Division of Rheumatology, Department of Internal Medicine and
Dr Chul-Soo Cho, Division of Rheumatology, Department of Internal Medicine, School of Medicine, The Catholic University of Korea, St Mary’s Hospital, 62 Youido-Dong, Youngdeungpo-Ku, Seoul 150–713, South Korea. E-mail: email@example.com
Dr Chul-Soo Cho, Division of Rheumatology, Department of Internal Medicine, School of Medicine, The Catholic University of Korea, St Mary’s Hospital, 62 Youido-Dong, Youngdeungpo-Ku, Seoul 150–713, South Korea. E-mail: firstname.lastname@example.org
Inadequate apoptosis may contribute to the synovial hyperplasia associated with rheumatoid arthritis (RA). The Fas-associated death domain protein (FADD)-like interleukin (IL)-1β-converting enzyme (FLICE)-inhibitory protein (FLIP), which is an apoptotic inhibitor, has been implicated in the resistance to Fas-mediated apoptosis of synoviocytes. This study investigated whether hydroxychloroquine (HCQ), an anti-rheumatic drug, induces the apoptosis of rheumatoid synoviocytes, and modulates the expression of FLIP. Fibroblast-like synoviocytes (FLS) were prepared from the synovial tissues of RA patients, and were cultured with various concentrations of HCQ in the presence or absence of the IgM anti-Fas monoclonal antibodies (mAb) (CH11). Treatment with HCQ, ranging from 1 to 100 µM, induced the apoptosis of FLS in a dose- and time-dependent manner. The increase in synoviocytes apoptosis by HCQ was associated with caspase-3 activation. A combined treatment of HCQ and anti-Fas mAb increased FLS apoptosis and caspase-3 activity synergistically, compared with either anti-Fas mAb or HCQ alone. The Fas expression level in the FLS was not increased by the HCQ treatment, while the FLIP mRNA and protein levels were decreased rapidly by the HCQ treatment. Moreover, time kinetics analysis revealed that the decreased expression of FLIP by HCQ preceded the apoptotic event that was triggered by HCQ plus anti-Fas mAb. Taken together, HCQ increases the apoptosis of rheumatoid synoviocytes by activating caspase-3, and also sensitizes rheumatoid synoviocytes to Fas-mediated apoptosis. Our data suggest that HCQ may exert its anti-rheumatic effect in rheumatoid joints through these mechanisms.
Rheumatoid arthritis (RA) is characterized by a tumour-like expansion of the synovium, inflammatory infiltrates and progressive destruction of the cartilage and bone . In the rheumatoid synovium, both the synovial fibroblasts and mononuclear cells express the functional Fas antigen, and these cells can undergo apoptosis by anti-Fas monoclonal antibody (mAb) [2,3]. In animal models of RA, the active induction of apoptosis in the rheumatoid synovium by anti-Fas mAb or Fas ligand (FasL) gene transfer improves the arthritis due to the elimination of both the proliferating synoviocytes and infiltrating lymphocytes in the inflamed synovium [4,5]. During Fas-mediated apoptosis of RA synoviocytes, the Fas-associated death domain protein (FADD) is recruited selectively to the Fas death domain , which suggests that the sensitivity to Fas-mediated apoptosis in synoviocytes may be regulated by the recruitment of FADD to the Fas death domain, resulting in the formation of a death-inducing signalling complex (DISC).
The FADD-like interleukin (IL)-1β-converting enzyme (FLICE) inhibitory protein, FLIP (also known as FLAME-1, a caspase-8 inhibitory molecule), has been identified as a regulator of caspase-8 activation at the DISC . It contains a caspase-like domain that shares significant homology with caspase 8 . FLIP interacts with the adaptor protein FADD and the protease FLICE, and inhibits potently the apoptosis induced by the human death receptors [7,8]. FLIP is expressed during the early stage of T cell activation, but disappears when the T cells become susceptible to Fas ligand-induced apoptosis . The expression of FLIP in macrophages confers resistance to Fas-mediated apoptosis , which may contribute to the development of inflammatory disease. In RA patients FLIP is expressed strongly in the synovium, particularly at the sites of cartilage invasion and bone destruction , suggesting that it contributes to joint destruction.
