Inhibition of protein geranylgeranylation induces apoptosis in synovial fibroblasts

The protocol for patient consent and the use of human tissues was approved by ethics review committees at both the University Health Network and St Michael's Hospital. All tissue was obtained with patient consent.

Synovial fibroblasts were isolated from synovial tissues of patients with RA or OA removed at the time of arthroplasty, as described previously [21]. Cells were maintained in Opti-MEM supplemented with 4% fetal bovine serum and 1% antibiotic-antimycotic (Life Technologies, Rockville, MD, USA) and were cultured at 37°C in a humidified chamber containing 95% air, 5% CO₂. Synovial fibroblasts were used at passages 2 to 4. Unless indicated otherwise, cells used for Western blots were plated at 10⁵ cells per well in a six-well plate for 48 hours before initiation of experiments.

Stock solutions of lovastatin (Sigma-Aldrich, Oakville, Ontario, Canada) and atorvastatin (Toronto Research Chemicals, North York, Ontario, Canada) were prepared in dimethylsulfoxide (DMSO). Cerivastatin (Toronto Research Chemicals) was prepared in PBS and simvastatin (Calbiochem, La Jolla, CA, USA) was activated and neutralized in accordance with the manufacturer's directions. Recombinant human TNF-α was purchased from R&D Systems Inc. (Minneapolis, MN, USA). GGPP and FPP, sodium 3-(1-(phenylamino-carbonyl)-3, 4-tetrazolium)-bis (4-methoxy-6-nitro) benzenesulfonic acid hydrate (XTT), and phenazine methosulfate (PMS) were obtained from Sigma-Aldrich. Synovial fibroblasts (n = 6 RA, n = 6 OA) were plated in 100 μl at a concentration of 3 × 10⁴ cells/ml in a 96-well plate. Before treatment with statin or carrier control, they were cultured for 24 hours in the presence or absence of 10 ng/ml TNF-α. Statins were then added to the wells and cells were continued to be cultured for the duration indicated in the figure legends. Where indicated, synovial fibroblast cultures were supplemented with 5 μM FPP or 5 μM GGPP at the time of statin addition [22]. Cellular viability was determined by the XTT assay on quadruplicate wells [23]. XTT (1 mg/ml; 50 μl) and 1 μl of 1.25 mmol (PMS) were added to each well and plates were cultured for a further 4 hours. Colour development was measured at 490 nm with a reference wavelength of 650 nm.

RA synovial fibroblasts were incubated for 72 hours with 3 μM simvastatin in the presence or absence of 5 μM FPP or 5 μM GGPP. Integral membrane proteins were extracted by Triton X-114 as described [24]. In brief, 200 μl of lysis solution (10 mM Tris-HCl pH 7.4, 150 mM NaCl, 1.0% Triton X-114) was added to each well of a six-well culture plate and incubated on ice for 30 minutes. Samples were scraped from the plate and overlaid on a sucrose cushion. After a 3-minute incubation at 30°C they were centrifuged at 300 g for 3 minutes to separate the phases. The upper aqueous phase was rinsed twice; then Triton X-114 and buffer were added to the aqueous and detergent phases, respectively, to yield equal volumes and approximately the same salt and detergent concentration in both fractions. Samples were separated on a 12% SDS-polyacrylamide gel. Fractionated proteins were transferred to a poly(vinylidene difluoride) membrane (Immobilon-P; Millipore Corp, Bedford, MA, USA). Membranes were blocked for 1 hour in blocking buffer (10 mM Tris pH 7.5, 2.5 mM EDTA, 100 mM NaCl, 0.1% Tween 20) containing 3% gelatin, and then probed with anti-RhoA (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) or anti-Rac1 (BD Biosciences, Mississauga, Ontario, Canada) for 1 hour. After being washed, membranes were incubated with a goat anti-mouse IgG conjugated to horseradish peroxidase secondary antibody (Sigma-Aldrich) for a further 1 hour, and bands were revealed by enhanced chemiluminescence (ECL; Amersham Pharmacia Biotech, Baie d'Urfe, Québec, Canada).

Cytoplasmic histone-associated DNA fragments associated with apoptotic cell death were determined quantitatively with the Cell Death Detection ELISA^PLUS (Roche Molecular Biochemicals, Mannheim, Germany). Cells were plated at 1.5 × 10⁴ cells per well in a 24-well plate and treated for 72 hours with DMSO or with 3 or 10 μM lovastatin in the presence or absence of 5 μM FPP or 5 μM GGPP as indicated. Cells were lysed with 200 μl of lysis buffer; 20 μl was then added to the microtiter plate followed by 80 μl of immunoreagent. After incubation for 2 hours at room temperature, the plate was washed and developed with ABTS (2,2'-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid)) solution. Each experimental condition was performed in duplicate.

