Activation of Microsomal Glutathione S-Transferase and Inhibition of

Nicholas J. Westwood , Graeme W. Rogers , Justin C. McArthur , Norman J. Haughey , Barbara S. Slusher , Camilo Rojas .... P.Blaise Smith , Charles...
0 downloads 0 Views 112KB Size
Download PDF
396

Chem. Res. Toxicol. 1999, 12, 396-402

Articles Activation of Microsomal Glutathione S-Transferase and Inhibition of Cytochrome P450 1A1 Activity as a Model System for Detecting Protein Alkylation by Thiourea-Containing Compounds in Rat Liver Microsomes Rob C. A. Onderwater,† Jan N. M. Commandeur,† Wiro M. P. B. Menge,‡ and Nico P. E. Vermeulen*,† Divisions of Molecular Toxicology and Medicinal Chemistry, Leiden/Amsterdam Center for Drug Research (LACDR), Vrije Universiteit, De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands Received August 19, 1998

The recent development of several promising new thiourea-containing drugs has renewed interest in the thiourea functionality as a potential toxicophore. Most adverse reactions of thiourea-containing compounds are attributed to the thionocarbonyl moiety. Oxidation of these thionocarbonyl compounds by flavin-containing monooxygenases (FMO) and cytochrome P450 isoenzymes (P450) to reactive sulfenic, sulfinic, or sulfonic acids leads to alkylation of essential macromolecules. To more rationally design thiourea-containing drugs, structure-toxicity relationships (STRs) must be derived. Since for the development of STRs a large number of thiourea-containing compounds must be investigated, it is important to develop rapid in vitro assays for alkylating potential. In this study, the utility of activation of microsomal glutathione S-transferase (mGST) and inactivation of P450 1A1 as markers of the alkylating potential of metabolites of thiourea-containing compounds was investigated. It was found that metabolites of thiourea-containing compounds inactivate P450 1A1 in a time-dependent manner, as evidenced by a decrease in 7-ethoxyresorufin O-dealkylation (EROD) activity. An extent of inactivation of P450 1A1 by 100 µM N-phenylthiourea (PTU) of 64% was found after 10 min. This inactivation was dependent on the presence of NADPH and the presence of the thionosulfur, since the carbonyl analogue of PTU was not found to inactivate P450 1A1, and was partially prevented by heat treatment of the microsomes which is known to selectively inactivate FMO enzymes. Inactivation of P450 1A1 could be reversed by treatment with dithiothreitol, indicating the formation of disulfide bonds. However, thiourea-containing compounds also inhibited the EROD activity of P450 1A1 in a competitive manner. This property complicates the usefulness of the EROD activity of P450 1A1 as a marker for the alkylating potential of thiourea-containing compounds. It was found that metabolites of thiourea-containing compounds could transiently activate the mGST. A maximal level of activation by 100 µM PTU of 162 ( 16% was found after 10 min. Activation of mGST by 100 µM PTU was dependent on the presence of NADPH and the presence of the thionosulfur, since the carbonyl analogue of PTU was not found to activate mGST. Activation was completely prevented by heat treatment of the microsomes, indicating involvement of FMO in the bioactivation process. Finally, a series of structurally diverse thiourea-containing compounds were tested for their ability to activate mGST. It appeared that their potency in alkylating mGST was inversely related to their Vmax/Km value for the FMO enzyme. From this study, it is concluded that, whereas activation of mGST in rat liver microsomes may be a useful system with which to investigate the relationship between structure and alkylating potential of thiourea-containing compounds in vitro, inactivation of P450 1A1 is not.

Introduction The recent development of some promising new thiourea-containing drugs, among which are centrally active histamine H3 antagonists (1) and HIV RT inhibitors (2), * To whom correspondence should be addressed. Telephone: (31)20-4447590. Fax: (31)-20-4447610. † Division of Molecular Toxicology. ‡ Division of Medicinal Chemistry.

has led to a renewed interest in the thiourea group as a pharmacophore and potential toxicophore. Thioureacontaining compounds are often goitrogenic and inhibit thyroid hormone biosynthesis, while some also cause hypersensitivity reactions, and are pulmonary toxins or hepatotoxins. Most adverse reactions are attributed to the thionocarbonyl functionality since in many cases the corresponding carbonyl-containing compounds did not cause similar toxic effects (3, 4).

10.1021/tx980198p CCC: $18.00 © 1999 American Chemical Society Published on Web 04/07/1999

