Collection of Tissue Specimens
Osteochondral specimens of knee joints were obtained from operations performed from 2020 to 2021 in the Department of Bone Microsurgery, Foot and Ankle Surgery, and Joint Surgery of Xi’an Honghui Hospital. All tissue samples had a final diagnosis provided by clinical specialists. The inclusion criteria were as follow: primary knee joint osteoarthritis and Kellgren-Lawrence grading of knee X-ray ⅲ or ⅳ. Patients with rheumatoid arthritis, infectious arthritis, traumatic arthritis and other immune system diseases were eliminated from the research samples. As a result, 12 patients were enrolled in the knee osteoarthritis group: all patients were of Han ethnicity Han and they included 5 males and 7 females with an average age of 66.1 years (range: 57-75). All patients underwent knee replacement surgery. Meanwhile, 11 tibia platform and distal femur comminuted fracture patients were recruited as the control group (7 males and 4 females, average age of 41.3 years (range: 24-56)). As the fixed weight-bearing area free bone cartilage block could not be reset in all cases, the pain symptoms of the wounded knee before seeking medical help were compared with the contralateral knee by X-ray and osteoarthritic performance.
Histologic analysis
To evaluate the degree of degradative changes in the cartilage, cartilaginous tissue from OA patients and rabbit models were embedded in paraffin and sectioned at 6 μm. After deparaffinization and hydration, the sections were stained with hematoxylin-eosin (HE) and Safranine-O (Solarbio) according to the manufacturer's instructions.
Cell culture and treatment
The human chondrocyte cell line C28/I2 (C28) was cultured in DMEM/F12 supplemented with 10% FCS (Gibco Thermo Fisher USA), 100 U/ml penicillin (Solarbio), and 100 μg/ml streptomycin (Solarbio) in a humidified incubator at 37 °C in the presence of 5% CO2. Before treatment, chondrocytes were serum-starved for 12 h and then stimulated with recombinant human TNF-α (active trimer, acrobiosystems, China) at a concentration of 40 ng/ml for 48 h, and unstimulated chondrocytes were used as controls synchronously [23]. To investigate the effect of fucosylation on the OA process, 2F-peracetyl-fucose (2FF, EMD Millipore, Germany) was used to inhibit the biosynthesis of fucosylation in chondrocytes [24]. Briefly, after serum starvation for 12 h, 2FF was added to complete medium at a concentration of 100 μM and incubated for 72 h. And then chondrocytes were collected and subjected to further analysis.
Construction of rabbit OA model
Hulth's modeling method was used to establish the rabbit model of knee osteoarthritis [25]. Briefly, twenty-four male New Zealand white rabbits (purchased from Laboratory Animal Center of Xi'an Jiaotong University, China) were divided into an OA group (n=12) and a sham-operation group (n=12) randomly. At the time of surgery, the rabbits were 3 to 4 months old and had body weights of 2.5 ± 0.4 kg and 2.7 ± 0.3 kg, respectively (mean ± SEM). Rabbits were anesthetized by 3% sodium pentobarbital (1 mL/kg; Sigma). Intravenous cefazolin (22 mg/kg; Harbin Pharmaceutical Group Pharmaceutical General Factory, China) was administered at the time of the surgical procedure and once every 24 h for 3 d postoperatively. A midline skin incision was made over the right knee and a medial parapatellar incision was made through the retinaculum. The medial collateral ligament was sharply divided. A medial parapatellar arthrotomy was performed and the patella dislocated laterally. Care was taken to protect and retract the vascular intraarticular fat pad. With the knee flexed, the anterior cruciate ligament and the posterior cruciate ligament were transected. The knee joint was then dislocated to excise the medial meniscus. The joint was irrigated with sterile saline solution. The capsule and the synovium were then closed together with a 4.0 interrupted Vicryl, and the skin was closed. A sham procedure was performed on the right hind limb to serve as a control. For sham surgical controls, right knees were opened as described. After dislocating the patella laterally, the knee was irrigated, but the ligaments and menisci were left intact. Postoperatively, the animals were permitted cage activity without immobilization. The animals were closely monitored for health and welfare. At weeks 0, 4, 8 and 12, rabbits from OA (n=3) and sham-operated group (n=3) were euthanized by an intravenous injection of overdose pentobarbital to obtain cartilage samples.
