- Email: [email protected]
Interactions of S-(2-haloethyl)-mercapturic acid analogs with plasmid DNA
TOXICOLOGY
AND
APPLIED
PHARMACOLOGY
Interactions
80,
of S-(2-Haloethyl)-mercapturic with Plasmid DNA
HELENA V. VADI,CHARLES Department
386-396 (1985)
of Biochemistry
Received
S. SCHASTEEN,' AND DONALDJ. REED'
and Biophysics,
November
Acid Analogs
Oregon
State University,
3. 1984; accepted
April
Corvallis,
Oregon
97331
23, 1985
Interactions of S(2-Haloethyl)-mercaptmic Acid Analogs with Plasmid DNA. VADI, H. V.. C. S., AND REED, D. J. (1985). Toxicol. Appl. Pharmacol. 80, 386-396. A series of related 9(2-haloethyl)+cysteine analogs were synthesized and their interaction with DNA was studied with plasmid pBR322. Both S-(2-chloroethyl)-L-cysteine (CEC) and S-(2-bromoethyl)L-cysteine (BrEC) rapidly induced relaxation of the supercoiled plasmid as determined by agarose gel electrophoresis and electron microscopy, whereas S-(2-fluoroethyl)-L-cysteine did not interact with DNA. The relaxation was most probably due to strand scission at alkylated labile sites in the DNA. When “S-labeled CEC or BrEC was used as the substrate, covalent binding of % to DNA was obtained; CEC displayed a somewhat higher binding than BrEC. No binding of “S was obtained with (Zhydroxyethyl~L-[%]cysteine, [“Slcysteine, or [“SJcystine, substrates which did not induce relaxation of the DNA. Esterification of the carboxyl group resulted in a somewhat lower rate of DNA strand scission, whereas N-acetylation prevented the cysteine analogs from inducing DNA strand breaks. S-(2-Chloroethyl)-glutathione (GSH) did not interact with DNA as determined by lack of effect on the superhelicity of DNA, a finding which is in agreement with the hypothesis that the primary amine groups of CEC or BrEC may participate in the formation of reactive intermediates which can interact with DNA. S-(2-Hydroxyethyl)-GSH and S(2-hydroxyethyl)-L-cysteine were unable to induce DNA strand breaks, Neutral denaturation of supercoiled pBR322 treated with the analogs revealed that compounds which were able to induce DNA strand breaks also interfered with denaturation of double-stranded circular DNA. No such interference was observed when double-stranded linear DNA (obtained by BamHI restriction digestion) was treated with the analogs prior to denaturation. These data indicate that a marked difference exists between S-(2-chloroethyl)-Lcysteine and ,I5(2-chloroethyI)-glutathione in their reaction with supercoiled plasmid DNA. Either a major difference exists in the reactivity of the corresponding episulfonium ions of these conjugates or a separate mechanism of alkylation based on a free or-amino of the cysteine conjugate is participating in DNA strand breakage and possible crosslinking. In vivo toxic effects of these S-(2-chloroethyl) conjugates are predicted to be distinctly different. o 1985 Academic SCHASTEEN,
Press. Inc.
Halogenated aliphatic two-carbon compounds such as the 1,2-dihaloethanes are presently widely used as industrial solvents, pesticides, fumigants, and gasoline additives (Fishbein, 1980). Both dichloro- and dibromoethane have been demonstrated to be mutagenic in
a variety of test systems (Fabricant and Chalmers, 1980; Hooper et al., 1980; Simmon, 1980) and produce tumors in experimental animals (National Cancer Institute, 1978; Weisburger, 1977). Dihaloethanes may be metabolized by the microsomal mixed function oxidase system, ’ National Institutes of Environmental Health Sciences and this activation seems to result mainly in irreversible binding to microsomal protein Postdoctoral Fellow 1 F32 ES-05192. ’ To whom requests for reprints should be sent. via 2-haloacetaldehydes (Guengerich et al.,
004 1-008x/85 $3.00 Copyright 0 1985 by Academic Press. Inc. All rights of reproduction in any form reserved.