Hydroxychloroquine (HCQ) has been used widely to treat RA for more than a century. Similar to most anti-rheumatic drugs, HCQ has a wide range of actions. It interferes with the cellular function in the compartments with an acidic microenvironment, such as the lysosomes . This may have different effects on the cellular function, including the inhibition of intracellular processing and protein secretion, the interference of autoantibody production, decreased lymphocyte proliferation and decreased cytokine production, such as tumour necrosis factor (TNF)-α. Recently, it has been observed that HCQ induces apoptosis in several cell types. It induces apoptosis in peripheral blood T lymphocyte through caspase cascades . The apoptosis of leukaemic cells is also induced by HCQ, which is associated with the activation of caspase-3 and the modulation of bcl-2/bax ratio . However, there is no information on the effect of HCQ on synoviocyte apoptosis in RA. In this study, we investigated whether HCQ induces the apoptosis of rheumatoid synoviocytes, and whether it modulates the Fas-mediated apoptosis and the expression of FLIP, the apoptosis inhibitor.
Materials and methods
Isolation and culture of RA synoviocytes
The fibroblast-like synoviocytes (FLS) were prepared from the synovial tissues of six RA patients, who had undergone total joint replacement surgery. The isolation of the FLS from the synovial tissues was performed according to a procedure described elsewhere . Briefly, the tissues were minced into 2–3-mm pieces, and treated for 4 h with 4 mg/ml of type I collagenase (Worthington Biochemical, Freehold, NJ, USA) in Dulbecco’s modified Eagle’s medium (DMEM) at 37°C in a 5% CO2 atmosphere. The dissociated cells were then resuspended in DMEM, supplemented with 10% fetal calf serum (FCS), 2 mM glutamine, penicillin and streptomycin, and then plated in 75 cm2 flasks. After culturing overnight, the non-adherent cells were removed and the adherent cells were cultivated in DMEM plus 10% FCS. The cultures were kept at 37°C in a 5% CO2 atmosphere, and the medium was replaced every 3 days. At confluence, the cells were passed by diluting them 1 : 3 with fresh medium and recultured until used. The synoviocytes, from passages 3–8, were used for each experiment. The cells were morphologically homogeneous and had the appearance of FLS, with a typical bipolar configuration under inverse microscopy. The purity of the cells was examined by flow cytometry analysis (> 95% CD90, < 2% CD14, < 1% CD3 and < 1% CD19 positive).
The FLS were seeded in 24-well plates (Nunc, Roskilde, Denmark) at 3 × 104 cells per well or in a 100 mm culture dish at 5 × 105 cells in 1 ml DMEM/5% FCS, supplemented with 5% FCS, 100 U/ml penicillin, 100 mg/ml streptomycin and 2 mM l-glutamine, and incubated at 37°C in the presence of various HCQ concentrations (kindly provided by Kyung-Poong Pharmaceuticals, Seoul, Korea) ranging from 1 to 100 µM. In some experiments, the FLS were cultured with HCQ in the presence of anti-Fas mAb IgM (CH11; Immnotech, Marseille, France) in order to determine the effect of HCQ on Fas-mediated apoptosis of FLS. Caspase activation was inhibited by preincubating the cells with caspase inhibitors, Z-VAD-FMK (R&D, Minneapolis, MN, USA) and Z-DEVD-FMK (R&D) 30 min before adding the HCQ or HCQ plus anti-Fas mAb. All the cultures were performed in either duplicate or triplicate.
The FLS (3 × 104 cells) undergoing apoptosis were assessed by the level of cellular DNA fragmentation enzyme-linked immunosorbent assay (ELISA). The cellular DNA fragmentation ELISA kit (Roche Applied Science, Indianapolis, IN, USA) is based on the quantitative sandwich ELISA principle using two mouse mAbs directed against the DNA and BrdU. Briefly, an anti-DNA antibody was fixed in the wells of a microtitre plate. The BrdU-labelled DNA fragments contained in the sample were bound to the immobilized anti-DNA antibody. The immune-complexed BrdU-labelled DNA fragments were denatured and fixed on the surface of the plate by microwave irradiation. In the final step, the anti-BrdU peroxidase conjugate was reacted with the BrdU incorporated into the DNA. After removing the unbound peroxidase conjugates, the quantity of peroxidase bound in the immune complex was determined photometrically with tetramethylbenzidine (TMB) as a substrate.
