Chem. Res. Toxicol. 1994, 7, 659-665
659
Thioredoxin Alkylation by a Dihaloethane-Glutathione Conjugate Marian Meyer,? Ole N. Jensen,? Elisabeth Barofsky,$?$ Douglas F. Barofsky,$pgand Donald J. Reed*lflg Department of Biochemistry and Biophysics, Department of Agricultural Chemistry, and Environmental Health Sciences Center, Oregon State University, Agricultural and Life Sciences Building, Corvallis, Oregon 97331-7305 Received May 6,1994@ Glutathione is a thiol-containing tripeptide which functions to protect cellular constituents from endogenous and xenobiotic electrophiles via conjugation and eventual excretion. In the case of compounds such as l,&-dihaloethanes, however, conjugate formation results i n bioactivation of the species rather than detoxification. The conjugate can then act as a n alkylating agent toward cellular constituents including DNA, proteins, or lipids. Alkylation of protein thiols in cells exposed t o dihaloethane may contribute substantially to the toxicity produced by these compounds. We examined the reactivity of the conjugate S-(2-chloroethyl)glutathione (CEG) toward the model protein Escherichia coli thioredoxin. At physiological pH, treatment of thioredoxin by CEG resulted in the production of several bands visible on isoelectric focusing, which were determined by matrix-assisted laser desorption ionization (MALDI) mass spectrometry to be mono-, di-, tri-, and tetra-alkylated forms of thioredoxin. A concomitant loss of in vitro enzymatic activity was observed. These products were also observed when reaction was allowed to take place at pH 11.4. Treatment a t pH 4.4 resulted in lesser alkylation of thioredoxin, with only the mono-and di-alkylated forms detected. Iodoacetic acid treatment of CEG-alkylated thioredoxin revealed that the iodoacetic acid-susceptible Cy932 was not carboxymethylated, suggesting that this is one of the sites alkylated by CEG.
Introduction In 1983, 1,2-dichloroethane was the fifteenth highest volume chemical produced in the United States. 1,2Dichloroethane is used as an industrial solvent as well as in the synthesis of vinyl chloride and several other chloro-containingcompounds ( I 1. 1,2-Dibromoethaneis used as a scavenger in leaded gasoline, as an insecticidal fumigant, and as a chemical intermediate in the production of some dyes and waxes (2). l,2-Dihaloethanes have been found to be hepatotoxic, mutagenic, and carcinogenic. Exposure to these compounds leads to a rapid, substantial loss of cellular glutathione accompanied by lipid peroxidation (3),formation of DNA adducts (4,5) and covalent protein adduct formation (6, 7). Glutathione (GSH)l is a thiol-containing tripeptide that protects cellular constituents from endogenous and xenobiotic electrophiles via conjugation and eventual excretion. In some instances, however, the formation of conjugates results in bioactivation of the species rather than detoxification. The dihaloethane conjugate of glutathione, S-(2-haloethyl)glutathione, can cyclize to form a reactive episulfonium ion, which can act as an alkylating agent. One of the possible mechanisms by which dihaloethanes exert their toxic effects is that of protein alkylation.
* Address correspondence to this author at Biochemistry and Biophysics Department, Agricultural and Life Sciences Building 2011, Oregon State University, Corvallis, OR 97331-7305.Phone: (503)7374438;FAX: (503)737-4371. + Department of Biochemistry and Biophysics. Department of Agricultural Chemistry. 5 Environmental Health Sciences Center. @Abstractpublished in Advance ACS Abstracts, August 15, 1994. Abbreviations: CEG, S-(2-chloroethyl)glutathione;IAA,iodoacetic acid; DTT, dithiothreitol; MALDI, matrix-assisted laser desorption ionization; GSH, glutathione.