386
S-(2-HALOETHYL)
ANALOGS
1980; Shih and Hill, 1981). In contrast, that activation of the dihaloethanes which occurs mainly via reaction with glutathione to yield S-(2-haloethyl)-glutathione conjugates (van Bladeren et al., 1979; Guengerich et al., 1980) may in turn be converted to ethylene (Livesey and Anders, 1979) or form highly reactive intermediates, episulfonium ions (thiiranium ions) (Smit et al., 1978). Glutathione-S-transferase-catalyzed activation of dihaloethanes results in covalent binding of the substrates mainly to cellular nucleophiles such as DNA and RNA (Guengerich et al., 1980; Sundheimer et al., 1982) and has been suggested to play an important role in the mutagenicity of dihaloethanes (Rannug et al., 1978; Guengerich et al., 1980). Overall activation, especially for 1,Zdibromoethane, may be greater via 2-bromoacetaldehyde with GSH and protein adduct formation being predominate (Shih and Hill, 198 1; Sundheimer et al., 1982). In order to better understand the mechanisms of toxicity of dihaloethanes, a series of S-(2-haloethyl)-cysteine analogs were recently synthesized and investigated for mutagenicity (van Bladeren et al., 1980, 198 1) and hydrolysis and alkylating activities (Schasteen and Reed, 1983). It was found that the mutagenicity of the analogs did not correlate with the halide order for leaving group ability (van Bladeren et al., 198 1) and that N-acetylS-(2-haloethyl)-cysteine methyl esters were more mutagenic than the parent compounds. Studies on the hydrolysis of several analogs at different pHs suggested that the amine moiety was responsible for an increased hydrolysis rate with alkaline conditions, and formation of 3-(thiomorpholine)-carboxylic acid was proposed as an alternative pathway to the generally accepted hydrolysis mechanism for 9(2-haloethyl)-L-cysteine analogs (Schasteen and Reed, 1983). This paper describes an extension of previous studies to include reactions with purified DNA. A series of related S-(2-haloethyl)-Lcysteines was synthesized and incubated with plasmid pBR322. The reactivity of the analogs
AND DNA ALKYLATION
387
with DNA was measured in terms of time dependent relaxation of the supercoiled form of DNA. The amount of covalent binding of certain analogs to DNA was determined. METHODS Chemicals. L-Cystine and N-acetyl-L-cysteine were purchased from Sigma (St. Louis, MO.). 1-Bromo-2fluoroethane was purchased from Columbia Organic Chemical Company (Columbia, SC.). Diazomethane was prepared from N-methyl-N-nitroso-N’nitroguanidine purchased from Aldrich (Milwaukee, Wise.). 2-Bromoethanol was also purchased from Aldrich Chemical Company. Sodium was obtained from Mallinckrodt (St. Louis, MO.). All other chemicals utilized in this study were ACS grade or the highest grade obtainable. Radiochemicals. [meth.vl-‘H]Thymidine (48 to 49 Ci/ mmol) was obtained from Amersham Int., Ltd. L[3SS]Cysteine (436.4 Ci/mmol) was obtained from New England Nuclear Corporation, Boston, Massachusetts. Synthesis of S-(2-haloethyl)-analogs. All compounds utilized in this investigation were prepared by the published procedures of Carson and Wong (1964) and van Bladeren et al. (1980) or du Vigneaud and Patterson (1936) and the modifications of McKinney et al. (1957). All products were considered pure by at least two thinlayer chromatographic systems, and the structures were consistent with the respective ‘H-NMR spectra obtained. Identification via mass spectrometry was obtained by either plasma desorption ionization mass spectrometry (Baldwin and McLafferty, 1973; Cotter. 1979) or fast atom bombardment (FAB) quadrupole mass spectrometry (Surman and Vickerman, 1981; Barber ef al., 1981). Synthesis of9(2-chloroethyl)-glutathione. To approximately 200 ml of liquid ammonia in a three-neck flask, which was fitted with a mechanical stirrer and sodalime tube, was added 7.68 g (0.025 mol) glutathione (GSH). The flask was kept at -35°C by cooling in a dry ice-ethanol bath. Sodium was introduced into the flask in small portions to the first permanent discharge of blue color. A total of 2.89 ml (0.035 mol) of I-bromo-2chloroethane was added in small portions over a I-hr period with the reaction temperature held at -35°C. After evaporation of the ammonia, the off-white solid was lyophilized. Thin-layer chromatography in n-butanol: acetic acid:HzO (25:4:10) Avicel of the solid residue showed a ninhydrin positive spot corresponding to Rf = 0.44. A sample placed in distilled water at room temperature for 25 hr showed a spot corresponding to synthetic .S-(2-hydroxyethyl)glutathione (R, = 0.21) prepared in a similar manner as above except with 2bromoethanol instead of I-bromo-2-chloroethane. MS(FAB): 396 (M + Na)‘, 2.0% 394 (M + Na)+, 6.5%. 330 (GSH + Na)+, 0.6%, 308 (GSH + H)+ 208, 11.4%.
388
VADI, SCHASTEEN.
Preparation and isolation of plasmid DNA. Plasmid pBR322 (Sutcliffe, 1978) was grown and amplified in Escherichia coli strain RRI as described (Norgard et al., 1979; Vadi and Reed, 1983). Radioactive plasmid was obtained by adding [methyl-SH]tbymidine to the medium (1 to 2 mCi/liter) just prior to the amplification step, which was initiated by addition of chloramphenicol (Clewell, 1972). Following 10 hr of amplification, the cells were sedimented by centrifugation, incubated with lysozyme, and treated with a salt detergent solution. Chromosomal DNA and proteins were removed by centrifugation, and remaining nucleic acids were precipitated with a mixture of sodium trichloroacetate:ethanol. The pellet of nucleic acids was resuspended in TENbuffer (10 mM Tris-HCl, I mM EDTA, 10 mM NaCl, pH 7.5) and treated with RNase A (Vadi and Reed, 1983) before reprecipitation with NaTCA:ethanol. The final product, consisting of pBR322, was