Inhibition of protein geranylgeranylation induces apoptosis in synovial fibroblasts

Abstract

Statins, competitive inhibitors of hydroxymethylglutaryl-CoA reductase, have recently been shown to have a therapeutic effect in rheumatoid arthritis (RA). In RA, synovial fibroblasts in the synovial lining, are believed to be particularly important in the pathogenesis of disease because they recruit leukocytes into the synovium and secrete angiogenesis-promoting molecules and proteases that degrade extracellular matrix. In this study, we show a marked reduction in RA synovial fibroblast survival through the induction of apoptosis when the cells were cultured with statins. Simvastatin was more effective in RA synovial fibroblasts than atorvastatin, and both statins were more potent on tumor necrosis factor-alpha-induced cells. In contrast, in osteoarthritis synovial fibroblasts, neither the statin nor the activation state of the cell contributed to the efficacy of apoptosis induction. Viability of statin-treated cells could be rescued by geranylgeraniol but not by farnesol, suggesting a requirement for a geranylgeranylated protein for synovial fibroblast survival. Phase partitioning experiments confirmed that in the presence of statin, geranylgeranylated proteins are redistributed to the cytoplasm. siRNA experiments demonstrated a role for Rac1 in synovial fibroblast survival. Western blotting showed that the activated phosphorylated form of Akt, a protein previously implicated in RA synovial fibroblast survival, was decreased by about 75%. The results presented in this study lend further support to the importance of elevated pAkt levels to RA synovial fibroblast survival and suggest that statins might have a beneficial role in reducing the aberrant pAkt levels in patients with RA. The results may also partly explain the therapeutic effect of atorvastatin in patients with RA.

Figure 1

Statins decrease the viability of synovial fibroblasts in a concentration-dependent manner. Synovial fibroblasts from a patient with rheumatoid arthritis were cultured for 96 hours in the presence of 1, 3 or 10 μM lovastatin, atorvastatin, simvastatin or cerivastatin. XTT assays were then performed on quadruplicate wells to determine the percentage cellular viability of statin-treated cultures in comparison with those treated with the vector. Results are representative of two experiments.

Figure 2

Statins decrease the viability of synovial fibroblasts in a time-dependent manner. Synovial fibroblasts were cultured for 24, 48, 72 and 96 hours in the presence of 10 μM atorvastatin, 10 μM simvastatin or 3 μM cerivastatin. XTT assays were then performed on quadruplicate wells to determine the percentage cellular viability of statin-treated cultures in comparison with those treated with the vector under identical conditions. Results are representative of two experiments.

Figure 3

Statins are most effective at inhibiting the viability of TNF-α-stimulated RA synovial fibroblasts. Synovial fibroblasts derived from patients with rheumatoid arthritis (RA) (n = 6) and osteoarthritis (OA; n = 6) were stimulated or not with 10 ng/ml tumor necrosis factor-α (TNF-α) for 24 hours before the addition of 3 or 10 μM atorvastatin or simvastatin. They were then maintained for a further 96 hours in the presence or absence of TNF-α and statin. XTT assays were performed on quadruplicate wells. The percentage cellular viability was determined by comparing statin-treated cultures with non-treated cultures in either the presence or absence of TNF-α as appropriate. Means and standard deviations are indicated by solid bars. *p < 0.05 for atorvastatin-treated versus simvastatin-treated RA fibroblasts. ⁺p < 0.05 for RA versus OA TNF-α-stimulated fibroblasts.

Figure 4

Geranylgeranylpyrophosphate (GGPP) restores the viability of synovial fibroblasts in the presence of statins. Synovial fibroblasts were cultured for 96 hours in the presence of 1, 3 or 10 μM lovastatin, atorvastatin or simvastatin supplemented with 5 μM farnesylpyrophosphate (FPP) or 5 μM GGPP, or neither FPP nor GGPP (control). XTT assays were performed on quadruplicate wells to determine the percentage cellular viability of treated in comparison with non-treated cultures. Each treatment was performed on two different synovial fibroblast lines. A representative experiment is shown. In a separate experiment, synovial fibroblasts were pretreated with 10 ng/ml tumor necrosis factor-α for 24 hours before the addition of 10 μM atorvastatin or simvastatin and 5 μM FPP or 5 μM GGPP. They were cultured for a further 96 hours before an XTT assay on quadruplicate wells.