Analysis of caspase activity
The enzymatic activity of caspase-3 was determined using the apotarget caspase-3/cpp32/colourimetric protease assay kit (Biosource, Camarillo, CA, USA), as suggested by the manufacturer. Briefly, 2 × 106 cells were resuspended in 50 µl of lysis buffer and reaction buffer, and a chromogenic cpp32 substrate DEVD-p-nitroanilide (DEVD-pNA) was added. Reactions were incubated at 37°C for 1 h and samples were measured at 405 nm. Fold increase in caspase-3 activity was determined by direct comparison with the level of untreated cells. In some experiments, caspase-8 activity in FLS was also determined by protease assay kit (Biosource).
Flow cytometry for the determination of Fas expression on FLS
After treatment of FLS (5 × 105 cells) with various concentrations of HCQ for 12 h, the cells were harvested, incubated for 20 min on ice in a blocking buffer [phosphate buffered saline (PBS) with 3% FCS] and 0·02% 1 M sodium azide), and subsequently stained for 30 min on ice with phycoerythrin (PE)-conjugated mouse anti-human CD95 antibody (Pharmingen, San Diego, CA, USA), which is specific against Fas. PE-conjugated mouse IgG1 (Pharmingen) was used as the isotype control antibody. The cells were washed and resuspended twice in a staining buffer (PBS containing 3% FCS and 0·02% 1 M sodium azide), and analysed on a fluorescence activated cell sorter (FACScan) cytometer (Becton Dickinson, Mountain View, CA, USA). At least 5000 events were acquired from each sample and subsequently analysed using Lysis II and cellQuest software (Becton Dickinson).
RNA isolation and semiquantitative reverse transcription–polymerase chain reaction (RT–PCR) analysis for FLIPL
The FLS (5 × 105 cells) were incubated with various concentrations of HCQ (0–100 µM) and analysed for FLIPL (the long form of FLIP) mRNA expression by semiquantitative RT–PCR. Briefly, after culture for 4 h, the mRNA was extracted using RNAzol B according to the manufacturer’s instruction (Biotec Laboratories, Houston, TX, USA). The RNA was converted to cDNA using SuperscriptII RT (Gibco brl, Gaithersburg, MD, USA), 10 mM dNTP, 0·1 M DTT, RNase inhibitor (Rnasin, Toyobo, Osaka, Japan) and random hexamer oligonucleotide priming (GibcoBRL). The PCR amplification of the cDNA aliquots was performed by adding 2·5 mM deoxyribonucleoside triphosphate (dNTPs), 2·5 U Taq DNA polymerase (Boehringer, Mannheim, Germany) and 0·25 µM each of the sense and anti-sense primers. The reaction was performed in a PCR buffer (1·5 mM MgCl2, 50 mM KCl, 10 mM Tris HCl, pH 8·3) in a total volume of 25 µl. The following sense and anti-sense primers for FLIPL and glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) were used (all written in 5′→3′ direction): FLIPL sense GTTAGGTAGCCAGTTGG, anti-sense CCTGCCT TGCTTCAGC; GAPDH sense CGATGCTGGGCGTGAG TAC, GAPDH anti-sense CGTTCAGTCCAGGGATGACC. The reactions were processed in a DNA thermal cycler (Hybaid, Teddington, UK). The cycling conditions were as follows: 1 min denaturation at 94°C for FLIPL, 30 s denaturation at 94°C for GAPDH; 1 min annealing at 56°C for FLIPL and at 55°C for GAPDH; and 1 min elongation at 72°C. The PCR rounds were repeated for 30 cycles for FLIPL and 25 cycles for GAPDH, which had been determined to fall within the exponential phase of amplification for each molecule. The PCR products were run on a 1·5% agarose gel and stained with ethidium bromide. The mRNA expression level is presented as a ratio of the cytokine product over the GAPDH product.
