Femoral condyles or tibial plateaus were selected as specimens. Human OA cartilage was harvested from patients who underwent total knee replacement surgery (10 males and 6 females; mean age, 70±4 years) at the Department of Orthopaedic Surgery, Peking Union Medical College Hospital (Beijing, China). OA was diagnosed according to the clinical and radiological evaluation criteria published by the American College of Rheumatology (ACR) (26). In addition, none of the patients from whom the samples were collected had received intra-articular steroid therapy in the 3 months prior to the surgery. Healthy (control) articular cartilage was obtained from donors who had died (by car accident or due to cardiovascular and cerebrovascular diseases such as severe heart failure, or malignant tumors), within 12 h of death (8 males and 5 females; mean age, 67±5 years). The normal healthny individuals had no previous history of joint disorder. This study was approved by the Medical Ethics Committee of Peking Union Medical College Hospital (Beijing, China) and informed consent was acquired from all participants.

After collecting the cartilage specimens, the tissues were diced and digested in 1.5 mg/ml collagenase type 2 (CLS-2; Worthington, Lakewood, NJ, USA) in Dulbecco's modified Eagle's medium (DMEM; Invitrogen, Carlsbad, CA, USA) for 16 h at 37°C. The DMEM solution was supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin and 100 μg/ml streptomycin (all from Sigma, St. Louis, MO, USA). The isolated cells were then passed through a 100-μm cell strainer (Falcon; Becton-Dickinson Labware GmbH, Heidelberg, Germany), pelleted and washed with medium. The cultures were maintained in DMEM containing 10% FBS in a humidified incubator containing 5% CO₂ at 37°C.

The cells were subsequently seeded in a 48-well prior to transfection. After post-culturing for 18–24 h, when the density of the chondrocytes had reached 70%, the cells were transfected using Lipofectamine 2000 (Invitrogen) according to a previously described method (27). In the present study, hsa-miR-15a mimics, pcDNA3/enhanced green fluorescent protein (EGFP)-ADAMTS5-3′-untranslated region (3′-UTR), antisense oligonucleotides (ASO)-hsa-miR-15a, ASO-NC (negative control), pcDNA3/EGFP-ADAMTS5-3′-UTR mutant and pDsRed2-N1, along with the plasmids were constructed and synthesized by Sangon Biotech Co., Ltd. (Shanghai, China) based on standard protocols (28). Furthermore, ADAMTS5 expression was suppressed using the small interfering RNA (siRNA) synthesized by Sangon Biotech Co., Ltd. in both human OA and normal chondrocytes. A negative control (siNC) was also performed concurrently on each specimen. The sequence of the ADAMTS5 siRNA was 5′-AAGAUAAGCG CUUAAUGUCUU-3′. Cells were collected 48 h after transfection for subsequent assays.

After post-culturing for 24 h, total RNA was extracted from the cells using an miRNA Isolation kit (Ambion, Austin, TX, USA) according to the manufacturer's instructions. The relative expression of levels miRNAs (specific for has-miR-15a) were determined using TaqMan^® MicroRNA assays (Applied Biosystems, Foster City, CA, USA) with the comparative 2^−ΔΔCT method. Total RNA was reverse transcribed into complementary DNA (cDNA) using MuLV reverse transcriptase (Life Technologies, Carlsbad, CA, USA), followed by quantification using the QuantiTect SYBR-Green real-time-polymerase chain reaction kit (Qiagen SA, Hilden, Germany). Primers for miR-15a were designed and obtained from the TaqMan miRNA assays. U6 snRNA (Applied Biosystems) was used as a loading control for miRNA expression. The primer sequence for stem-loop RT-PCR hsa-miR-15a was 5′-CTCAACTGGTGTCGTGGAGTCGGC AATTCAGTTGAGCACAAACC-3′. The following primers were used for PCR: hsa-miR-15a forward, 5′-ACACT CCAGCTGGGTAGCAGCACATAATGG-3′ and reverse, 5′-TGGTGTCGTGGAGTCG-3′; U6 forward, 5′-CTCGCTT CGGCAGCACA-3′ and reverse, 5′-AACGCTTCACGAATTTGCGT-3′.