routinely purified on a cesium chloride gradient containing EtdBr (ethidium bromide) (Radloff et al., 1967). All steps involving EtdBr were performed under red light, thus ensuring a high yield of supercoiled plasmid DNA. The isolation procedure has been described in detail elsewhere (Vadi and Reed, 1983). Agarose gel electrophoresis. The various forms of pBR322 were separated by agarose gel electrophoresis with 0.7% agarose in a horizontal slab gel system (McDonnell et al., 1977). DNA bands were visualized by staining the gel in 5 Irglml EtdBr solution and illuminating from below with ultraviolet light. The amount of DNA in each band was quantified by liquid scintillation counting, and the proportion of supercoiled DNA was calculated as the percentage of total DNA in that channel (Vadi and Reed, 1983). Incubation conditions. pBR322 was incubated with 2haloethylcysteine analogs or glutathione conjugates at 37°C for the times indicated. The reaction mixtures contained 0.1 M cacodylate, pBR322 (0.2 rdgl), 7 mM Tris, pH 7.5, 0.7 mM EDTA, 7 mM NaCI, and 10 mM concentration of the analog studied, unless otherwise stated. S-(2-Haloethyl)-L-cysteine analogs were added in DMSO. The final concentration of DMSO as 10% in all samples including control incubations. Glutathione conjugates were added in TEN-buffer. S-(2-Haloethyl)-glutathione was checked by thin-layer chromatography for purity [<5% of s-(2-hydroxy-ethyl)-glutathione as the major contaminant] prior to each incubation. Reactions were stopped by spin dialysis through Sephacryl S-200 equilibrated with TEN-buffer (Neal and Florini, 1973; Vadi and Reed, 1983). DNA was precipitated by addition of l/IO vol 3 M sodium acetate and 5 vol of ethanol, and samples were kept at -50°C for at least 1.5 hr. Pelleted DNA was resuspended in TENbuffer and analyzed for supercoiled plasmid by agarose gel electrophoresis as described above. Denaturation ojDNA. In some experiments the DNA was denatured by heating at neutral pH (Vadi and Reed, 1983). Following spin dialysis and precipitation, the
AND REED
DNA was resuspended in 5 mM EDTA and kept at 4°C for 4 to 6 hr. Sucrose was added to a final concentration of IO%, and the samples were heated in a boiling waterbath for 70 sec. The denatured samples were cooled in ice and were immediately analyzed by gel electrophoresis. This procedure has been demonstrated to cause complete denaturation of double stranded linear and nicked circular forms of non-alkylated pBR322 (Vadi and Reed, 1983). Quantification of extent of alkylation. [3H]pBR322 was incubated with (2-chloroethyl)-t.-[3SS]cysteine or (2bromoethyl)-L-[35S]cysteine as described above. Reactions were stopped by precipitation of the DNA. The pellets were washed with a mixture of 15% TEN-buffer, 1.5% 3 M sodium acetate, and 83.5% ethanol, resuspended in TEN-buffer, and reprecipitated. This cycle of precipitating, washing, and resuspending the DNA was repeated at least three times or until the ratio 35S:3Hbecame constant. By knowing the specific activities of the substrates and the DNA, the amount of alkylation per unit DNA was calculated. Electron microscopy. Certain DNA bands were isolated from agarose gels as described (Vadi and Reed. 1983) and prepared for electron microscopy by an aqueous staining technique (Kleinschmidt, 1968). The grids were examined in a Zeiss 10 electron microscope.
RESULTS [3H]pBr322 was incubated with 2-haloethyl-L-cysteine analogs, purified by spin dialysis, and analyzed by agarose gel electrophoresis. Treatment with S-(2-chloroethyl)L-cysteine (CEC) or 9(2-bromoethyl)-L-cysteine (BrEC) resulted in a time-dependent relaxation of the supercoiled form of DNA as determined by a decrease in the amount of DNA migrating as band d in agarose gels (Fig. 1). Band b, which in non-treated samples consists of relaxed circular DNA, disappeared by 120 min (Fig. l), and there was an increase in the proportion of DNA migrating as band c. We have previously shown that in nontreated samples of DNA, the double-stranded linear form of DNA migrates as band c, whereas when DNA has been alkylated with nitrosoureas or dimethyl sulfate prior to gel electrophoresis, band c consists of relaxed circular DNA (Vadi and Reed, 1983). Similar results were obtained when plasmid DNA was treated with 4(2-chloroethyl)-L-cysteine or S-(2-bromoethyl)-L-cysteine (results not
s-(2-HALOETHYL)
Time (min)
0
30
60
ANALOGS
120
FIG. 1. Agarose gel electrophoresis of pBR322 treated with S-(2-bromoethyl)-L-cysteine. Incubations were performed as described under Methods for the times indicated. Bands are designated: (a) upper band; (b) relaxed circle: (c) doubte stranded linear; (d) supercoiied DNA.
shown). Electron microscopy of band c purified from agarose gels of S-(Zchloroethyl)L-cysteine- or S-(2-bromoethyl)-L-cysteinetreated DNA showed that this band consisted of relaxed circles (Fig. 2). S(2-Fluoroethyl)L-cysteine did not interact with plasmid DNA since no change in electrophoretic pattern of the DNA could be detected after treatment with this analog (results not shown). Figures 3 and 4 show the effects of some of the haloethyl+cysteine analogs on the superhelicity of DNA. S-(2-Chloroethyl)-Lcysteine induced a rapid relaxation of plasmid DNA (Fig. 3). Blocking the carboxyl group of CEC by esterification reduced the rate of relaxation to some extent, whereas acetylation of the free amine of CEC dramatically decreased interaction with DNA. Thus, N-acetyl-S-(2-chloroethyl)-L-cysteine (CENAcC) was able to lower only slightly the proportion of superhelical DNA present. The methyl ester of this analog (CENAc M.E.) did not interact with DNA. Similar results were obtained with the S (2-bromoethyl)-L-cysteine analogs (Fig. 4).
