D- cooperates with sulfate to enhance the surface expression of functional Fas antigen in rheumatoid synovial fibroblasts via the generation of hydrogen peroxide

S. Harada, E. Sugiyama, H. Taki, K. Shinoda, T. Fujita, M. Maruyama, M. Kobayashi

First Department of Internal Medicine, Toyama Medical and Pharmaceutical University, Toyama, Japan.

Abstract Objective D-penicillamine (DP) has been shown to cooperate with copper to inhibit cell growth in a variety of cell types. To determine whether this inhibitory action is involved in Fas-mediated apoptosis, we examined the effect of DP and copper sulfate on the expression and function of Fas antigen in rheumatoid synovial fibroblasts (RSFs). Methods The expression of Fas antigen on the cell surface was determined by flow cytometric analysis. Western blot analysis was performed to examine the protein expressions of Fas and Fas-. In addition, the amounts of apoptotic cells were determined by 4’, 6-diamidino-2’-phenylindol dihydrochloride (DAPI) and propidium iodide (PI) staining. Results Although DP and copper sulfate alone did not affect the surface expression of Fas antigen on RSFs, both in combination augmented the Fas expression in dose- and time-dependent manners. The enhanced expression of Fas antigen on their surface was also observed in interleukin-1 (IL-1 ) and/or tumor necrosis factor (TNF ) stimulated RSFs. On the other hand, the combination of DP and copper sulfate did not increase the amounts of cellular Fas protein, as determined by Western blot analysis. To determine whether the induced Fas antigen is functional, we examined the effect of DP and copper sulfate on Fas-mediated apoptosis, using an agonistic anti- Fas antibody. The treatment of this antibody induced the apoptosis in untreated RSFs, as determined by DAPI staining. The combination of DP and copper sulfate further enhanced the Fas-mediated apoptosis. The enhanced apoptosis and cell surface expression of Fas was completely prevented by catalase, indicating that hydrogen peroxide may be involved in these effects of DP and copper sulfate. The protein expression of Fas-ligand, a natural ligand for Fas antigen, in RSFs. was expressed in untreated RSFs. However, the protein levels were not modulated by DP and copper sulfate. Conclusions Our data demonstrated that DP cooperated with copper sulfate to enhance the cells surface expression of func- tional Fas antigen in RSFs. In addition, Fas-ligand was expressed in the RSFs. These findings suggested that DP might regress rheumatoid synovial hyperplasia via Fas-mediated apoptosis. Key words Rheumatoid arthritis, synovial fibroblast, D-penicillamine, Fas-mediated apoptosis, copper sulfate.

Clinical and Experimental Rheumatology 2002; 20: 469-476. Fas expression by D-penicillamine in RA / S. Harada et al.