We extracted total RNA from the chondrocytes using TRIzol reagent (Invitrogen) according to the manufacturer's instructions. The purity of the total RNA was assessed with the A260/A280 ratio, and values of 1.8 and 2.0 were considered good. Total RNA was reverse transcribed into cDNA and amplified, and was subsequently quantified using RT-qPCR with SYBR-Green Ex Taq on a LightCycler 480 (Roche Applied Science, Indianapolis, IN, USA). Primers for ADAMTS5 were designed and obtained from Invitrogen. The glyceraldehyde 3-phosphate dehydrogenase (GAPDH) gene was used as a reference. The PCR conditions included 1 prede-naturation cycle of 5 min at 95°C, 40–50 cycles of 95°C for 30 sec, 58–62°C for 30 sec, and 72°C for 30 sec, and a final extension at 72°C for 5 min. Relative quantification was calculated using the 2^−ΔΔCT method.

Target genes of hsa-miR-15a were predicted by bioinformatics analysis using TargetScan 6.2 (http://www.targetscan.org) and/or microRNA.org (http://www.microrna.org).

The 3′-UTR of ADAMTS5 was PCR-amplified, cloned into the psiCHECK-2 vector and co-transfected with the control or hsa-miR-15a precursor into the chondrocytes. After 48 h of transfection, luciferase activity was evaluated using the Dual-luciferase activity assay system (DLR; Promega, Madison, WI, USA).

Purified total DNA was quantified using Quant-iT™ PicoGreen^® dsDNA reagent (Invitrogen) according to the manufacturer's instructions. Fluorescence was measured using a fluorescence microplate reader (Tecan Polarion, Stevenage, UK) at an excitation/emission wavelength of 480/520 nm. Lambda DNA (Sigma) was used as marker for determination of the DNA quantity.

Total glycosaminoglycans (GAGs) were quantified using a 1,9-dimethylmethylene blue (DMMB) spectrophotometric analysis according to a previously described method (29). Briefly, following the addition of DMMB, the mixture was assayed for GAG at an absorbance of 525 nm. Chondroitin sulfate C (Sigma) was used as a reference.

After collecting the culture medium, collagenase activity assay was performed to determine the collagenase activity using an Enzchek Gelatinase/Collagenase assay kit (Invitrogen) according to the manufacturer's instructions. The assays were carried out in reaction buffer containing DQ Collagen Fluorescein conjugate at room temperature. Fluorescence was determined using a fluorescence micro-plate reader (SpectraMAX Gemini XS; Molecular Devices, Sunnyvale, CA, USA) at 485 nm excitation and 538 nm emission. Collagenase activity was determined using the formula provided with the assay kit and normalized to the weight of the explant. Collagenase type IV (Clostridium) was used as a reference standard.

Following culture for 24 h, cell protein was extracted using protein lysis buffer. Protein concentration was assessed using the Bio-Rad DC protein assay kit (Bio-Rad, Ivry-sur-Seine, France). The protein was separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and was then transferred to polyvinylidene difluoride (PVDF) membranes (Bio-Rad, Hercules, CA, USA). The membranes were blocked in a sealed in 5% fresh non-fat dry milk for 2 h and were then incubated with the primary anti-type II collagen antibody (1:1,000 dilution; MAB1330; Chemicon, Millipore, Billerica, CA, USA) and primary anti-ADAMTS5 antibody (1:100 dilution; ab45042; Abcam, Cambridge, UK) for 2 h at room temperature or overnight at 4°C, followed with horseradish peroxidase-conjugated anti-mouse secondary antibody (accession numbers ab195239). An anti-GAPDH antibody (sc-365062; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) was used as a loading control. Finally, the samples were tested using enhanced chemiluminescence and densitometric analysis. For relative expression analysis, NIH Image J software (NIH Image, Bethesda, MD, USA) was performed to determine band intensity.

In the present study, the statistical package for the social sciences (SPSS) (version 17.0; SPSS Inc., Chicago, IL, USA) was used to perform statistical analysis. A student's t-test was performed to study statistical comparisons. A p-value <0.05 was considered to indicate a statistically significant difference.

To dertermine the expression levels of hsa-miR-15a in human OA chondrocytes, we determined the levels of hsa-miR-15a in human OA cartilage specimens and normal cartilage tissues by RT-qPCR. As shown in Fig. 1, the expression level of hsa-miR-15a was significantly decreased in the OA chondrocytes compared with the normal cartilage tissue (P<0.05).

To predict whether the ADAMTS5 gene mRNA 3′-UTR contains a hsa-miR-15a binding site, TargetScan 6.2 and microRNA.org were used to confirm the prediction. The results revealed that a binding site for hsa-miR-15a was in the ADAMTS5 mRNA 3′-UTR (Fig. 2A).

