- Email: [email protected]
Gold-induced reactive oxygen species (ROS) do not mediate suppression of monocytic mitochondrial or secretory function
Toxicology in Vitro 20 (2006) 625–633 www.elsevier.com/locate/toxinvit
Gold-induced reactive oxygen species (ROS) do not mediate suppression of monocytic mitochondrial or secretory function Yo Omata a, Jill B. Lewis b, Petra E. Lockwood b, Wan Y. Tseng c, Regina L. Messer b, Serge Bouillaguet d, John C. Wataha b,¤ b
a Hokkaido University, Sapporo, Japan Medical College of Georgia, Oral Biology and Maxillofacial Pathology, Augusta, GA 30912-1160, United States c National Taiwan University, Taipei, Taiwan, ROC d School of Medicine, University of Geneva, Geneva, Switzerland
Received 24 August 2005; accepted 14 November 2005 Available online 27 December 2005
Abstract The toxicity of anti-rheumatic gold compounds has limited their use and development, yet both the toxicological and therapeutic actions of these compounds remain unclear. In the current study, we tested the hypothesis that intracellular reactive oxygen species (ROS) induced by Au(I) or Au(III) compounds mediate their ability to suppress mitochondrial activity. Methods: Human THP1 monocytes were exposed to HAuCl4 · 3H2O (Au(III)), or the anti-rheumatic compounds auranoWn (AF) or gold sodium thiomalate (GSTM) for 6–72 h, after which mitochondrial activity (succinate dehydrogenase) was measured. To assess the role of cellular redox status as a mediator of mitochondrial suppression, monocytes were pre-treated with a pro-oxidant (t-butyl hydroquinone, t-BHQ) or antioxidant (N-acetyl cysteine, NAC ). ROS levels were measured 0–24 h post-gold addition to determine their role as mediators of mitochondrial activity suppression. Results: AF was the most potent inhibitor of mitochondrial activity, followed by Au(III) and GSTM. Only Au(III) induced intracellular ROS; no ROS formation was observed in response to AF or GSTM exposure. Although anti- and pro-oxidants had some eVects on mitochondrial suppression of Au compounds, collectively the data do not support redox eVects or ROS formation as major mediators of Au-compound mitochondrial suppression. Conclusions: Our results do not indicate that ROS and redox eVects play major roles in mediating the cytotoxicity of AF, GSTM or Au(III). © 2005 Elsevier Ltd. All rights reserved. Keywords: Rheumatoid arthritis; Metals; Redox; Mitochondrial activity
1. Introduction Gold compounds have been used therapeutically for over a century to treat a variety of diseases, but most commonly to treat rheumatoid and several other arthrides (Parish, 1992; Fricker, 1996; Simon, 2000). The most common gold-based anti-arthritic drugs today are gold sodium thiomalate (GSTM, injected intramuscularly) and auranoWn
*
Corresponding author. Tel.: +1 706 721 2991; fax: +1 706 721 3992. E-mail address: [email protected] (J.C. Wataha).
0887-2333/$ - see front matter © 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.tiv.2005.11.001
(AF, oral), both of which contain Au(I) bound to organic ligands (Fig. 1, Schmidbaur, 1992; Simon, 2000). The therapeutic value of these compounds in limiting the progression of rheumatoid arthritis is unquestioned, but they have serious toxic side eVects including diarrhea, skin rashes, blood dyscrasias, and proteinurea (Kean et al., 1997; Eisler, 2003). These side eVects often are life-threatening and force patients to stop therapy despite a signiWcant reduction of pain and joint destruction (Simon, 2000). In contrast, Au(III) compounds (Fig. 1, Messori et al., 2000) are not used clinically because of the reactivity and cytotoxicity of the Au(III) cation, but numerous new Au(III) compounds
626
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
Fig. 1. Gold compounds tested in the current study. AuranoWn (AF, upper left) is a Au(I) compound with thiol- and phosphine-linked ligands. Gold sodium thiomalate (GSTM, below) is a Au(I) compound with a single thiol-linked ligand. Both AF and GSTM have been successfully used to treat arthrides, but have signiWcant toxicity clinically. HAuCl4 · 3H2O (Au(III), upper right) is an inorganic Au(III) compound, not currently used medicinally because of its toxicity. However, its use has been proposed for treatment of cancer.
have been proposed for the treatment of arthritis, cancer, or other diseases (Novelli et al., 1999; Messori et al., 2000; Tiekink, 2002; Che et al., 2003; Bendek, 2004). In spite of their long history of medicinal use, the mechanisms by which gold compounds are therapeutic or toxic are not known. This ignorance has severely limited the full clinical utility of Au(I) compounds, the development of new Au(I) or Au(III) compounds as drugs, and strategies to limit side eVects (Simon, 2000; Wong et al., 2003). Historically, GSTM and AF have been considered immunosuppressive agents, thought to limit the activation of monocytes and other blood cells. These compounds have been reported to limit secretion of pro-inXammatory mediators such as IL-1, IL6, or TNF (Drakes et al., 1987; Remvig et al., 1988; Schmidt and Abdulla, 1988; Barrerra et al., 1996; Bondeson, 1997; Yoshida et al., 1999). By ‘suppressing’ activation of inXammatory cells, these compounds purportedly limit synovial hyperplasia, B- and T-cell activation, and erosion of cartilage and bone (Lipsky, 1994). However, recent evidence indicates that gold compounds are not simply suppressors of immune function, but complex modulators of cytokine secretion (Barrerra et al., 1996; Bondeson, 1997; Lampa et al., 2002; Seitz et al., 2003). Auinduced modulation may result in up- or down-regulation of cytokine secretion, particularly when the gold compounds are present in conjunction with other cellular activators such as TNF or lipopolysaccharide (LPS). Similar results have been reported for Au(III) chloride (Stern et al.,
2005). The mechanisms by which gold-based compounds induce these complex, cytokine-speciWc responses also are under-explored and unclear. Yet understanding these mechanisms is essential to optimizing the use of existing compounds and developing new ones. Several lines of evidence suggest that Au-based compounds act, at least in part, by modifying cellular redox balance. In general, metal ions are thought to elevate levels of intracellular reactive oxygen species (ROS) (Kasprzak, 1995; Lloyd and Phillips, 1999; Kudrin, 2000). Elevated ROS levels are suspected triggers of several transcription factors such as AP1 and NFB (Byun et al., 2002; Li and Stark, 2002) that mediate the secretion of many inXammatory cytokines and factors (Barnes and Karin, 1997). Other evidence suggests that gold compounds (e.g., AF) may activate Nrf2, a transcription factor that recruits an antioxidative cellular response (Katoaka et al., 2001). Furthermore, Au-induced oxidative stress may cause a buildup of oxidized thioredoxin, a key intracellular redox protein that mediates transcription factor activation, nuclear translocation, and DNA binding (Handel, 1997). The ability of Au-based compounds and Au ions to elevate or modulate intracellular ROS formation at sublethal levels has not been reported in spite of the potential importance of this mechanism to the therapeutic and toxicological proWle of these compounds. In the current study, we test the hypothesis that Au compounds generate reactive
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
oxygen species (ROS) that mediate cellular eVects. We have focused here on the ability of Au-based compound to suppress mitochondrial activity.
627
pounds were added a second time before ROS levels were measured. 2.3. Cell activators, antioxidants, and pro-oxidants
2. Material and methods 2.1. Au compounds Three gold compounds were evaluated (Fig. 1) based on their use or proposed use as therapeutic agents. Gold sodium thiomalate (GSTM, Sigma-Aldrich) and auranoWn (AF, Alexis Corporation) have been used for over 70 years to treat rheumatoid and other arthrides (Fricker, 1996; Simon, 2000). Au(III) is the metallic center of a wide variety of compounds proposed as anticancer agents (Messori et al., 2000; Che et al., 2003). For each of these compounds, the ability to alter cellular redox balance has been proposed as a possible mechanism of action (Ercal et al., 2001). AF is a Au(I) compound with thiol- and phosphine-linked organic ligands in a linear conWguration. GSTM also is a Au(I) compound, but with a single thiol-linked organic ligand. HAuCl4 · 3H2O (Au(III), Sigma-Aldrich) is an inorganic compound with square planar coordination chemistry. Compounds were added to cell-cultures from sterile stock solutions in water (GSTM, Au(III)) or ethanol (AF, 0.5% Wnal concentration).