Shuji Harada, MD; Eiji Sugiyama, MD, Introduction tion was also seen in RA patients tak- PhD; Hirofumi Taki, MD, PhD; Kouichiro Rheumatoid arthritis (RA) is an inflam- ing DMARDs including DP for longer Shinoda, MD; Tadashi Fujita, MD, PhD; matory joint disease in which perpetua- pe r iods (12). The re fo r e, aggres s i ve ther- Muneharu Maruyama, MD, PhD; Masashi tion of the chronic synovitis to apy with standard DMARDs according Kobayashi, MD, PhD. bone and cartilage degradation (1). Ini- to “sawtooth” strategy (13) or combi- Please address correspondence and tial histological features of RA are cha- n ation therapy ap p e a rs to still have reprint requests to: Dr. Eiji Sugiyama, First Department of Internal Medicine, racterized by synovial lining hyperpla- some impact on the RA treatment. Toyama Medical and Pharmaceutical sia, and the synovial cells play critical Although a number of pharmacological University, Sugitani 2630 Toyama 930- roles in the pathogenesis of RA. These actions of DP were reported, the mech- 0194, Japan. rheumatoid synovial cells have trans- anisms underlying its beneficial effects E-mail: [email protected] formed phenotypes, which express the h ave not been adequat e ly ex a m i n e d. Received on July 17, 2001; accepted protooncogenes ras, myc, fos, or jun Previous studies reported that DP syn- in revised form on March 1, 2002. (2). These protooncogenes up-regulate thesized hydrogen peroxide in the pres- © Copyright CLINICAL AND the proliferation of synovial cells and ence of copper ion, and that the gener- EXPERIMENTAL RHEUMATOLOGY 2002. induce the production of cysteine pro- ated hy d rogen peroxide inhibited the teinase, collagenase, and stromelysin, pro l i fe r ation of T lym p h o c ytes (14, 15 ) , which are involved in cartilage destruc- endothelial cells (16) and fi b ro bl a s t s tion (3). In addition, the synovial cells (17). However, the mechanism of these also produce large amounts of inflam- inhibitory actions still remains unclear. matory cytokines such as interleukin-1 The oxidative stress has been shown to (I L - 1 ) , tumor necrosis factor (T N F ), be critical in triggering cell death by and interleukin-6 (IL-6), which perpet- various stimuli (18), and that the reac- uate rheumatoid inflammation (4, 5). tive oxygen intermediate such as super- Th e re fo re, these synovial cells are oxide (19) or hydrogen peroxide (20) thought to be main targets for the treat- may be involved in apoptosis. These ment of RA. findings prompted us to examine the Dysregulation of apoptosis (program- effect of DP on Fas-mediated apoptosis med cell death) may be one of the rea- in rheumatoid synovial cells. In this sons for synovial hyperplasia in RA. study, we evaluated the effect of DP in The apoptosis of cells is generally initi- the presence of copper ion on the ex- ated by molecular and cellular interac- pression and function of Fas antigen tions such as the binding of Fas to Fas- and Fas-ligand in rheumatoid synovial ligand or TNF to its TNF receptor. fibroblasts. Recent studies have shown that the Fas/Fas-ligand pathway may be dimin- Patients and methods ished by pro i n fl a m m at o ry cy t o k i n e s , Cell preparation s o l u ble Fa s , or soluble Fa s - l i gand in C u l t u red synovial cells used in this RA (6-8), and the decreased signaling s t u dy we re derived from 15 pat i e n t s resulted in imbalance between prolifer- with RA who we re underwent total ation and cell death. Th e re fo re, t h e knee replacement or synovectomy. In- induction of Fas-mediated apoptosis of fo rmed consent was obtained fro m rheumatoid synovial cells is believed to e a ch patient. All patients we re diag- be useful for the treatment of RA (9). nosed as having RA according to the The treatment of patients with RA has 1987 criteria of American College of ch a n ged dra m at i c a l ly in the past 3 Rheumatology (21). The synovial tis- years, and the new drugs targeting in- sues were cut into fragments of 1-3 mm fl a m m at o ry cytokines such as T N F in diameter, and incubated in 0.5-1 and IL-1 were introduced for the treat- mg/ml of collagenase and 5-10 g/ml ment. However, standard disease modi- of deoxyribonuclease I for 2-3 hours. fying anti-rheumatic drugs (DMARDs) After digestion, the resultant single cell such as gold or D-penicillamine (DP) suspension was wa s h e d, fi l t rat e d seems to be beneficial for patients with through sterile gauze and nylon mesh, mild RA in regard to their cost effec- and finally resuspended in Dulbecco’s tiveness and side effects. DP has been modified Eagle’s medium supplement- reported to improve the symptoms of ed with 10% heat inactivated fetal calf RA and to bring to clinical remission serum (FCS), penicillin 100 U/ml, gen- (10, 11). In addition, improved func- tamicin 60 mg/ml, N - 2 - hy d rox ye t h l-

470 Fas expression by D-penicillamine in RA / S. Harada et al. p i p e razine-N-2-ethanesulphonic acid fl u o ride). The cell ly s ates we re cen- To determine the amounts of the apop- 12.5 mM, and L-glutamine 2 mM trifuged at 12,000 g for 10 min at 4¡C, totic cells, we measured the amounts of (DMEM medium). After overnight in- and the supernatants were used for the hypodiploid nuclear DNA contents by c u b at i o n , n o n a d h e rent cells we re re- subsequent experiments. Protein con- flow cytometric analysis of PI staining. m ove d. When adherent cells became centrations were measured by Bradford After incubation with various reagents, confluent, the cells were harvested by methods (Bio-Rad protein assay; Bio- the cells wer e harves t e d , and immediate- trypsinisation, washed, and one third of rad, Richmond, CA). Thirty g of the ly immobilized by 70% cold ethanol them were resuspended in culture dish- protein preparations were separated by overnight. After washing with PBS, the es. The synovial cells at three or more electrophoresis on a 7.5% sodium do- pellet was resuspended in 40 mM cit- p a s s ages became morp h o l ogi c a l ly fi- decyl sulfate-polyacrylamide gel, and rate buffer for 30 min, incubated with broblast-like, and all negative for CD14 transferred onto an Immobilon-P mem- 100 g/ml RNase in PBS for 30 min at antigen and HLA-DR antigen on their brane (Millipore, Bedford, MA). After 37¡C, and stained with 50 g/ml PI in cell surface.Therefore, synovial cells blocking with 5% skim milk, the mem- the dark. The quantity of cells with at 3-6 passages were used as rheuma- brane was incubated with monoclonal hypodiploid DNA was measured on a toid synovial fibroblasts (RSFs). antibodies against Fas (cl o n e 13) or FACScan at the FL-2 channel. Te n Fas-ligand (clone33: Transduction La- thousand cells were examined for each Flow cytometric detection of Fas b o rato ri e s , L ex i n g t o n , K T ) , and then determination. antigen on cell surface with a go at anti-mouse immu n og l o- After treatment with DP (Sigma Chem- bulin antibody conjugated with horse- Statistical analysis ical Co. St Louis, MO) and/or copper radish peroxidase (Santa Cruz Biotech- Values are presented as the mean ± s u l fate (Wa ko Pure Chemical Indus- nology Inc. Santa Cruz, CA). Specific S.D. Data were analyzed by a paired tries, Osaka, Japan), the cells were har- bands were visualized on an enhanced Student t-test. Differences were consid- vested with 0.05% trypsin containing chemiluminescence detection system ered significant at P < 0.05. 0.53 mM EDTA , and washed twice ( E C L , A m e rsham Buck i n h a m s h i re, with staining solution containing 1.0% UK). Each band of blotting film was Results b ovine serum albumin (BSA) and densitometrically quantified by a NIH Effect of DP and copper sulfate on

