924
Chem. Res. Toxicol. 1999, 12, 924-930
Methylarsenicals and Arsinothiols Are Potent Inhibitors of Mouse Liver Thioredoxin Reductase Shan Lin,† William R. Cullen,‡ and David J. Thomas*,§ Curriculum in Toxicology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599, Department of Chemistry, University of British Columbia, Vancouver, British Columbia, Canada, and Pharmacokinetics Branch, Experimental Toxicology Division, National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711 Received May 11, 1999
Thioredoxin reductase (TR, EC 1.6.4.5) was purified 5800-fold from the livers of adult male B6C3F1 mice. The estimated molecular mass of the purified protein was about 57 kDa. The activity of the purified enzyme was monitored by the NADPH-dependent reduction of 5,5′dithiobis(2-nitrobenzoic acid) (DTNB); this activity was fully inhibited by 1 µM aurothioglucose. Arsenicals and arsinothiols, complexes of AsIII-containing compounds with L-cysteine or glutathione, were tested as inhibitors of the DTNB reductase activity of the purified enzyme. Pentavalent arsenicals were much less potent inhibitors than trivalent arsenicals. Among all the arsenicals, CH3AsIII was the most potent inhibitor of TR. CH3AsIII was found to be a competitive inhibitor of the reduction of DTNB (Ki ∼ 100 nM) and a noncompetitive inhibitor of the oxidation of NADPH. The inhibition of TR by CH3AsIII was time-dependent and could not be reversed by the addition of a dithiol-containing molecule, 2,3-dimercaptosuccinic acid, to the reaction mixture. The inhibition of TR by CH3AsIII required the simultaneous presence of NADPH in the reaction mixture. However, unlike other pyridine nucleotide disulfide oxidoreductases, there was no evidence that mouse liver TR was inactivated by exposure to NADPH. Treatment with CH3AsIII did not increase the NADPH oxidase activity of the purified enzyme. Thus, CH3AsIII, a putative intermediate in the pathway for the biomethylation of As, is a potent and irreversible inhibitor of an enzyme involved in the response of the cell to oxidative stress.
Introduction Thioredoxin reductase (TR,1 NADPH:oxidized thioredoxin oxidoreductase, EC 1.6.4.5) is a selenoprotein that catalyzes the reduction of a variety of disulfidecontaining substrates. A major function of TR is to catalyze the NADPH-dependent reduction of thioredoxin (Trx), a 12 kDa protein with redox-active Cys residues. As summarized by Holmgren (1), the linked functions of TR, Trx, and NADPH in the reduction of protein disulfides can be described by the following reaction scheme:
A variety of other substrates are also reduced in reactions catalyzed by TR. These include alloxan, 5,5′-dithiobis(2nitrobenzoic acid) (DTNB), dehydroascorbate, selenite, selenodiglutathione, lipid hydroperoxides, and protein substrates such as protein disulfide isomerase and * To whom correspondence should be addressed: Pharmacokinetics Branch, ETD, NHEERL, MD-74, U.S. EPA, Research Triangle Park, NC 27711-2055. Telephone: (919) 541-4974. Fax: (919) 541-5394. E-mail:
[email protected]. † University of North Carolina at Chapel Hill. ‡ University of British Columbia. § U.S. Environmental Protection Agency.
thioredoxin peroxidase (2-8). The reduction of disulfides by TR is thought to play a critical role in the maintenance of the intracellular redox state, and the subcellular distribution of TR and Trx is consistent with this function (9). Because Trx is a modulator of cell proliferation and DNA synthesis (10) and is overexpressed in some multidrug-resistant tumors (11), the regulation of the activity of TR may be a potential target for chemotherapy of cancer (12, 13). TR is a member of the pyridine nucleotide-disulfide oxidoreductase family of flavoenzymes which includes glutathione (GSH) reductase, trypanothione reductase, mercuric reductase, and lipoamide dehydrogenase (14). Like other NAD(P)H-dependent oxidoreductases, TR functions as a dimeric protein in which each monomer contains a tightly bound FAD molecule and a redoxsensitive pair of Cys residues at the active site (1). Electron transfer from bound NADPH needed to reduce the disulfide at the active site of the oxidized enzyme is 1 Abbreviations: TR, thioredoxin reductase; Trx, thioredoxin; DTNB, 5,5′-dithiobis(2-nitrobenzoic acid); Secys, selenocysteine; CH3AsIIII2, methylarsonous diiodide; (CH3)2AsI, dimethylarsonous iodide; CH3AsIII(Cys)2, dicysteinylmethyldithioarsenite; (CH3)2AsIIIGSH, bis(γ-glutamylcysteinylglycine)dimethylthioarsenite; Na2HAsVO4, arsenic acid, disodium salt; NaAsIIIO2, sodium m-arsenite; CH3AsVO(ONa)2, methylarsonic acid, disodium salt; (CH3)2AsVO(OH), dimethylarsinic acid; (CH3)3AsO, trimethylarsine oxide; AsIII(Cys)3, tricysteinyltrithioarsenite; (CH3)2AsIII(Cys), cysteinyldimethylthioarsenite; DMSA, 2,3dimercaptosuccinic acid; DNCB, 1-chloro-2,4-dinitrobenzene; ASK 1, apoptosis kinase 1; ROS, reactive oxygen species.