Activation of mGST and Inhibition of P450 1A1

Pulmonary toxicity and hepatotoxicity have been suggested to be the result of metabolic activation of the thiourea moiety. The thionocarbonyl moiety is known to be oxidized by flavin-containing monooxygenases (FMOs)1 to a sulfenic acid (5-7). In previous studies, the toxic potential of thiourea-containing compounds was related to the extent of their metabolic desulfuration. A mechanism of toxicity was proposed for the release of atomic sulfur, similar to that proposed for carbon disulfide and parathion (8). However, since then, several studies have shown that alkylation of macromolecular structures, either by the initially formed sulfenic acid or by sulfinic or sulfonic acids resulting from further oxidation steps, could also possibly explain the observed structuretoxicity relationships (9-11). Thiourea S-oxides would react with protein sulfhydryl groups either to form mixed disulfides or, via an addition-elimination reaction, to form thioethers. To investigate the relationships between the structure and alkylating potential of thiourea-containing compounds, a large number of compounds will have to be tested. Because metabolic activation is required for alkylating activity, either the reactive metabolites would have to be synthesized or the toxicological test system would have to include the potential to bioactivate the thiourea-containing compounds. Since the FMO and P450 isoenzymes, which are most likely responsible for thiourea bioactivation, are membrane-bound proteins, the test system would also have to contain microsomal membranes. The test system should also contain a macromolecular target to be alkylated at specific sites, e.g., at a sulfhydryl group, and alkylation of this target should result in a clear “signal”. Ideally, such a target would have an enzymatic activity that is unrelated to the bioactivation of the thiourea-containing compounds. Moreover, its enzymatic activity would not be inhibited by the tested compounds or metabolites of the tested compounds in any way other than by alkylation of the target site. In this study, we have investigated the applicability of two enzyme activities as markers for alkylation potential of thiourea-containing compounds, notably, the 7-ethoxyresorufin O-dealkylation (EROD) activity of P450 1A1 and the conjugation of 1-chloro-2,4-dinitrobenzene (CDNB) to glutathione (GSH) by the microsomal glutathione S-transferase (mGST) (see Figure 1). Alkylation of P450 1A1 is known to lead to an inactivation of the enzyme and therefore a decreased EROD activity (10, 11). Alkylation of mGST is known to lead to activation of the enzyme, resulting in an increased CDNB conjugation activity (12). Activation of the mGST by metabolites of thiol-reactive substances has been shown to occur before both in vitro (13-15) and in vivo (16). The aim of this study was, therefore, to develop and evaluate a novel system for the evaluation of the alkylating potential of compounds which require bioactivation to reactive metabolites. The utility of the activation of the mGST and the inactivation of P450 1A1 as markers of the alkylation potential was investigated. In this study, the alkylating potential of a number of thiourea-contain1 Abbreviations: BNF, β-naphthoflavone; CDNB, 1-chloro-2,4-dinitrobenzene; DEX, dexamethasone; DTNB, 5,5′-dithiobis(2-nitrobenzoic acid); EROD, 7-ethoxyresorufin O-dealkylation; FMO, flavin-containing monooxygenase; GSH, reduced glutathione; mGST, microsomal glutathione S-transferase; NEM, N-ethylmaleimide; PB, phenobarbital; PU, N-phenylurea; PTU, N-phenylthiourea; PETU, N-phenylethylthiourea; PBTU, N-phenylbutylthiourea.

Chem. Res. Toxicol., Vol. 12, No. 5, 1999 397

Figure 1. Schematic representation of the concept of bioactivation of thiourea-containing compounds by FMO and of detecting the alkylating potential of metabolites of thiourea-containing compounds by measuring two marker enzyme activities in rat liver microsomes.

ing compounds was investigated since it is proposed that thiourea-containing compounds are bioactivated by FMO and/or P450 to reactive sulfenic acid intermediates which have a high reactivity toward protein thiols.

Materials and Methods Isolation of Rat Liver Microsomes. Male Wistar rats (200-250 g) were obtained from Harlan (Zeist, The Netherlands) and were housed for at least 1 week prior to the experiment. The animals were fed a standard laboratory diet from Hope Farms (Woerden, The Netherlands) and had access to food and water ad libitum. Rats were killed by decapitation, and livers were excised and homogenized at 4 °C in 2 volumes of 50 mM potassium phosphate buffer (pH 7.4) with 0.9% sodium chloride. The homogenates were centrifuged for 20 min at 12000g, and the resulting supernatant was centrifuged for 60 min at 100000g. To remove all residual cytosolic glutathione S-transferases and glutathione, the resulting pellet was washed twice, by resuspension in 50 mM potassium phosphate buffer (pH 7.4) with 0.9% sodium chloride using a Potter-Ehlvejem tube and centrifugation for 60 min at 100000g. The washed pellet was resuspended again, and the microsomal preparation was stored in aliquots at -80 °C. Activation of mGST. All experiments involving the activation of microsomal glutathione S-transferase (mGST) by thioureacontaining compounds were performed with the same microsomal preparation. Incubations were performed in 1.5 mL polypropylene vials at 37 °C. Incubation mixtures (1 mL) contained 1 mg of microsomal protein in 0.1 M potassium phosphate buffer (pH 7.4) with 0.5 mM NADPH and 1 mM EDTA. All thiourea- and urea-containing compounds tested were dissolved in methanol and were added in a volume of 10 µL. N-Phenylthiourea (PTU), N,N′-diphenylthiourea, and N,N′diethylthiourea were obtained from Merck (Munich, Germany). N-Phenylurea (PU) was obtained from Fluka Chemie AG (Buchs, Switzerland). N-(R-Methylbenzyl)thiourea, N-phenylethylthiourea (PETU), and N-phenylbutylthiourea (PBTU) were synthesized according to the methods of Van der Goot et al. (17). At specific time points, a 100 µL aliquot was taken from the incubation mixture and the activity of the mGST was tested according to the methods of Habig and Jacoby (18). Briefly, the aliqout was added to 50 µL of 100 mM GSH (final concentration of 5 mM) and 840 µL of 50 mM potassium phosphate buffer (pH 6.50) at 37 °C. After addition of 10 µL of 100 mM 1-chloro-2,4dinitrobenzene (CDNB, final concentration of 1 mM) in ethanol, the rate of the increase of the absorbance at 340 nm was determined using a Pharmacia LKB Ultrospec plus spectrophotometer. To investigate the role of FMO in the activation of mGST by thiourea-containing compounds, microsomes were divided into two portions, one of which was rapidly heated to 50 °C and held at this temperature for 2 min and the other of which was held at 4 °C as a control. Inhibition of 7-Ethoxyresorufin O-Dealkylation (EROD) Activity. Experiments were performed with microsomes from

