Atorvastatin (Pfizer, New York, NY, USA) was prepared as a suspension in dimethyl sulfoxide (DMSO; Sigma, St. Louis, MO, USA). Mevalonate (Sigma) was dissolved in 1 N NaOH (pH 7.1). SB2035820, p38 inhibitor, was purchased from Cell Signaling Technology (Danvers, MA, USA).

Synovial tissues were obtained from five patients undergoing joint-replacement surgery. All five patients fulfilled the 2010 rheumatoid arthritis classification criteria of RA by the American College of Rheumatology/European League Against Rheumatism collaborative initiative 20. This study was approved by the Institutional Review Board, and informed consent was obtained from all patients. Their clinical characteristics are shown in Table 1.

Synovial tissues were washed with phosphate-buffered saline (PBS), minced, and digested for 2 hours at 37°C in Dulbecco Modified Eagle's Medium-High Glucose (DMEM-HG; JBI, Daejeon, Korea) containing 1 mg/ml type II collagenase (Worthington Biochemical Corporation, NJ, USA). The digested tissues were filtered through a 70-μm cell strainer (Becton Dickinson, Franklin Lakes, NJ, USA). Cell suspensions were centrifuged at 1,400 rpm for 20 minutes at room temperature, and cell pellets were resuspended in DMEM-HG containing 1% penicillin-streptomycin (Gibco/BRL, Grand Island, NY, USA) and 10% heat-inactivated fetal bovine serum (FBS; Gibco/BRL). The cells were then plated in 100-mm culture dishes (Becton Dickinson) and incubated in a humidified 5% CO₂ atmosphere. On reaching confluence, cells were detached with 0.1% trypsin-EDTA (Gibco/BRL) and split into 5 × 10⁵ cells per dish. For all experiments, three- to five-passage synovial fibroblasts were used.

CD14⁺ cells were purified by MACS according to the manufacturer's instructions (Miltenyi Biotech, Auburn, CA, USA). In brief, 10⁷ PBMCs were resuspended in 80 μl of cold PBS containing 0.5% BSA and 2 mM EDTA (MACS buffer). Microbeads (20 μl) conjugated with mouse anti-human CD14IgG2a (Miltenyi Biotech) were then added to cells and incubated for 15 minutes at 4°C. Cells were then washed and resuspended in 3 ml of MACS buffer and then applied to a preequilibrated sterile LS separation column (Miltenyi Biotech) placed in a MidiMACS magnet. The column was then washed 3 times with MACS buffer, and CD14⁺ cells were eluted with 5 ml of MACS buffer after removing the column from the magnet.

FLSs or PBMCs were cultured to confluence in 96-well plates (Becton Dickinson) and treated with M-CSF (2 ng/ml), 1,25-dihydroxyvitamin D₃ (10^-7 M), atorvastatin for 24 hours (10 to 100 μM) or 3 weeks (0.001 to 0.1 μM) in a 5% CO₂ incubator. Methylthiazol tetrazolium (MTT) labeling reagent (500 μg/ml; Sigma) was added to FLSs or PBMCs in wells and incubated for 2 hours at 37°C. Isopropanol supplemented with 0.4 M HCl was then added to release formazan dye from cells. Formazan absorbance in solution was measured by using a microtiter-plate enzyme-linked immunosorbent assay (ELISA) reader at 570 nm.

FLSs were cultured in 100-mm culture dishes and treated with TNF-α (20 ng/ml) and atorvastatin (10 to 100 μM) for 24 hours. Cells were harvested by using 0.1% trypsin-EDTA, washed twice in PBS containing 0.1% (wt/vol) NaN₃, and then 1 ml of cold 70% (vol/vol) ethanol was added to cell pellets. Cells were then vortexed and incubated at 4°C for 1 hour, washed twice, and resuspended in 0.1 ml of PBS containing 0.1% NaN₃. For DNA staining, cells were treated with 1 mg/ml of RNase A (Sigma), resuspended in 0.5 ml of PBS containing 50 mg/ml propidium iodide (PI; Sigma), and incubated for 30 minutes at 4°C in the dark. They were then analyzed with flow cytometry (FACS Caliber, BD Science), and the sub-G₀/G₁ portion (the M1 fraction) was considered to be the apoptotic fraction.