0.02% sodium azide (NaN3) in PBS. image software. the surface expression of Fas antigen The cell pellet was incubated with 10 on RSFs g/ml a mouse anti-Fas Ab (clone 13: Fluorescence microscopic analysis The surface expressions of Fas antigen Transduction Laboratories, Lexington, with DAPI staining on RSFs were determined by flow cy- KT) for 30 min at 4¡C. Then, the cells To confirm morphologically the devel- tometric analysis. The flow peak levels were washed twice, and resuspended in opment of apoptosis in cell preparation, of the cells stained with FITC labeled 20 g/ml fl u o rescent isothiocya n at e we carried out nuclear staining with 4’, anti-mouse immunoglobulins antibody ( F I T C ) - c o n j u gated rabbit anti-mouse 6 - d i a m i d i n o - 2 ’ - p h e nylindol dihy d ro- were identical between those with or

Ig G + Ig A + IgM F(ab’)2 fragment ch l o ride (DAPI) (Boehri n ger Mann- without treatment of DP and/or copper (Zymed Laboratories Inc., San Francis- heim Biochemicals, Indianapolis, IN). sulfate (data not shown). As shown in co, CA), and incubated in the dark for RSFs were cultured with or wthout DP Figure 1a, Fas antigen was expressed 30 min at 4¡C. The cells stained with and copper sulfate for 24 hours, and the on untreated RSFs, and DP or copper the FITC-conjugated secondary A b cells were further cultured in the pres- sulfate alone did not affect the amounts alone were run in parallel as negative ence of 1 g/ml of an agonistic anti- of Fas antigen. In contrast, the combi- controls. Data acquisition and analysis Fas Ig M Ab (clone 7C11, nation of them enhanced the expression were performed on a flow cytometer I m mu n o t e ch , M a rs e i l l e, France) or of Fas antigen. This enhancement was (FACScan using the Cell Quest soft- control IgM for 24 hours. The stock consistently observed in the RSFs ob- wa re; Becton Dick i n s o n , M o u n t a i n DAPI was diluted to 10 g/ml with tained from 3 different synovial sam- Vi ew, CA). Ten thousand cells we re methanol. To each cell preparation, 30 ples (Fig. 1b). In the presence of copper examined for each determination. l of diluted DAPI solutions we re sulfate, DP enhanced the expression of added, and incubated for 30 min in the Fas antigen in a dose-dependent man- Western blot analysis dark. Then the cells were seeded on a ner (Fig. 1c). The effect of DP was ob- To determine the effect of DP and cop- glass slide, and photographed with a served at a low concentration of 12 g/ per sulfate on the expressions of Fas fl u o rescence microscope (Niko n , ml, which was within ranges reached in and Fas-ligand proteins in RSFs, we To kyo , Japan). The cells with ch ro- blood in vivo during DP treatment (22, carried out Western blot analysis. The matin condensation and nuclear frag- 23). Time-course study showed that the treated cells were lysed in a buffer (10 mentation were considered as apoptotic enhancement of Fas ex p ression wa s mM TRIS hydrochloric acid, pH 7.4, cells. o b s e rved at 12-hour incubation (Fi g. 1% Nonidet P-40, 150 mM sodium 1d). These dose- and time-dependent chloride, 1 mM EGTA, 1 mM EDTA, Flow cytometric DNA content analysis effects of DP and copper sulfate were and 0.2 mM phenyl methylsulphonyl with propidium iodide (PI) c o n s i s t e n t ly observed in 3 diffe re n t