Western blotting analysis for FLIP protein
The FLS (5 × 105 cells) were treated with 100 µM of HCQ for a different culture time. The total cellular protein extracts were obtained by washing the cells twice in PBS and resuspending them in a lysis buffer (0·5% Triton X-100, 300 mM NaCl, 50 mM Tris HCl, pH 7·6, containing 1 mM phenylmethylsulphonyl fluoride, 2 µg/ml aprotinin and 10 µg/ml leupeptin). The cells were kept on ice for 30 min and then centrifuged at 10 000 g for 10 min. The amount of the cellular protein present in the clarified supernatant was evaluated using a Bradford protein assay (Bio-Rad Laboratories, Hercules, CA, USA). Equal amounts of the cellular protein (20 µg) from each sample were then separated by 10% sodium dodecyl sulphate-polyacrylamide gel electrophoresis and transferred to polyvinyl difluoride membranes (Amersham Pharmacia, Roosendaal, the Netherlands). The blots were hybridized with rat anti-human FLIP mAb (clone Dave 3, recognizing both FLIPL and FLIPS isoforms; Alexis Biochemicals, Lausanne, Switzerland) or mouse anti-human β-actin mAb (Sigma), followed by horseradish peroxidase-conjugated anti-rat IgG (Alexis Biochemicals) or anti-mouse IgG (Amersham Biosciences, Little Chalfont, UK). The proteins were visualized using SuperSignal West Pico chemiluminescent substrate (Pierce, Rockford, IL, USA) and exposed to X-ray films (Hyperfilm ECL, Amersham Biosciences) according to the manufacturer’s instructions.
The data are expressed as mean ± standard deviation (s.d.). Comparisons of the numerical data between the groups were performed by using a Mann–Whitney U-test. P-values less than 0·05 were considered statistically significant.
Induction of apoptosis by HCQ in rheumatoid synoviocytes
As shown in Fig. 1a, a rapid increase in apoptosis of FLS was observed as early as 1 h after treatment with 100 µM of HCQ, as determined by cellular DNA fragmentation ELISA. The percentage of apoptotic cells was time-dependently increased after the HCQ treatment, compared with the level in the untreated cells. The HCQ-induced increase in the synoviocyte apoptosis was dose-dependent, and the maximum increase was 8·4-fold over the spontaneous levels, as determined 12 h after the treatment with 100 µM of HCQ (Fig. 1b). In contrast, another immunosuppressive agent cyclosporin, ranging from 40 to 4000 nM, failed to increase the degree of apoptosis of RA synoviocytes (data not shown). The dose-dependent induction of apoptosis by HCQ was also evident on phase-contrast microscopy (Fig. 1c). The HCQ-treated FLS became spherical, shrunk and detached from the bottom of the culture plates, which is in contrast to the appearance of the untreated cells with a typical bipolar and attached configuration.
HCQ-induced synoviocytes apoptosis is caspase-3 dependent
It has been reported that HCQ activates caspase-3 in some cell types [14,16], whereas it induces the apoptosis of HL-60 cells independently of caspase-3 . Therefore, we next examined whether HCQ-induced apoptosis of FLS is dependent or independent on the casepase-3 activity. As shown in Fig. 2a, caspase-3 protease activity was increased dose-dependently following treatment with HCQ, which is in parallel with the data from DNA fragmentation ELISA. Moreover, pretreatment of the FLS with a caspase-3 inhibitor, Z-DEVD-FMK, prevented the apoptosis in response to 100 µM of HCQ (Fig. 2b), suggesting that HCQ increased apoptosis of FLS through the activation of capase-3. However, the activity of caspase-8, a critical enzyme for Fas-mediated apoptosis, was not changed by treating cells with HCQ (1–100 µM) for 12 h (data not shown).