Total RNA was extracted from FLSs or CD14⁺ cells by using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer's instructions. In brief, FLSs or CD14⁺ cells were cultured in six-well plates in the presence or absence of TNF-α (20 ng/ml), atorvastatin for FLSs (10 to 100 μM), for CD14⁺ cells (0.01 to 1 μM) for 24 to 72 hours. The cells were then lysed by adding 1 ml TRIzol reagent, and lysates were collected in 1.5-ml microtubes, to which was added 0.2 ml of chloroform (Sigma). The microtubes were then centrifuged at 12,000 g for 15 minutes at 4°C, and supernatants were transferred to new microtubes. Isopropyl alcohol (0.5 ml) was then added to precipitate RNA, and microtubes were centrifuged at 12,000 g for 8 minutes at 4°C. Pellets were washed with 75% ethanol, dried at room temperature, and resuspended in 10 to 20 μl nuclease-free water (Applied Biosystems/Ambion, Austin, TX, USA). RNA concentrations were determined by measuring absorbance at 260 nm.

First-strand cDNA was synthesized by using a Power cDNA Synthesis Kit (Intron Biotechnology, Seongnam, South Korea). The reaction was conducted in 20 μl of buffer containing 1 μg of total RNA, 0.2 mM oligo (dT)₁₅ primer, 5× RT buffer, 40 mM DTT, 10 units of RNase inhibitor, 2.5 mM deoxynucleotide triphosphate (dNTP) mixture, and 5 units of AMV reverse transcriptase. After incubation at 37°C for 1 hour, the reaction was stopped by heating at 70°C for 15 minutes. To remove the remaining RNA, 1 μl of Escherichia coli RNase H (4 mg/ml) was added to reaction mixtures and this was followed by incubation at 37°C for 30 minutes. The housekeeping gene, glyceraldehyde 3-phosphate dehydrogenase (GAPDH), was used as an internal control for determining gene expression. The primers used to detect RANKL (S, 5' GCC AGT GGG AGA TGT TAG 3'; AS, 5' TTA GCT GCA AGT TTT CCC 3'); RANK (S, 5' TTA AGC CAG TGC TTC ACG GG 3'; AS, 5' ACG TAG ACC ACG ATG ATG TCG C 3'); OPG (S, 5' TGC TGT TCC TAC AAA GTT TAC C 3'; AS, 5' CTT TGA GTG CTT TAG TGC GTG 3'), and GAPDH (S, 5' GCT CTC CAG AAC ATC ATC CC 3'; AS, 5' CGT TGT CAT ACC AGG AAA TG 3').

PCR amplification of cDNA was performed in an automated thermal cycler (GeneAmp PCR System 2400, Applied Biosystems) in a final volume of 20 μl containing 1 to 4 μl of cDNA, 20 mM Tris-HCl (pH 8.4), 50 mM KCl, 1.5 mM MgCl₂, 0.1% Triton X-100, 0.2 mM dNTP mixture, 0.5 pmol of each primer, and 5 units of Taq DNA polymerase (Promega). After PCR, the amplified products were analyzed with electrophoresis in 1.5% agarose gel and visualized with ethidium bromide staining under UV illumination. Band intensities of PCR products were measured by using a UV/VIS Vilber Lourmat digital camera and the software BioCaptMW (Vilber Lourmat, Marne-la-Vallée, France).

Real-time PCR was performed with the ABI 7500 real-time PCR system (Applied Biosystems). Total RNA was extracted by using an RNeasy kit (Qiagen, Valencia, CA, USA). Synthesis of cDNA was performed by using an SuperScript III First-Strand Synthesis System (Invitrogen) according to the manufacturer's protocol. Quantitative real-time PCR was performed by running a QuantiTect SYBR Green PCR Kit (Qiagen). Primer sequences were as follows; GAPDH, forward: CAATGACCCCTTCATTGACC; GAPDH, reverse: TGGACTCCACGACGTACTCA; and RANKL, forward GCTTGAAGCTCAGCCTTTTG:; and RANKL, reverse: CGAAAGCAAATGTTGGCATA. The reactions were incubated at 94°C for 15 minutes for one cycle, and then at 94°C (15 seconds), 59°C (30 seconds), and 72°C (30 seconds) for 40 cycles. The quantity of mRNA was calculated by using the threshold cycle (C_t) value for amplification of human RANKL and for human GAPDH as a reference gene. Relative gene expression was determined by the 2^ΔΔCt method.