AND DNA ALKYLATION
389
Once again it was observed that methylation of the carboxyl group slightly reduced the rate of relaxation, whereas acetylation of the primary amine prevented the analogs from interacting with DNA. Bromo-containing analogs were always less reactive with DNA than their chloro-containing counterparts. As expected from the above results with CENAcC, S-(2-chloroethyl)-glutathione did not interact with plasmid DNA as determined by lack of effect on the superhelicity of DNA (Table 1). s-(2-HydroxyethyB-glutathione and 9(2-hydroxyethyl)-L-cysteine also failed to interact with DNA (Table 1). Neutral denaturation of non-alkylated pBR322 resulted in complete loss of doublestranded linear DNA (Fig. 5, channel 1, band c) and formation of single-stranded linear DNA (Fig. 5, channel 2, band e). Electrophoretie patterns identical to control patterns were obtained when pBR322 was treated with N-acetyl-S-(2-chloroethyl)-L-cysteine methyl ester (CENAc M.E.), N-acetyl-S-(2bromoethyl)-L-cysteine (BrENAcC), its methyl ester (BrENAcC M.E.), S(Zchloroethyl)-glutathione, s-(2-hydroxyethyl)-glutathione, or S(2-hydroxyethyl)-L-cysteine followed by neutral denaturation (Fig. 5, channels 6, 9, and 10, or results not shown). These analogs did not induce a relaxation of supercoiled plasmid (Figs. 3 and 4; Table 1) and did not interfere with denaturation. In contrast, the 2-haloethyi+cysteine analogs that were able to induce relaxation of supercoiled plasmid (Figs. 3 and 4) were also able to interfere with denaturation and prevent the double-stranded DNA species in band c from denaturing (Fig. 5, channels 3 to 5, 7 and 8). In these experiments the plasmid was treated for 15 min with 5 mM concentration of the analogs prior to denaturation. When higher concentrations and/ or longer incubations were employed, denaturation resulted in varying degrees of fragmentation of DNA treated with DNA-interacting analogs CEC, CEC M.E., CENAcC, BrEC, or BrEC M.E. (results not shown). pBR322 that was digested with BamHl
390
VADI, SCHASTEEN,
AND REED
FIG. 2. Electron microscopy of S-(2-chloroethyl)-L-cysteine treated pBR322. Plasmid DNA (0.35 rg/ ~1) was incubated with 10 mM S-(2-chloroethyl)+cysteine at 37°C for 2 hr. Gel eiectrophoresis was performed, band c was purified from the gel as described (Vadi and Reed, 1983) and stained (Kleinschmidt, 1968). Magnification X24.480. The presence of only relaxed circular and not linear or supercoiled DNA is based on previous electron microscopic studies with pBR322 (Vadi and Reed, 1983).
S-(2-HALOETHYL)
D
ANALOGS
CENAcC
ME.
CENAcC
391
AND DNA ALKYLATION
(Fig. 7). Approximately 60 to 70% (CEC) or 50% (BrEC) of the radioactivity that precipitated with DNA comigrated with DNA in agarose gels, and when treated DNA was denatured by the neutral sucrose technique prior to gel electrophoresis, 30 to 35% of the radiolabel still corn&rated with DNA. These findings suggest that the radiolabel was covalently bound to DNA rather than associated by either electrostatic interactions or intercalation. The degree of binding was concentration dependent and was higher for CEC than for BrEC at 10 mM substrate concentration (Fig. 7). 35S labeled cysteine or cystine failed to show any detectable binding of
ME BrENAcC
ME
1
loo -BrBrENAcC
CEC
Time,
min
FIG. 3. Effect of S-(2-chloroethyl)-L-cysteine analogs on the superhelicity of pBR322. Incubations were performed as described under Methods for the times indicated in the graph. Values were normalized to 100% superhelicity for non-treated DNA and linear regression was used for data analysis. CEC, S(Z-chloroethyl)-L-cysteine; CEC M.E., S-(2-chloroethyl)-L-cysteine methyl ester; CENAcC, N-acetyl-S-(2-chloroethyl)-Lcysteine; CENAcC M.E., N-acetyl-s-(2-chloroethyl)-Lcysteine methyl ester.
40BrEC
% B yj
x\
30-
ME
BrEC
5 cn’
z 20
restriction enzyme (Vadi and Reed, 1983) prior to CEC treatment and denaturation gave no indication of CEC-induced DNA interstrand crosslink formation in linear DNA (Fig. 6, channels 5 and 6). Thus, linear DNA did not display the same alteration by CEC treatment as supercoiled plasmid which induced to relax and also showed interference of denaturation (Fig. 5, Channels 3 to 5, 7 and 8). Treatment of pBR322 with (2-chloroethyl)L-[35S]-cysteine or (2-bromoethyl)-L-[35S]cysteine resulted in binding of 35Slabel to DNA
t
100 0
30
60
120 Time,
mm
FIG. 4. Effect of s-(2-bromoethyl)-L-cysteine analogs on the superhelicity of pBR322. See legend to Fig. 3 for details. BrEC, 4(2-bromoethyl)-L-cysteine; BrEC M.E., .S-(2-bromoethyl)-L-cysteine methyl ester; BrENAcC, Nacetyl-S-(2-bromoethyl)-L-cysteine; BrENAcC M.E., Nacetyl-S-(2-bromoethyI)-L-cysteine methyl ester.