10.1021/tx9900775 CCC: $18.00 © 1999 American Chemical Society Published on Web 09/09/1999
Trivalent Arsenicals Inhibit Thioredoxin Reductase
mediated by FAD. Unique among members of this enzyme family, TR is a selenoprotein that contains a selenocysteine (Secys) residue in the pentultimate position at the carboxyl terminus (15, 16). A critical role for Secys in the catalytic function of TR has been demonstrated (17-19). Furthermore, the bioavailability of Se affects TR activity in cultured cells (20, 21) and in animals (22). Arsenicals are a useful class of inhibitors of the NAD(P)H-dependent flavoprotein oxidoreductases. Trivalent arsenicals and arsinothiols [complexes of trivalent arsenicals and thiol-containing molecules such as Cys and GSH (23)] inhibit GSH reductase (GR) (24, 25). The most potent inhibitors of purified yeast GR were methylarsonous diiodide (CH3AsI2), dimethylarsonous iodide [(CH3)2AsI], dicysteinylmethyldithioarsenite [CH3As(Cys)2], and bis(γ-glutamylcysteinylglycine)methylthioarsenite [CH3As(GSH)2] with Ki values of 74, 56, 18, and 9 µM, respectively. The metabolic pathway for As involves alternating steps of reduction of As from a pentavalent state to a trivalent state and of oxidative methylation (26, 27) as well as the formation of complexes of trivalent arsenicals with intracellular thiols such as GSH. The reduction and oxidation of As in this methylation scheme can be summarized as follows:
AsVO43- + 2e f AsIIIO33- + CH3 f CH3AsVO32- + 2e f CH3AsIIIO32- + CH3 f (CH3)2AsVO3- + 2e f (CH3)2AsIIIO- + CH3 f (CH3)3AsVO + 2e f (CH3)3AsIII Pentavalent arsenicals can be reduced to a trivalent state in a reaction that converts GSH to GSSG (28, 29). In bacteria, this reaction is catalyzed by the ArsC protein (30, 31). The resulting inorganic and methylated trivalent arsenicals are complexed with GSH in cells (32, 33). The methylation of arsenicals is enzymatically catalyzed, and in mammals, methyl and dimethyl As are the principal products of methylation that are excreted in urine (34). Because trivalent arsenicals are putative intermediates in this methylation scheme and given the high affinity of trivalent arsenicals for intracellular thiols, it is probable that TR, like GR, would be exposed to these metabolites in the course of the intracellular metabolism of As. In the work described here, inorganic and methylated arsenicals and arsinothiols were examined as potential inhibitors using TR purified from mouse liver. This design tested the effects of both the oxidation state of As and the degree of methylation of As on its activity. All compounds containing trivalent arsenicals, including arsinothiols, proved to be more potent inhibitors than compounds that contained pentavalent As. CH3AsI2 was found to be a potent competitive and irreversible inhibitor of enzyme activity. These results are consistent with the concept that methylation of inorganic As is a means of activating it to more reactive and toxic forms.
Experimental Procedures Caution: Inorganic arsenic is classified as a human carcinogen (35). Toxicities of trivalent methylated arsenicals have not been examined. These arsenicals should be handled as highly toxic and potentially carcinogenic compounds.
