as follows: TNF-α (forward: 5′-ATGGGCTCCCTCTCATCAGT-3′; reverse: 5′-GCTTGGTGGTTTGCTACGAC-3′), IL-1β (forward: 5′-AGCAGCTTTCGACAGTGAGG-3′; reverse: 5′-CTCCACGGGCAAGACATAGG-3′), IL-6 (forward: 5′-CTCTCCGCAAGAGACTTCCAG-3′; reverse: 5′-TTCTGACAGTGCATCATCGCT-3′), and β-actin (forward: 5′-ACGGTC AGGTCATCACTATCG-3′; reverse: 5′-GGCATAGAGGTCTTTACGGATG-3′). Quantitative real-time PCR was performed on a 7500 Real-Time PCR System (Thermo Fisher Scientific) according to the manufacturer’s instructions. The cycling program was conducted as follows: predenaturation at 95°C for 30 minutes, then 40 cycles of denaturation at 95°C for 5 seconds and annealing and extension at 60°C for 34 seconds. Target gene expression was determined relatively to the housekeeping gene as a 2^−ΔCt value, where the Δ threshold cycle (Ct) was calculated with the comparative Ct between β-actin, as the endogenous control, and the target gene (Anderson et al., 2015; Yeo et al., 2016).

Western Blot Analysis.

For each experiment, 3 × 10⁵ cells/well were seeded in six-well plates (Corning Glassworks). When subconfluence (approximately 70%) was reached, the FLSs were stimulated with 10 ng/ml IL-1β at serial time points (15, 30, 45, and 60 minutes) to determine an appropriate time point, and they were then pretreated with serial concentrations of DMY for 24 hours and exposed to 10 ng/ml IL-1β for the desired time. Total cellular proteins were extracted by adding ice-cold radioimmunoprecipitation assay lysis buffer (Beyotime, Shanghai, China) along with the protease inhibitor phenylmethanesulfonyl fluoride (100 mM; Biosharp, Wuhan, China) along with a phosphatase inhibitor cocktail (100×; Thermo Fisher Scientific). The FLSs were placed without shaking for 30 minutes; then the cells were scraped off the plate, and the whole-cell lysates were centrifuged at 14,000 rcf for 15 minutes at 4°C (Xin et al., 2014). The obtained supernatants were mixed with 5× protein loading dye (Sangon Biotech, Shanghai, China) with a ratio of 5:1 and boiled for 5 minutes. The blots were probed with primary antibodies (1:1000 dilution; Cell Signaling Technology, Beverly, MA) against NF-κB p65, phospho–NF-κB p65, inhibitor of NF-κB α (IκBα), phospho-IκBα, phospho-IκB kinase (IKK) α/β, IKKα, IKKβ, and glyceraldehyde-3-phosphate dehydrogenase at 4°C overnight, followed by washing with Tris-buffered saline Tween 20 (Epizyme Scientific, Shanghai, China) and incubation with the corresponding secondary antibodies (1:10,000 dilution; Cell Signaling Technology) for 1 hour at room temperature (Zhou et al., 2014; Li et al., 2017). The expression of each signal protein was shown as the ratio of the staining intensity for the target protein to that for glyceraldehyde-3-phosphate dehydrogenase (Qiao et al., 2016). The band intensity was quantified by ImageJ software.

Immunofluorescence Staining.

FLSs cultured on confocal dishes were treated in the same way as for Western blotting. Then they were washed and fixed with 4% PFA for 30 minutes, permeabilized with 1% Triton X-100 for 30 minutes, blocked with 3% BSA for 30 minutes at room temperature, and incubated with primary anti–NF-κB-p65 antibody (1:400 dilution; Cell Signaling Technology) overnight at 4°C and with CY3-labeled goat anti-rabbit IgG as secondary antibody (1:500 dilution;

Santa Cruz Biotechnology) at room temperature for 1 hour. The nuclei were visualized using 4′,6-diamidino-2-phenylindole (Xin et al., 2014; Li et al., 2017).

Molecular Modeling.

DMY and the rat IKK proteins were preprocessed by LigPrep 3.4 software (Schrödinger, New York, NY) and Protein Preparation Wizard software (Schrödinger), respectively. In silico docking was subsequently performed using the standard precision scoring function of Glide 5.5 software (Schrödinger) with default values for other parameters.