392
VADI, SCHASTEEN, TABLE
EFFECT OFS(2-CHLOROETHYL)-GSH,
AND REED 1
&(&HYDROXYETHYL)-GSH, ONTHESUPERHELICITYOF
AND 9('i'-HYDROXYETHYL)-L-CYSTEINE
vBR322
% Supercoiled DNA Time (mm)
s-(2-0HEt)GSH
S-(2-CIEt)GSH
0 30 60 120
78.7 76.5 77.0 75.8
f + + k
3.2 4.4 4.0 3.9
78.7 77.9 77.0 78.2
f + + +
s-(2-0HEt)cys
3.2 4.3 5.9 3.3
78.7 77.5 75.5 74.2
rt k * t
3.2 2.1 3.9 1.3
Note. Incubations contained 10 tIIM substrate and were performed as described under Methods for the times indicated. N values are X + SD of three experiments.
radiolabel to DNA (results not shown). (2Hydroxyethyl)-L-[35S]cysteine coprecipitated with DNA to a certain extent, but the inter-
1
2
3
4
5
action with DNA decreased with incubation time, and the radiolabel did not corn&ate with DNA in agarose gels.
b
7
8
9
10
FIG. 5. Agarose gel electrophoresis of pBR322 treated with S(2-haloethyl)-L-cysteine analogs and denatured by a neutral sucrose denaturation technique. pBR322 was incubated with 5 mM concentration of the analogs for I5 min. The experimental conditions were as described under Methods. Channels I and 2 represent control pBR322 before and after denaturation, respectively. Channels 3 to IO represent pBR322 which was treated with haloethyl cysteine analogs as indicated in the figure prior to neutral sucrose denaturation. See legends to Figs. 3 and 4 for explanation of abbreviations.
.S-(Z-HALOETHYL)
Barn
ANALOGS
Hl
AND DNA ALKYLATION
393
CONTROL
FIG. 6. Denaturation of BumHI-digested pBR322 treated with S-(Z-chloroethyI)-L-cysteine. Supercoiled and BamHl digested pBR322 were treated with 3 mM CEC prior to denaturation and agarose gel electrophoresis. Channels 1 and 2: Control BamH 1-digested pBR322; native and denatured, respectively. Channels 3 and 4: BarnH l-digested pBR322 treated with CEC for 15 and 30 min. respectively. Channels 5 and 6: Same samples as in channels 3 and 4, after denaturation. Channels 7 and 8: Control pBR322; native and denatured, respectively. Channels 9 and IO: pBR322 treated with CEC for 15 and 30 min, respectively. Channels I1 and 12: Same samples as in channels 9 and 10 after denaturation.
DISCUSSION In this study we investigated the interactions between a series of S-(2-haIoethyl)-Lcysteine analogs and plasmid DNA. When incubated with supercoiled pBR322, S-(2chloroethyl)-L-cysteine and 9(2-bromoethyl)L-cysteine induced relaxation of the DNA as determined by agarose gel electrophoresis and electron microscopy (Figs. 1 and 2). We suggest that the relaxation is due to strand scission at alkylated labile sites in the DNA. When 35S labeled 9(2-chloroethyl)+cysteine or s-(2-bromoethyl)-L-cysteine was used as the substrate, covalent binding of 35Sto DNA was obtained; S(2-chloroethyl)-L-cysteine resulted in a somewhat higher 35Sbinding than s-(2-bromoethyl)-L-cysteine. This difference could be a reflection of the faster hydroIysis
rate previously noted for the bromo analogs compared to the chloro analogs (Schasteen and Reed, 1983). At least 30 to 35% of this binding was resistant to neutral denaturation of the DNA. No binding of 35S to DNA was obtained with (2-hydroxyethyl)-L-[35S]cysteine, [35S]cysteine, or [35S]cystine substrates which did not induce relaxation of the DNA. As pBR322 was treated with increasing concentrations of DNA-interacting analogs, neutral denaturation resulted in increased fragmentation of the DNA. This finding supports our suggestion that certain analogs were able to create potentially labile sites in the DNA, thus generating sites for spontaneous strand scission. Findings by other workers that support this contention include in vivo studies in which 1,2-dibromoethane (DBE) was administered to mice and a dose-depen-
394
VADI, SCHASTEEN,
AND REED
(2-bromoethyl)GSH (Ozawa and Guengerich, 1983). However, since we now show that CEC relaxed pBR322 rapidly and 5’-(2-chloroethyl)GSH (CEG) essentially had no effect, then metabolism of CEG to CEC could be important relative to the total DNA damage resulting from exposure to 1,2-dichloroethane (DCE) and DBE. Our finding that N-acetylation of the analogs prevented them from interacting with DNA is also in contrast to a report by van Bladeren et al. (198 1) who recorded mutagenic activity with N-acetyl-S-(2-haloethyl)TIME,min L-cysteine methyl esters. A possible explanaFIG. 7. Alkylation of [3H]pBR322 with [35S]CEC or tion for this finding may be the presence of [35S]BrE~. Incubations were as described under Methods for the times indicated in the graph. The amounts of bacterial deacetylases in their system whereas alkylation were calculated from the amount of “S asso- they were absent in the experimental system ciated with the DNA after three precipitations. used in the present studies. The finding that DNA-interacting analogs prevent neutral denaturation of doubledent increase in DNA single strand breaks stranded circular DNA, although no interferand alkali labile sites was observed (White et ence was obtained in the denaturation of al., 1981; Storer and Conolly, 1983). analog-treated linear DNA, can be explained Of particular interest was the finding that in two ways: (1) some analogs do, in fact, S-(2-chloroethyl)-glutathione did not affect interact with both strands of circular DNA, the superhelicity of pBR322 DNA (Table 1). but no such interaction takes place in linear This finding is in contrast with the