Cell activators, antioxidants, or pro-oxidants were added to cultures to test the hypothesis that gold compounds alter mitochondrial function via redox-related mechanisms. Lipopolysaccharide (LPS, Ecoli, serotype
2.2. Cell-culture and experimental strategy The monocyte is one key target of anti-arthritic compounds (Bondeson, 1997) and a major orchestrator of inXammatory responses (Auger and Ross, 1992). Thus, human monocytic cells (THP1, ATCC TIB 202) were used to measure Au-induced ROS levels or changes in cellular mitochondrial function or ROS levels. THP1 were maintained in RPMI 1640 supplemented with 10% FBS, 2 mM glutamine, 100 g/mL streptomycin, 100 U/mL penicillin (all from Gibco BRL), and 50 M -mercaptoethanol (Sigma-Aldrich). The -mercaptoethanol, a reducing agent, was withheld 3-d prior to the start of experiments to avoid masking any oxidative eVects of the gold compounds or enhancing the eVects of added antioxidants. For cytotoxicity experiments, cells were plated at 50,000 cells/well in 200 L in 96-well round-bottom format (n D 8/condition). Activators, antioxidants, pro-oxidants, or the metal compounds were added as described below to provide a 24, 48, or 72 h total exposure to the gold compounds. For experiments that measured mitochondrial suppression, succinate dehydrogenase (SDH) activity of the cells was measured using the MTT method (see below). For measurement of reactive oxygen species (ROS), cells were plated at 10,000 cells/well in 200 L in 96-well Xat-bottom format (n D 3/condition). Antioxidants were added in selected experiments as described below. Gold compounds were then added and incubated with the cells for 0–24 h, after which intracellular ROS levels were measured using the DFDA assay (see below). In selected cases, the Au com-
Fig. 2. Mitochondrial suppression by auranoWn (AF), gold sodium thiomalate (GSTM), and gold chloride (Au(III)) in THP1 monocytes, estimated by measuring succinate dehydrogenase (SDH) activity. Gold compounds were exposed to monocytes for 24, 48, or 72 h. Results were expressed as a percentage of controls exposed only to vehicle (water for GSTM and Au(III), 0.5% ethanol for AF). Error bars represent one standard deviation of the mean (n D 8/condition). Asterisks indicate a statistical diVerence from vehicle controls for each concentration (t-tests, 2-sided, D 0.05).
628
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
Fig. 3. Mitochondrial suppression 72 h after exposure to auranoWn (AF), gold sodium thiomalate (GSTM), or gold chloride (Au(III)) in monocytes treated with activators (LPS, 1 g/mL; TNF 10 ng/mL; t-BHQ, 100 mM, last 24 h) or pretreated (2 h) with antioxidants (NAC, 5 mM). The control graph is identical to Fig. 2 (bottom) and is shown here for comparison. Mitochondrial suppression was estimated by succinate dehydrogenase (SDH) activity. Results were expressed as a function of controls exposed only to NAC, LPS, t-BHQ, or TNF. Error bars represent one standard deviation of the mean (n D 8/ condition). Asterisks indicate a statistical diVerence from control for each concentration (t-tests, 2-sided, D 0.05).
0111:B4, 1 g/mL for last 24 h, Sigma) and tumor necrosis factor-alpha (TNF, recombinant human, 10 ng/mL for last 24 h, R&D Systems) were two activators relevant to monocytes in arthrides or infection (Simon, 2000). LPS also has been described as a pro-oxidant (Suliman et al., 2004). Selenium (Se, Na2SeO3, 50 nM, Sigma), N-acetyl cysteine (NAC, 5 mM, Sigma), and vitamin E (Vit E, -tocopherol, 10 M, Sigma) were antioxidants added 2 h before metal addition to attempt to mitigate any redox-mediated, Auinduced eVects. t-BHQ (tert-butyl hydroquinone, 100 M, added last 24 h, Sigma) was used as a pro-oxidant. Pilot studies ensured that none of the activators, antioxidants, or pro-oxidants aVected monocytic mitochondrial function or induced ROS levels by themselves at the above concentrations and times of addition.
2.4. Cell succinate dehydrogenase (SDH) activity The activity of succinate dehydrogenase (SDH), an enzyme unique to the mitochondria and central to mitochondrial function, was measured (n D 8/condition) using MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide, Sigma) reduction as described previously (Wataha et al., 1995). This assay was structured to estimate cellular mitochondrial activity (Wataha et al., 1991). Control wells received only vehicle (water for Au(III) and GSTM, 0.5% ethanol for AF), and were used to normalize SDH activity of Au-treated wells. SDH activity experiments were used to select sublethal concentrations of gold compounds for use in the assessment of intracellular ROS.
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
629
2.5. Measurement of reactive oxygen species (ROS) The dihydroXuorescein diacteate (DFDA) assay was used to estimate intracellular ROS levels induced by gold compounds. DFDA is taken up by cells, then activated by esterase-mediated cleavage of acetate, which traps dihydroXuorescein (DF) in the cells. DF may be oxidized by various ROS to Xuorescein, which has measurable Xuorescence 530 nm (excitation @ 485 nm)(Hempel et al., 1999). The assay therefore estimated only intracellular ROS levels. After incubation with gold compounds, cells were washed with 200 L/well of Hallam’s buVer, then 100 L of Hallam’s was re-added to maintain cell viability. During all measurements, cells were kept in the dark as much as possible. Diamide (5 mM, Sigma) was then added to control wells (t D 0) as a positive control (Lockwood et al., 2005); negative controls received no gold compounds. DFDA (20 M, Molecular Probes) was then added 0–60 min after time zero, and the cultures were incubated for a total of 90 min, after which Xuorescence was measured. In this manner, the lifetime of the ROS could be estimated (Lockwood et al., 2005) because the DFDA only reacted with ROS that remained from the time of DFDA addition to 90 min. 3. Results 3.1. Mitochondrial suppression AuranoWn (AF) was the most potent suppressor of mitochondrial function, inhibiting succinate dehydrogenase (SDH) activity completely above 2 M, regardless of exposure time (Fig. 2, 24–72 h). The suppression by HAuCl4 · 3H2O (Au(III)) was less than that of AF, but was increased as the exposure time increased from 24 h to 72 h. Au(III) concentrations that inhibited SDH activity were generally above 100 M. Gold sodium thiomalate (GSTM) was the least potent of the three compounds, inhibiting SDH activity signiWcantly only after a 72 h exposure to 100–200 M. For every duration of exposure, the rank of the cytotoxicities of these compounds was AF > Au(III) > GSTM (Fig. 2). These data were used to select concentrations for measuring goldinduced production of reactive oxygen species. For Au(III) and GSTM, concentrations below 100 M were used; for AF, concentrations below 1 M were used. Mitochondrial suppression by AF was not aVected by the pre-addition of the antioxidant N-acetyl-cysteine (NAC) (Fig. 3). However, GSTM or Au(III) suppression was somewhat reduced with NAC. Lipopolysaccharide (LPS) did not aVect SDH activity levels in response to any of the gold compounds. However, addition of t-BHQ markedly increased the potency of all three compounds, and particularly that of GSTM. The addition of TNF did not aVect AF or GSTM-modulated SDH activity, but tended to increase SDH activity (10–20%) when Au(III) was added concurrently. These increases occurred at Au(III) concentrations of 90–110 M, but were not statistically signiWcant.