471 Fas expression by D-penicillamine in RA / S. Harada et al.

(a) (b) synovial samples. Effect of IL-1 and TNF on the enhanced expression of Fas antigen by DP and copper sulfate It is known that IL-1 and TNF , major cytokines in rheumatoid synovium (4, 5), modulated Fas-mediated apoptosis (8,24, 25). Therefore, we examined the effect of these inflammatory cytokines on the enhanced expression of Fas anti- gen by DP and copper sulfat e. A s shown in Figure 2, IL-1 , TNF or in combination did not affect the surface expression of Fas antigen with or with- RA-3 out DP and copper sulfate.

Effect of DP and copper sulfate on the protein expression of Fas antigen in RSFs We next examined the effect of DP and copper sulfate on the protein expres- sion of Fas antigen, using Western blot- t i n g. As shown in Fi g u re 3 a,b, t h e combination of DP and copper sulfate did not affect the protein synthesis in 4 different synovial samples. Thus, the enhancement of surface expression of Fas protein cannot be explained by the (c) (d) increasing biosynthesis of Fas protein.

Effect of DP and copper sulfate on the Fas-mediated apoptosis in RSFs To clarify whether Fas antigen induced by DP and copper sulfate is functional, we analyzed the Fas-mediated apopto- s i s , using an agonistic anti-Fas anti- body (7C11). First, we confirmed that

Fig. 1. Effect of D-penicillamine (DP) and cop- per sulfate on the cell surface expression of Fas a n t i gen on rheumatoid synovial fi b ro bl a s t s (RSFs). (a, b) RSFs were cultured with DP (50 g/ml),copper sulfate (2 g/ml) or in combina- tion for 24 hours. Then, the cells were harvested after tri p s i n i z at i o n , and stained with a mouse anti-human Fas Ab, followed by FITC-conjugat- ed rabbit anti-mouse immu n og l o bulins. Th e treated cells were analyzed by flow cytometry. The fi g u res shows the Fas ex p ression of the RSFs with tre atment (bl a ck line) or without treatment (gray line), and control FITC staining (dashed line). (c) RSFs we re cultured with increasing concentrations of DP in the presence of copper sulfate (2 g/ml) for 24 hours. The treated RSFs were harvested, stained with the anti-Fas Ab, and analyzed by flow cytometry. (d) RSFs were cultured with DP (50 g/ml) and copper sulfate (2 g/ml) for the indicated time. The treated cells were analyzed for Fas expres- sion as above. All data are representative of 3 separate experiments using 3 different synovial

472 Fas expression by D-penicillamine in RA / S. Harada et al.

DP and copper sulfate in combination did not affect the cell viability of the R S F s , as determined by trypan bl u e dye - ex clusion test (data not show n ) . Although the combination of DP and copper sulfate did not affect the mor- phological shape in RSFs, typical fea- tures of apoptosis, such as chromatin c o n d e n s ation and nu clear frag m e n- tation, were clearly seen after the addi- tion of the agonistic anti-Fas antibody (Fig. 4a). The induction of apoptosis was consistently observed in 3 different s y n ovial samples, i n d i c ating that the RSFs have the susceptibility to Fa s - mediated apoptosis. A flow cytometric a n a lysis of PI staining was used fo r quantifying the amounts of endonucle- olytic cleavage of DNA in the cells. As shown in Figures 4b,c, the treatment of DP and copper sulfate signifi c a n t ly increased the amounts of hypodiploid population. This effect of DP and cop- per sulfate was consistently observed in four diffe rent synovial samples (Fi g. 4c). The enhanced apoptosis by DP and copper sulfate was also observed in IL- 1 or T N F t re ated RSFs (data not Fig. 2. Effect of interleukin-1 (IL-1 ) and tumor necrosis factor (TNF ) on the enhanced expres- shown). sion of Fas antigen in rheumatoid synovial fibroblasts (RSFs) by D-penicillamine (DP) and copper sul- fate. RSFs were cultured with IL-1 (10 ng/ml), TNF (20 ng/ml) or in combination for 24 hours. Involvement of hydrogen peroxide in Then,the RSFs were further cultured with or without DP (50 g/ml) and copper sulfate (2 g/ml) for 24 hours. The treated cells were harvested, stained with a mouse anti-human Fas Ab, followed by the enhanced expression of Fas anti- FITC-conjugated rabbit anti-mouse immunoglobulins. The stained cells were analyzed by flow cytom- gen by DP and copper sulfate etry. The figure shows the expression of Fas antigen on the RSFs with treatment (black line), without Since DP and copper inhibit cell growt h treatment (gray line), and control FITC staining (dash line). The data are representative of 3 separate via the ge n e ration of hy d rogen pe- experiments using 3 different synovial samples. roxide (15-17), we examined the effect of catalase, a scavenger of hydrogen (a) peroxide, on the enhanced expression of Fas antigen by DP and copper sul- fate. As shown in Figure 5a,b, catalase completely abolished the enhanced ex- pression of Fas. Moreover, the enhanc- ed apoptosis induced by DP and copper sulfate was also prevented by catalase. (b) In addition, hydrogen peroxide enhanc- Fig. 3. Effect of D-penicillamine (DP) and copper sulfate on the protein expressions of ed the expression of Fas antigen, and Fas antigen in rheumatoid synovial fibrob- increased Fas-mediated apoptosis. This lasts (RSFs). (a) RSFs were cultured with induction by hy d rogen peroxide wa s DP (50 g/ml), copper sulf ate (2 g/ml) consistently observed in 3 different sy- or in combination for 24 hours. After incu- bation, the treated cells were harvested for novial samples, indicating that hydro- Western blot analysis using a anti-Fas Ab gen peroxide generated from DP and as described in Patients and Methods. C = copper sulfate, is involved in the induc- positive control. (b)The intensity of bands tion of functional Fas expression. of Fas protein in sep a rate ex p e ri m e n t s using 4 diffe rent synovial samples we re quantified by densitometric analysis using Effect of DP and copper sulfate on the a NIH Image software, and each bar r epre- expression of Fas-ligand in RSFs sents% control (Mean ± SD). We next examined the effect of DP and