RA synovial fibroblast apoptotic nuclei were detected by a TdT-FragEL DNA fragmentation kit in accordance with the manufacturer's directions (Oncogene Research Products, Cambridge, MA, USA). Cells were plated at 2 × 10⁴ per chamber of a two-well chamber slide and treated for 72 hours with DMSO or 10 μM lovastatin in the presence or absence of 5 μM FPP or 5 μM GGPP as indicated in the figure legends. After TUNEL staining, cells were counterstained with hematoxylin. This was performed on two different lines.

RA synovial fibroblasts were treated with DMSO or 3 μM lovastatin for 48 hours and subsequently cultured for 15 minutes with 10 ng/ml TNF-α or 2 ng/ml IL-1 (R&D Systems Inc, Minneapolis, MN, USA). Cells were then rinsed twice with PBS and lysed in 20 mM Tris pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM β-glycerolphosphate, 0.5 μg/ml leupeptin, 0.2 mM phenylmethylsulphonyl fluoride. Protein concentrations were determined by the micro bicinchoninic acid (BCA) protein assay kit (Pierce, Rockford, IL, USA) to enable comparable sample loading on a 12% SDS-polyacrylamide gel. SDS-PAGE was performed as described for the phase partitioning. pJNK (Thr183/Tyr185), pERK and pAkt (Ser473) antibodies were obtained from Cell Signaling Technology (Beverly, MA, USA). Prehybridization and hybridization were performed in accordance with the manufacturer's protocol. Blots were scanned and band intensities were measured with the UN-SCAN-IT version 5.1 software (Silk Scientific, Orem, UT, USA). Relative amounts of protein were expressed as multiples of the solvent control.

Blots were stripped between probings with Re-Blot-Plus (Chemicon International, Inc., Temecula, CA, USA).

To confirm the involvement of geranylgeranylated proteins in FLS survival and pAkt expression, siRNA for Rac1 (target sequence 5'-AACCGGTGAATCTGGGCTTAT-3') was transfected into RA synovial fibroblasts (n = 3) plated at 6.0 × 10⁴ cells per well of a six-well plate or 3 × 10³ cells per well of a 96-well plate, at an siRNA to RNAiFect Reagent (Qiagen Inc., Mississauga, Ontario, Canada) ratio of 1:4.5. In preliminary experiments, a non-silencing control siRNA labeled with Alexa Fluor 488 was monitored by fluorescence microscopy. The conditions used in this study resulted in a cell transfection efficiency of greater than 95%. The medium was changed 18 hours after transfection. XTT assays were performed and cell lysates were prepared 96 hours after transfection as described above.

Results are presented as means ± SD. Significant differences between groups were analysed by using a two-tailed unpaired Student t test; p < 0.05 was considered statistically significant.

To determine whether statins influenced the viability of RA synovial fibroblasts, RA synovial fibroblasts were cultured in the presence of increasing concentrations of statins ranging from 1 to 10 μM. After 96 hours, cell viability was determined by the XTT assay. The results showed that all of the statins led to reduced viability of RA synovial fibroblasts in a concentration-dependent manner (Figure 1). However, it is noteworthy that simvastatin and cerivastatin reduced FLS viability to a greater extent than lovastatin or atorvastatin. Time-course experiments with atorvastatin (10 μM), simvastatin (10 μM) and cerivastatin (3 μM) revealed that reduced viability occurred in a time-dependent manner (Figure 2). We chose to compare the effects of atorvastatin and simvastatin because they had exhibited different degrees of potency, both on RA synovial fibroblast viability in our preliminary studies and on the CIA mouse model [14]. As shown in Figure 3, in comparison with non-treated cells, simvastatin exhibited a greater effect on the viability of RA synovial fibroblasts (n = 6) than atorvastatin both at 3 μM (51 ± 18% versus 77 ± 13% viability, respectively; p = 0.014) and at 10 μM (24 ± 13% versus 46 ± 19%, respectively; p = 0.04). The effects of the statins were also examined on OA synovial fibroblasts (n = 6). Although both simvastatin and atorvastatin reduced the viability of OA synovial fibroblasts, although to a smaller extent than that of RA synovial fibroblasts, no significant differences were observed between the statins at either concentration.