Chem. Res. Toxicol., Vol. 12, No. 10, 1999 925 Arsenicals. Arsenic acid, disodium salt (Na2HAsVO4), and sodium m-arsenite (NaAsIIIO2) were obtained from Sigma (St. Louis, MO), and stock solutions were prepared in deionized water immediately before use. Methylarsonic acid, disodium salt [CH3AsVO(ONa)2], was obtained from Chem Services (West Chester, PA), and dimethylarsinic acid [(CH3)2AsVO(OH)] was from Strem (Newburyport, MA). CH3AsIIII2 was synthesized as previously described, and a stock solution was prepared in 70% ethanol (24). Trimethylarsine oxide [(CH3)3AsO] was synthesized as described previously (26). Syntheses of the arsinothiols, tricysteinyltrithioarsenite [AsIII(Cys)3], CH3AsIII(Cys)2, cysteinyldimethylthioarsenite (CH3)2AsIII(Cys), and bis(γ-glutamylcysteinylglycine)dimethylthioarsenite [(CH3)2AsIIIGSH], have also been previously described (24). The solubilities of arsinothiols in stock solutions were increased by dropwise addition of 1 N HCl. A stock solution of 2,3-dimercaptosuccinic acid (DMSA, Sigma) was prepared as previously described (24). Purification of TR. Male C6B3F1 mice (4-6 weeks old) were obtained from Charles River Laboratories (Raleigh, NC) and kept at 23 °C in a 12 h light-12 h dark photocycle with free access to rodent chow (Purina Mills, Richmond, IN) and tap water, and these mice provided livers that were used for the purification scheme. The purification of TR from mouse liver homogenate followed the general scheme outlined by Luthman and Holmgren (2) and Zhong and associates (36). The DTNB reductase activity of various fractions obtained in the purification scheme was determined as described below. Various fractions from the purification scheme were electrophoresed on 4 to 20% polyacrylamide gels by the method of Laemmli (37), using the Mark 12 wide range protein standard set (Novex, San Diego, CA) for a calibration, and visualized by staining with Coomassie Brilliant Blue G. The details of the purification scheme are described in the Supporting Information. DTNB Reductase Assay. The standard assay of DTNB reductase activity followed the method of Luthman and Holmgren (2), containing (in a final volume of 1 mL) 100 mM potassium phosphate (pH 7.4) with 10 mM Na4EDTA, 0.2 mg of bovine serum albumin (fraction 9, Sigma) per milliliter, 1% (v/v) ethanol, 2 mM DTNB, and 1 mM NADPH. Addition of the source of the enzyme activity started the reaction. The NADPHdependent reduction of DTNB to 2-nitro-5-thiobenzoic acid was monitored at 25 °C by the change in absorbance at 412 nm. On the basis of the consumption of 1 mol of NADPH per 2 mol of 2-nitro-5-thiobenzoic acid produced, activities were calculated as micromoles of NADPH consumed per minute per milligram of protein. The NADPH oxidase activity of purified TR was measured by the method of Arne´r and co-workers (38). Protein contents of samples were determined by the method of Smith and associates (39) using bovine serum albumin (fraction 9) as a standard. Screening of Arsenicals as Inhibitors of TR. Various arsenicals were tested as inhibitors of the DTNB reductase activity of purified TR. Here, arsenicals were added to the standard assay mixture immediately before the addition of purified enzyme (final concentration of ∼0.112 nM). Comparisons were made between the rate of DTNB reduction in assay mixtures that contained the arsenicals in the nanomolar to micromolar concentration range with the activities in assay mixtures that contained the same final concentration of TR without arsenicals. Activities in reaction mixtures that contained arsenicals were expressed as percentages of those observed in concurrently run assay mixtures without added inhibitors. Inhibition of TR by Methylarsenic. The reversibility of inhibition of purified TR by CH3AsI2 was tested in two ways. In the first, the purified enzyme (final concentration of ∼0.112 nM), 100 nM CH3AsI2 and 1 mM NADPH were incubated for 4 min at 25° C in the standard reaction mixture without DTNB. One-half of this mixture was taken, DTNB added to a final concentration of 2 mM, and activity measured immediately under standard conditions. The remaining one-half of the mixture was diluted serially at a 1:2 ratio, yielding assay
926
Chem. Res. Toxicol., Vol. 12, No. 10, 1999
Lin et al.