Protein expressions were measured with Western blotting. In brief, cells were washed twice with ice-cold PBS, scraped into microfuge tubes, pelleted by centrifugation, and suspended in PRO-PREP protein-extraction buffer (Intron). After 30 minutes of incubation on ice, samples were centrifuged at 14,000 rpm for 30 minutes at 4°C. Protein concentrations were measured by using Bio-Rad Protein Assay reagent (Bio-Rad) with bovine serum albumin as a standard. Sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was carried out by using a mini-protein system (Bio-Rad). Dissolved protein samples were combined with sample buffer (125 mM Tris (pH 6.8), 5% glycerol, 2% SDS, 1% 2-mercaptoethanol, and 0.006% bromophenol blue) boiled for 5 minutes, and immediately cooled. Equivalent amounts of protein samples (50 μg/lane) were loaded onto 12% SDS gels. Electrophoresis was carried out in a running buffer at 100 V for 2 hours. Proteins were transferred from gels onto polyvinylidene difluoride (PVDF) membranes (Millipore, Bedford, MA, USA; 90 minutes at 30 V) in transfer buffer (192 mM glycine, 25 mM Tris-HCl, pH 8.3, 0.02% SDS, and 20% vol/vol TBS-Tween buffer (20 mM Tris, pH 8.0, 150 mM NaCl, and 0.1% Tween 20)) for 1 hour at 4°C. After blocking with 5% nonfat dried milk in TBS-Tween buffer for 1 hour, rabbit polyclonal antibody to human RANKL (R&D), rabbit monoclonal anti-Akt/phospho-Akt, anti-ERK/phospho-ERK, anti-JNK/phospho-JNK, anti-p38/phospho-p38 antibodies (all from Cell Signaling Technology), and rabbit monoclonal anti-actin antibody (Sigma) were added and incubated for 12 hours. Membranes were then washed and incubated with HRP-conjugated goat anti-rabbit IgG polyclonal antibodies (1:1,000; Jackson Immunoresearch, Philadelphia, PA, USA). Chemiluminescent substrate (Amersham Life Science, Arlington Heights, IL, USA) was added, and incubation was continued for 10 minutes. Blots were then exposed to radiographic film.

The secretion of sRANKL was detected by using RANKL ELISA kit (Peprotech), according to the manufacturer's directions.

Blood was collected from healthy volunteers, and PBMCs were isolated by centrifugation over Ficoll/Paque at 1,700 rpm for 30 minutes. PBMCs were resuspended in α-MEM containing 10% FBS and 2 ng/ml M-CSF (Sigma) and seeded at 2 × 10⁵ cells/well in 96-well culture plates. On the following day, FLSs were added at 2 × 10⁴ cells/well to adherent PBMCs and cocultured for 3 weeks in α-MEM containing 10% FBS, 100 IU/ml benzyl penicillin, 100 mg/ml streptomycin, 2 ng/ml M-CSF, and 10^-7 M 1,25-dihydroxyvitamin D₃. The culture medium was replaced every 3 days. Adherent cells were stained for tartrate-resistant acid phosphatase (TRAP) by using a commercially available kit (Sigma), as previously described (10). TRAP-positive multinucleated cells containing three or more nuclei were identified as osteoclasts and counted with light microscopy. All experiments were carried out 3 times in triplicate.

Data are presented as mean ± SD. Statistical analysis was performed by using SPSS version 12 (SPSS Inc., Chicago, IL, USA), and P values of < 0.05 were considered significant.