Fig. 4. Levels of intracellular reactive oxygen species (ROS) generated in monocytes 0, 3, and 24 h after exposure to HAuCl4 · 3H2O. After exposure to Au(III), cells were labeled with dihydroXuorescein diacetate (DFDA, 20 M) and ROS levels were determined for up to 90 min, measuring Xuorescence emission at 530 nm (excitation at 485 nm). Levels of ROS were expressed as a percentage of ROS induced by the pro-oxidant diamide (5 mM, 30 min, control level shown with dashed line). Diamide controls induced a change of approximately 7500 Xuorescence units over background levels. Error bars indicate one standard deviation (n D 3). Asterisks indicate statistical diVerences from background levels for each time after gold compound addition (t-tests, 2-sided, D 0.05).
3.2. Production of reactive oxygen species (ROS) Au(III) induced ROS formation to levels Wve times those of the diamide positive control (Fig. 4). ROS induction was concentration dependent, and induced ROS were detected after exposure to the lowest concentration tested (5 M). Induced ROS decreased with time and returned to baseline levels by 90 min (Fig. 4, top); after 3 or 24 h, no evidence of
630
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
Fig. 5. Levels of intracellular reactive oxygen species (ROS) generated in monocytes when HAuCl4 · 3H2O was added a second time 3, 6, or 24 h after the Wrst addition. After the second exposure to Au(III), cells were labeled with dihydroXuorescein diacetate (DFDA, 20 M) and ROS levels were determined for up to 90 min, measuring Xuorescence emission at 530 nm (excitation at 485 nm). Levels of ROS were expressed as a percentage of ROS induced by the pro-oxidant diamide (5 mM, control level shown with dashed line). Diamide controls induced a change of approximately 7500 Xuorescence units over background levels. Error bars indicate one standard deviation (n D 3). Asterisks indicate statistical diVerences from background levels for each time after gold compound addition (ttests, 2-sided, D 0.05).
secondary or persistent ROS formation was evident. When Au(III) was added a second time, ROS formation depended on the time of secondary addition (Fig. 5). At 3 h, the response was identical to that induced by the Wrst addition (Figs. 4 and 5). However, if the second addition occurred at 6 h, the ROS response was mitigated signiWcantly. When Au(III) was added a second time at 24 h, the mitigation seen at 6 h persisted.
Fig. 6. EVect of the antioxidants N-acetyl cysteine (NAC, 5 mM, top), selenium (Se, from Se2O3, 50 nM, middle) or vitamin E (-tocopherol, 10 mM, bottom), on levels of reactive oxygen species generated in monocytes after addition of HAuCl4 · 3H2O (Au(III)). Controls contained only the antioxidants (0 M). ROS were assessed as in Figs. 4,5. Error bars indicate one standard deviation (n D 3). Asterisks indicate statistical diVerences from background levels for each time after gold compound addition (t-tests, 2-sided, D 0.05).
Compounds commonly employed as cellular antioxidants all signiWcantly reduced Au(III)-induced ROS formation (Fig. 6). Treatment with NAC, Se, or Vit E all reduced the peak ROS levels induced by 75 M Au(III) by 4–5 times. There were no apparent diVerences in the ability of these three compounds to mitigate Au(III)-induced ROS formation. Unlike Au(III), AF and GSTM did not induce detectable ROS formation. Compound-induced ROS were not observed immediately after addition (Fig. 7), nor after 3 or 24 h (data not shown). Furthermore, a second addition of the compounds at 24 h did not induce ROS formation (Fig. 7).
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
631
Fig. 7. Levels of intracellular reactive oxygen species (ROS) generated in monocytes immediately after addition of AF (left) or GSTM (right), or after second addition at 24 h. ROS were measured as described in Figs. 4 and 5. Error bars indicate one standard deviation (n D 3). There were no statistically signiWcant levels of ROS detected.
4. Discussion Collectively, our data do not support our hypothesis that ROS and redox eVects play major roles in mediating the cytotoxicity of AF, GSTM or Au(III). Furthermore, each of these compounds appears to act via distinct mechanisms despite our anticipation that they would behave similarly. AuranoWn (AF) was the most potent mitochondrial suppressor (Fig. 2), and mitochondrial suppression was uneVected by pro- or antioxidants (Fig. 3). Furthermore, AF induced no formation of ROS either initially or after addition a second time (Fig. 7). Although the relatively potent cytotoxicity of AF was expected (Simon, 2000), we found no evidence to support redox mechanisms or ROS as mediators of AF toxicity. It is possible that the method to determine ROS formation was not sensitive enough to detect AF-induced ROS, but this seemed unlikely given the ability of the assay to detect ROS induced by Au(III) (see discussion below). Gold sodium thiomalate (GSTM) also induced no intracellular ROS (Fig. 7). The addition of NAC slightly reduced, and the pro-oxidant t-BHQ increased GSTMinduced mitochondrial suppression (Fig. 3), suggesting at least some role for redox-mediated eVects. However, the absence of ROS generation eliminated a deWnitive role of this mechanism. Because both NAC and t-BHQ may act by altering (increasing or decreasing, respectively) reserves of thiol-based reducing power in the cell, it is possible that
NAC and t-BHQ either bolstered or limited thiol reserves, thereby limiting or facilitating GSTM eVects. Although speculative, this mechanism could be assessed by measuring glutathione redox balance, which is the most common source of cellular thiol-mediated reducing power. Mitochondrial suppression induced by Au(III) was largely uneVected by NAC, but was enhanced by t-BHQ (Fig. 3). The ability of Au(III) to induce a potent but transient increase in ROS levels initially suggested that Au(III) caused cellular toxicity, at least in part, via ROS formation. However, the inability of NAC to limit cytotoxicity, coupled with the continued development of the Au(III)-induced cytotoxicity response long-after the ROS increases had returned to baseline levels, does not support a deWnitive role for ROS as mediators of Au(III) eVects. Our data do not support ROS generation as a mechanism by which Au(I) compounds such as AF or GSTM alter cytokine secretion (Barrerra et al., 1996; Bondeson, 1997; Lampa et al., 2002; Seitz et al., 2003; Stern et al., 2005). Our results dispel a long-espoused mechanism of action of these compounds. The previously reported ability of AF to activate Nrf2, a transcription factor responsive to pro-oxidants (Katoaka et al., 2001), suggests that Nrf2 must be activated in a ROS-independent manner, perhaps by binding the Keap-1 Nrf2 inhibitor directly. The current results leave open the possibility that Au(III) compounds alter monocytic cytokine secretion via
632
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