data DNA; and (2) at the initial stages of alkylobtained with S-(2-chloroethyl)-L-cysteine but ation, DNA becomes relaxed. but no strand not with N-acetyl-S-(2-chloroethyl)-L-cysteine. scission takes place. Relaxed circular DNA Blocking of the amine moiety is known to with both strands covalently closed would affect the hydrolysis enhancement noted with not be denatured to single-stranded DNA alkaline conditions for S-(2-chloroethyl)-Lunder the conditions used. cysteine and eliminates the formation of 3(thiomorpholine)-carboxylic acid (Schasteen ACKNOWLEDGMENTS and Reed, 1983). Incubation of calf thymus DNA with DBE This research was supported by NIEHS Grants ESand GSH in the presence of GSH S-transfer00040 and 1 F32 ES-05192. The authors acknowledge ase results in a product which yields N7the use of the Prophet computer system in molecular ethylguanine when reductively desulfurized modeling and data manipulation. The FAB mass spectra (Ozawa and Guengerich, 1983). Since the analyses by Don Griffin are greatly appreciated. parent DNA adduct contains radiolabel from GSH, the authors assigned it as being 9[2REFERENCES (N7-guanyl)ethyl]GSH. They also concluded that their findings support the hypothesis of BALDWIN, M. A., AND MCLAFFERTY, S. W. (1973). the major interaction of DBE with DNA Direct CI of relatively involatile samples. Org. Mass involving covalent modification by a preSpectrom. 7, 1353-l 356. formed complex of DBE and GSH, i.e., S- BARBER, M., BORDOLI, R. S., SEDGWICK, R. D., AND
S-(2-HALOETHYL)
ANALOGS
TYLER, A. N. (1981). Fast atom bombardment of solids (FAB): A new ion source for mass spectrometry. J. C. S. Chem. Commun. 325-327. CARSON, J. F., AND WONG, F. (1964). The synthesis of L- 1,4-thiazane-3-carboxylic acid 1-oxide. J. Org. Chem. 29, 2203-2205. CLEWELL, D. B. (1972). Nature of ColEl plasmid replication in the presence of chloramphenicol. J. Bucteriol. 110, 667-676.
COTTER, R. J. (1979). Probe for direct solid sample exposure to reagent gas in Cl mass spectrometry. Anal. Chem. 51, 317-318. Du VIGNEAUD, V.. AND PATTERSON, W. I. (1936). The synthesis of djenkolic acid. J. Biol. Chem. 114, 533538. FABRICANT, J. D., AND CHALMERS. J. H., JR. (1980). Evidence of the mutagenicity of ethylene dichloride and structurally related compounds. In Banbury Reporf 5; EthJplene Dichloride: A Poientiai Health Risk? (8. Ames, P. Infante, and R. Reitz. eds.). Cold Spring Harbor. FISHBEIN,L. ( 1980). Potential carcinogenic and mutagenic industrial chemicals. 1. Alkylating agents. J. Toxicol. Environ. Health 6, 1133-I 177. GUENGERICH, F. P., CRAWFORD. W. M., DOMORADZKI, J. Y.. MACDONALD, T. L., AND WATANABE. P. G. (1980). In vlfro activation of 1,2-dichloroethane by microsomal and cytosolic enzymes. Toxicol. Appl. Pharmacol. 55, 303-3 17. HOOPER. K., GOLD, L. S., AND AMES, B. N. (1980). The carcinogenic potency of ethylene dichloride in two animal bioassays: A comparison of inhalation and gavage studies. In Banbury Report 5: Ethylene Dichloride: A Potential Health Risk? (B. Ames, P. Infante, and R. Reitz. eds.), pp. 65-81. Cold Spring Harbor Laboratory. KLEINSCHMIDT, A. K. (1968). Monolayer techniques in electron microscopy of nucleic acid molecules. In Methods in En~ymolugv (L. Grossman and K. Moldave. eds.), Vol. 12. pp. 36 l-377. Academic Press, New York. LIVESEY, J. C.. AND ANDERS, M. W. (1979). In vitro metabolism of 1,2-dihaloethanes to ethylene. Drug Metab. Dis. 7, 199-203. MCDONNELL, M. W.. SIMON, M. N., AND STUDIER, F. W. (1977). Analysis of restriction fragments of T7 DNA and determination of molecular weights by electrophoresis in neutral and alkaline gels. J. Mol. Biol. 110, 119-146. MCKINNEY, L. L.. WEAKLEY, F. B., ELDRIDGE, A. C., CAMPBELL, R. E., COWAN, J. C., PICKEN, J. C., JR., AND BEISTER, H. E. (1957). S-(Dichlorovinyl)-L-cysteine: An agent causing fatal aplastic anemia in calves. J. Amer. Chem. Sot. Commun. 79, 3932-3933. National Cancer Institute. (1978). Technical Background Irzformation Report #55 on Carcinogenesis Bioassay
AND DNA ALKYLATION
395
of 1,2-Dichloroethane (EDC). U.S. DHEW Publication number (NIH) 78- 136 1. Govt. Printing Office, Washington, D.C. NEAL, M. W., AND FLORINI, J. R. (1973). A rapid method for desalting small volumes of solution. Anal. Biochem. 55, 328-330. NORGARD, M. V., EMIGHOLTZ. K., AND MONAHAN, J. J. (1979). Increased amplification of pBR322 plasmid deoxyribonucleic acid in Escherichia coli K- 12 strains RR1 and X 1776 grown in the presence of high concentrations of nucleoside. J. Bacterial 138, 270-272. OZAWA, N., AND GUENGERICH, F. P. (1983). Evidence for formation of an S-[2-(N-guanyl)ethyl]glutathione adduct in glutathione-mediated binding of the carcinogen 1,2-dibromoethane to DNA. Proc. Natl. Acad. Sci. USA 80, 5266-5270. RADLOFF, R.. BAUER, W., AND VINCCRAD, J. (1967). A dye-buoyant density method for the detection and isolation of closed circular duplex DNA: The closed circular DNA in HeLa Cells. Proc. Nat/. Acad. Sci. USA 57, 1514-1521. RANNUG, U., SUNDVALL, A., AND RAMEL, C. (1978). The mutagenic effect of 1,2-dichloroethane on Salmonella typhimurium. I. Activation through conjugation with glutathione in vitro. Chem. Biol. Interact. 20, l-16. SCHASTEEN.C. S., AND REED, D. J. (1983). The hydrolysis and alkylation activities of S-(2-haloethyl)-L-cysteine analogs-Evidence for extended half-life. To,xico/. Appl. Pharmacol. 