398 Chem. Res. Toxicol., Vol. 12, No. 5, 1999 rats which were pretreated intraperitoneally with β-naphthoflavone (BNF) (60 mg/kg) for two consecutive days prior to isolation of the microsomal fraction. Incubations were performed in 1.5 mL polypropylene vials at 37 °C. Incubation mixtures (1 mL) contained 1 mg of microsomal protein in 0.1 M potassium phosphate buffer (pH 7.40) with 0.5 mM NADPH and 1 mM EDTA. All thiourea-containing compounds tested were dissolved in methanol and were added in a volume of 10 µL. At specific time points, a 100 µL aliquot was taken from the incubation and added to a cuvette containing 3 mL of 0.1 M potassium phosphate buffer (pH 7.80), 100 µL of 7-ethoxyresorufin (Sigma Chemical Co., St. Louis, MO) in ethanol (final concentration of 1.5 µM), and 100 µL of NADPH (final concentration of 0.15 mM). The formation of resorufin (λex ) 530 nm, λem ) 586 nm) was assessed kinetically on a Shimadzu RF 5001 spectrofluorometer. To investigate the role of FMO in the inhibition of the EROD activity, microsomes were divided into two portions, one of which was rapidly heated to 50 °C and held at this temperature for 2 min and the other of which was held at 4 °C as a control. To investigate the nature of the inactivation of P450 1A1, 100 µL of the incubation mixture was added to 100 µL of icecold DTT (40 mM) in 0.1 M potassium phosphate buffer (pH 7.40) or to 100 µL of ice-cold potassium phosphate buffer (pH 7.40) as a control. After 10 min at 4 °C, the EROD activity of 100 µL of these treated samples was measured. Thiourea-Dependent Oxidation of Thiocholine. The thiourea-dependent oxidation of thiocholine assay is based on the principle that thiourea-containing compounds are oxidized by FMOs to sulfenic acids which in turn rapidly oxidize the small thiol thiocholine to a disulfide, thereby recycling the original thiourea-containing compound (6, 7). The rate of oxidation of thiocholine is a direct measure of the amount of sulfenic acids that is formed. Thiocholine was prepared from acetylthiocholine chloride (Sigma Chemical Co.) in acidic methanol. The extent of thioureadependent oxidation of thiocholine was determined according to the methods of Guo and Ziegler (6). Briefly, 0.5 mg of microsomal protein was preincubated at 37 °C for 5 min in a 0.1 M potassium phosphate buffer (pH 7.50) containing 1 mM EDTA, 5 mM glucose 6-phosphate, 0.5 mM NADPH, 1 unit/mL glucose-6-phosphate dehydrogenase, 2 units/mL catalase, 2 mM metyrapone (Aldrich, Milwaukee, WI), and 75 µM thiocholine. The thiourea-containing compounds, dissolved in methanol, were added to the reaction mixture (resulting in a final concentration of 1% methanol), and samples were taken at fixed time points. The amount of thiocholine remaining in the samples was determined spectrophotometrically with 5,5′-dithiobis(2nitrobenzoic acid) (DTNB) (Fluka Chemie AG) at 405 nm using a Packard Argus 400 microplate reader. Statistical Evaluation of the Results. Statistical evaluation of the results was performed with a Student’s t test. Differences were considered significant if p was less than 0.05.

Results Activation of mGST. Rat liver microsomes were incubated with 1 mM N-ethylmaleimide (NEM) to test whether the isolation procedure of the microsomal fraction had been successful in removing residual cytosolic glutathione S-transferases (GSTs). Treatment with NEM led to a more than 7-fold increase in 1-chloro-2,4dinitrobenzene (CDNB) conjugation activity of the microsomal preparation, indicating that no significant residual cytosolic GST activity remained (12). As can be seen in Figure 2, 100 µM N-phenylthiourea (PTU) leads to a time-dependent increase in the activity of the microsomal glutathione S-transferase (mGST) as evidenced by a significantly increased rate of conjugation of CDNB to glutathione (GSH). The extent of activation reached a maximal value of 193 ( 33% of the control

Onderwater et al.

Figure 2. Time-dependent activation of mGST by N-phenylthiourea (PTU) in rat liver microsomes (n ) 4). Incubation mixtures (1 mL) contained 1 mg of microsomal protein in 0.1 M potassium phosphate buffer (pH 7.4) with 0.5 mM NADPH and 1 mM EDTA. All thiourea-containing compounds tested were dissolved in methanol and were added in a volume of 10 µL. At specific time points, a 100 µL aliquot was taken from the incubation mixture and the activity of the mGST was tested according to the methods of Habig and Jacoby (18). Two asterisks indicate significant differences between treatment and control values (p < 0.001).