ROS-mediated mechanisms. Au(III) generated a potent but transient increase in intracellular ROS (Fig. 4). The ability of NAC to reduce Au(III)-induced ROS suggests that Au(III) acts at some point by changing cellular thiol redox balance. ROS levels were transient and did not induce secondary ROS (Fig. 4), but did alter cell status such that ROS induced by second exposures were reduced (Fig. 5). We observed this mitigation after Au(III) exposures of 6 or 24 h, but not 3 h, suggesting that phase 2 protein synthesis is involved. Previous studies indicate that Au(III) may induce Nrf2, but concentrations used in these studies were much higher than those used in the current study (Lewis et al., 2005). Finally, the ability of Au(III) to induce ROS may be related to its high reactivity in biological systems that has limited its use as a therapeutic agent. In conclusion, the current results do not support ROS as mediators of mitochondrial suppression of Au-based compounds in monocytes. Furthermore, ROS generation does not appear to be related to the ability of Au(I) compounds to alter monocytic cytokine secretion, a mechanism of suspected import for Au(I)-anti-rheumatic drugs. This possibility is unresolved for Au(III). Mechanisms of all Aubased drugs may involve more complex cellular redox responses, but they must be independent of ROS. Based on the unique response of monocytes to AF, GSTM and Au(III) in the current study, these oxidative cellular responses are likely to be compound-speciWc. ConXict of Interest Statement No outside Wnancial arrangements or interests exist that would constitute a conXict of interest. Acknowledgement The authors thank the Medical College of Georgia Biocompatibility program for their support of this work. References Auger, M.J., Ross, J.A., 1992. The biology of the macrophage. In: Lewis, C.E., McGee, J.O. (Eds.), The Macrophage. Oxford University Press, Oxford, pp. 3–74. Barnes, B.J., Karin, M., 1997. Mechanisms of disease: nuclear factor(kappa) B-a pivotal transcription factor in chronic inXammatory diseases. New Engl. J. Med. 336 (15), 1066–1071. Barrerra, P., Boerbooms, A.M., van de Putte, L.B.A., van der Meer, J.W.M., 1996. EVects of antirheumatic agents on cytokines. Semin. Arthritis Rheu. 25, 234–253. Bendek, T.G., 2004. History of gold therapy for tuberculosis. J. His. Med. All. Sci. 59, 50–89. Bondeson, J., 1997. The mechanisms of action of disease-modifying antirheumatic drugs: a review with emphasis on macrophage signal transduction and the induction of proinXammatory cytokines. Gen. Pharmacol. 29, 127–150. Byun, M.-S., Jeon, K.-I., Choi, J.-W., Shim, J.-Y., Jue, D.-M., 2002. Dual eVect of oxidative stress on NF-B activation in HeLa cells. Exp. Mol. Med. 34, 332–339. Che, C.M., Sun, R.W., Yu, W.Y., Ko, C.B., Shu, N., Sun, H., 2003. Gold(III) porphyrins as a new class of anticancer drugs: cytotoxicity
DNA binding and induction of apoptosis in human cervix epitheloid cancer cells. Chem. Commun. (Camb.) 21 (Jul), 1718–1719. Drakes, M.L., Harth, M., Galsworthy, S.A., McCain, G.A., 1987. EVects of gold on the production and response to human interleukin-1. J. Rheumatol. 14, 1123–1127. Eisler, R., 2003. Chrysotherapy: a synoptic review. InXamm. Res. 52, 487– 501. Ercal, N., Gurer-Orhnan, H., Aykin-Burns, N., 2001. Toxic metals and oxidative stress part 1: mechanisms involved in metal-induced oxidative damage. Curr. Top. Med. Chem. 1, 529–539. Fricker, S.P., 1996. Medical uses of gold compounds: past, present, and future. Gold Bull. 29 (2), 53–60. Handel, M.L., 1997. Transcription factors AP-1 and NF-B: where steroids meet the gold standard of anti-rheumatic drugs. InXamm. Res. 46, 282–286. Hempel, S.L., Buettner, G.R., O’Malley, Y.Q., Wessels, D.A., Flaherty, D.M., 1999. DihydroXuorescein diacetate is superior for detecting intracellular oxidants: comparison with 2⬘7⬘-dichlorodihydroXuorescein diactetate, 5 (and 6)-carboxyl-2⬘7⬘-dichlorodihydroXuroescein diacetate, and dihydrorhodamine 123. Free Radical Biol. Med. 27, 146– 159. Kasprzak, K.S., 1995. Possible role for oxidative damage in metal-induced carcinogenesis. Cancer Invest. 13, 411–430. Katoaka, K., Handa, H., Nishizawa, M., 2001. Induction of cellular antioxidative stress genes through heterodimeric transcription factor Nrf2/ small Maf by antirheumatic gold(I) compounds. J. Biol. Chem. 276, 34074–34081. Kean, W.F., Hart, L., Buchanan, W.W., 1997. AuranoWn. Br. J. Rheumatol. 36, 560–572. Kudrin, A.V., 2000. Trace elements in regulation of NFB activity. J. Trace Elem. Med. Biol. 14, 129–142. Lampa, J., Kareskog, L., Rönnelid, J., 2002. EVects of gold on cytokine production in vitro; increase of monocyte dependent interleukin 10 production and decrease of interferon- levels. J. Rheumatol. 29, 21–28. Lewis, J.B., Wataha, J.C., McCloud, V., Lockwood, P.E., Messer, R.L.W., Tseng, W.-Y., 2005. Au(III), Pd(II), Ni(II), and Hg(II) alter NFB signaling in THP1 monocytic cells. J. Biomed. Mater. Res. 74A, 471–481. Li, X., Stark, G.R., 2002. NfB-dependent signaling pathways. Exp. Hematol. 30, 285–296. Lipsky, P.E., 1994. Rheumatoid arthritis. In: Harrison’s Principles of Internal Medicine, 13th ed. McGraw-Hill, New York, pp. 1648–1655. Lloyd, D.R., Phillips, D.H., 1999. Oxidative DNA damage mediated by copper(II), iron(II), and nickel(II). Fenton reactions: evidence of sitespeciWc mechanisms in the formation of double-strand breaks, 8hydroxydeoxyguanosine and putative intrastrand cross-links. Mutat. Res. 424, 23–36. Lockwood, D.B., Wataha, J.C., Lewis, J.B., Tseng, W.Y., Messer, R.L.W., Hsu, S.D., 2005. Blue light generates reactive oxygen species (ROS) diVerentially in tumor vs. normal epithelial cells. Dent. Mater. 21, 683– 688. Messori, L., Abbate, F., Marcon, G., Orioli, P., Fontani, M., Mini, E., Mazzei, T., Carottie, S., O’Connell, T., Zanello, P., 2000. Gold(III) complexes as potential antitumor agents: solution chemistry and cytotoxic properties of some selected gold(III) compounds. J. Med. Chem. 43, 3541–3548. Novelli, F., Recine, M., Sparatore, F., Juliano, C., 1999. Gold(I) complexes as antimicrobial agents. II. Farmaco 54, 232–236. Parish, R.V., 1992. Gold in medicine—chrysotherapy. Interdiscipl. Sci. Rev. 17, 221–228. Remvig, L., Erik, C., Bligaard, N., 1988. EVect of auranoWn and sodium aurothiomalate on interleukin-1 production from human monocytes in vitro. Scand. J. Rheumatol. 17, 255–262. Schmidbaur, H., 1992. Gold chemistry is diVerent. Interdiscipl. Sci. Rev. 17, 213–220. Schmidt, J.A., Abdulla, E., 1988. Down-regulation of IL1 biosynthesis by inducers of the heat-shock response. J. Immunol. 141, 2027–2034. Seitz, M., Valbracht, J., Quach, J., Lotz, M., 2003. Gold sodium thiomalate and chloroquine inhibit cytokine production in monocytic THP-1 cells
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633 through distinct transcriptional and post-translational mechanisms. J. Clin. Immunol. 23, 477–484. Simon, L.S., 2000. DMARDS in the treatment of rheumatoid arthritis: current agents and future developments. Int. J. Clin. Pract. 54, 243–249. Stern, I., Wataha, J.C., Lewis, J.B., Messer, R.L.W., Lockwood, P.E., Tseng, W.Y., 2005. Anti-rheumatic gold compounds as sublethal modulators of monocytic LPS-induced cytokine secretion. Toxicol. In Vitro 19, 365–371. Suliman, H.B., Welty-Wolf, K.E., Carraway, M.S., Tatro, L., Piantadosi, C.A., 2004. Lipopolysaccharide induces oxidative cardiac mitochondrial damage and biogenesis. Cardiovasc. Res. 64, 279–288. Tiekink, E.R.T., 2002. Gold derivatives for the treatment of cancer. Crit. Rev. Oncol. Hematol. 42, 225–248.
633
Wataha, J.C., Hanks, C.T., Craig, R.G., 1991. The in vitro eVects of metal cations on eukaryotic cell metabolism. J. Biomed. Mater. Res. 25, 1133–1149. Wataha, J.C., Hanks, C.T., Sun, Z.L., 1995. In vitro reaction of macrophages to metal ions from dental biomaterials. Dent. Mater. 11, 239– 245. Wong, P.K.K., Campbell, I.K., Egan, P.J., Ernst, M., Wicks, I.P., 2003. The role of the interleukin-6 family of cytokines in inXammatory arthritis and bone turnover. Arthritis Rheum. 48, 1177–1189. Yoshida, S.-I., Kato, T., Sakurada, S., Kurono, C., Yang, J.-P., Matsui, N., Soji, T., Okamoto, T., 1999. Inhibition of IL-6 and IL-8 induction from cultured rheumatoid synovial Wbroblasts by treatment with aurothioglucose. Int. Immunol. 11, 151–158.