Extraction of Cell/Tissue Protein
The proteins from chondrocytes and cartilage tissue were extracted by using RIPA Lysis Buffer (Millipore, Billerica, MA) and T-PER Tissue Protein Extraction Reagent (Thermo Fisher Scientific Inc., Rockford, IL, USA) according to the manufacturer's instructions, and 1% (v/v) of protease inhibitor cocktail (Sigma-Aldrich) was added. The protein concentration was determined by BCA assay (Beyotime Biotechnology, Nantong, China).
Lectin Microarray and Data Analysis
The manufacture of lectin microarray and data acquisition were performed as described previously [26-28]. The proteins isolated from cells or tissue were labeled with Cy3 fluorescent dye (GE Healthcare, Biosciences, Piscataway, NJ, USA) and purified using a Sephadex-G25 column (GE Healthcare). Subsequently, 4 μg of labeled protein was applied to the lectin microarrays and incubated in the chamber at 37 ℃ for 3 h. After washing and centrifugation, the slides were scanned using a confocal scanner (4000B, AXON Instruments, USA). The fluorescence intensities were extracted by GenePix 7.0 software (Axon). After filtration and normalization, the parallel datasets were compared with each other based upon fold-changes according to the following criteria: fold changes ≥ 1.50 or ≤ 0.67 and p < 0.05 indicated upregulation or downregulation, respectively. Significant difference in lectin between samples were tested by Student’s t test.
Lectin blotting
Lectin blotting was performed as described previously [28-30]. 40 μg of protein from OA and control samples was separated by 10% SDS–PAGE and transferred to PVDF membranes (0.22 μm Millipore, Bedford, MA, USA). After blocking, the membranes were incubated with Cy5 labeled lectins overnight at 4 °C and imaged by a STROM FluorImager (Molecular Dynamics, Sunnyvale, CA, USA). The gray value of the selected protein bands was measured by ImageJ software (NIH).
Immunoblotting
Briefly, 10 μg of protein was separated by 10% SDS–PAGE, and then transferred to PVDF membranes (Millipore, Bedford, MA), and blocked with 5% (w/v) skim milk (Becton Dickinson, Franklin Lakes, NJ) in TBST (TBS buffer with 0.05% Tween-20, pH 7.6) or 3% BSA in PBST (for phosphorylated antibodies) for 1 h at room temperature. The membranes were probed with primary antibodies overnight at 4 °C with shaking. The primary antibodies used in this study were as follows: (i) rabbit polyclonal anti-FUT10 (Proteintech, Wuhan, China); (ii) rabbit polyclonal anti-MMP-13 (Proteintech); (iii) rabbit polyclonal anti-COL2A1 (Proteintech); (iv) mouse monoclonal anti-IL-1β (Proteintech); (v) mouse monoclonal anti- NF-κB p65 (Proteintech); (vi) mouse monoclonal anti- phospho-NF-κB p65 (Ser536, CST); (vii) rabbit polyclonal anti-IKBA (Proteintech); (viii) rabbit polyclonal anti-phospho-IKBA (Ser32, CST); (ix) mouse monoclonal anti-P38 MAPK (Proteintech); (x) rabbit monoclonal anti-Phospho-p38 MAPK (Thr180/Tyr182, CST); (xi) mouse monoclonal anti-JNK (Proteintech); (xii) mouse monoclonal phospho-SAPK/JNK (Thr183/Tyr185, CST); (xiii) mouse monoclonal anti-Caspase-8 (Proteintech); (xiv) rabbit polyclonal anti-Caspase-3 (Proteintech); (xv) rabbit polyclonal anti-TNFR1 (Proteintech); (xvi) mouse monoclonal anti-β tubulin as internal control (Abways Biotechnology, Shanghai, China). After washing three times with TBST, the membranes were incubated with horseradish peroxidase (HRP)-labeled second antibody (Immunoway, Jiangsu, China) for 2 h at room temperature with shaking. The membranes were visualized with Immobilon Western chemiluminescent HRP substrate (Millipore, Billerica, MA, USA).
Isolation of RNA and semi-qPCR
Total RNA from chondrocytes and cartilage tissue was extracted by TRI Reagent (Sigma) according to manufacturer’s protocol. Then, 1 μg of total RNA was reverse-transcribed using PrimeScript™ RT Master Mix (TaKaRa, Japan), and qPCR was performed using a ViiA 7 Real-Time PCR System (Applied Biosystems, USA). SYBR Green based three-step RT–qPCR was performed using TB Green® Premix Ex Taq™ II (TaKaRa, Japan). The primer sequences were retrieved from the online PrimerBank database (https://pga.mgh.harvard.edu/primerbank/index.html). The information of primers is summarized in Table S1.