value after 10 min and then decreased to a level which was not significantly different from the respective control value after 15 min. This transient activation of mGST by 100 µM PTU was dependent on the presence of NADPH in the incubation mixture. Since it has been proposed in the literature that both hepatic flavincontaining monooygenases (FMOs) and P450 isoenzymes (in the presence of certain compounds) can give rise to the formation of H2O2 (19, 20), which was shown to activate mGST in vitro (21), 10 units/mL catalase was added to the incubation mixture. As can be seen from Figure 3, the addition of catalase did not reduce the transient activation of mGST induced by 100 µM PTU. The involvement of the FMO enzyme in the activation of mGST by 100 µM PTU was demonstrated by heating the microsomal preparation for 2 min at 50 °C prior to the start of the incubation (Figure 4). This pretreatment is known to completely inhibit FMO activity of rat liver microsomes (19). Heat pretreatment did not significantly decrease the basal activity of mGST, but did completely prevent the activation of mGST induced by 100 µM PTU. The activation of mGST by PTU was concentrationdependent. A concentration of 1 µM PTU, which is below the Km value for the formation of sulfenic acids by hepatic FMO (Table 1), gave rise to a delayed activation of mGST (Figure 3) when compared to that stimulated by 100 µM PTU. Indirect evidence for the involvement of the thionosulfur moiety of PTU in the transient activation of mGST was obtained by studying the effects of 100 µM Nphenylurea (PU). As can be seen from Figure 4, 100 µM PU did not give rise to an activation of the mGST. Various structurally diverse thiourea-containing compounds can give rise to the activation of mGST (Table 2).

Activation of mGST and Inhibition of P450 1A1

Chem. Res. Toxicol., Vol. 12, No. 5, 1999 399

Figure 3. Activation of mGST by N-phenylthiourea (PTU) in rat liver microsomes (n ) 4). Incubation mixtures (1 mL) contained 1 mg of microsomal protein in 0.1 M potassium phosphate buffer (pH 7.4) with 0.5 mM NADPH and 1 mM EDTA. All thiourea-containing compounds tested were dissolved in methanol and were added in a volume of 10 µL. To all incubation mixtures was added 10 units/mL catalase. At specific time points, a 100 µL aliquot was taken from the incubation mixture and the activity of the mGST was tested according to the methods of Habig and Jacoby (18). One asterisk indicates significant differences between treatment and control values (p < 0.05). Two asterisks indicate significant differences between treatment and control values (p < 0.001).

Figure 4. Activation of mGST by N-phenylthiourea (PTU) and N-phenylurea (PU) in rat liver microsomes (n ) 4). Microsomes were heated for 2 min at 50 °C to inhibit FMO activity. Incubation mixtures (1 mL) contained 1 mg of microsomal protein in 0.1 M potassium phosphate buffer (pH 7.4) with 0.5 mM NADPH and 1 mM EDTA. All thiourea-containing compounds tested were dissolved in methanol and were added in a volume of 10 µL. To all incubation mixtures was added 10 units/ mL catalase. At specific time points, a 100 µL aliquot was taken from the incubation mixture and the activity of the mGST was tested according to the methods of Habig and Jacoby (18). Two asterisks indicate significant differences between treatment and control values (p < 0.001).

Inhibition of 7-Ethoxyresorufin O-Dealkylation (EROD) Activity. As can be seen in Figure 5, 100 µM PTU leads to a significant decrease in the activity of P450 1A1 as evidenced by a decrease in the activity of the O-dealkylation of 7-ethoxyresorufin (EROD). After 10 min, 100 µM PTU caused a 64% decrease in the EROD activity as compared to control value (Table 3). PTU also caused a concentration-dependent inhibition of P450 1A1 (Table 3). The inactivation of P450 1A1 could completely be reversed by treatment with dithiothreitol (DTT), indicating that the inactivation was due to the formation of a disulfide bond. The involvement of FMO in the inhibition of P450 1A1 by 100 µM PTU was shown by heating the microsomal preparation for 2 min at 50 °C prior to the start of the incubation (Table 3) which is known to inhibit FMO activity (19). Heat pretreatment did not significantly decrease the basal activity of P450 1A1, but partially prevented the inactivation of P450 1A1 by 100 µM PTU. Indirect evidence for the involvement of the thionosulfur moiety of PTU in the inactivation of P450 1A1 was obtained when 100 µM PU was added to the incubation mixture. As can be seen from Figure 5, 100 µM PU does not give rise to an inactivation of P450 1A1. Because in this case (EROD) a decrease in activity is measured, it is important to determine whether the compound being investigated for alkylating potential directly causes a decrease in the marker enzyme activity. In Figure 6, the concentration-dependent inhibition of EROD activity by an analogous series of N-phenylalkylthiourea, notably PTU, N-phenylethylthiourea (PETU), and N-phenylbutylthiourea (PBTU) is shown, without preincubation. An IC50 for EROD activity of 50 µM was

Table 1. Thiourea-Dependent Oxidation of Thiocholinea compound

Vmax (nmol min-1 mg of protein-1)

Km (µM)

N-phenylthiourea (PTU) N-phenylethylthiourea N-phenylbutylthiourea N-(R-methylbenzyl)thiourea N,N′-diethylthiourea N,N′-diphenylthiourea thioperamideb

2.8 ( 0.8 2.0 ( 0.3 6.5 ( 4.8 2.2 ( 0.3 1.7 ( 0.1 1.9 ( 0.1 4.7 ( 0.3

5.5 ( 2.4 2.5 ( 0.6 7.8 ( 7.1 3.0 ( 0.7 8.9 ( 1.1 3.7 ( 0.5 191 ( 24

a Values for V max and Km were determined using a concentration range of 1 order of magnitude around a Km estimated from a concentration range of 1-1000. b Values determined in a different microsomal batch.