mixtures that contained 50 and 25 nM CH3AsI2. Aliquots of the serially diluted samples were adjusted to a DTNB concentration of 2 mM and assayed under standard conditions. Parallel assays were performed with samples that matched the diluted samples in the concentrations of purified TR, CH3AsI2, NADPH, and DTNB. In the second method, the purified enzyme (final concentration of ∼0.063 nM), 100 nM CH3AsI2 and 1 mM NADPH were incubated for 4 min at 25 °C in the standard reaction mixture without DTNB. The mixture was then divided into three parts. One-third of the mixture was immediately assayed for TR activity under standard assay conditions. Onethird of the mixture was subjected to filtration on an Amicon 30 microconcentrator with a nominal molecular mass cutoff of 30 kDa. The retentate was washed three times with 50 mM Tris and 1 mM Na4EDTA (pH 7.5) and then recovered for determination of TR activity under standard assay conditions. The remaining one-third of the mixture was adjusted to a final concentration of 10 mM DMSA in 50 mM Tris and 1 mM Na4EDTA (pH 7.5) and incubated for 10 min at room temperature. This sample was filtered on an Amicon 30 microconcentrator. The retentate was washed once with 10 mM DMSA in 50 mM Tris and 1 mM Na4EDTA (pH 7.5) and twice with 50 mM Tris and 1 mM Na4EDTA (pH 7.5). This retentate was recovered for determination of TR activity under standard assay conditions. To examine the effect of NADPH on the activity of purified TR, the enzyme was preincubated at 25 °C for various periods of time in the presence of 1 mM NADPH with and without 100 nM CH3AsI2. At various times, aliquots of these mixtures were taken and adjusted to a final DTNB concentration of 2 mM, and activities were determined under standard assay conditions. As a control, TR was preincubated at 25 °C without NADPH and CH3AsI2. Aliquots of this mixture were taken at various times and adjusted to a final DTNB concentration of 2 mM and a final NADPH concentration of 1 mM, and activities were determined under standard assay conditions. Activities of all assays were expressed as percentages of the activity determined for TR preincubated without NADPH and CH3AsI2 for the same period of time. The effect of CH3AsIII on the NADPH oxidase activity of purified TR was also examined. Earlier work has shown that dinitrohalobenzenes not only irreversibly inhibit the reductase activity of TR but also induce the NADPH oxidase activity of the enzyme (38, 40). Here, the potency of CH3AsIII as an inducer of NADPH oxidase activity was compared with the effect of 1-chloro-2,4-dinitrobenzene (DNCB, Sigma). Here, purified TR (∼0.014 nM) was preincubated in a reaction mixture that contained 250 µM DNCB or 2 µM CH3AsI2 and 200 µM NADPH but did not contain DTNB as a substrate. The oxidation of NADPH was monitored by the change in absorbance at 340 nm.
Results Purification and Characterization of Mouse Liver TR. Purification of TR from mouse liver yielded a highly purified enzyme in which DTNB reductase activity was enriched 5800-fold compared to that found in the 10000g supernate prepared from mouse liver. This active fraction migrated as a single major protein band on SDS-PAGE with an estimated molecular mass of about 57 kDa. This fraction (designated purified TR) was used for all studies of the effects of arsenicals and arsinothiols on the kinetic behavior of TR. Details about the yield, fold purification, and kinetic characteristics of purified TR used in these studies are provided in the Supporting Information. Inhibition of TR by Arsenicals and Arsinothiols. The potencies of various arsenicals and arsinothiols as inhibitors of DTNB reductase activity were examined in assays that contained purified TR at a final concentration of ∼0.014 nM. Figure 1 shows the effect of various concentrations of trivalent or pentavalent arsenicals and
Figure 1. Potency of arsenicals and arsinothiols as inhibitors of DTNB reduction by purified TR. (a) Effects of trivalent arsenicals and arsinothiols on TR activity: CH3AsI2 (b), As(Cys)3 (O), (CH3)2As(GS) (+), CH3As(Cys)2 (9), (CH3)2As(Cys) (0), and NaAsO2 (2). (b) Effects of pentavalent arsenicals on TR activity: Na2HAsO4 (b), CH3AsO(ONa)2 (O), (CH3)2AsO(OH) (+), and (CH3)3AsO (9).
arsinothiols on the rate of DTNB reduction. Here, the rates are expressed as a percentage of that found in a standard assay that contained purified TR, NADPH, DTNB, and no inhibitor. CH3AsI2 (IC50 ∼ 700 nM) was the most potent inhibitor (Figure 1a). To ensure that inhibition by CH3AsI2 was not due to I, the potency of NaI as an inhibitor was tested. Addition of up to 10 µM NaI did not affect DTNB reductase activity (data not shown). Other trivalent arsenicals and arsinothiols were also inhibitors with IC50 values of