*
Alkylation of cellular proteins as a result of dihaloethane exposure is k n o w n to occur, but the role of this protein modification in the events leading to cell death is not known. Extensive protein alkylation by 1,a-dihaloethanes has been reported both in vivo (8)and with hepatocytes in vitro (6, 7). Investigation of the alkylating activity of the bioactive dihaloethane-derived species S-(2-chloroethyl)glutathione(CEG) against model compounds such as dipeptides, nucleosides, and glutathione revealed that CEG preferentially alkylates cysteinyl thiol groups (9). Escherichia coli (E. coli) thioredoxin was selected as a model protein in this study of how protein thiols, and possibly other amino acid residues, can be modified by dihaloethane-glutathione conjugates. Thioredoxin, a ubiquitous protein, is active in the oxidoreduction biochemistry of a large number of biological systems, including acting as a reducing agent for ribonucleotide reductase, functioning as a subunit of DNA polymerase of bacteriophage T4,and playing a role in the assembly of filamentous viruses (10, 11). Thioredoxin is important in the light-dependent regulation of a number of enzymes in photosynthetic bacteria and plants (12)and has been purified and characterized from algae (13). Recent evidence suggests that thioredoxin can regenerate proteins inactivated by oxidative stress (14). The thioredoxin from E. coli possesses two redox-active Cys residues, in the sequence Trp-Cys32-Gly-Pro-Cyss5-, that have different pK, values. Here, we examine the extent of CEG alkylation of amino acid residues in E. coli thioredoxin and the effect of this alkylation on the enzymatic activity of the protein.
0893-228x/94/2707-0659$04.50/00 1994 American Chemical Society
Meyer et al.
660 Chem. Res. Toxicol., Vol. 7, No. 5, 1994
E
E
f
e4 c)
m
Time Figure 1. Capillary electrophoresis of thioredoxin treated with S-(2-~hloroethyl)glutathione andlor iodoacetic acid. Separation took place for 30 min. (A) Thioredoxin. (B) Thioredoxin treated with CEG. (C) Thioredoxin treated with CEG followed by IAA.(D) Thioredoxin treated with IAA.(E) Mixture of A and B. (F) Mixture of A and C. (G) Mixture of A and D.Peaks denoted with 1 represent unreacted thioredoxin. Peaks denoted with 2 represent derivatized thioredoxin.
Experimental Procedures Materials. E. coli thioredoxin was purchased from Calbiochem-Novabiochem Corp. (La Jolla, CAI. 1-Bromo-2-chloroethane was obtained from Aldrich Chemical Co. (Milwaukee, WI). Reduced glutathione, iodoacetic acid (IAA),isoelectric focusing protein standards, dithiothreitol (D'IT), and bovine pancreas insulin were all purchased from Sigma Chemical Co. (St. Louis, MO). Isoelectric focusing gels were obtained from Serva Chemical Co. (Hauppauge, NY). Preparation of S-(2-Chloroethyl)glutathione (CEG). CEG was prepared and purified as described by Reed and Foureman (15). For some experiments, CEG was synthesized by the method of Humphreys et al. (16) and used without subsequent purification. Fast atom bombardment mass spectrometric analysis of the product of the latter method demonstrated that most of the final material was sodium salts of CEG, with very little unreacted glutathione present. Sodium was estimated to constitute approximately 50%of the final mass of CEG prepared by this method. Alkylation of Thioredolrin with CEG and/or Iodoacetic Acid (IAA). Thioredoxin was mixed with between 10- and 100fold molar excess of D'IT in 0.15 M potassium phosphate buffer (pH 7.4) and incubated at room temperature for 1h. The sample was then split into four aliquots. To two of these, at least 5-fold molar excess (over D'IT) of CEG was added and allowed to react for 1h at room temperature. Finally, 2-fold molar excess (over CEG) of IAA was added to one CEG-containing tube and one non-CEG tube and allowed to react for 1h. The reaction was terminated by the addition of excess D'IT. The samples were dialyzed against water, and then lyophilized. Treatments: (A) thioredoxin, (B) thioredoxin CEG, (C) thioredoxin + CEG + IAA, (D) thioredoxin IAA. In some experiments, the alkylation reaction was also performed in phosphate buffer at pHs of 4.4 and 11.4. Isoelectric Focusing of Alkylated