Table 2. Activation of mGST by Thiourea-Containing Compoundsa compound (100 µM)

% of control activity

N-phenylthiourea (PTU) N-(R-methylbenzyl)thiourea N,N′-diethylthiourea N,N′-diphenylthiourea thioperamide

262 ( 16 262 ( 33 212 ( 21 271 ( 59 166 ( 38

a Incubation mixtures contained 10 units/mL catalase to prevent mGST activation by microsomal H2O2 production. Values of mGST activity after 10 min are reported (n ) 4).

calculated for PTU and PBTU and of 1000 µM for PETU. Since the samples taken from the incubation mixture are diluted more than 30-fold in the EROD assay, the direct effect of PTU, PETU, and PBTU on the EROD activity of P450 1A1 is not likely to interfere with the EROD activity of the microsomal samples. PBTU and PETU both were found to inhibit EROD activity in a time-

400 Chem. Res. Toxicol., Vol. 12, No. 5, 1999

Figure 5. Time-dependent inhibition of P450 1A1 activity by 100 µM N-phenylthiourea (PTU) and 100 µM N-phenylurea (PU) (n ) 3). Samples (100 µL) were taken from an incubation mixture and assayed for EROD activity in a cuvette containing 100 µL of 7-ethoxyresorufin in ethanol (50 µM) and 100 µL of NADPH (5 mM) in 3 mL of 0.1 M potassium phosphate buffer (pH 7.80). Table 3. Concentration-Dependent Inhibition of EROD Activity by PTU after 10 mina

PTU (µM)

EROD activity (% of control)

EROD activity (% of control) after DTT treatment

0 0.5 1 5 10 100 100 (2 min at 50 °C)

100b 95 71 57 38 36 62

ndc nd 108 nd 105 110 98

a EROD activity can be restored to the control level by treatment with DTT (n ) 2). Heat pretreatment of the microsomes for 2 min at 50 °C attenuates the inhibition of EROD activity. b Values are the mean of two separate experiments. c Not determined.

dependent manner with t1/2 values of 5.1 ( 0.7 and 7.4 ( 1.2 min, respectively.

Discussion The aim of this study was to develop and evaluate a novel system for evaluation of the alkylating potential of compounds that require bioactivation to reactive metabolites. In this study, the utility of the activation of the microsomal glutathione S-transferase (mGST) and the inactivation of P450 1A1 as markers of the alkylating potential of thiourea-containing compounds was investigated. Since it is known that thiourea-containing compounds are bioactivated by flavin-containing monooxygenases (FMOs) to metabolites which can potentially alkylate protein thiols, this was investigated in rat liver microsomes. The mGST is a membrane-bound protein consisting of three individually catalytically active subunits (22). The mGST efficiently catalyzes the conjugation of chlorodinitrobenzene (CDNB) to glutathione (GSH), an activity which can be tested under conditions in which thioureacontaining compounds do not interfere with this conjugation activity as can be seen in Figure 2. Alkylation of the sulfhydryl group of Cys-49 of these subunits is known to

Onderwater et al.

Figure 6. Direct inhibition of EROD activity by N-phenylalkylthiourea: (0) N-phenylthiourea (n ) 3), (b) N-phenylethylthiourea (average of n ) 2), and (9) N-phenylbutylthiourea (n ) 3). Thirty microliters of a thiourea-containing compound dissolved in methanol was added directly to a cuvette containing 100 µL of microsomes from β-naphthoflavone-pretreated rats (10 mg/mL), 100 µL of 7-ethoxyresorufin in ethanol (50 µM), and 100 µL of NADPH (5 mM) in 3 mL of 0.1 M potassium phosphate buffer (pH 7.80). The increase in fluorescence was followed for 2 min.

result in an increased catalytic activity (23). An increase in activity of mGST has been shown, both in microsomal preparations and in the purified protein, with compounds which specifically alkylate sulfhydryl groups such as N-ethylmaleimide (NEM) and iodoacetamide (24), acrolein (13), metabolites of phenol (14), and R-methyldopa (15). A 1.7-fold increase in activity after 10 min has been shown to occur after alkylation of mGST in the presence of tert-butyl hydroperoxide and hydrogen peroxide in microsomes (21, 25) as well as in perfused rat liver (26). In this study, it is shown that metabolites of thioureacontaining compounds are also capable of activating mGST. The observed activation is transient in nature and reaches its maximum after 10 min. Results from a study by Moorhouse and Casida (27) of the alkylating potential of several pesticides suggested that activation of mGST is not a suitable marker for alkylating potential of compounds which require metabolic activation. However, in their study, the activity of mGST was determined after 30 min of incubation of microsomes of mouse liver with the compounds being investigated. Our results, as well as those of others (21, 25), have shown that, because of the transient nature of the activation and the increased basal activity of mGST in the prolonged presence of NADPH at 37 °C, the extent of activation of mGST should be determined within the first 15 min of incubation. Furthermore, whereas in this study 1 mM NEM activated the mGST more than 7-fold, Moorhouse and Casida (27) reported an extent of activation of only 2-fold. Since a 2-fold extent of activation by the sulfydryl selective reagent NEM represents the maximally achievable activation of mGST in their microsomal preparation, it would have been difficult to detect activation by less selective activators under the conditions applied. Whereas thiol-alkylating compounds such as NEM and iodoacetamide cause a stable activation of mGST, the metabolites of thiourea-containing compounds do not. This may result from alkylation of other nucleophilic sites in the enzyme besides Cys-49 by reactive metabolites of

