Synovial fibroblasts were isolated from surgical samples from 13 rheumatoid arthritis and 8 osteoarthritis patients. Informed patient consents were obtained for isolation of fibroblasts. Cells were obtained by enzymatic digestion as described before 18. Cells were grown in Dulbecco's modified Eagle's medium (DMEM) (Sigma-Aldrich, St. Louis, MO, USA) with 10% fetal calf serum (Gibco-BRL, Grand Island, NY, USA). The fibroblasts were cultured for six to eight passages. All studies were approved by the Chonbuk National Hospital ethics committee.

Fibroblasts were assessed microscopically for dead cells by trypan blue exclusion. Cell viability was calculated by dividing the non-stained (viable) cell count by the total cell count. The number of cells was determined by averaging the number of cells in four squares and multiplying this average by a dilution factor.

Autophagy was analyzed as described before 19. Synovial fibroblasts from osteoarthritis and rheumatoid arthritis patients were plated at 2 × 10⁵ on glass coverslips in six-well plates and cultured to 70% confluence. Cells were transfected with GFP-LC3 plasmid DNA (kindly provided by Dr. T. Yoshimori, Osaka University, Japan) for 16 h and then treated with thapsigargin or tunicamycin for various times. Transfection was performed using an Amaxa Nucleofector apparatus (Amaxa, Cologne, Germany). Five μg of plasmid DNA were mixed with 0.1 ml of cell suspension, transferred to a 2.0-mm electroporation cuvette, and transfected using an Amaxa Nucleofector apparatus (Amaxa, Cologne, Germany) according to the manufacturer's protocol. The DNA quantity, cell concentration and buffer volume were kept constant throughout the experiments. After electroporation, the cells were transferred immediately to 2.0 ml of complete medium and cultured in six-well plates at 37°C until needed. Microphotographs of GFP-LC3 fluorescence were obtained with a fluorescence microscope. The detection of punctuated staining of GFP-LC3 from diffuse staining indicated the formation of autophagosomes. The punctuated stained cells were compared to the total number of GFP-transfected cells to calculate percents.

Fibroblasts (3 × 10⁶) were washed with phosphate buffered saline (PBS) and incubated for 30 minutes on ice with 100 ml of lysis buffer (10 mM Tris-HCl, 10 mM NaH₂PO₄/NaHPO₄, pH 7.5, 130 mM NaCl, 1% Triton1 X-100, and 10 mM sodium pyrophosphate). Cell lysates were spun down, supernatants were collected, and protein concentrations were determined using the BCA method. For each reaction, 30 μg of protein was added to 1 ml of freshly prepared protease assay buffer (20 mM HEPES pH 7.5, 10% glycerol, 2 mM dithiothreitol) containing 20 mM of AC-DEVD-AMC (Sigma-Aldrich). Reaction mixtures without cellular extracts were used as negative controls. Reaction mixtures were incubated for 1 h at 37°C and the aminomethyl-coumarin liberated from AC-DEVD-AMC was determined by spectrofluorometry (Hitachi F-2500, Hitachi, Tokyo, Japan) at 380 nm^excitation and 400 to 550 nm^emission. Readings were corrected for background fluorescence.

Western blotting was performed using the protocol described previously 19. The total protein was resolved in pre-casted 4 to 12% SDS-PAGE gradient gels. Immunoblotting was performed using the indicated antibodies. ECL reagents (Amersham Biosciences, Piscataway, NJ, USA) were used to visualize signals.

siRNAs were synthesized in duplex and purified using Bioneer technology (Daejon, South Korea). Double-stranded small interfering RNA (siRNA) targeting CHOP (SC-35437) was obtained from Santacruz company (Santa Cruz, California, USA) with control siRNA (SC-37007). For Beclin siRNA, 5'-CAGUUACAGAUGGAGCUAAtt-3' and for non-specific siRNA, 5'-CUUACGCUGAGUACUUCGAtt-3' were transfected into OASF and RASF using Amaxa Nucleofector (Amaxa, Gaithersburg, MD, USA). Briefly, confluent cells were trypsinized and resuspended in Amaxa Nucleofector solution at a density of 2 × 10⁵ cells per 100 μl of solution, and each siRNA was added. Cells were transfected by electroporation using the A24 pulsing program.

The data were analyzed by analysis of variance (ANOVA) in dose-response experiments, or by two-tailed Student's t-tests. A P value < 0.05 was considered significant. In each case, the statistical test used is indicated, and the number of experiments is stated in figure legends.