Figure 5

Geranylgeranylated proteins are redistributed to the cytoplasm in the presence of statin. Synovial fibroblasts were cultured for 72 hours in the presence of 3 μM simvastatin supplemented or not with 5 μM farnesylpyrophosphate (FPP) or 5 μM geranylgeranylpyrophosphate (GGPP). Phase separation of proteins was achieved by partitioning with Triton X-114. The entire membrane and cytoplasmic fractions were subjected to PAGE on a 12% gel. Blots were probed with mouse anti-human RhoA or Rac1 followed by goat anti-mouse IgG conjugated to horseradish peroxidase secondary antibody, and were revealed by enhanced chemiluminescence.

Figure 6

Cellular lysates of synovial fibroblasts cultured in lovastatin contain histone-bound DNA fragments. Cell lysates and corresponding culture supernatants were assayed by Cell Death Detection ELISA^PLUS for the presence of histone-bound DNA fragments. (a) Synovial fibroblasts were cultured for 72 hours in the presence of 0, 3 or 10 μM lovastatin. The dark bars represent fragments present in the cell lysate indicative of apoptosis; the light bars represent fragments present in the supernatant. Shown is a representative of three different experiments. (b) Synovial fibroblasts were cultured for 72 hours in 10 μM lovastatin supplemented or not with 5 μM farnesylpyrophosphate (FPP) or 5 μM geranylgeranylpyrophosphate (GGPP).

Figure 7

TUNEL assay confirms the presence of DNA fragments in synovial fibroblasts treated with lovastatin. DNA fragments in synovial fibroblasts treated or not with 10 μM lovastatin for 72 hours were detected by a TdT-FragEL DNA fragmentation kit. After TdT-mediated dUTP nick end labelling (TUNEL) staining, cells were counterstained with hematoxylin. TUNEL-positive cells have dark nuclear staining, whereas the nuclei of TUNEL negative cells stain blue. Asterisk, 5 μM geranylgeranylpyrophosphate (GGPP) or 5 μM farnesylpyrophosphate (FPP) were included with the lovastatin.

Figure 8

The steady-state and activated Akt pathway is affected by lovastatin. Rheumatoid arthritis synovial fibroblasts (2 lines shown) were cultured in the presence of 3 μM lovastatin or DMSO (solvent control) for 48 hours. They were then treated with 2 ng/ml IL-1 or 10 ng/ml tumor necrosis factor-α (TNF-α) for 15 minutes. (a) Cellular lysates were prepared, subjected to 12% PAGE and analysed by immunoblotting for pAkt, pJNK, pERK and actin. D, dimethylsulfoxide (DMSO); 3L, 3 μM lovastatin; D, IL-1, cells cultured with DMSO then induced with IL-1; D, TNF, cells cultured with DMSO then induced with TNF-α; 3L, IL-1, cells cultured with 3 μM lovastatin then induced with IL-1; 3L, TNF, cells cultured with 3 μM lovastatin then induced with TNF-α (b-d) Bands were quantified by UN-SCAN-IT and corrected for loading by actin (b) pAkt, (c) pJNK, (d) pERK.

Figure 9

siRNA-directed inhibition of Rac1 reduces synovial fibroblast viability and pAkt levels. Synovial fibroblasts were transfected with siRNA against Rac1 or a non-silencing RNA as control, and cultured for 96 hours at which point XTT assays were performed to measure cell viability and lysates were prepared from wells that had been transfected in parallel. (a, b) Lysates were subjected to immunoblotting and probed with anti-Rac (a) or anti-pAkt or actin (b) for normalization. (c) Effect of siRNA on cell viability. ** p < 0.01 for viability of synovial fibroblasts transfected with non-siRNA control versus Rac1 siRNA. Shown is a representative of three independent experiments. siRNA transfection efficiency was greater than 95%.