Activation of mGST and Inhibition of P450 1A1

thiourea-containing compounds. An initial increase in the mGST activity would then be followed by a subsequent decrease in the enzyme activity. In fact, it has been shown in our laboratory that acrolein, which is known to alkylate both sulfhydryl moieties and amine functionalities, exhibits a similar transient activation of mGST, which ultimately results in a total loss of mGST activity (E. M. Van der Aar, personal communication). In this study, 100 µM PTU activated the mGST by 162 ( 16% after 10 min. The parent thiourea-containing compounds are not capable of alkylkating mGST directly, but require bioactivation to alkylating metabolites, since NADPH was required for the activation of mGST by thiourea-containing compounds (Figure 2). The alkylating metabolites are formed by hepatic FMO enzymes since heat treatment of the microsomes completely abated the activation of mGST (Figure 4). The alkylating metabolites are likely to be sulfenic acids, since N-phenylurea (PU), which cannot give rise to the formation of sulfenic acids, did not activate mGST (Figure 3). Furthermore, addition of 1 mM GSH to the incubation mixture completely prevented the activation of mGST by PTU (Figure 3). Presumably, GSH reduces the sulfenic acids back to the original thiourea moiety, thereby preventing them from alkylating Cys-49 of the mGST, or forms stable mixed disulfides with the sulfenic acid. In microsomes of nonpretreated rats, bioactivation of the thiourea-containing compounds by FMO appears to be responsible for the activation of mGST since pretreatment of the microsomes for 2 min at 50 °C inhibited PTU-dependent activation of mGST whereas it did not inhibit the mGST basal activity. As can be seen from Table 2, various structurally diverse thiourea-containing compounds can give rise to the activation of mGST. However, in the case of 100 µM thioperamide [N-cyclohexyl-N′-[4-[imidazol-4(5)-yl]piperidyl]thiourea, a histamine H3 receptor antagonist] after 10 min this activation (166 ( 33% of control value; n ) 4) is significantly lower than that obtained with 100 µM PTU (262 ( 16% of control value; n ) 4). This is likely because of the larger Km value of thioperamide (191 ( 24 µM; n ) 3) for the formation of sulfenic acids by FMO, as determined by the thiourea-dependent oxidation of thiocholine (Table 1), when compared to the corresponding Km value of PTU (5.5 ( 2.4 µM; n ) 3). In the latter case, the rate of formation of sulfenic acids would almost have reached its maximal value, whereas in the case of thioperamide, it would still be far below its half-maximal rate of formation. Theoretically, due to the unusual catalytic cycle of FMO, whereas Km values will differ, Vmax values for the rate of formation of sulfenic acids from thiourea-containing compounds by an FMO isoenzyme should be identical (7, 19). The significantly higher Vmax value for thioperamide compared to those of the other thiourea-containing compounds in Table 1 could result from differences in microsomal recovery of FMO, i.e., a higher FMO concentration, because a different microsomal preparation was used. As was shown by Kedderis and Rickert (28), methimazole inactivates cytochrome P450 chromophore in microsomes of non-pretreated rats after bioactivation by FMO. A loss of cytochrome P450 (P450) of 25% was observed after incubation for 15 min with 1 mM methimazole. Losses of P450 were also observed after incubation of microsomes from PB- (31%) and BNF-pretreated rats (44%) with methimazole and NADPH. Kedderis and

Chem. Res. Toxicol., Vol. 12, No. 5, 1999 401

Rickert found that heat treatment of the microsomes resulted in a complete protection from inhibition of P450s in microsomes from non-pretreated rats, but not in microsomes of phenobarbital (PB)- or β-naphthoflavone (BNF)-pretreated rats, indicating that the induced P450s might be involved in thiourea bioactivation. Decker and Doerge (10, 11) have shown that ethylenethiourea (ETU) and several benzimidazole-2-thiones are also capable of inhibiting P450 activities by covalent binding. In their study with radiolabeled [14C]ETU, Decker and Doerge found the extent of protein binding to be maximal after 15 min. The ETU-mediated loss of P450 activity (up to 68%) was found to be attenuated by heat inactivation in microsomes of phenobarbital-pretreated rats (by 43%) but not in microsomes of DEX- or BNF-pretreated rats. ETUmediated loss of P450 activity could be prevented by addition of GSH to the incubation, in which case GSSG formation was observed, and could be partially restored by addition of DTT afterward. The urea analogue of ETU was found not to inhibit P450 activities. From their results, Decker and Doerge concluded that the formation of an ETU-GSH disulfide subsequently followed by disulfide exchange with another GSH molecule could possibly explain the protective effect of GSH. When no GSH is present, or when the level of GSH is depleted, the formation of a mixed disulfide with protein thiols would lead to the inactivation of P450. In this study, P450 1A1 inhibition was used as a second marker for the alkylating potential of metabolites of thiourea-containing compounds. PTU was found to inactivate P450 1A1 in a time- and concentration-dependent manner (Figure 5). PTU (100 µM) was found to inactivate 64% of the 7-ethoxyresorufin dealkylation activity of P450 1A1 in 10 min (Table 3), whereas the urea analogue PU (100 µM) was found not to inactivate P450 1A1, indicating involvement of the thionosulfur moiety in the inactivation process. The inactivation of P450 1A1 by PTU was found to be totally reversible upon treatment with DTT (Table 3). This suggests that the thiourea metabolite forms a disulfide bond with a sulfhydryl group of P450 1A1, consistent with the action of a sulfenic acid. Since the inhibition of P450 activity can be attenuated by addition of GSH to the incubation mixture, the sulfhydryl-reactive metabolite is diffusable, and therefore, none of the eight cysteine residues of P450 1A1 can be ruled out as the target. However, chemical modification experiments and thiol modification with rabbit liver P450 1A1 have indicated that Cys-161 and Cys-461 are crucial for enzyme action (29). Decker and Doerge (10) found in their study that all bound ETU was released upon treatment of microsomes of dexamethasone (DEX)-pretreated rats after incubation with DTT. However, the activity of the P450 isoenzymes induced by DEX was not totally restored. In the study presented here, heat treatment resulted in an attenuation of the PTUmediated loss of P450 1A1 activity (from 64 to 38%; Table 3). Apparently, the induced P450 1A1 or P450 1A2 is capable of forming reactive metabolites of PTU which also inactivate P450 1A1. This might present a problem in the use of P450 1A1 activity as a marker for alkylating potential. Ideally, the marker enzyme activity should not itself interfere with the response. Even without pretreatment, PTU directly inhibits the EROD activity of P450 1A1 (Figure 6) with an IC50 of 50 µM. Since PTU is diluted from the samples to a concentration which does not have a direct effect on EROD activity, this did not