Gold-induced reactive oxygen species (ROS) do not mediate suppression of monocytic mitochondrial or secretory function Yo Omata a, Jill B. Lewis b, Petra E. Lockwood b, Wan Y. Tseng c, Regina L. Messer b, Serge Bouillaguet d, John C. Wataha b,¤ b
a Hokkaido University, Sapporo, Japan Medical College of Georgia, Oral Biology and Maxillofacial Pathology, Augusta, GA 30912-1160, United States c National Taiwan University, Taipei, Taiwan, ROC d School of Medicine, University of Geneva, Geneva, Switzerland
Received 24 August 2005; accepted 14 November 2005 Available online 27 December 2005
Abstract The toxicity of anti-rheumatic gold compounds has limited their use and development, yet both the toxicological and therapeutic actions of these compounds remain unclear. In the current study, we tested the hypothesis that intracellular reactive oxygen species (ROS) induced by Au(I) or Au(III) compounds mediate their ability to suppress mitochondrial activity. Methods: Human THP1 monocytes were exposed to HAuCl4 · 3H2O (Au(III)), or the anti-rheumatic compounds auranoWn (AF) or gold sodium thiomalate (GSTM) for 6–72 h, after which mitochondrial activity (succinate dehydrogenase) was measured. To assess the role of cellular redox status as a mediator of mitochondrial suppression, monocytes were pre-treated with a pro-oxidant (t-butyl hydroquinone, t-BHQ) or antioxidant (N-acetyl cysteine, NAC ). ROS levels were measured 0–24 h post-gold addition to determine their role as mediators of mitochondrial activity suppression. Results: AF was the most potent inhibitor of mitochondrial activity, followed by Au(III) and GSTM. Only Au(III) induced intracellular ROS; no ROS formation was observed in response to AF or GSTM exposure. Although anti- and pro-oxidants had some eVects on mitochondrial suppression of Au compounds, collectively the data do not support redox eVects or ROS formation as major mediators of Au-compound mitochondrial suppression. Conclusions: Our results do not indicate that ROS and redox eVects play major roles in mediating the cytotoxicity of AF, GSTM or Au(III). © 2005 Elsevier Ltd. All rights reserved. Keywords: Rheumatoid arthritis; Metals; Redox; Mitochondrial activity
1. Introduction Gold compounds have been used therapeutically for over a century to treat a variety of diseases, but most commonly to treat rheumatoid and several other arthrides (Parish, 1992; Fricker, 1996; Simon, 2000). The most common gold-based anti-arthritic drugs today are gold sodium thiomalate (GSTM, injected intramuscularly) and auranoWn
*
Corresponding author. Tel.: +1 706 721 2991; fax: +1 706 721 3992. E-mail address: [email protected] (J.C. Wataha).
0887-2333/$ - see front matter © 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.tiv.2005.11.001
(AF, oral), both of which contain Au(I) bound to organic ligands (Fig. 1, Schmidbaur, 1992; Simon, 2000). The therapeutic value of these compounds in limiting the progression of rheumatoid arthritis is unquestioned, but they have serious toxic side eVects including diarrhea, skin rashes, blood dyscrasias, and proteinurea (Kean et al., 1997; Eisler, 2003). These side eVects often are life-threatening and force patients to stop therapy despite a signiWcant reduction of pain and joint destruction (Simon, 2000). In contrast, Au(III) compounds (Fig. 1, Messori et al., 2000) are not used clinically because of the reactivity and cytotoxicity of the Au(III) cation, but numerous new Au(III) compounds
626
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
Fig. 1. Gold compounds tested in the current study. AuranoWn (AF, upper left) is a Au(I) compound with thiol- and phosphine-linked ligands. Gold sodium thiomalate (GSTM, below) is a Au(I) compound with a single thiol-linked ligand. Both AF and GSTM have been successfully used to treat arthrides, but have signiWcant toxicity clinically. HAuCl4 · 3H2O (Au(III), upper right) is an inorganic Au(III) compound, not currently used medicinally because of its toxicity. However, its use has been proposed for treatment of cancer.
have been proposed for the treatment of arthritis, cancer, or other diseases (Novelli et al., 1999; Messori et al., 2000; Tiekink, 2002; Che et al., 2003; Bendek, 2004). In spite of their long history of medicinal use, the mechanisms by which gold compounds are therapeutic or toxic are not known. This ignorance has severely limited the full clinical utility of Au(I) compounds, the development of new Au(I) or Au(III) compounds as drugs, and strategies to limit side eVects (Simon, 2000; Wong et al., 2003). Historically, GSTM and AF have been considered immunosuppressive agents, thought to limit the activation of monocytes and other blood cells. These compounds have been reported to limit secretion of pro-inXammatory mediators such as IL-1, IL6, or TNF (Drakes et al., 1987; Remvig et al., 1988; Schmidt and Abdulla, 1988; Barrerra et al., 1996; Bondeson, 1997; Yoshida et al., 1999). By ‘suppressing’ activation of inXammatory cells, these compounds purportedly limit synovial hyperplasia, B- and T-cell activation, and erosion of cartilage and bone (Lipsky, 1994). However, recent evidence indicates that gold compounds are not simply suppressors of immune function, but complex modulators of cytokine secretion (Barrerra et al., 1996; Bondeson, 1997; Lampa et al., 2002; Seitz et al., 2003). Auinduced modulation may result in up- or down-regulation of cytokine secretion, particularly when the gold compounds are present in conjunction with other cellular activators such as TNF or lipopolysaccharide (LPS). Similar results have been reported for Au(III) chloride (Stern et al.,
2005). The mechanisms by which gold-based compounds induce these complex, cytokine-speciWc responses also are under-explored and unclear. Yet understanding these mechanisms is essential to optimizing the use of existing compounds and developing new ones. Several lines of evidence suggest that Au-based compounds act, at least in part, by modifying cellular redox balance. In general, metal ions are thought to elevate levels of intracellular reactive oxygen species (ROS) (Kasprzak, 1995; Lloyd and Phillips, 1999; Kudrin, 2000). Elevated ROS levels are suspected triggers of several transcription factors such as AP1 and NFB (Byun et al., 2002; Li and Stark, 2002) that mediate the secretion of many inXammatory cytokines and factors (Barnes and Karin, 1997). Other evidence suggests that gold compounds (e.g., AF) may activate Nrf2, a transcription factor that recruits an antioxidative cellular response (Katoaka et al., 2001). Furthermore, Au-induced oxidative stress may cause a buildup of oxidized thioredoxin, a key intracellular redox protein that mediates transcription factor activation, nuclear translocation, and DNA binding (Handel, 1997). The ability of Au-based compounds and Au ions to elevate or modulate intracellular ROS formation at sublethal levels has not been reported in spite of the potential importance of this mechanism to the therapeutic and toxicological proWle of these compounds. In the current study, we test the hypothesis that Au compounds generate reactive
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
oxygen species (ROS) that mediate cellular eVects. We have focused here on the ability of Au-based compound to suppress mitochondrial activity.
627
pounds were added a second time before ROS levels were measured. 2.3. Cell activators, antioxidants, and pro-oxidants
2. Material and methods 2.1. Au compounds Three gold compounds were evaluated (Fig. 1) based on their use or proposed use as therapeutic agents. Gold sodium thiomalate (GSTM, Sigma-Aldrich) and auranoWn (AF, Alexis Corporation) have been used for over 70 years to treat rheumatoid and other arthrides (Fricker, 1996; Simon, 2000). Au(III) is the metallic center of a wide variety of compounds proposed as anticancer agents (Messori et al., 2000; Che et al., 2003). For each of these compounds, the ability to alter cellular redox balance has been proposed as a possible mechanism of action (Ercal et al., 2001). AF is a Au(I) compound with thiol- and phosphine-linked organic ligands in a linear conWguration. GSTM also is a Au(I) compound, but with a single thiol-linked organic ligand. HAuCl4 · 3H2O (Au(III), Sigma-Aldrich) is an inorganic compound with square planar coordination chemistry. Compounds were added to cell-cultures from sterile stock solutions in water (GSTM, Au(III)) or ethanol (AF, 0.5% Wnal concentration).