473 Fas expression by D-penicillamine in RA / S. Harada et al.

(a) Medium DP/CuSO4 (b)

Anti-Fas Ab DP/CuSO Anti-Fas Ab

(c)

Fig. 4. Effect of D-penicillamine (DP) and copper sulfate on the Fas-mediated apoptosis in rheumatoid synovial fibroblasts (RSFs). (a) RSFs were cultured with or wthout DP (50 g/ml) and copper sulfate (2 g/ml) for 24 hours, and the cells were further cultured in the presence of 1 g/ml of an agonistic anti-Fas Ig M Ab or control Ig M for 24 hours. The treated cells were harvested, stained with 10 g/ml of 4’, 6-diamidino-2’- phenylindol dihydrochloride (DAPI) for 30 min in the dark, and the morphological changes of cells were observed by a fluorescence microscopy. The data are representative of 3 separate experiments. (b)The treated cells with DP and copper sulfate as above were immobilized, stained with 50 g/ml propidium iodide (PI), and analyzed by flow cytometry for the quantification of hypodiploid DNA. (c) RSFs were cultured with or without DP (50 g/ml) and copper sulfate (2 g/ml) for 24 hours,and further cultured in the presence of the anti-Fas Ab for 24 hours. The amounts of hypodiploid DNA was assessed by PI staining.The analysis was performed in 4 different synovial samples. Differences in 2 groups were analyzed by a paired student t-test. copper sulfate on the expression of Fas- served in the RSFs even after treatment c e n t rations within the therap e u t i c ligand, a natural ligand for Fas antigen of these cytokines. On the basis of the range. in the RSFs. As shown in Figure 6 a,b, effect of catalase, this combined effect The mechanism(s) by which DP and the expressions of Fas-ligand were con- was mediated by the generated hydro- copper sulfate enhanced the surface ex- sistently observed in 4 different synovi- gen peroxide. DP significantly enhanc- pression of Fas in RSFs is still unclear. al samples, and the combination of DP ed the Fas expression at the concentra- We showed that hy d rogen perox i d e and copper sulfate did not modulate the tion of 12 g/ml. The concentrations was involved in the combined effect of expression. of DP in the sera of treated patients DP and copper sulfate. In fact,the addi- ra n ges up to 20 g/ml (22, 23). In tion of low dose hydrogen peroxide (50 Discussion addition, the concentration of copper M) to culture resulted in the We demonstrated here that DP enhanc- used in this study, was approximately i n c reased surface ex p ression of Fa s ed the Fas expression on cell surface, half that found in normal serum and a n t i gen in RSFs. Hydrogen perox i d e and augmented Fas-mediated apoptosis much less than those in the sera of RA exerts toxic effects on susceptible cells. in RSFs in the presence of copper sul- patients (26). Moreover, copper ion has However, low concentrations of hydro- fat e. Although IL-1 and T N F a re been also detected in rheumatoid syn- gen peroxide alter cellular functions by known to modulate the Fas-mediated ovial fluid (27). Therefore, it appears m o d u l ating signal transduction. Fo r apoptosis (8, 24, 25), this effect of DP that the enhanced Fas expression by DP example, hydrogen peroxide activates and copper sulfate was consistently ob- and copper sulfate is occurring at con- c ritical tra n s c ription fa c t o rs such as