HCQ synergistically increases Fas-mediated apoptosis of synoviocyte
The Fas antigen (CD95) is a representative molecule that transmits the apoptotic signal to several types of cells. It has been documented that the FLS express the functional Fas antigen on their surface and undergo apoptosis that is mediated by IgM anti-Fas mAb in vitro. Our next goal was to determine if HCQ could affect the sensitivity of Fas-mediated apoptosis in FLS. As shown in Fig. 3a, the combined treatment of HCQ and IgM anti-Fas mAb enhanced synergistically the degree of FLS apoptosis, compared with either anti-Fas mAb or HCQ alone. For example, when compared with the spontaneous level, the level of apoptosis of the FLS was increased 1·2-fold by 50 ng/ml of anti-Fas mAb, 2·7-fold by 50 µM of HCQ and 5·8-fold by anti-Fas mAb (50 ng/ml) plus HCQ (50 µM). The caspase-3 activity was also increased synergistically by anti-Fas mAb plus HCQ (Fig. 3b). In addition, pretreatment of the FLS with a caspase inhibitor, Z-VAD-FMK (pan-caspase inhibitor), prevented apoptosis in response to either 100 µM of HCQ alone or anti-Fas mAb (100 ng/ml) plus HCQ (100 µM) (Fig. 3c). The caspase-3 inhibitor, Z-DEVD-FMK, showed a similar pattern (data not shown). Taken together, these results indicate that HCQ may sensitize the FLS to Fas-mediated apoptosis via a caspase-3-dependent pathway.
HCQ down-regulates FLIP expression in synoviocytes
Down-regulation of FLIP sensitizes the FLS to Fas-mediated apoptosis . Based on the knowledge that HCQ increases synergistically Fas-mediated apoptosis of the synoviocytes, two possibilities for the mechanism of this synergism were examined. First, HCQ might affect Fas expression on the FLS. Secondly, HCQ could potentiate the Fas-mediated signalling pathway. The results showed that Fas expression, as determined by flow cytometry, was almost unaffected by any HCQ concentration tested (1–50 µM) (Fig. 4a,b). Indeed, the Fas expression level was decreased slightly by the treatment of FLS with 100 µM of HCQ for 12 h, down to approximately 75% of the basal expression level (data not shown). This suggests that the activation of the intracellular signalling pathway, rather than an alteration of the Fas expression level, may be the major mechanism responsible for the increased susceptibility to Fas ligation. Therefore, an experiment was conducted to determine the effect of HCQ on the expression of FLIP, which is an inhibitory signalling molecule of Fas-mediated apoptosis. As shown in Fig. 5a, FLIP mRNA was highly expressed in the unstimulated FLS, but was decreased dose-dependently by the HCQ treatment. The protein expression of both FLIPL and FLIPs was also decreased rapidly as early as 1 h after the treatment with 100 µg of HCQ, and remained down-regulated to 6 h (Fig. 5b). Moreover, time kinetics analysis revealed that the apoptosis triggered by HCQ plus anti-Fas mAb was followed by a decrease in the FLIP protein expression level by HCQ (Fig. 5c), and both were correlated inversely with each other. These results suggest that HCQ may increase the Fas-mediated apoptosis of FLS by modulating FLIP expression.
The synovial fibroblasts contribute to the chronic inflammatory responses of RA as a major part of the invasive pannus . They express strongly a variety of activation markers including surface molecules (e.g. MHC-II, VCAM-1), and thereby can present antigens efficiently to the T cells. The fibroblast cell lines isolated from RA patients have the potential to produce matrix-degrading enzymes and several cytokines, such as IL-1, IL-6 and IL-8 [20–22]. Moreover, synovial fibroblasts proliferate abnormally, and invade the local environments and exhibit the characteristics of tumour cells such as somatic mutations in H-ras and p53 [23,24]. Even though synoviocytes express the functional Fas antigen, most cells escape the process of apoptosis in vivo, which may contribute to the activation and hyperplasia of FLS [25,26]. Therefore, the development of agents that enhance the apoptosis of synoviocytes, such as anti-Fas mAb , would represent a new therapeutic strategy for human RA. In the present study we have demonstrated that an anti-rheumatic drug, HCQ, could induce the apoptosis of FLS. The apoptotic effect of HCQ was observed very early (at 1 h after treatment), and increased dose- and time-dependently. The HCQ concentration used in our culture system was relevant physiologically in that HCQ 100 µM, the maximum dose tested, was similar to the concentration achieved in the plasma of patients receiving anti-rheumatic therapy with 400 mg HCQ daily . Together, these results suggest that the enhancement of synoviocyte apoptosis is one of the major mechanisms explaining the therapeutic benefit of HCQ in RA patients.