Cell activators, antioxidants, or pro-oxidants were added to cultures to test the hypothesis that gold compounds alter mitochondrial function via redox-related mechanisms. Lipopolysaccharide (LPS, Ecoli, serotype
2.2. Cell-culture and experimental strategy The monocyte is one key target of anti-arthritic compounds (Bondeson, 1997) and a major orchestrator of inXammatory responses (Auger and Ross, 1992). Thus, human monocytic cells (THP1, ATCC TIB 202) were used to measure Au-induced ROS levels or changes in cellular mitochondrial function or ROS levels. THP1 were maintained in RPMI 1640 supplemented with 10% FBS, 2 mM glutamine, 100 g/mL streptomycin, 100 U/mL penicillin (all from Gibco BRL), and 50 M -mercaptoethanol (Sigma-Aldrich). The -mercaptoethanol, a reducing agent, was withheld 3-d prior to the start of experiments to avoid masking any oxidative eVects of the gold compounds or enhancing the eVects of added antioxidants. For cytotoxicity experiments, cells were plated at 50,000 cells/well in 200 L in 96-well round-bottom format (n D 8/condition). Activators, antioxidants, pro-oxidants, or the metal compounds were added as described below to provide a 24, 48, or 72 h total exposure to the gold compounds. For experiments that measured mitochondrial suppression, succinate dehydrogenase (SDH) activity of the cells was measured using the MTT method (see below). For measurement of reactive oxygen species (ROS), cells were plated at 10,000 cells/well in 200 L in 96-well Xat-bottom format (n D 3/condition). Antioxidants were added in selected experiments as described below. Gold compounds were then added and incubated with the cells for 0–24 h, after which intracellular ROS levels were measured using the DFDA assay (see below). In selected cases, the Au com-
Fig. 2. Mitochondrial suppression by auranoWn (AF), gold sodium thiomalate (GSTM), and gold chloride (Au(III)) in THP1 monocytes, estimated by measuring succinate dehydrogenase (SDH) activity. Gold compounds were exposed to monocytes for 24, 48, or 72 h. Results were expressed as a percentage of controls exposed only to vehicle (water for GSTM and Au(III), 0.5% ethanol for AF). Error bars represent one standard deviation of the mean (n D 8/condition). Asterisks indicate a statistical diVerence from vehicle controls for each concentration (t-tests, 2-sided, D 0.05).
628
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
Fig. 3. Mitochondrial suppression 72 h after exposure to auranoWn (AF), gold sodium thiomalate (GSTM), or gold chloride (Au(III)) in monocytes treated with activators (LPS, 1 g/mL; TNF 10 ng/mL; t-BHQ, 100 mM, last 24 h) or pretreated (2 h) with antioxidants (NAC, 5 mM). The control graph is identical to Fig. 2 (bottom) and is shown here for comparison. Mitochondrial suppression was estimated by succinate dehydrogenase (SDH) activity. Results were expressed as a function of controls exposed only to NAC, LPS, t-BHQ, or TNF. Error bars represent one standard deviation of the mean (n D 8/ condition). Asterisks indicate a statistical diVerence from control for each concentration (t-tests, 2-sided, D 0.05).
0111:B4, 1 g/mL for last 24 h, Sigma) and tumor necrosis factor-alpha (TNF, recombinant human, 10 ng/mL for last 24 h, R&D Systems) were two activators relevant to monocytes in arthrides or infection (Simon, 2000). LPS also has been described as a pro-oxidant (Suliman et al., 2004). Selenium (Se, Na2SeO3, 50 nM, Sigma), N-acetyl cysteine (NAC, 5 mM, Sigma), and vitamin E (Vit E, -tocopherol, 10 M, Sigma) were antioxidants added 2 h before metal addition to attempt to mitigate any redox-mediated, Auinduced eVects. t-BHQ (tert-butyl hydroquinone, 100 M, added last 24 h, Sigma) was used as a pro-oxidant. Pilot studies ensured that none of the activators, antioxidants, or pro-oxidants aVected monocytic mitochondrial function or induced ROS levels by themselves at the above concentrations and times of addition.
2.4. Cell succinate dehydrogenase (SDH) activity The activity of succinate dehydrogenase (SDH), an enzyme unique to the mitochondria and central to mitochondrial function, was measured (n D 8/condition) using MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide, Sigma) reduction as described previously (Wataha et al., 1995). This assay was structured to estimate cellular mitochondrial activity (Wataha et al., 1991). Control wells received only vehicle (water for Au(III) and GSTM, 0.5% ethanol for AF), and were used to normalize SDH activity of Au-treated wells. SDH activity experiments were used to select sublethal concentrations of gold compounds for use in the assessment of intracellular ROS.
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
629
2.5. Measurement of reactive oxygen species (ROS) The dihydroXuorescein diacteate (DFDA) assay was used to estimate intracellular ROS levels induced by gold compounds. DFDA is taken up by cells, then activated by esterase-mediated cleavage of acetate, which traps dihydroXuorescein (DF) in the cells. DF may be oxidized by various ROS to Xuorescein, which has measurable Xuorescence 530 nm (excitation @ 485 nm)(Hempel et al., 1999). The assay therefore estimated only intracellular ROS levels. After incubation with gold compounds, cells were washed with 200 L/well of Hallam’s buVer, then 100 L of Hallam’s was re-added to maintain cell viability. During all measurements, cells were kept in the dark as much as possible. Diamide (5 mM, Sigma) was then added to control wells (t D 0) as a positive control (Lockwood et al., 2005); negative controls received no gold compounds. DFDA (20 M, Molecular Probes) was then added 0–60 min after time zero, and the cultures were incubated for a total of 90 min, after which Xuorescence was measured. In this manner, the lifetime of the ROS could be estimated (Lockwood et al., 2005) because the DFDA only reacted with ROS that remained from the time of DFDA addition to 90 min. 3. Results 3.1. Mitochondrial suppression AuranoWn (AF) was the most potent suppressor of mitochondrial function, inhibiting succinate dehydrogenase (SDH) activity completely above 2 M, regardless of exposure time (Fig. 2, 24–72 h). The suppression by HAuCl4 · 3H2O (Au(III)) was less than that of AF, but was increased as the exposure time increased from 24 h to 72 h. Au(III) concentrations that inhibited SDH activity were generally above 100 M. Gold sodium thiomalate (GSTM) was the least potent of the three compounds, inhibiting SDH activity signiWcantly only after a 72 h exposure to 100–200 M. For every duration of exposure, the rank of the cytotoxicities of these compounds was AF > Au(III) > GSTM (Fig. 2). These data were used to select concentrations for measuring goldinduced production of reactive oxygen species. For Au(III) and GSTM, concentrations below 100 M were used; for AF, concentrations below 1 M were used. Mitochondrial suppression by AF was not aVected by the pre-addition of the antioxidant N-acetyl-cysteine (NAC) (Fig. 3). However, GSTM or Au(III) suppression was somewhat reduced with NAC. Lipopolysaccharide (LPS) did not aVect SDH activity levels in response to any of the gold compounds. However, addition of t-BHQ markedly increased the potency of all three compounds, and particularly that of GSTM. The addition of TNF did not aVect AF or GSTM-modulated SDH activity, but tended to increase SDH activity (10–20%) when Au(III) was added concurrently. These increases occurred at Au(III) concentrations of 90–110 M, but were not statistically signiWcant.
Fig. 4. Levels of intracellular reactive oxygen species (ROS) generated in monocytes 0, 3, and 24 h after exposure to HAuCl4 · 3H2O. After exposure to Au(III), cells were labeled with dihydroXuorescein diacetate (DFDA, 20 M) and ROS levels were determined for up to 90 min, measuring Xuorescence emission at 530 nm (excitation at 485 nm). Levels of ROS were expressed as a percentage of ROS induced by the pro-oxidant diamide (5 mM, 30 min, control level shown with dashed line). Diamide controls induced a change of approximately 7500 Xuorescence units over background levels. Error bars indicate one standard deviation (n D 3). Asterisks indicate statistical diVerences from background levels for each time after gold compound addition (t-tests, 2-sided, D 0.05).