474 Fas expression by D-penicillamine in RA / S. Harada et al.

(a) (b) (a)

(b)

Fig. 6. Effect of D-penicillamine (DP) and copper sulfate on the protein expressions of Fas-ligand in rheumatoid synovial fi b ro blasts (RSFs). (a) RSFs we re cultured with DP (50 g/ml), copper sulfate (2 g/ml) or hydrogen peroxide (50 M) for 24 hours. After incubation, the treated cells were har- vested for Western blot analysis using an antibody to Fas-lig- and. (b)The intensity of bands of Fas-ligand protein in separate experiments using 4 different synovial samples were quantified by densitometric analysis using a NIH Image software, and each bar represents % control (Mean ± SD).

Fig. 5. Effect of catalase on the enhancement of Fas expression and Fas-mediated apoptosis by D- of hy d rogen peroxide used. A l t e rn a- penicillamine (DP) and copper sulfate. (a) RSFs were cultured with DP (50 g/ml) and copper sulfate t ive ly, since RSFs alre a dy ex p re s s e d (2 g/ml),catalase (1000 U/ml) or hydrogen peroxide (50 M) for 24 hours. Then the cells were har- the Fas protein without stimu l i , we vested, and stained with an anti-human Fas Ab, followed by FITC-conjugated rabbit anti-mouse might not see the effect of DP and cop- immunoglobulins. The treated cells were analyzed by flow cytometry. The figure shows the expression of Fas on the RSFs with treatment (black line),without treatment (gray line),and control FITC stain- per sulfate on the ex p ression of Fa s ing (dashed line). The data are representative of 3 separate experiments using 3 different synovial sam- antigen. Our data indicated that the en- ples. (b) RSFs were cultured with or without DP (50 g/ml),copper sulfate (2 g/ml),catalase (1000 hanced Fas expression by DP and cop- U/ml), or hydrogen peroxide (50 M) for 24 hours. Then the treated cells were further cultured in the per sulfate might be due to the traffick- presence of 1 g/ml an agonistic anti-Fas Ab for 24 hours. The amounts of hypodiploid DNA was assessed by propidium iodide (PI) staining. The data are representative of 3 separate experiments. ing of Fas protein from cytosol to the membrane. Recent study demonstrated that p53 activation transiently increas- N F - B (28) and AP-1 (29). DP and hydrogen peroxide modulated the sig- ed surface expression of Fas by trans- copper sulfate at used concentrations in nal transduction, and enhanced surface p o rt from the Golgi complex (32). this study did not alter cell viability, as Fas expression. Hydrogen peroxide has Since it is known that hydrogen perox- determined by trypan blue dye-exclu- been shown to induce up-regulation of ide induce the activation of P53 (33), it sion test. Production of hydrogen per- Fas in human endothelial cell, and that is likely that the P53 pathway may be oxide by DP and copper ion has been it also induces the Fas mRNA in a involved in the enhanced Fas expres- ob-served in a cell-free system, and the dose-dependent manner (from 200 M sion by DP and copper sulfate. amounts of hydrogen peroxide was as to 1 mM) (31). In contrast, our data What type of cells in rheumatoid syn- low as 8-9 n molar levels at the concen- indicated that the DP and copper sul- ovium are responsible for the induction tration of 50 g/ml DP in the presence fate did not increase the cellular ex- of Fas-mediated apoptosis ? Immuno- of copper sulfate (30). These findings pression of Fas antigen. This discrep- histochemical studies have shown that s u ggested that low concentrations of ancy might be due to the concentration Fa s - l i gand is ex p ressed in synov i a l