Because TNF-α is an important driver of RA pathology, we investigated the effects of statins on TNF-α-stimulated proliferation of synovial fibroblasts. Preliminary time-course experiments ranging from 24 to 96 hours (not shown) revealed that similarly to non-treated synovial fibroblasts, those pretreated with TNF-α exhibited decreased viability in a time-dependent manner. As depicted in Figure 3, both statins were able to decrease the viability of TNF-α-stimulated RA synovial fibroblasts (n = 6). In TNF-α-stimulated RA synovial fibroblasts, simvastatin caused a marked reduction in synovial fibroblast viability compared with atorvastatin both at 3 μM (35 ± 11% versus 67 ± 13%, respectively; p = 0.0008) and at 10 μM (2 ± 5% versus 23 ± 6%, respectively; p = 8 × 10^-5). Reduced RA synovial fibroblast viability was observed in TNF-α-stimulated cells compared with unstimulated cells at 10 μM for both simvastatin (p = 0.003) and atorvastatin (p = 0.02). As with unstimulated OA synovial fibroblasts, simvastatin reduced the viability of TNF-α-stimulated OA synovial fibroblasts (n = 6) to a greater extent than atorvastatin, but no statistically significant differences were observed at 3 μM. In contrast with RA synovial fibroblasts, there were no differences between TNF-α-stimulated and unstimulated OA synovial fibroblasts with either statin at 3 μM or 10 μM. There was a significant difference in reduction in viability between TNF-α-stimulated RA synovial fibroblasts compared with OA synovial fibroblasts at 10 μM, both with simvastatin (p = 0.02) and with atorvastatin (p = 0.007). These differences were not explained by differences in the degree of stimulation of OA and RA synovial fibroblasts to TNF-α (p = 0.72 for the atorvastatin control; p = 0.33 for the simvastatin control). Taken together, these results show that although TNF-α stimulates OA and RA synovial fibroblasts equally, RA synovial fibroblasts show a further decrease in viability in the presence of statins.

The decrease in cellular viability caused by statins has been attributed to a lowering of intracellular FPP and GGPP concentrations [25]. This concept was examined in FLS by evaluating the ability of exogenously added FPP and GGPP in the presence of statins to rescue their viability. Our results demonstrated that the statin-induced reduction in RA synovial fibroblast viability could be partly abrogated with 5 μM FPP and completely abrogated with 5 μM GGPP (Figure 4). The complete restoration of viability with GGPP suggests that the inhibition of a geranylgeranylated protein in RA synovial fibroblasts accounts for the reduced viability in the presence of statins.

To explore the mechanism by which the inhibition of geranylgeranylation leads to decreased RA synovial fibroblast viability, proteins that normally associate with the cell membrane as a consequence of geranylgeranylation were examined for redistribution into the cytoplasm in the presence of statins. Potential candidates for such redistribution are geranylgeranylated members of the Rho subfamily of the Ras superfamily. Members of the Rho family control several cellular processes including transcriptional regulation, cell-cycle progression, cell adhesion and apoptosis. Two geranylgeranylated members of the Rho family that have been implicated in cell survival are RhoA and Rac1. With the use of Triton X-114 partitioning, we demonstrated a decrease in membrane-associated RhoA and Rac1 and an increase in cytoplasm-associated RhoA and Rac1 in the presence of simvastatin (Figure 5). The inclusion of geranylgeranylpyrophosphate in the culture medium restored RhoA and Rac1 to the membrane. The results suggest that statins reduce RA synovial fibroblast viability by affecting the normal cellular distribution of geranylgeranylated proteins such as RhoA and Rac1.

To determine whether apoptosis accounts for the decrease in RA synovial fibroblast viability we measured histone-bound DNA fragments in cell lysates and used a TUNEL assay to measure DNA fragments. Lysates prepared from RA synovial fibroblasts grown in increasing concentrations of lovastatin contained increasing amounts of histone bound DNA fragments, suggesting the induction of apoptosis (Figure 6a). Histone-bound DNA fragments decreased about sixfold with the addition of 5 μM FPP and returned to baseline with the addition of 5 μM GGPP (Figure 6b). Similar results were obtained with the TUNEL assay (Figure 7). Taken together, these results suggest that the reduced viability of synovial fibroblasts induced by statins results from apoptosis brought about primarily by the decrease in membrane-associated geranylgeranylated proteins.

We next examined intracellular signaling pathways implicated in RA synovial fibroblast survival. Lovastatin decreased the steady-state pAkt levels at least threefold, and TNF-α or IL-1 signalling through pAkt was also prevented (Figure 8). In contrast, neither the steady-state levels nor the TNF-α or IL-1β induction of pJNK or pERK were affected by culturing RA synovial fibroblasts with 3 μM lovastatin. This suggests that the Akt survival pathway is inhibited by statins in RA synovial fibroblasts.

To determine the relationship between the decrease in geranylgeranylated proteins and pAkt levels, we used siRNA against Rac1. Rac1 was selected because the expression of a dominant-negative form of Rac1 has been shown to sensitize cells to TNF-α-mediated apoptosis [26, 27]. RA synovial fibroblasts transfected with siRNA against Rac1 exhibited a decrease in both Rac1 and pAkt protein levels in comparison with a control siRNA against no known protein coding sequence (Figure 9a,b). Furthermore, synovial fibroblasts transfected with siRNA against Rac1 exhibited decreased viability as determined in an XTT assay (n = 3; p < 0.01; Figure 9c). These results suggest that the inhibition of Rac1 affects pAkt levels and that Rac1 inhibition influences RA synovial fibroblast viability.