3.2. Production of reactive oxygen species (ROS) Au(III) induced ROS formation to levels Wve times those of the diamide positive control (Fig. 4). ROS induction was concentration dependent, and induced ROS were detected after exposure to the lowest concentration tested (5 M). Induced ROS decreased with time and returned to baseline levels by 90 min (Fig. 4, top); after 3 or 24 h, no evidence of
630
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
Fig. 5. Levels of intracellular reactive oxygen species (ROS) generated in monocytes when HAuCl4 · 3H2O was added a second time 3, 6, or 24 h after the Wrst addition. After the second exposure to Au(III), cells were labeled with dihydroXuorescein diacetate (DFDA, 20 M) and ROS levels were determined for up to 90 min, measuring Xuorescence emission at 530 nm (excitation at 485 nm). Levels of ROS were expressed as a percentage of ROS induced by the pro-oxidant diamide (5 mM, control level shown with dashed line). Diamide controls induced a change of approximately 7500 Xuorescence units over background levels. Error bars indicate one standard deviation (n D 3). Asterisks indicate statistical diVerences from background levels for each time after gold compound addition (ttests, 2-sided, D 0.05).
secondary or persistent ROS formation was evident. When Au(III) was added a second time, ROS formation depended on the time of secondary addition (Fig. 5). At 3 h, the response was identical to that induced by the Wrst addition (Figs. 4 and 5). However, if the second addition occurred at 6 h, the ROS response was mitigated signiWcantly. When Au(III) was added a second time at 24 h, the mitigation seen at 6 h persisted.
Fig. 6. EVect of the antioxidants N-acetyl cysteine (NAC, 5 mM, top), selenium (Se, from Se2O3, 50 nM, middle) or vitamin E (-tocopherol, 10 mM, bottom), on levels of reactive oxygen species generated in monocytes after addition of HAuCl4 · 3H2O (Au(III)). Controls contained only the antioxidants (0 M). ROS were assessed as in Figs. 4,5. Error bars indicate one standard deviation (n D 3). Asterisks indicate statistical diVerences from background levels for each time after gold compound addition (t-tests, 2-sided, D 0.05).
Compounds commonly employed as cellular antioxidants all signiWcantly reduced Au(III)-induced ROS formation (Fig. 6). Treatment with NAC, Se, or Vit E all reduced the peak ROS levels induced by 75 M Au(III) by 4–5 times. There were no apparent diVerences in the ability of these three compounds to mitigate Au(III)-induced ROS formation. Unlike Au(III), AF and GSTM did not induce detectable ROS formation. Compound-induced ROS were not observed immediately after addition (Fig. 7), nor after 3 or 24 h (data not shown). Furthermore, a second addition of the compounds at 24 h did not induce ROS formation (Fig. 7).
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
631
Fig. 7. Levels of intracellular reactive oxygen species (ROS) generated in monocytes immediately after addition of AF (left) or GSTM (right), or after second addition at 24 h. ROS were measured as described in Figs. 4 and 5. Error bars indicate one standard deviation (n D 3). There were no statistically signiWcant levels of ROS detected.
4. Discussion Collectively, our data do not support our hypothesis that ROS and redox eVects play major roles in mediating the cytotoxicity of AF, GSTM or Au(III). Furthermore, each of these compounds appears to act via distinct mechanisms despite our anticipation that they would behave similarly. AuranoWn (AF) was the most potent mitochondrial suppressor (Fig. 2), and mitochondrial suppression was uneVected by pro- or antioxidants (Fig. 3). Furthermore, AF induced no formation of ROS either initially or after addition a second time (Fig. 7). Although the relatively potent cytotoxicity of AF was expected (Simon, 2000), we found no evidence to support redox mechanisms or ROS as mediators of AF toxicity. It is possible that the method to determine ROS formation was not sensitive enough to detect AF-induced ROS, but this seemed unlikely given the ability of the assay to detect ROS induced by Au(III) (see discussion below). Gold sodium thiomalate (GSTM) also induced no intracellular ROS (Fig. 7). The addition of NAC slightly reduced, and the pro-oxidant t-BHQ increased GSTMinduced mitochondrial suppression (Fig. 3), suggesting at least some role for redox-mediated eVects. However, the absence of ROS generation eliminated a deWnitive role of this mechanism. Because both NAC and t-BHQ may act by altering (increasing or decreasing, respectively) reserves of thiol-based reducing power in the cell, it is possible that
NAC and t-BHQ either bolstered or limited thiol reserves, thereby limiting or facilitating GSTM eVects. Although speculative, this mechanism could be assessed by measuring glutathione redox balance, which is the most common source of cellular thiol-mediated reducing power. Mitochondrial suppression induced by Au(III) was largely uneVected by NAC, but was enhanced by t-BHQ (Fig. 3). The ability of Au(III) to induce a potent but transient increase in ROS levels initially suggested that Au(III) caused cellular toxicity, at least in part, via ROS formation. However, the inability of NAC to limit cytotoxicity, coupled with the continued development of the Au(III)-induced cytotoxicity response long-after the ROS increases had returned to baseline levels, does not support a deWnitive role for ROS as mediators of Au(III) eVects. Our data do not support ROS generation as a mechanism by which Au(I) compounds such as AF or GSTM alter cytokine secretion (Barrerra et al., 1996; Bondeson, 1997; Lampa et al., 2002; Seitz et al., 2003; Stern et al., 2005). Our results dispel a long-espoused mechanism of action of these compounds. The previously reported ability of AF to activate Nrf2, a transcription factor responsive to pro-oxidants (Katoaka et al., 2001), suggests that Nrf2 must be activated in a ROS-independent manner, perhaps by binding the Keap-1 Nrf2 inhibitor directly. The current results leave open the possibility that Au(III) compounds alter monocytic cytokine secretion via
632
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633