475 Fas expression by D-penicillamine in RA / S. Harada et al. infiltrating lymphocytes but not in syn- tumor necrosis factor in rheumatoid arthri- penicillamine in serum and urine of patients ovial lining cells (34). These data sug- tis. Arthritis Rheum 1995; 38: 151-60. with rheumatoid art h ritis. Clin Chim A c t a 6.HASUNUMA T, KAYAGAKI N, ASAHARA H et 1979; 94: 173-80. gest that the interactions between lym- al.: Accumulation of soluble Fas in inflamed 23. MUIJSERS AO, VAN DE STADT RJ, HENRICHS phocytes and synovial cells are impor- joints of patients with rheumatoid arthritis. AM, AMENT HJ, VAN DER KORST JK: D-peni- tant for the induction of apoptosis. Inte- Arthritis Rheum 1997; 40: 80-6. cillamine in patients with rheumatoid arthri- restingly, recent studies demonstrated 7.HASHIMOTO H, TANAKA M, SUDAT et al.: tis. Serum levels, pharmacokinetic aspects, Soluble Fas ligand in the joints of patients and correlation with clinical course and side that hydrogen peroxide induced the ex- with rheumatoid arthritis and osteoarthritis. effects. Arthritis Rheum 1984; 27: 1362-9. pression of Fas-ligand in Jurkat T cell Arthritis Rheum 1998; 41: 657-62. 24. TSUBOI M, EGUCHI K, KAWAKAMI A et al.: line (35) and microglial cells (36). The s e 8.WAKISAKA S, SUZUKI N,TAKEBA Y, et al.: Fas antigen expression on synovial cells was findings prompted us to examine the Modulation by proinflammatory cytokines of down-regulated by interleukin 1 . Biochem Fas/Fas ligand-mediated apoptotic cell death Biophys Res Commun 1996; 218: 280-5. effect of DP and copper sulfate on the of synovial cells in patients with rheumatoid 25. TSUBOI M, KAWAKAMI A, NAKASHIMA T et expression of Fas-ligand.The RSFs ex- arthritis (RA). Clin Exp Immunol 1998; 114: al.:Tumor necrosis factor- and interleukin- pressed the Fas-ligand protein without 119-28. 1 i n c rease the Fa s - m e d i ated apoptosis of s t i mu l ation although DP and copper 9.NISHIOKA K, HASUNUMA T, KATO T, SUMI- human osteoblasts. J Lab Clin Med 1 9 9 9 ; DA T, KOBATA T:Apoptosis in rheumatoid 134: 222-31. sulfate did not modulate their expres- arthritis. A novel pathway in the regulation of 26. SCUDDER PR, AL-TIMIMI D,MCMURRAY W, sions. These ex p ressions we re consi- sy n o vial tissue. Art h r itis Rheum 1998; 41:1-9. WHITE AG, ZOOB BC, D O R M A N DY T L : stently observed in four RSFs obtained 10. B E R RY H, LI YA N AGE SP, DURANCE RA, Serum copper and related variables in rheum- BARNES CG, BERGER LA,EVANS S: Azathio- atoid arthritis. Ann Rheum Dis 1978; 37: 67- f rom diffe rent donors. Consideri n g prine and penicillamine in treatment of rheu- 70. dual expressions of Fas and Fas-ligand matoid arthritis: A controlled trial. Br Med J 27. NAUGHTON DP, KNAPPITT J,FAIRBURN K, antigens in RSFs, it is likely that Fas- 1976; 1: 1052-4. GAFFNEY K, BLAKE DR, GRO O TVELD M: mediated apoptosis could be occurring 11. DIXON AJ, DAVIES J, DORMANDY TL et al.: Detection and investigation of the molecular Synthetic D(-)penicillamine in rheumat o i d nature of low-molecular- mass copper in as a result of cell to cell interaction in arthritis. Double-blind controlled study of a isolated rheumatoid knee-joint synovial fluid. patients with the treatment of DP. high and low dosage regimen. Ann Rheum FEBS Lett 1995; 361: 167-72. In summary, we demonstrated that DP Dis 1975; 34: 416-21. 28. SCHMIDT KN, A M S TAD P, C E RUTTI P, in the presence of copper ion increased 12. PORTER DR,MELNNERS I, CAPELL HA: Out- BAEUERLE PA:The roles of hydrogen perox- come of second line therapy in rheumatoid ide and superoxide as messengers in the acti- the surface expression of Fas antigen in arthritis. Ann Rheum Dis 1994; 53: 812-5. vation of transcription factor NF- B. Chem RSFs via the generation of hydrogen 13. FRIES JF: R e eva l u ating the therapeutic ap- Biol 1995; 2: 13-22. peroxide. Our data suggested that DP proach to rheumatoid arthritis:the ‘sawtooth’ 29. STAUBLE B, BOSCOBOINIK D,TASINATO A, inhibited the cell growth of rheumatoid strategy. J Rheumatol 