FLIP contains two death effector domains and an inactive caspase domain, binds to FADD and caspase-8, and thereby inhibits the death receptor-mediated apoptosis [7,8]. In adjuvant-induced arthritis, FLIP was strongly expressed at the erosion sites and was localized to the pannus . In RA patients, FLIP expression was also found in both the lining and sublining layers, and was identified at the sites of cartilage invasion and bone destruction . The synovial fibroblasts of RA patients had a 50% higher FLIP expression level than those of osteoarthritis patients . Moreover, synovial macrophages in the RA synovium express the FLIP and are refractory to Fas-mediated apoptosis . Overall, it can be postulated that in rheumatoid synovial cells FLIP may regulate the Fas-mediated apoptosis by interfering with the interaction between FADD and caspase-8 [7,8,31]. In this respect, the inhibition of FLIP may augment synovial apoptosis , thereby ameliorating the RA inflammation. In the present study, we demonstrated first that HCQ synergistically increased the level of Fas-mediated apoptosis of synoviocytes. Moreover, HCQ (0·1–50 µM) inhibited the mRNA and protein expression levels of FLIP, but not Fas. This suggests that HCQ may sensitize the Fas-mediated apoptosis of rheumatoid synoviocytes by down-regulating FLIP expression.
HCQ activates caspase-3 in some cell types [14,16], which is a critical step for DNA fragmentation and apoptotic cell death. In this study, HCQ was able to increase the caspase-3 activity in FLS. Moreover, the caspase inhibitors, Z-VAD-FMK (pan caspase inhibitor) and Z-DEVD-FMK (caspase-3 inhibitor), block synoviocyte apoptosis almost completely by either HCQ or HCQ plus anti-Fas IgM. Therefore, it is conceivable that the increased apoptosis of FLS by HCQ may have been caused by the cumulative action of it on at least two different pathways, e.g. the direct activation of caspase-3 and the potentiation of intracellular Fas signalling by suppressing FLIP expression, which may trigger caspase activity indirectly. However, because there are many more molecules participating in the apoptotic death of FLS, other molecules such as bcl-2 and bax, whose expression level are significantly modified by HCQ in other types of cells [14,16], are probably also involved in this process.
It is unclear how HCQ regulates FLIP mRNA expression. It has been proposed that HCQ interferes with the post-transcriptional events [11,12]. HCQ is a weak base that is known to affect the acid vesicles leading to a dysfunction of the enzymes essential for the post-translational modifications of proteins. Therefore, HCQ might alter the function of some proteins critical for maintaining the stability of FLIP mRNA, just as it modified the gp120 protein of human immunodeficiency virus-1 (HIV-1) [32,33]. Otherwise, it is possible that FLIP mRNA transcription may be disrupted by HCQ without interfering with the mRNA stability. The relevant evidence was obtained from the earlier findings that chloroquine inhibited TNF-α mRNA expression in a macrophage cell line and peripheral blood mononuclear cells [34,35], and that the activation of some transcriptional factors, such as AP-1, are dependent on a chloroquine-sensitive step . Further studies to evaluate these possibilities are currently under way.
In summary, HCQ induced the apoptosis of the FLS in dose- and time-dependent fashion. A combined treatment of HCQ and anti-Fas mAb increased the apoptosis of FLS synergistically, compared with either anti-Fas mAb or HCQ alone. Moreover, treatment with HCQ plus anti-Fas mAb resulted in a rapid decrease in FLIP expression in the FLS, which was followed by the apoptotic event. These results suggest that HCQ sensitizes rheumatoid synoviocytes to Fas-mediated apoptosis by down-regulating FLIP expression, and may exert its therapeutic effect against RA via this mechanism.
This work was supported by grants from the Korea Health 21 R&D Project, Ministry of Health and Welfare, Republic of Korea (0405-DB01-0104–0006) and St Vincent’s Hospital in Suwon.