ROS-mediated mechanisms. Au(III) generated a potent but transient increase in intracellular ROS (Fig. 4). The ability of NAC to reduce Au(III)-induced ROS suggests that Au(III) acts at some point by changing cellular thiol redox balance. ROS levels were transient and did not induce secondary ROS (Fig. 4), but did alter cell status such that ROS induced by second exposures were reduced (Fig. 5). We observed this mitigation after Au(III) exposures of 6 or 24 h, but not 3 h, suggesting that phase 2 protein synthesis is involved. Previous studies indicate that Au(III) may induce Nrf2, but concentrations used in these studies were much higher than those used in the current study (Lewis et al., 2005). Finally, the ability of Au(III) to induce ROS may be related to its high reactivity in biological systems that has limited its use as a therapeutic agent. In conclusion, the current results do not support ROS as mediators of mitochondrial suppression of Au-based compounds in monocytes. Furthermore, ROS generation does not appear to be related to the ability of Au(I) compounds to alter monocytic cytokine secretion, a mechanism of suspected import for Au(I)-anti-rheumatic drugs. This possibility is unresolved for Au(III). Mechanisms of all Aubased drugs may involve more complex cellular redox responses, but they must be independent of ROS. Based on the unique response of monocytes to AF, GSTM and Au(III) in the current study, these oxidative cellular responses are likely to be compound-speciWc. ConXict of Interest Statement No outside Wnancial arrangements or interests exist that would constitute a conXict of interest. Acknowledgement The authors thank the Medical College of Georgia Biocompatibility program for their support of this work. References Auger, M.J., Ross, J.A., 1992. The biology of the macrophage. In: Lewis, C.E., McGee, J.O. (Eds.), The Macrophage. Oxford University Press, Oxford, pp. 3–74. Barnes, B.J., Karin, M., 1997. Mechanisms of disease: nuclear factor(kappa) B-a pivotal transcription factor in chronic inXammatory diseases. New Engl. J. Med. 336 (15), 1066–1071. Barrerra, P., Boerbooms, A.M., van de Putte, L.B.A., van der Meer, J.W.M., 1996. EVects of antirheumatic agents on cytokines. Semin. Arthritis Rheu. 25, 234–253. Bendek, T.G., 2004. History of gold therapy for tuberculosis. J. His. Med. All. Sci. 59, 50–89. Bondeson, J., 1997. The mechanisms of action of disease-modifying antirheumatic drugs: a review with emphasis on macrophage signal transduction and the induction of proinXammatory cytokines. Gen. Pharmacol. 29, 127–150. Byun, M.-S., Jeon, K.-I., Choi, J.-W., Shim, J.-Y., Jue, D.-M., 2002. Dual eVect of oxidative stress on NF-B activation in HeLa cells. Exp. Mol. Med. 34, 332–339. Che, C.M., Sun, R.W., Yu, W.Y., Ko, C.B., Shu, N., Sun, H., 2003. Gold(III) porphyrins as a new class of anticancer drugs: cytotoxicity
DNA binding and induction of apoptosis in human cervix epitheloid cancer cells. Chem. Commun. (Camb.) 21 (Jul), 1718–1719. Drakes, M.L., Harth, M., Galsworthy, S.A., McCain, G.A., 1987. EVects of gold on the production and response to human interleukin-1. J. Rheumatol. 14, 1123–1127. Eisler, R., 2003. Chrysotherapy: a synoptic review. InXamm. Res. 52, 487– 501. Ercal, N., Gurer-Orhnan, H., Aykin-Burns, N., 2001. Toxic metals and oxidative stress part 1: mechanisms involved in metal-induced oxidative damage. Curr. Top. Med. Chem. 1, 529–539. Fricker, S.P., 1996. Medical uses of gold compounds: past, present, and future. Gold Bull. 29 (2), 53–60. Handel, M.L., 1997. Transcription factors AP-1 and NF-B: where steroids meet the gold standard of anti-rheumatic drugs. InXamm. Res. 46, 282–286. Hempel, S.L., Buettner, G.R., O’Malley, Y.Q., Wessels, D.A., Flaherty, D.M., 1999. DihydroXuorescein diacetate is superior for detecting intracellular oxidants: comparison with 2⬘7⬘-dichlorodihydroXuorescein diactetate, 5 (and 6)-carboxyl-2⬘7⬘-dichlorodihydroXuroescein diacetate, and dihydrorhodamine 123. Free Radical Biol. Med. 27, 146– 159. Kasprzak, K.S., 1995. Possible role for oxidative damage in metal-induced carcinogenesis. Cancer Invest. 13, 411–430. Katoaka, K., Handa, H., Nishizawa, M., 2001. Induction of cellular antioxidative stress genes through heterodimeric transcription factor Nrf2/ small Maf by antirheumatic gold(I) compounds. J. Biol. Chem. 276, 34074–34081. Kean, W.F., Hart, L., Buchanan, W.W., 1997. AuranoWn. Br. J. Rheumatol. 36, 560–572. Kudrin, A.V., 2000. Trace elements in regulation of NFB activity. J. Trace Elem. Med. Biol. 14, 129–142. Lampa, J., Kareskog, L., Rönnelid, J., 2002. EVects of gold on cytokine production in vitro; increase of monocyte dependent interleukin 10 production and decrease of interferon- levels. J. Rheumatol. 29, 21–28. Lewis, J.B., Wataha, J.C., McCloud, V., Lockwood, P.E., Messer, R.L.W., Tseng, W.-Y., 2005. Au(III), Pd(II), Ni(II), and Hg(II) alter NFB signaling in THP1 monocytic cells. J. Biomed. Mater. Res. 74A, 471–481. Li, X., Stark, G.R., 2002. NfB-dependent signaling pathways. Exp. Hematol. 30, 285–296. Lipsky, P.E., 1994. Rheumatoid arthritis. In: Harrison’s Principles of Internal Medicine, 13th ed. McGraw-Hill, New York, pp. 1648–1655. Lloyd, D.R., Phillips, D.H., 1999. Oxidative DNA damage mediated by copper(II), iron(II), and nickel(II). Fenton reactions: evidence of sitespeciWc mechanisms in the formation of double-strand breaks, 8hydroxydeoxyguanosine and putative intrastrand cross-links. Mutat. Res. 424, 23–36. Lockwood, D.B., Wataha, J.C., Lewis, J.B., Tseng, W.Y., Messer, R.L.W., Hsu, S.D., 2005. Blue light generates reactive oxygen species (ROS) diVerentially in tumor vs. normal epithelial cells. Dent. Mater. 21, 683– 688. Messori, L., Abbate, F., Marcon, G., Orioli, P., Fontani, M., Mini, E., Mazzei, T., Carottie, S., O’Connell, T., Zanello, P., 2000. Gold(III) complexes as potential antitumor agents: solution chemistry and cytotoxic properties of some selected gold(III) compounds. J. Med. Chem. 43, 3541–3548. Novelli, F., Recine, M., Sparatore, F., Juliano, C., 1999. Gold(I) complexes as antimicrobial agents. II. Farmaco 54, 232–236. Parish, R.V., 1992. Gold in medicine—chrysotherapy. Interdiscipl. Sci. Rev. 17, 221–228. Remvig, L., Erik, C., Bligaard, N., 1988. EVect of auranoWn and sodium aurothiomalate on interleukin-1 production from human monocytes in vitro. Scand. J. Rheumatol. 17, 255–262. Schmidbaur, H., 1992. Gold chemistry is diVerent. Interdiscipl. Sci. Rev. 17, 213–220. Schmidt, J.A., Abdulla, E., 1988. Down-regulation of IL1 biosynthesis by inducers of the heat-shock response. J. Immunol. 141, 2027–2034. Seitz, M., Valbracht, J., Quach, J., Lotz, M., 2003. Gold sodium thiomalate and chloroquine inhibit cytokine production in monocytic THP-1 cells
Y. Omata et al. / Toxicology in Vitro 20 (2006) 625–633 through distinct transcriptional and post-translational mechanisms. J. Clin. Immunol. 23, 477–484. Simon, L.S., 2000. DMARDS in the treatment of rheumatoid arthritis: current agents and future developments. Int. J. Clin. Pract. 54, 243–249. Stern, I., Wataha, J.C., Lewis, J.B., Messer, R.L.W., Lockwood, P.E., Tseng, W.Y., 2005. Anti-rheumatic gold compounds as sublethal modulators of monocytic LPS-induced cytokine secretion. Toxicol. In Vitro 19, 365–371. Suliman, H.B., Welty-Wolf, K.E., Carraway, M.S., Tatro, L., Piantadosi, C.A., 2004. Lipopolysaccharide induces oxidative cardiac mitochondrial damage and biogenesis. Cardiovasc. Res. 64, 279–288. Tiekink, E.R.T., 2002. Gold derivatives for the treatment of cancer. Crit. Rev. Oncol. Hematol. 42, 225–248.
633
Wataha, J.C., Hanks, C.T., Craig, R.G., 1991. The in vitro eVects of metal cations on eukaryotic cell metabolism. J. Biomed. Mater. Res. 25, 1133–1149. Wataha, J.C., Hanks, C.T., Sun, Z.L., 1995. In vitro reaction of macrophages to metal ions from dental biomaterials. Dent. Mater. 11, 239– 245. Wong, P.K.K., Campbell, I.K., Egan, P.J., Ernst, M., Wicks, I.P., 2003. The role of the interleukin-6 family of cytokines in inXammatory arthritis and bone turnover. Arthritis Rheum. 48, 1177–1189. Yoshida, S.-I., Kato, T., Sakurada, S., Kurono, C., Yang, J.-P., Matsui, N., Soji, T., Okamoto, T., 1999. Inhibition of IL-6 and IL-8 induction from cultured rheumatoid synovial Wbroblasts by treatment with aurothioglucose. Int. Immunol. 11, 151–158.