Thioredoxin. Isoelectric focusing was performed with the use of a Hoefer HE900 Series horizontal slab gel unit on a Serva Precote pH 3-6 gel, over a 3-h period of increasing voltage from 200 to 2000 V. Between 5 and 10 pg of protein was loaded in each lane. The gel was stained with Coomassie Blue R. Matrix-AssistedLaser DesorptionlIonization (MALDI) Mass Spectrometry. Approximately 50 pg protein samples were redissolved in 50 pL of water. An aliquot of sample was mixed 1:9 (v/v) with matrix solution [lo g/L 3,5-dimethoxy-4hydroxycinnamic acid (sinapinic acid) in aqueous 0.1% trifluoroacetic acid and acetonitrile (2:1)]to a final protein concentration of approximately 0.1 g/L. One microliter of this solution was deposited on the mass spectrometric probe and dried under a stream of air at ambient temperature. The dried sample was washed by dipping the probe tip into cold (4 "C) water followed by air-drying. A custom-built time-of-flight mass spectrometer equipped with a frequency-tripled (355-nm) Nd:YAG laser (SpectraPhysics GCR-11) was used for mass analysis (17).The ion
+
+
source was operated at 24 kV. Mass spectra were collected by summing the data generated from 50 individual laser pulses. Mass calibration was accomplished by using equine myoglobin (MW16 951) as an internal molecular weight standard, and the accuracy of the mass calibration was determined from repeated analyses to be f0.05% at mlz 12 000 (i.e., f 6 m / z units). Data analysis was performed with m-over-z software provided by Dr. R. C. Beavis at Memorial University, Newfoundland. Capillary Electrophoresis. A Beckman P/ACE System 2050 capillary electrophoresis unit with a 20 cm x 50 pm capillary was used. Samples dissolved in water were mixed 1 : l O with electrophoresis buffer (0.10 M phosphate buffer, pH 7.6, containing 0.25 M o-phosphorylethanolamine)and loaded by pressure for 10 s. Separation took place at 5 kV for 30 min. Detection was at 214 nm. Thioredoxin Enzyme Activity Assay. The activity of E. coli thioredoxin was assayed according to the insulin disulfide reduction method described by Holmgren (18). Protein Concentration. Concentration of protein was determined using the microtiter plate protocol with the Pierce (Rockford, IL) BCA protein assay reagent. Plates were read on a Biotek Instruments, Inc., EL340 microplate reader using KinetiCalc EIA Application Software Version 2.03.
Results Capillary electrophoresis of the four samples described above revealed the following electrophoretic patterns (Figure 1). Thioredoxin (A) produced a single peak. When thioredoxin was treated with IAA (D), the retention time of the predominant peak was slightly longer, as can be seen when samples A and D were mixed a n d subjected to capillary electrophoresis (G). The first peak in chromatogram G represents unreacted thioredoxin from sample A, a n d the second peak arises from sample D. Thioredoxin treated with CEG (B) produced two peaks in the same retention time region, t h e first of which corresponds to unreacted thioredoxin, as revealed when samples A a n d B were mixed together a n d rerun (E). The second, slower eluting peak represented the principal product of CEG alkylation of the protein. When CEGtreated thioredoxin was further reacted with IAA (C), the peak corresponding to unreacted thioredoxin disappeared, b u t no other major changes were visible, suggesting that CEG-alkylated a n d LAA-alkylated thioredoxin have similar retention times under these electrophoretic conditions. The other peaks visible are residual reagents, established by controls (data not shown). Isoelectric focusing of the same samples gave more information (Figure 2). Treatment of thioredoxin with IAA (D)(lane 4) resulted in the loss of the band
Thioredoxin Alkylation by a Glutathione Conjugate
Chem. Res. ToxicoZ., VoZ. 7, No. 5, 1994 661
1 2 3 4 5 6 7 8 Figure 2. Isoelectric focusing of thioredoxin alkylated at pH 7.4. Lane 1: Thioredoxin. Lane 2: Thioredoxin treated with CEG. Lane 3: Thioredoxin treated with CEG followed by IAA. Lane 4: Thioredoxin treated with IAA.Lane 5: Glucose oxidase, pZ = 4.2. Lane 6: Trypsin inhibitor, pZ = 4.6. Lane 7: /3-lactoglobulin, pZ = 5.1. Lane 8: Carbonic anhydrase, PI= 5.4.