In order to confirm that has-miR-15a binds to this region and leads to translational repression, we cloned the putative binding site into the pcDNA3/EGFP plasmid downstream of the EGFP reporter construct and co-transfected this plasmid into the cells with ASO-hsa-miR-15a or ASO-negative control (NC). We found that the relative EGFP intensity was significantly higher in the cells that were co-transfected with ASO-miR-15a compared with the cells co-transfected with ASO-NC. In addition, we cloned the putative binding site into pcDNA3/EGFP downstream of the EGFP reporter construct and co-transfected this plasmid with hsa-miR-15a mimics or NC. The results revealed that the relative EGFP intensity was significantly lower in the cells co-transfected with has-miR-15a mimics compared with those co-transfected with NC (Fig. 2B). Additionally, we performed RT-qPCR and western blot analysis to measure the mRNA and protein expression levels of ADAMTS5, respectively. Both RT-qPCR (Fig. 2C) and western blot analysis (Fig. 2D and E) revealed that the ADAMTS5 expression levels were significantly lower in the has-miR-15a mimics-transfected group compared with the NC-transfected group (P<0.05); however these levels were significantly higher in the ASO-miR-15a-transfected group compared with the ASO-NC-transfected group (P<0.05), respectively. These results suggest that hsa-miR-15a binds directly to the 3′-UTR of ADAMTS5 mRNA and inhibits gene expression. These results also indicate that ADAMTS5 is a direct target of hsa-miR-15a.

The amount of proteoglycans was significantly decreased in the cell substrate generated by the OA chondrocytes transfected with ASO-miR-15a compared to the ASO-NC-transfected group (Fig. 3A), while the amount of proteoglycans was significantly increased in the culture medium compared to the ASO-NC-transfected group (Fig. 3B). However, as regards the total amount of proteoglycans, there were no significant differences observed between the ASO-miR-15a-transfected group and the controls (ASO-NC) (Fig. 3C). The total amount of proteoglycans was measured by combining the amount found in the cell substrate and that released into the medium at the end of the culture period.

The amount of collagen was significantly decreased in the cell substrate generated by the OA chondrocytes transfected with ASO-miR-15a compared to the ASO-NC-transfected group (Fig. 4A), while the amount of collagen was significantly increased in the culture medium of the cells transfected with ASO-miR-15a compared to the ASO-NC-transfected group (Fig. 4B). In addition, the total amount of collagen was significantly decreased in the ASO-miR-15a-transfected group compared with the controls (ASO-NC) (Fig. 4C). However, collagenase activity was significantly increased in ASO-miR-15a-transfected group compared with the controls (Fig. 4D). Moreover, the protein expression of collagen II was significantly lower (0.75 ± 0.14 vs. 1.88 ± 0.16, with a 40% decrease) in the ASO-miR-15a-transfected group compared with the ASO-NC-transfected group (Fig. 4E). These results indicate that the decreased expression of hsa-miR-15a inhibits the aggregation of collagen, while the overexpression of hsa-miR-15a promotes the aggregation of collagen.

We performed RT-qPCR to measure the expression levels of ADAMTS5 in human OA chondrocytes, as well as in normal cartilage tissue. As shown in Fig. 5A, the expression level of ADAMTS5 was significantly increased in the OA chondrocytes compared with the normal cartilage tissue (P<0.05). To confirm a more definitive function of ADAMTS5 in the development of OA, we used ADAMTS5 siRNA to transfect human OA chondrocytes. The results revealed that the amount of proteoglycan was significantly increased in the chondrocyte substrate of the siR-ADAMTS5 group compared with that of the silenced negative control (siNC) group (Fig. 5B), whereas the amount released into the medium was significantly decreased compared with siNC group (Fig. 5C). No significant differences were observed in the total amount of proteoglycans between the 2 groups (Fig. 5D).

The amount of collagen was significantly increased in the chondrocyte substrate of the siR-ADAMTS5 group compared with that of the siNC group (Fig. 6A), whereas the amount released into the medium was significantly decreased compared with the siNC group (Fig. 6B). In addition, the total amount of collagen was significantly increased in the siR-ADAMTS5 group compared with the siNC group (Fig. 6C), while the collagenase activity was significantly lower in the siR-ADAMTS5 group than that in the siNC group (Fig. 6D). Furthermore, the protein expression of collagen II was significantly higher (2.22 ± 0.04 vs. 1.04 ± 0.02, with a 50% increase) in the siR-ADAMTS5 group compared with the siNC group (Fig. 6E).