FLSs were cultured for 24 hours with atorvastatin in the presence of TNF-α (20 ng/ml), and assayed for RANKL expression with RT-PCR and real-time PCR. Atorvastatin significantly and dose-dependently inhibited the expression of RANKL mRNA in the FLSs of RA patients (P < 0.05) (Figure 1A, B). Western blotting of these cells showed that atorvastatin also dose-dependently suppressed RANKL protein levels (Figure 1C). Simvastatin has been reported to reduce arthritis incidence, activity, and histologic scores in a collagen-induced arthritis model, and simvastatin has been shown to have antiinflammatory potential in RA patients (18,22). Therefore, we determined whether simvastatin inhibits the expression of RANKL. Simvastatin also significantly and dose-dependently inhibited the expression of RANKL mRNA and protein in the FLSs of RA patients (see Additional file 1, Figure S1). To determine the most effective time for TNF-α stimulation, RANKL expression was analyzed with RT-PCR after treating the FLSs of RA patients for 24 to 48 hours with TNF-α (20 ng/ml). RANKL mRNA expression was found to peak after 24 hours of stimulation with TNF-α (20 ng/ml) (Figure 1D). Because mevalonate is synthesized from 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) by HMG-CoA reductase, HMG-CoA reductase inhibitors, like statins, reduce the entry of mevalonate into the cholesterol synthesis pathway. To determine whether mevalonate prevents the inhibition of RANKL expression by atorvastatin, mevalonate (100 μM) was cotreated with atorvastatin. RT-PCR and Western blotting showed that mevalonate prevented the suppression of RANKL expression by atorvastatin (Figure 2). Also, mevalonate prevented the inhibition of RANKL expression by simvastatin (see Additional file 2, Figure S2).

The effect of atorvastatin on cell viability was determined by using an MTT-based assay (Figure 3A). Apoptosis was measured by staining with propidium iodide (Figure 3B). Neither cell viability nor apoptosis was affected by atorvastatin at concentrations of 10 to 100 μM. Simvastatin did not affect cell viability and apoptosis (see Additional file 3, Figure S3).

The expressions of OPG mRNA were examined in FLSs from patients cultured for 24 hours with RT-PCR. Atorvastatin did not affect OPG expression in these cells (Figure 4A). We also assayed the secretion of OPG into the conditioned media of FLSs cultured in the presence of atorvastatin, and the statin was found to affect OPG secretion (data not shown).

The effect of atorvastatin on RANK in osteoclast precursor cells was evaluated by purifying CD14⁺ cells from total PBMCs by using a magnetic cell-separation (MACS) system. Atorvastatin did not affect RANK mRNA expression in CD14⁺ cells (Figure 4B), and the viabilities of CD14⁺ cells were not affected by atorvastatin at concentration up to 1 μM (Figure 4C).

To investigate the mechanism by which atorvastatin suppresses the expression of TNF-α-induced RANKL, we examined its effects on the phosphorylations of p38 MAPK, JNK, ERK, and Akt. Western blotting revealed that atorvastatin inhibited TNF-α-induced p38 phosphorylation after 12 hours, but that it did not affect the phosphorylations of JNK, ERK, and Akt (Figure 5A). When we pretreated cells with SB203580 (a p38 inhibitor), TNF-α-induced RANKL mRNA expression was decreased, which was similar to the suppression of RANKL mRNA expression by atorvastatin (Figure 5B). We also measured soluble RANKL levels in FLSs culture media, and we found that its levels were decreased by SB203580 and by atorvastatin (Figure 5C).

It was reported that RANKL is produced by RA FLS. We added 1,25-dihydroxyvitamin D₃ and M-CSF to media in which FLSs or cocultured cells were grown, according to a previously described method 3. We could produce functionally active osteoclasts, and they were determined by TRAP staining (Figure 6A). When atorvastatin was added to cocultures of PBMCs and FLSs for 3 weeks, atorvastatin at ≥0.01 μM significantly decreased the numbers of TRAP-positive multinucleated osteoclasts (P < 0.01 versus control) (Figure 6A, B). To ensure that the inhibitory effect of atorvastatin on osteoclast formation was not due to an effect on cell viability, we performed MTT assay. Cell viabilities were assayed in separate cultures of FLSs and PBMCs. The viability of PBMCs after 3 weeks of treatment in the presence of M-CSF was not affected by atorvastatin (Figure 6C). Furthermore, atorvastatin had no effect on FLS viability after 3 weeks of treatment in the presence of vitamin D₃ (Figure 6D).