representing unreacted thioredoxin (A) (lane 1)and the appearance of two major bands, the upper of which predominated. These two bands reflect the expected high yield of monocarboxymethylated protein and a smaller yield of the doubly carboxymethylated product. In comparison, treatment of thioredoxin with CEG (B) (lane 2) yielded bands for unreacted thioredoxin and three major products, one of which focused at a position comparable to the first IAA reaction product. Further treatment of CEG-treated thioredoxin with IAA (C) (lane 3) led to the disappearance of the unreacted thioredoxin band but no other change. The MALDI mass spectrum (Figure 3) of thioredoxin (A) revealed a peak corresponding to the protein at m l z 11673. A minor peak corresponding to the covalent
mlz
14000
protein-sinapinic acid photoadduct was also observed. The mass spectrum of thioredoxin treated with CEG (B) yielded a peak at m l z 11673, corresponding to unreacted thioredoxin, and four other peaks at m l z 12 009,12 343, 12 675, and 13 018. Since the episulfonium ion alkylating species of CEG has a mass of 334, these four peaks correspond to thioredoxin alkylated by one, two, three, and four CEG episulfonium ion molecules, respectively. A secondary series of peaks at m l z 12 211,12 548, and 12 876 are matrix photoadducts of the alkylated protein species. Mass spectra of thioredoxin treated with CEG followed by treatment with IAA (C) showed no peak corresponding to unreacted thioredoxin but did exhibit a peak at m / z 11736 that corresponds to the addition of the carboxymethyl group from IAA (mass of 59). The m l z for the remaining peaks were unchanged, indicating that these protein forms did not react with IAA. The mass spectrum of thioredoxin treated with IAA (D) showed a major peak at m l z 11735, representing monocarboxymethylated thioredoxin. The smaller peak at m l z 11939 was the photoadduct. The small shoulder on the high-mass side of the m l z 11735 peak could be evidence of the formation of a small amount of dicarboxymethylated thioredoxin. Further experiments were performed with CEG prepared by the method of Humphreys et al. (161, to examine the extent of alkylation at various CEG/protein ratios (Figure 4). Because of the relative simplicity of this method of synthesis of CEG, it was of interest to examine the efficacy of alkylation with reagent prepared by this method. DTT (10-fold molar excess) was added to E. coli thioredoxin to ensure the protein was maintained in the reduced form. CEG was added in amounts of approximately 250-fold, 125-fold, 50-fold, 25-fold, and 5-fold excess over DTT concentrations. At 250-fold excess CEG, mass spectrometry analysis showed an intense signal at m l z 11673 representing unmodified thioredoxin, a sec-
mlz
0
C
11000
mlt
14000
Figure 3. Matrix-assisted laser desorptiodionization(MALDI) mass spectrometryof thioredoxinalkylated at pH 7.4. (A) Thioredoxin. (B) Thioredoxin treated with CEG. (C) Thioredoxin treated with CEG followed by IAA.(D)Thioredoxin treated with IAA.
662 Chem. Res Foxicol., Vol. 7, No. 5, 1994
Meyer et al.
I
Figure 4. hL4LDI mass spectrometry of thioredoxin alkylated at pH 7.4 with unpurified CEG at various CEG/protein ratios. 10fold molar excess of DTT was added to thioredoxin, and CEG was added in the following amounts over DTT concentrations. (A) 250-fold excess CEG. (B)125-fold excess CEG. (C) 50-fold excess CEG. (D)25-fold excess CEG. (E) 5-fold excess CEG. ond, almost equally intense signal suggesting extensive mono-alkylation, lesser amounts of di-alkylation, and low but still clearly visible tri-alkylation of thioredoxin (A). At 125-fold excess CEG, these three alkylated forms of the protein were still visible, although slightly smaller in intensity relative to the unreacted form of the protein (B). When CEG was reacted at 50-fold excess over DTT, the mono-alkylated form was the only major product, although a small amount of di-alkylated protein may be visible (C). At 25-fold and 5-fold molar excess, only the mono-alkylated form of thioredoxin was detectable (D and E). When thioredoxin was treated with CEG and/or IAA at other than physiologic pHs (Figure 51, different distributions of bands were seen on isoelectric focusing gels. At pH 4.4, no substantial derivatization of thioredoxin by CEG took place (B, lane 21, although IAA treatment resulted in a change of band migration, indicating carboxymethylation (B,lane 4). At pH 11.4, thioredoxin was seen to be extensively adducted by CEG, with four bands visible in addition to the band represent-
ing unalkylated thioredoxin (A, lane 2). Reaction with IAA at pH 11.4 (A, lane 4) also yielded more visible bands than reaction at pH 7.4 (A, lane 10). MALDI mass spectra of these samples (Figure 6 ) showed that some mono- and dialkylation occurred with CEG treatment of thioredoxin at pH 4.4 (A), but confirmed that the alkylation was much more extensive at pH 11.4 (D). Although several bands were seen for IAAtreated thioredoxin a t pH 11.4 on isoelectric focusing, mass spectrometry reveals that the dicarboxymethylated protein species was the only detectable product (F). Assay of the ability of these samples to catalyze the reduction of insulin demonstrated that thioredoxin treated with CEG at pH 7.4 and 11.4 exhibited a loss of activity, whereas thioredoxin treated with CEG at pH 4.4 retained enzymatic activity (Figure 7).