402 Chem. Res. Toxicol., Vol. 12, No. 5, 1999

have any consequences for the utility of the EROD assays as a marker for P450 1A1 inactivation. However, it does complicate the interpretation of the results, since, in theory, PTU could competitively inhibit the covalent modification of P450 1A1 by reactive metabolites of PTU (which were formed by FMO). Furthermore, the large difference in the IC50 values for direct inhibition of P450 1A1 among a homologous series of N-phenalkylthioureas (Figure 6) indicates a further problem with the use of P450 1A1 as a target for protein alkylation. Whereas PTU and PBTU will inhibit P450 1A1 at the concentration used to investigate their alkylating potential, PETU will not, thus complicating the use of P450 1A1 inactivation as a means of evaluating the relationship between the structure and alkylating potential of thiourea-containing compounds. The aim of this study was to develop and evaluate a novel system for the evaluation of the relationship between the structure and alkylating potential of compounds which require bioactivation to reactive metabolites. In this study, the utility of the activation of the microsomal glutathione S-transferase (mGST) and the inactivation of P450 1A1 as markers for the alkylating potential was investigated. It can be concluded from this study that activation of the mGST is a suitable marker for studying the relationship between the structure and alkylating potential of thiourea-containing compounds in vitro; however, inactivation of the P450 1A1 activity is only suitable to a lesser extent. Furthermore, this study has shown that covalent modification of mGST likely occurs as a disulfide between the thiourea moiety and Cys-49 of the mGST formed by the reaction of the Cys49 thiol and the sulfenic acid formed by FMO-mediated oxidation of the thiourea-containing compounds.

References (1) Vollinga, R. C., Menge, W. M. P. B., Leurs, R., and Timmerman, H. (1995) New analogs of burimamide as potent and selective histamine H3 receptor antagonists: the effect of chain length variation of the alkyl spacer and modifications of the N-thiourea substituent. J. Med. Chem. 38, 2244-2250. (2) Cantrell, A. S., Engelhardt, P., Ho¨gberg, M., Jaskunas, S. R., Gunnar Johansson, N., Jordan, C. L., Kangasmetsa¨, J., Kinnick, M. D., Lind, P., Morin, J. M., Jr., Muesing, M. A., Noree´n, R., O ¨ berg, B., Pranc, P., Sahlberg, C., Ternansky, R. J., Vasileff, R. T., Vrang, L., West, S. J., and Zhang, H. (1996) Phenethylthiazolylthiourea (PETT) compounds as a new class of HIV-1 reverse transcriptase inhibitors. 2. Synthesis and further structurereactivity relationship studies of PETT analogs. J. Med. Chem. 39, 4261-4274. (3) Skellern, G. G. (1989) Thiocarbamides. In Sulphur containing drugs and related organic compounds. Chemistry, Biology, Toxicity (Damani, L. A., Ed.) Part 1B, p 49, Ellis Horwood Ltd., Chichester, U.K. (4) Onderwater, R. C. A., Commandeur, J. N. M., Groot, E. J., Sitters, A., Menge, W. M. P. B., and Vermeulen, N. P. E. (1998) Cytotoxicity of a series of mono- and di-substituted thiourea in freshly isolated rat hepatocytes: a preliminary structure-toxicity relationship study. Toxicology 125, 117-129. (5) Poulsen, L. L., Hyslop, R. M., and Ziegler, D. M. (1979) SOxygenation of N-substituted thioureas catalyzed by the pig liver microsomal FAD-containing monooxygenase. Arch. Biochem. Biophys. 198, 78-88. (6) Guo, W.-X. A., and Ziegler, D. M. (1991) Estimation of flavincontaining monooxygenase activities in crude tissue preparations by thiourea dependent oxidation of thiocholine. Anal. Biochem. 198, 143-148. (7) Guo, W.-X. A., Poulsen, L. L., and Ziegler, D. M. (1992) Use of thiocarbamides as selective substrate probes for isoforms of flavincontaining monooxygenases. Biochem. Pharmacol. 44, 2029-2037. (8) Lee, P. W., Arnau, T., and Neal, R. A. (1980) Metabolism of R-Naphthylthiourea by rat liver and rat lung microsomes. Toxicol.