1990; 17: 12-5 A Z ZIA: M o d u l ation of activator pro t e i n - 1 14. LIPSKY PE, ZIFF M: Inhibition of human (AP-1) transcription factor and protein kinase synovial cells via the induction of Fas- helper T cell function in vitro by D-penicil- C by hydrogen peroxide and D- -tocopherol mediated apoptosis, and that these ef- lamine and CuSO4. J Clin Invest 1980; 65: in vascular smooth muscle cells. Eur J Bio - fects of DP might contribute to the re- 1069-76. chem 1994; 226: 393-402. gression of synovial hyperplasia in RA 15. LIPSKY PE: Immunosuppression by D-peni- 30. STAITE ND, MESSNER RP, ZOSCHKE DC: In cillamine in vitro. Inhibition of human T lym- vitro production and scavenging of hydrogen synovium. phocyte proliferation by copper- or cerulopla- peroxide by D-penicillamine. Relationship to smin-dependent generation of hydrogen per- copper ava i l ab i l i t y. A rt h ritis Rheum 1 9 8 5 ; Acknowledgements oxide and protection by monocytes. J Clin 28: 914-21. 31. SUHARA T, FUKUO K, SUGIMOTO T et al.: The authors grat e f u l ly ack n ow l e d ge Invest 1984; 73: 53-65. 16. MATSUBARA T, SAURA R,HIROHATA K,ZIFF Hydrogen peroxide induces up-regulation of Dainippon Pharmaceutical for provid- M: Inhibition of human endothelial cell proli- Fas in human endothelial cells. J Immunol ing recombinant human IL-1 a n d fe ration in vitro and neova s c u l a ri z ation i n 1998; 160: 4042-7. TNF , Drs. Hitoshi Yamada, Kouichi vivo by D-penicillamine. J Clin Invest 1989; 32. BENNETT M, MACD O NALD K, CHAN SW, 83: 158-67. LUZIO JP, SIMARI R,WEISSBERG P: Cell sur- Kanekasu for rheumatoid synov i a l 17. MATSUBARA T, HIROHATA K: Suppression of face trafficking of Fas: a rapid mechanism of specimens. human fibroblast proliferation by D-penicil- p53-mediated apoptosis. Science 1998; 282: lamine and copper sulfate in vitro. Arthritis 290-3. References Rheum 1988; 31: 964-72. 33. CHEN QM, BARTHOLOMEW JC, CAMPISI J, 18. KAMATA H, HIRATA H: Redox regulation of ACOSTA M, REAGAN JD, AMES BN: Molecu- 1. HARRIS ED JR: Rheumatoid arthritis: Patho- lar analysis of H O -induced senescent-like physiology and implications for therapy. N cellular signalling. Cell Signal 1999; 11:1-14. 2 2 growth arrest in normal human fibroblasts: Engl J Med 1990; 322: 1277-89. 19. CAI J, JONES DP: Superoxide in apoptosis. Mitochondrial generation triggered by cyto- p53 and Rb control G1 arrest but not cell re- 2. GAY S, GAY RE: Cellular basis and oncogene plication. Biochem J 1998; 332: 43-50. expression of rheumatoid joint destruction. chr ome c loss. J Biol Chem 1998; 273: 11 4 0 1 - 34. ASAHARA H,HASUNUMA T, KOBATA T et al.: Rheumatol Int 1989; 9: 105-13. 4. 20. GARDNER AM, XU FH, FADY C et al.:Apop- In situ ex p ression of pro t o o n c ogenes and 3. T R A BANDT A , AICHER W K , G AY RE, totic vs. nonapoptotic cytotoxicity induced by Fas/Fas ligand in rheumatoid arthritis synovi- S U K H ATME V P, FASSBENDER HG, G AY S: um. J Rheumatol 1997; 24: 430-5. Spontaneous ex p ression of immediat e ly - hy d rogen perox i d e. Free Radic Biol Med 35. BAUER M, VOGT M, LOS M, SIEGEL J, WES- early response genes c-fos and egr-1 in colla- 1997; 22: 73-83. SELBORG S , SCHULZE-OSTHOFF K: Role of genase-producing rheumatoid synovial fibro- 21. ARNETT F, EDWORTHY S, BLOCH D et al.: reactive oxygen intermediates in activation- blasts. Rheumatol Int 1992; 12: 53-9. The American Rheumatism Association 1987 revised criteria for the classification of rheu- induced CD95 (APO- 1/Fas) ligand expres- 4. AREND W, DAYER J-M: Cytokine and cyto- sion. J Biol Chem 1998; 273: 8048-55. kine inhibitors or antagonists in rheumatoid matoid arthritis. Arthritis Rheum 1988; 31: 36. VOGT M, BAUER MK,FERRARI D, SCHULZE- arthritis. Arthritis Rheum 1990; 33: 305-15. 315-24. 22. MUIJSERS AO, VAN DE STADT RJ, HENRICHS OSTHOFF K: Oxidative stress and hypoxia- 5. AREND W, DAYER J-M: Inhibition of the pro- / re ox y ge n ation tri gger CD95 (APO-1/Fa s ) duction and effects of interleukin-1 and AM, VAN DER KORST JK: Determination of D-

476