70, 423-432. SHIH, T.-W., AND HILL, D. L. (1981). Metabolic activation of 1.Zdibromoethane by glutathione transferase and by microsomal mixed function oxidase: Further evidence for formation of two reactive metabolites. Res. Commun. Chem. Pathol. Pharmacol. 33, 449461. SIMMON, V. F. (1980). Review of nonbacterial tests of the genotoxic activity of ethylene dichloride. In Banbuty Report 5. ethylene Dichloride: A Potential Health Risk? (B. Ames, P. Infante, and R. Reitz, eds.), pp. 97-103. Cold Spring Harbor Laboratory. SMIT, W. A., ZEFIROV, N. S., BODRIKOV, I. V., AND KIRMER. M. Z. (1978). Episulfonium ions: Myth and reality. Acct. Chem. Res. 12, 282-288. STORER, R. D., AND CONOLLY, R. B. (1983). Comparative in viva genotoxicity and acute hepatotoxicity of three 1,2-dihaloethanes. Curcinogenesis 4( 1 l), 14911494. SUNDHEIMER, D. W.. WHITE, R. D., BRENDEL, K., AND SIPES,I. G. (1982). The bioactivation of I ,2-dibromoethane in rat hepatocytes: covalent binding to nucleic acids. Carcinogenesis 3, 1129-I 133. SURMAN, D. .l., AND VICKERMAN, J. C. (i981). Fast atom bombardment quadrupole mass spectrometry. J. C. S. Chem. Commun. 324-325. SUTCLIF’FE,J. G. (1978). pBR322 restriction map derived
VADI, SCHASTEEN,
396
from the DNA sequence: Accurate DNA size markers up to 436 1 nucleotide pairs long. Nucleic Acid Rex 5, 2121-2728. VADI, H. V., AND REED, D. J. (1983). Effect of 2chloroethylnitrosoureas on plasmid DNA, including formation of strand breaks and interstrand crosslinks. Chem.
Biol.
Interact.
46, 67-84.
VAN BLADEREN, P. J., BREIMER, D. D., ROTTEVEELSMIJS, G. M. T., DE KNIJFF, P., MOHN, G. R., VAN MEETEREN-WALCHLI, B., BUIJS, W., AND VAN DER GEN, A. (1981). The relation between the structure of vicinal dihalogen compounds and their mutagenic activation via conjugation to glutathione. Carcinogenesis 2(6), 499-505.
VAN BLADEREN, P. J., BUIJS, W.. BREIMER, D. C., AND
AND REED
VAN DER GEN, A. (1980). The synthesis of mercapturic acids and their esters. Eur. J. Med. Chem. 15, 49% 497. VAN BLADEREN, P. J., VAN DER GEN, A., BREIMER, D. D., AND MOHN, G. R. (1979). Stereoselective activation of vicinal dihalogen compounds to mutagens by glutathione conjugation. Biochem. Pharmacol. 28, 2521-2524. WEISBURGER, E. K. (1977). Carcinogenicity studies on halogenated hydrocarbons. Environ. Health Perspect. 21, 2804-28 14. WHITE, R. D., SIPES, I. G., GANDOLFI, A. J., AND BOWDEN, G. T. (198 1). Characterization of the hepatic DNA damage caused by 1,2-dibromoethane using the alkaline elution technique. Carcinogenesis 2, 839-844.
AND
APPLIED
PHARMACOLOGY
Interactions
80,
of S-(2-Haloethyl)-mercapturic with Plasmid DNA
HELENA V. VADI,CHARLES Department
386-396 (1985)
of Biochemistry
Received
S. SCHASTEEN,' AND DONALDJ. REED'
and Biophysics,
November
Acid Analogs
Oregon
State University,
3. 1984; accepted
April
Corvallis,
Oregon
97331
23, 1985
Interactions of S(2-Haloethyl)-mercaptmic Acid Analogs with Plasmid DNA. VADI, H. V.. C. S., AND REED, D. J. (1985). Toxicol. Appl. Pharmacol. 80, 386-396. A series of related 9(2-haloethyl)+cysteine analogs were synthesized and their interaction with DNA was studied with plasmid pBR322. Both S-(2-chloroethyl)-L-cysteine (CEC) and S-(2-bromoethyl)L-cysteine (BrEC) rapidly induced relaxation of the supercoiled plasmid as determined by agarose gel electrophoresis and electron microscopy, whereas S-(2-fluoroethyl)-L-cysteine did not interact with DNA. The relaxation was most probably due to strand scission at alkylated labile sites in the DNA. When “S-labeled CEC or BrEC was used as the substrate, covalent binding of % to DNA was obtained; CEC displayed a somewhat higher binding than BrEC. No binding of “S was obtained with (Zhydroxyethyl~L-[%]cysteine, [“Slcysteine, or [“SJcystine, substrates which did not induce relaxation of the DNA. Esterification of the carboxyl group resulted in a somewhat lower rate of DNA strand scission, whereas N-acetylation prevented the cysteine analogs from inducing DNA strand breaks. S-(2-Chloroethyl)-glutathione (GSH) did not interact with DNA as determined by lack of effect on the superhelicity of DNA, a finding which is in agreement with the hypothesis that the primary amine groups of CEC or BrEC may participate in the formation of reactive intermediates which can interact with DNA. S-(2-Hydroxyethyl)-GSH and S(2-hydroxyethyl)-L-cysteine were unable to induce DNA strand breaks, Neutral denaturation of supercoiled pBR322 treated with the analogs revealed that compounds which were able to induce DNA strand breaks also interfered with denaturation of double-stranded circular DNA. No such interference was observed when double-stranded linear DNA (obtained by BamHI restriction digestion) was treated with the analogs prior to denaturation. These data indicate that a marked difference exists between S-(2-chloroethyl)-Lcysteine and ,I5(2-chloroethyI)-glutathione in their reaction with supercoiled plasmid DNA. Either a major difference exists in the reactivity of the corresponding episulfonium ions of these conjugates or a separate mechanism of alkylation based on a free or-amino of the cysteine conjugate is participating in DNA strand breakage and possible crosslinking. In vivo toxic effects of these S-(2-chloroethyl) conjugates are predicted to be distinctly different. o 1985 Academic SCHASTEEN,
Press. Inc.