Discussion There are two possible routes for the metabolism of dihaloethanes: microsomal oxidation by the P450 enzyme
Thioredoxin Alkylation by a Glutathione Conjugate
j
1
2
3
4
5
6
1
2
3
4
5
7
6
8
7
9
1
0
8
Figure 5. Isoelectric focusing of thioredoxin alkylated at pH 11.4,pH 7.4,and pH 4.4.(Panel A) Lane 1: Thioredoxin, pH 11.4.Lane 2: Thioredoxin treated with CEG, pH 11.4.Lane 3: Thioredoxin treated with CEG followed by IAA,pH 11.4.Lane 4: Thioredoxin treated with IAA,pH 11.4.Lane 5: Glucose oxidase, pZ = 4.2.Lane 6: /%Lactoglobulin, pZ = 5.1.Lane 7: Thioredoxin, pH 7.4.Lane 8: Thioredoxin treated with CEG, pH 7.4.Lane 9: Thioredoxin treated with CEG followed by MA, pH 7.4.Lane 10: Thioredoxin treated with IAA,pH 7.4.(Panel B) Lane 1: Thioredoxin, pH 4.4.Lane 2: Thioredoxin treated with CEG, pH 4.4. Lane 3: Thioredoxin treated with CEG followed by IAA,pH 4.4.Lane 4: Thioredoxin treated with IAA, pH 4.4.Lane 5: Glucose oxidase, pZ = 4.2. Lane 6: Trypsin inhibitor, pZ = 4.6.Lane 7: /3-Lactoglobulin, p1 = 5.1.Lane 8: Carbonic anhydrase, pZ = 5.4.
system, or direct conjugation to glutathione catalyzed by glutathione S-transferase, resulting in the formation of an episulfonium ion (7). The product of DNA alkylation by dibromoethane in hepatocytes was determined to be the result of episulfonium ion formation (19). Alkylation by the dibromoethane-glutathione conjugate occurs primarily a t the W guanyl position (20, 16). van Bladeren et al. (21) found that oxidation and conjugation urinary metabolites of 1,2-dibromoethane occurred at a ratio of 4:l in the rat. It has been suggested that the oxidation pathway leads to products that bind preferentially to protein, whereas the episulfonium ion resulting from GSH conjugation reacts more readily with DNA (5). More recent findings indicate, however, that although the 2-bromoacetaldehyde product of microsomal oxidation has been reported to bind to protein thiols, the
Chem. Res. Toxicol., Vol. 7, No. 5, 1994 663 glutathione S-transferase-catalyzed conjugation of 1,2dihaloethanes with glutathione is probably more important in protein alkylation (6,22). The major macromolecular product of 1,2-dihaloethane exposure both in vivo or to hepatocytes in vitro is alkylated protein (6-8). The number and identity of the dihaloethane-modified proteins has not been established. Whether this protein alkylation is an important factor in the cytotoxicity of these compounds is not known. In order to begin to address these questions, the behavior of the synthetic conjugate S-(2-chloroethyl)glutathione toward a possible target protein, thioredoxin, was investigated. All three methods of analysis demonstrated that CEG treatment of thioredoxin a t physiological pH resulted in modification of the protein. Isoelectric focusing suggested the production of three different protein adducts by CEG alkylation. MALDI mass spectrometry established that these are the mono-, di-, and tri-alkylated forms of thioredoxin and additionally revealed the production of a tetra-alkylated product. When E. coli thioredoxin was treated with IAA a t pH 7.4, the predominant product was determined by MALDI mass spectrometry to be monocarboxymethylated protein. Kallis and Holmgren (23) reported that only one of the two sulfhydryl groups, Cys32, of E. coli thioredoxin is alkylated by IAA a t less than pH 8. The pK,'s of the two Cys residues of this protein have been reported as approximately 6.7 and 9.0 (23), and 7.1 and 7.9 (24). The MALDI mass spectrum of thioredoxin treated with CEG followed by IAA is essentially identical to the mass spectrum of thioredoxin reacted with CEG alone (Figure 3). The only protein species that reacted with IAA after treatment with CEG was the unalkylated component. This implies that, for the alkylated protein species, the IAA-susceptible Cys32 was not available for reaction with IAA because it was already covalently bound by CEG. The identity of the other sites alkylated by CEG have not been ascertained, but the most likely second alkylation site is C Y S ~ If ~ . so, this amino acid residue is more reactive with the CEG episulfonium ion than IAA a t pH 7.4. The alkylation of thioredoxin by CEG at third and fourth sites was unexpected, since Jean and Reed (9) found that Cys-containing dipeptides are a t least 10-fold more reactive to CEG than other