Transfection of small interfering RNA
Small interfering RNA (siRNA) specific to FUT3, FUT9 and FUT10 were designed with the coding sequences of human by online software DSIR (http://biodev.extra.cea.fr/DSIR/DSIR.html). The information on the siRNAs is shown in Table S2. 2'Ome modified siRNAs and scramble siRNA (negative control) were obtained from GenePharma (Shanghai GenePharma, China). The transfection was performed using HiPerFect reagent (Qiagen, Chatsworth, CA). The cells were harvested 24 h after transfection, and knockdown effects were evaluated by semi-qPCR. In addition, after transfection for 12 h, the TNF-α (at a final concentration of 40 ng/mL) was added to the medium and culture for 48h. After that, the cells were collected and subjected to follow-up analysis.
Cell proliferation assay
Cell proliferation was determined by a cell counting kit (CCK-8, YEASEN, Shanghai, China). Briefly, C28 cells were seeded in 96-well plates at concentration of 5,000 cells per well and incubated for 12 h. After treatment, 10 μL of CCK8 reagent was added to each well and the cells were incubated at 37 ℃ for 1 h. The absorbance at 450 nm of each well was measured by a microplate reader (Bio-Tek Instruments Inc., Winooski, VT, USA). The cell proliferation rates were recorded every 12 h.
SA-β-gal staining
Senescence cells were stained using senescence β-galactosidase staining kit (Beyotime). Briefly, after treatment, the cells were fixed for 15 min by using stain-fixative. After washing for three times, 1 mL of staining solution was added to each well, and the plates were incubated at 37 ℃ overnight. The SA-β-gal positive chondrocytes in three random fields of each well were calculated using bright-field microscopy.
Apoptosis assay
An AnnexinV Alexa Fluor 647/PI apoptosis detection kit (Solarbio) was used to analyze apoptosis. Briefly, the chondrocytes were digested by 0.25% trypsin (without EDTA, Solarbio). Then, 1 mL of complete medium was added to neutralize trypsin and the cells were collected and resuspended by cold PBS. After centrifugation, the cells were resuspended in binding buffer, 5 µl of Annexin V/Alexa Fluor 647 was added and incubated at room temperature for 5 min in the dark 10 µL of PI was added and the cells were analyzed on a flow cytometer (ACEA Biosciences, San Diego, USA) using NovoExpress software to detect apoptotic cells.
Immunoprecipitation
Immunoprecipitation was performed using protein A/G PLUS-Agarose (Santa Cruz, Santa Cruz Biotechnology, Santa Cruz, CA, USA) according to manufacturer’s protocol with modifications. In order to protect the activity of TNFR1, the low pH elution buffer (100 mM glycine, 50 mM Tris-HCl, and 500 mM NaCl, pH 2.0) was used to elute TNFR1, and 10 μL of Tris buffer (1 M, pH 9.5) was added to neutralize the low pH condition.
Manufacture of antibody microarray
To investigate the effect of fucosylation on the binding capability of TNFR1 to TNF-α, a TNFR1 antibody microarray was fabricated. The antibodies were diluted with printing buffer (PBST with 0.01% (w/v) BSA) to a concentration of 200 ng/μL and spotted using an arrayer (SmartArrayer 48, Capitalbio). After blocking, the enrichment of TNFR1 from different sources was diluted to 0, 5, 10 and 20 ng/μL and applied to microarrays and incubated at 25 °C overnight. Then, the slides were incubated with TNF-α (20 ng/μL) and 10 ng/μL of primary antibody against TNF-α (Cy3 labeled, Bioss) for 3 h respectively. The net fluorescence intensities (the raw fluorescence intensities - background) of each sopt was acquired by Genepix 7.0 software (Axon Instruments, Inc. USA). To compare the difference in binding signals, we set the binding signals at 0 ng/μL TNFR1 as the baseline, and the differences in binding signals of TNFR1 Ab between TNF-α treated chondrocytes, siRNA-FUT10 transfected chondrocytes and controls were tested by one-way ANOVA.