Onderwater et al. Appl. Pharmacol. 53, 164-173. (9) Miller, A. E., Biscoff, J. J., and Pae, K. (1988) Chemistry of aminoiminomethanesulfinic and -sulfonic acids related to the toxicity of thioureas. Chem. Res. Toxicol. 1, 169-174. (10) Decker, C. J., and Doerge, D. R. (1991) Rat hepatic microsomal metabolism of ethylenethiourea. Contributions of the flavincontaining monooxygenase and cytochrome P-450 isoenzymes. Chem. Res. Toxicol. 4, 482-489. (11) Decker, C. J., and Doerge, D. R. (1992) Covalent binding of 14Cand 35S-labelled thiocarbamides in rat hepatic microsomes. Biochem. Pharmacol. 43, 881-888. (12) Morgenstern, R., DePierre, J. W., and Ernster, L. (1979) Activation of microsomal glutathione S-transferase activity by sulfhydryl reagents. Biochem. Biophys. Res. Commun. 87, 657-663. (13) Haenen, G. R. M. M., Vermeulen, N. P. E., Tai Tin Tsoi, J. N. L., Ragetli, H. M. N., Timmerman, H., and Bast, A. (1988) Activation of the microsomal glutathione-S-transferase and reduction of the gluatathione dependent protection against lipid peroxidation by acrolein. Biochem. Pharmacol. 37, 1933-1938. (14) Wallin, H., and Morgenstern, R. (1990) Activation of microsomal glutathione transferase activity by reactive intermediates formed during the metabolism of phenol. Chem.-Biol. Interact. 75, 185199. (15) Haenen, G. R. M. M., Jansen, F. P., Vermeulen, N. P. E., and Bast, A. (1991) Activation of the microsomal glutathione Stransferase by metabolites of R-methyldopa. Arch. Biochem. Biophys. 287, 48-52. (16) Yonamine, M., Aniya, Y., Yokomakura, T., Koyama, T., Nagamine, T., and Nakanishi, H. (1996) Acetaminophen-derived activation of liver microsomal glutathione S-transferase of rats. Jpn. J. Pharmacol. 72, 175-181. (17) Van der Goot, H., Schepers, M. J. P., Sterk, G. J., and Timmerman, H. (1992) Isothiourea Analogues of Histamine as Potent Agonists or Antagonists of the Histamine H-3-Receptor. Eur. J. Med. Chem. 27, 511-517. (18) Habig, W. H., and Jacoby, W. B (1981) Assays for differentiation of glutathione S-transferases. Methods Enzymol. 77, 398-405. (19) Ziegler, D. M. (1993) Recent studies on the structure and function of multisubstrate flavin-containing monooxygenases. Annu. Rev. Pharmacol. Toxicol. 33, 179-199. (20) Goeptar, A. R., Scheerens, H., and Vermeulen, N. P. E. (1995) Oxygen and xenobiotic reductase activities of cytochrome P450. Crit. Rev. Toxicol. 25, 25-65. (21) Aniya, Y., and Anders, M. W. (1992) Activation of rat liver microsomal glutathione S-transferas by hydrogen peroxide: role for protein-dimer formation. Arch. Biochem. Biophys. 296, 611616. (22) Lundqvist, G., Yu¨cel-Lindberg, T., and Morgenstern, R. (1992) The oligomeric structure of rat liver microsomal glutathione transferase studied by chemical cross-linking. Biochim. Biophys. Acta 1159, 103-108. (23) Andersson, C., Weinander, R., Lundqvist, G., DePierre, J. W., and Morgenstern, R. (1994) Functional and structural membrane topology of rat liver microsomal glutathione transferase. Biochim. Biophys. Acta 1204, 298-304. (24) Andersson, C., Piemonte, F., Mosialou, E., Weinander, R., Sun, T.-H., Lundqvist, G., Adang, A. E. P., and Morgenstern, R. (1995) Kinetic studies on rat liver microsomal glutathione transferase: consequences of activation. Biochim. Biophys. Acta 1247, 277283. (25) Aniya, Y., and Daido, A. (1993) Organic hydroperoxide-induced activation of liver microsomal glutathione S-transferase of rats in vitro. Jpn. J. Pharmacol. 62, 9-14. (26) 26) Aniya, Y., and Daido, A. (1994) Activation of microsomal glutathione S-transferase in tert-butyl hydroperoxide-inducd oxidative stress of isolated rat liver. Jpn. J. Pharmacol. 66, 123130. (27) Moorhouse, K. G., and Casida, J. E. (1992) Pesticides as activators of mouse liver microsomal glutathione S-transferase. Pestic. Biochem. Physiol. 44, 83-90. (28) Kedderis, G. L., and Rickert, D. E. (1984) Loss of rat liver microsomal cytochrome P-450 during methimazole metabolism. Role of flavin-containing monooxygenase. Drug Metab. Dispos. 13, 58-61. (29) Black, S. D., and Coon, M. J. (1985) Studies on the identity of the heme-binding cysteinyl residue in rabbit liver microsomal cytochrome P-450 isozyme 2. Biochem. Biophys. Res. Commun. 128, 82-88.

TX980198P