functional groups. A possible candidate for a third CEG alkylation site a t this pH is the sole His residue. Histidine is the next most CEG-reactive amino acid (9). At pH 4.4, isoelectric focusing gave barely perceptible evidence of thioredoxin alkylation by CEG. However, MALDI mass spectrometry distinctly shows a detectable degree of mono- and di-alkylation. The enzymatic activity of thioredoxin treated with CEG a t this pH was unaffected by this level of alkylation. Interestingly, IAA reacted extensively with thioredoxin a t this pH, as was shown both by isoelectric focusing and loss of enzyme activity, although at pH 4.0 the Cys residues of E. coli thioredoxin are reported to be fully in the thiol (as opposed to thiolate anion) form. As expected, a t pH 11.4 alkylation of thioredoxin by CEG took place at several sites, although loss of enzyme activity was still not as great as that experienced after IAA treatment. MALDI mass spectrometry of the pH 11.4 samples did not detect any more than the four CEG-alkylated species (also. produced a t pH 7.4) and also detected unalkylated thioredoxin, which accounts for the residual enzymatic activity seen. IAA treatment of thioredoxin at pH 11.4 resulted in dicarboxymethylated protein, as determined by MALDI mass spectrometry. The loss of thioredoxin
664 Chem. Res. Toxicol., Vol. 7, No. 5, 1994
Meyer et al.
11000
14000
0
I1000
mlz
I4000
Figure 6. MALDI mass spectrometry of thioredoxin alkylated at pH 4.4 and 11.4.(A)Thioredoxin treated with CEG, pH 4.4.(B) Thioredoxin treated with CEG followed by IAA, pH 4.4.(C) Thioredoxin treated with IAA,pH 4.4.(D)Thioredoxin treated with CEG, pH 11.4. (E) Thioredoxin treated with CEG followed by IAA,pH 11.4.(F) Thioredoxin treated with IAA,pH 11.4. activity upon treatment with CEG was consistent with the amount of mono-alkylation visible on isoelectric focusing. Since the catalytic mechanism of thioredoxin is via the reversible oxidation of the active site dithiol, alkylation of a single cysteine would be sufficient to cause inactivation (23). CEG preparation by the method described by Humphreys et al. (16) was simpler and less time-consuming than the method of synthesis and purification described by Reed and Foureman (15) although the resulting product gave a lower total yield of alkylated thioredoxin. This was probably due to the high amounts of sodium salt present in the product. CEG prepared by either method, however, seems to produce the same relative amounts of mono-, di-, and tri-alkylated thioredoxin. The yield from the CEG prepared by the Humphreys et al. method (16)was too low to produce any observable tetraalkylated product. From both Figures 3 and 4, it is evident that monoalkylated thioredoxin was the predominant product when thioredoxin was reacted with CEG. As the ratio of CEG to protein was reduced, less and less
of the other products were seen, but the mono-adduct was observable even at very low concentrations of CEG. Although mammalian thioredoxins have only 26-27% homology with the E. coli thioredoxin amino acid sequence, the three-dimensional structures (derived from computer modeling) are similar, and the active site sequences of all thioredoxins are conserved (11). In addition, the structure of human thioredoxin in solution has been analyzed by NMR and shown to have a high degree of similarity with that of E. coli thioredoxin, particularly in the conformation of the active site (25). It is expected, therefore, that thioredoxins from mammalian species would also exhibit susceptibility to thioredoxin alkylation by CEG. We conclude from this work that the glutathionederived metabolite of 1,2-dihaloethane exposure, S42chloroethyl)glutathione, is capable of alkylating and inactivating the multifunctional ubiquitous protein thioredoxin. This alkylation leads to a loss of enzymatic activity. Our evidence strongly suggests that the primary site of alkylation is at the Cys~psite of the protein, which
Thioredoxin Alkylation by a Glutathione Conjugate a
a
4.4
7.4
Reaction
11.4 pH
TRX+CEG TRX+CEG+IAA TRX+IAA
Figure 7. Enzymatic activity of thioredoxin alkylated a t pH 4.4,pH 7.4,and pH 11.4. Insulin disulfide reduction activity is expressed as a percentage of maximum activity of a standard at pH 7.4. Data are means f SEM, and values with different letters are different ( p < 0.05).