Decreased susceptibility to apoptosis in rheumatoid arthritis might contribute to resistance to anti-rheumatoid arthritis medications 20. To confirm the characteristics of this resistance, OASF and RASF were exposed to apoptotic stimuli, namely anti-Fas antibody, KCN and thapsigargin. RASF showed a higher resistance to thapsigargin than to other stresses (Figure 1a). Thapsigargin is a Ca²⁺ disturbance agent that leads to the accumulation of unfolded proteins in the ER 21. We therefore questioned whether RASF resists apoptosis induced by excess unfolded proteins in the ER. We compared the dose-dependent sensitivities of RASF to thapsigargin with that of OASF. RASF were relatively resistant to ER stress but OASF were not (Figure 1b). ER stress also decreased caspase-3 activity, the executive caspase, in RASF more than in OASF (Figure 1c), showing ER stress-induced apoptosis is negatively regulated in RASF.

ER stress increases levels of stress proteins such as GRP78 or CHOP, as well as adaptation or cell death pathways 22. In this study, we determined the expression levels of proteins involved in ER-stress-induced cell death in OASF and RASF. Intriguingly, the expression level of the pro-apoptotic protein, CHOP, was significantly decreased in RASF (Figure 2a). However, the expression of glucose response protein 78 (GRP78), also involved in ER stress responses, was similar in OASF and RASF. In addition, elongation initiation factor-1α (eIF-1α), the downstream protein of PERK, a PKR (RNA-dependent protein kinase)-like ER kinase that attenuates protein translation in response to ER stress, was similar in OASF and RASF (data not shown). When stress was prolonged more than 24 h, CHOP expression remained lower in RASF than OASF (Figure 2b). These results indicate that CHOP expression is regulated in RASF, which could explain resistance to ER stress-induced cell death.

When misfolded proteins accumulate in the ER, this stress activates the unfolded protein response (UPR) to induce expression of chaperones and proteins involved in the recovery process 23. Under conditions of ER stress, pre-autophagosomal structures are assembled, and autophagosomal transport to vacuoles is stimulated 24. In this study, we examined whether ER stress induces autophagy in either OASF or RASF. RASF showed high levels of autophagy at relatively low doses of thapsigargin (1 μM) (Figure 3a) and forms autophagosomes (Supplementary Figure 1). ER stress also increased the expression of beclin, an autophagy marker protein, and LC3-II more in RASF than in OASF (Figure 3b). When autophagy is induced, intra-lumenal LC3 is degraded by lysosomal proteases, forming an 18 kDa form (LC3-I) and subsequently being processed to a membrane-bound form (LC3-II, 16 kDa) 25. To verify autophagy, we measured crystal violet-stained vacuoles under a light microscope (Figure 3c). A GFP-LC3 (LC3, mammalian homolog of yeast Atg8) fusion gene was transfected into OASF and RASF to measure changes in autophagosome numbers after treatment. The classical expression pattern of processed LC3-II was more evident in RASF, indicating autophagy vesicle formation. The expression was quantified in the lower panel of Figure 3c.

Autophagy is induced under ER stress conditions to protect against cell death 2627. In this study, we examined the role of ER stress-induced autophagy via knock-down of the autophagy marker, beclin. In OASF and RASF, transfection of beclin siRNA inhibited the expression of beclin, showing efficient transfection (Figure 4a). In RASF, the beclin siRNA decreased autophagy (Figure 4b) and increased cell death (Figure 4c). Beclin siRNA transfection did not affect cell death in OASF (Figure 4c). To understand the pathological meaning of autophagy, we tested its regulatory effect in RASF. In RASF, Ca²⁺-induced autophagy is regulated by hydroxychloroquine, a routinely used Disease-Modifying Anti-Rheumatic Drug (DMARD) that inhibits autophagy 28. Hydroxychloroquine also increased susceptibility to cell death in thapsigargin-treated RASF (data not shown).

Autophagy plays an important role in the characteristics of RASF. In light of this, we examined other ER stress agents in OASF and RASF. First, we treated cells with tunicamycin, an N-acetyl glycosylation inhibitor. In RASF, tunicamycin increased autophagy in a similar manner as thapsigargin (Figure 5a). Using this model, we studied the effect of CHOP. First, CHOP siRNA was transfected into OASF and RASF. CHOP expression was barely detected in either OASF or RASF (Figure 2a). To show transfection efficiency, immunoblots were performed under ER stress-treated conditions. The siRNA knock-down approach successfully inhibited CHOP expression in both thapsigargin and tunicamycin-treated OASF and RASF (Figure 5b). CHOP inhibition significantly increased autophagy in RASF (Figure 5c) and increased cell viability (Figure 5d). Inhibition of CHOP increases cell viability in OASF, clearly showing the pro-apoptotic role of CHOP in OASF as well as in RASF (Figure 5d). These data suggest an inverse relation between CHOP expression and autophagy induction to increase cell resistance against ER stress in RASF.