Osteoarthritis cartilage tissues were obtained from 10 patients (all female; age, 63±5 years) with OA that underwent total knee joint replacement between July and December 2014, at the Affiliated Hospital of Shandong University (Jinan, China). The present study was performed following a protocol approved by Medical Ethical Committee of the Institute of Orthopedics, Xijing Hospital, Fourth Military Medical University (Xi'an, China) and all examinations were performed after written informed consent was obtained from the patients.

Chondrocytes were isolated from cartilage as previously described (15). The whole articular cartilage was minced into pieces ~0.5 cm². Following enzymatic digestion with 2 mg/ml pronase in serum-free Dulbecco's modified Eagle's medium (DMEM; Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) containing 100 U/ml penicillin and 100 µg/ml streptomycin (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) for 2 h at 37°C in 5% CO₂ atmosphere, the specimens were then digested with 0.2% (v/v) collagenase I for 4 h at 37°C. Primary chondrocytes were cultured at a density of 1×10⁵ cells/ml in petri dishes in monolayer culture for a period of 24 h at 37°C with 5% CO₂, and chondrocytes in passages 5 to 9 were used.

Cell viability was quantified by a Cell Counting Kit-8 assay (CCK-8; Dojindo Molecular Technologies, Inc., Kumamoto, Japan). Human chondrocytes (1×10⁴ cells/well) in 96-well plates were pretreated with 10, 50 and 100 µg/ml PF (Sigma-Aldrich; Merck KGaA) for 2 h at 37°C and subsequently co-incubated in the absence or presence of 10 ng/ml IL-1β (Sigma-Aldrich; Merck KGaA) at room temperature for 24 h. Then, CCK-8 reagents were added and incubated with the cells for 1 h at 37°C. The spectrophotometric absorbance was quantified at 490 nm on a multifunctional microplate reader (Tecan Group Ltd., Durham, NC, USA).

NO levels in the culture medium were detected by Griess reaction, as previously described (16). PGE2 levels were investigated using a commercially available ELISA kit, according to the manufacturer's protocol (KHL1701; Invitrogen; Thermo Fisher Scientific, Inc.).

Total RNA was isolated from human chondrocytes using TRIzol (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. Total RNA (5 µg) for each sample were reverse transcribed into first strand cDNA using the iScript cDNA synthesis kit (Bio-Rad Laboratories, Inc., Hercules, CA, USA) according to manufacturer's protocol for subsequent qPCR analysis. qPCR was performed in a final volume of 10 µl, which contained 5 µl of SsoFast™ EvaGreen Supermix (Bio-Rad Laboratories, Inc.), 1 µl of cDNA (1:50 dilution) and 2 µl of the forward and reverse primers (1 mM). The following specific primer sequences were used: iNOS, sense 5′-TTTCCAAGACACACTTCACCA-3′, antisense 5′-ATCTCCTTTGTTACCGCTTCC-3′; sense COX-2 5′-GAGAGATGTATCCTCCCACAGTCA-3′, antisense 5′-GACCAGGCACCAGACCAAAG-3′; and for β-actin, sense 5′-AGAAGGCTGGGGCTCATTTG-3′, antisense 5′-AGGGGCCATCCACAGTCTTC-3′. Thermocycling conditions were as follows: 94°C for 5 min; subsequently 35 cycles at 94°C for 1 min, 59°C for 15 sec and 72°C for 30 sec; followed by 72°C for 10 min. qPCR data was calculated by the 2^−ΔΔCq method (17).

Chondrocytes were washed twice with ice-cold PBS and sonicated with lysis buffer (Takara Biotechnology Co., Ltd., Dalian, China), the cell lysate supernatants were harvested by centrifugation at 10,000 × g for 10 min at 4°C. The protein concentrations of the cell supernatants were then quantified using Bradford protein dye reagent (Bio-Rad Laboratories). Proteins (30 µg/lane) were separated by 12% SDS-PAGE and transferred onto polyvinylidene difluoride membranes (EMD Millipore, Boston, MA, USA). The membranes were incubated with appropriate primary antibodies for iNOS (1:2,000; sc-7271), COX-2 (1:3,000; sc-19,999), nuclear factor (NF)-κB p65 (1:1,500; sc-8008), inhibitor of κB (IκBα) (1:3,000; sc-1643) and GAPDH (1:3,000; sc-59540) (Santa Cruz Biotechnology, Inc., Dallas, TX, USA) overnight at 4°C. The membranes were subsequently washed with PBS containing 0.1% (v/v) Tween 20 and incubated with horseradish peroxidase-conjugated secondary antibodies (1:2,500; sc-2005; Santa Cruz Biotechnology, Inc.) for 1 h at room temperature, followed by exposure using enhanced chemiluminescence detection reagents (Gibco; Thermo Fisher Scientific, Inc.). BandScan version 5.0 (Glyko, Inc., Novato, CA, USA) was used for the quantification of all the proteins following western blot analysis.

Culture medium was collected for ELISA assay. MMP-3 and MMP-13 levels were quantified using ELISA kits for human MMP3 and MMP13 (KAC1541 and EHMMP13; both from Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol.

The quantitative data are expressed as the mean ± standard deviation. Statistical significance was assessed by the one-way analysis of variance followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

Human OA chondrocytes were incubated with 0, 10, 50 and 100 µg/ml PF for 24 h, and the cell viability was determined using a CCK-8 assay (Fig. 1). At concentrations of 10, 50 and 100 µg/ml, PF induced no cytotoxic effects in human OA chondrocytes.

Subsequently the effects of PF on NO and PGE2 production in IL-1β-stimulated human OA chondrocytes were investigated. As indicated in Fig. 2, IL-1β significantly increased the production of NO and PGE2 compared with the control group (P<0.05). However, treatment with PF significantly inhibited NO and PGE2 production induced by IL-1β in human OA chondrocytes (P<0.05).

The effect of PF on iNOS and COX-2 mRNA and protein expression levels in IL-1β-induced chondrocytes were evaluated by RT-qPCR and western blot analysis. As presented in Fig. 3A and B, IL-1β treatment significantly increased the mRNA expression levels of iNOS and COX-2 (P<0.05) compared with the control. However, treatment with PF significantly inhibited IL-1β-induced expression of iNOS and COX-2 in chondrocytes (P<0.05). Western blot analysis demonstrated that PF also inhibited the protein expression levels of iNOS and COX-2 induced by IL-1β in human OA chondrocytes (P<0.05; Fig. 3C and D).

It is established that MMPs perform a critical role in bone remodeling and OA; therefore, the present study investigated the effect of PF on MMP-3 and MMP-13 production in IL-1β-induced chondrocytes. As presented in Fig. 4, chondrocytes stimulated with IL-1β exhibited significantly increased levels of MMP-3 and MMP-13 compared with the control group (P<0.05). However, treatment of chondrocytes with PF significantly reduced the production of MMP-3 and MMP-13 induced by IL-1β in human OA chondrocytes (P<0.05).

In order to further understand the anti-inflammatory mechanisms of PF, the present study examined the level of NF-κB p65 expression in IL-1β-stimulated chondrocytes. As indicated in Fig. 5, the protein expression level of NF-κB p65 in the nucleus of chondrocytes was significantly increased by IL-1β compared with the untreated control (P<0.05). However, PF significantly suppressed the expression of NF-κB p65 (P<0.05; Fig. 5B). Additionally, PF inhibited IL-1β-induced IκB-α degradation in human OA chondrocytes (P<0.05; Fig. 5C).