would be predicted from studies with CEG and dipeptides, as well as the known pK,'s of the two Cys residues. Work is planned to utilize electrospray mass spectrometry to attempt to identify the other sites of alkylation by CEG. Currently, studies are underway to assess the effect of dihaloethane exposure on enzyme systems in rat hepatocytes.
Acknowledgment. We thank Marda Brown and Corwin Willard for laboratory assistance and Roxanne Bodine for word processing. This work was supported by Grant ES05612-02, Grant ES00210-26, and Grant ES00040 from the National Institute of Environmental Health Sciences, NIH.
References World Health Organization(1987)Environmental Health Criteria '62; 1,2-Dichloroethane, World Health Organization, Geneva. US. Department of Health and Human Services;Agency for Toxic Substances and Disease Registry (1992) Toxicologicalprofile for 1,2-dibromoethane, Public Health Service, U.S. Department of Health and Human Services, Washington, DC. Albano, E., Poli, G., Tomasi, A., Bini, A., Vannini, V., and Dianzani, M. U. (1984) Toxicity of 1,Zdibromoethane in isolated hepatocytes: role of lipid peroxidation. Chem.-Biol.Interact. 50, 255-265. Cmarik, J. L., Humphreys, W. G., Bruner, K. L., Lloyd, R. S., Tibbetts, C., and Guengerich, F. P. (1992)Mutation spectrum and sequence alkylation selectivity resulting from modification of bacteriophage M13mp 18 DNA with S-(2-chloroethyl)glutathione. J. Biol. Chem. 267,6672-6679. Sipes, I. G., Wiersma, D. A., and Armstrong, D. J. (1986) The role of glutathione in the toxicity of xenobiotic compounds: metabolic activation of 1,2-dibromoethane by glutathione. In Biological Reactive Intermediates III (Kocsis,J. J., Jollow, D. W., Witmer, C. M., Nelson, J. O., and Snyder, R., Eds.) pp 457-467, Plenum Press, New York and London.
Chem. Res. Toxicol., Vol. 7, No. 5, 1994 665 (6) Jean, P. A., and Reed, D. J. (1992) Utilization of glutathione during l,2-dihaloethane metabolism in rat hepatocytes. Chem. Res. Toxicol. 5, 386-391. (7) Hill, D. L., Shih, T.-W., Johnston, T. P., and Struck, R. F. (1978) Macromolecular binding and metabolism of the carcinogen 1,2dibromoethane. Cancer Res. 38, 2438-2442. (8) Igwe, 0. J., Que Hee, S. S., Wagner, W. D. (1986) Effect of disulfiram pretreatment on the tissue distribution, macromolecuin the rat. lar binding, and excretion of [U-1,2-14Cldichloroethane Drug Metab. Dispos. 14, 65-72. (9) Jean, P. A., and Reed, D. J. (1989) In vitro dipeptide, nucleoside, and glutathione alkylation by S-(2-~hloroethyl)-~-cysteine. Chem. Res. Toxicol. 2, 455-460. (10) Holmgren, A. (1985)Thioredoxin.Annu. Rev. Biochem. 54,237271. (11) Eklund, H., Gleason, F. K., and Holmgren, A. (1991) Structural and functional relations among thioredoxins of different species. Proteins: Struct., Funct., Genet. 11, 13-28. (12) Crawford, N. A., Yee, B. C., Droux, M., Carlson, D. E., and Buchanan, B. B. (1988) Ferridoxiflhioredoxin system. Methods Enzymol. 167, 415-427. (13) Decottignies, P., Schmitter, J.-M., Jacquot, J.-P., Dutka, S., Picaud, A., and Gadal, P. (1990) Purification, characterization, and complete amino acid sequence of a thioredoxin from a green alga, Chlamydomonas reinhardtii. Arch. Biochem. Biophys. 280, 112-121. (14) Fernando, M. R., Nanri, H., Yoshitake, S., Nagata-Kuno, K., and Minakami,S. (1992)Thioredoxin regenerates proteins inactivated by oxidativestress in endothelialcells. Eur. J. Biochem. 209,917922. (15) Reed, D. J., and Foureman, G. L. (1986) A comparison of the alkylating capabilitiesof the cysteinyl and glutathionylconjugates of 1,2-dichloroethane. In Biological Reactive Intermediates III (Kocsis, J. J., Jollow, D. J., Witmer, C. M., Nelson, J. O., and Snyder, R., Eds.) pp 469-475, Plenum Press, New York and London. (16) Humphreys, W. G., Kim, D.-H., Cmarik, J. L., Shimada, T., and Guengerich, F. P. (1990) Comparison of the DNA-alkylating properties and mutagenic responses of a series of S-(2-haloethyl)substituted cysteine and glutathione derivatives. Biochemistry 29, 10342-10350. (17) Jensen, 0. N., Barofsky, D. F., Young, M. C., von Hippel, P. H., Swenson, S., and Seifried, S. E. (1993)Direct observation of Wcrosslinked protein-nucleic acid complexes by Matrix-assisted Laser Desorption Ionization Mass Spectrometry. Rapid Commun. Mass Spectrom. 7, 496-501. (18) Holmgren,A. (1979) Thioredoxin catalyzes the reduction of insulin disulfides by dithiothreitol and dihydrolipoamide. J. Biol. Chem. 254,9627-9632. (19) Ozawa, N., and Guengerich, F. P. (1983) Evidence for formation of an S-[2-(N'-guanyl)ethyl]glutathione adduct in glutathionemediated binding of the carcinogen 1,2-dibromoethaneto DNA. Proc. Natl. Acad. Sci. U S A . 80, 5266-5270. (20) Foureman, G. L., and Reed, D. J. (1987) Formation of S-[2-(Wguany1)ethyll adducts by the postulated S-(2-chloroethyl)cysteinyl and S-(2-chloroethyl)glutathionylconjugates of 1,2-dichloroethane. Biochemistry 26,2028-2033. (21) van Bladeren, P. J., Breimer, D. D., van Huijgevoort, J. A. T. C. M., Vermeulen, N. P. E., and van der Gen, A. (1981)The metabolic formation of N-acetyl-S-2-hydroxyethylyethyl-L-cysteine from tetradeutero-lY2-dibromoethane.Relative importance of oxidation and glutathione conjugation in vivo. Biochem. Pharmacol. 30,24992502. (22) Khan, S., Sood, C., and O'Brien, P. J. (1993) Molecular mechanisms of dibromoalkane cytotoxicity in isolated rat hepatocytes. Biochem. Pharmacol. 45,439-447. (23) Kallis, G.-B., and Holmgren, A. (1980) Differential reactivity of the functional sulfhydryl groups of cysteine-32 and cysteine-35 present in the reduced form of thioredoxin from Escherichia coli. J. Biol. Chem. 255, 10261-10265. (24) Li, H., Hanson, C., Fuchs, J. A., Woodward, C., and Thomas, G. J., Jr. (1993) Determination of the pKa values of active-center cysteines, cysteines32 and -35, in Escherichia coli thioredoxin by Raman spectroscopy.Biochemistry 32, 5800-5808. (25) Forman-Kay,J. D., Clore, G. M., Wingfield, P. T., and Gronenborn, A. M. (1991) High-resolution three-dimensional structure of reduced recombinant human thioredoxin in solution.Biochemistry 30,2685-2698.