Halogenated aliphatic two-carbon compounds such as the 1,2-dihaloethanes are presently widely used as industrial solvents, pesticides, fumigants, and gasoline additives (Fishbein, 1980). Both dichloro- and dibromoethane have been demonstrated to be mutagenic in
a variety of test systems (Fabricant and Chalmers, 1980; Hooper et al., 1980; Simmon, 1980) and produce tumors in experimental animals (National Cancer Institute, 1978; Weisburger, 1977). Dihaloethanes may be metabolized by the microsomal mixed function oxidase system, ’ National Institutes of Environmental Health Sciences and this activation seems to result mainly in irreversible binding to microsomal protein Postdoctoral Fellow 1 F32 ES-05192. ’ To whom requests for reprints should be sent. via 2-haloacetaldehydes (Guengerich et al.,
004 1-008x/85 $3.00 Copyright 0 1985 by Academic Press. Inc. All rights of reproduction in any form reserved.
386
S-(2-HALOETHYL)
ANALOGS
1980; Shih and Hill, 1981). In contrast, that activation of the dihaloethanes which occurs mainly via reaction with glutathione to yield S-(2-haloethyl)-glutathione conjugates (van Bladeren et al., 1979; Guengerich et al., 1980) may in turn be converted to ethylene (Livesey and Anders, 1979) or form highly reactive intermediates, episulfonium ions (thiiranium ions) (Smit et al., 1978). Glutathione-S-transferase-catalyzed activation of dihaloethanes results in covalent binding of the substrates mainly to cellular nucleophiles such as DNA and RNA (Guengerich et al., 1980; Sundheimer et al., 1982) and has been suggested to play an important role in the mutagenicity of dihaloethanes (Rannug et al., 1978; Guengerich et al., 1980). Overall activation, especially for 1,Zdibromoethane, may be greater via 2-bromoacetaldehyde with GSH and protein adduct formation being predominate (Shih and Hill, 198 1; Sundheimer et al., 1982). In order to better understand the mechanisms of toxicity of dihaloethanes, a series of S-(2-haloethyl)-cysteine analogs were recently synthesized and investigated for mutagenicity (van Bladeren et al., 1980, 198 1) and hydrolysis and alkylating activities (Schasteen and Reed, 1983). It was found that the mutagenicity of the analogs did not correlate with the halide order for leaving group ability (van Bladeren et al., 198 1) and that N-acetylS-(2-haloethyl)-cysteine methyl esters were more mutagenic than the parent compounds. Studies on the hydrolysis of several analogs at different pHs suggested that the amine moiety was responsible for an increased hydrolysis rate with alkaline conditions, and formation of 3-(thiomorpholine)-carboxylic acid was proposed as an alternative pathway to the generally accepted hydrolysis mechanism for 9(2-haloethyl)-L-cysteine analogs (Schasteen and Reed, 1983). This paper describes an extension of previous studies to include reactions with purified DNA. A series of related S-(2-haloethyl)-Lcysteines was synthesized and incubated with plasmid pBR322. The reactivity of the analogs
AND DNA ALKYLATION
387
with DNA was measured in terms of time dependent relaxation of the supercoiled form of DNA. The amount of covalent binding of certain analogs to DNA was determined. METHODS Chemicals. L-Cystine and N-acetyl-L-cysteine were purchased from Sigma (St. Louis, MO.). 1-Bromo-2fluoroethane was purchased from Columbia Organic Chemical Company (Columbia, SC.). Diazomethane was prepared from N-methyl-N-nitroso-N’nitroguanidine purchased from Aldrich (Milwaukee, Wise.). 2-Bromoethanol was also purchased from Aldrich Chemical Company. Sodium was obtained from Mallinckrodt (St. Louis, MO.). All other chemicals utilized in this study were ACS grade or the highest grade obtainable. Radiochemicals. [meth.vl-‘H]Thymidine (48 to 49 Ci/ mmol) was obtained from Amersham Int., Ltd. L[3SS]Cysteine (436.4 Ci/mmol) was obtained from New England Nuclear Corporation, Boston, Massachusetts. Synthesis of S-(2-haloethyl)-analogs. All compounds utilized in this investigation were prepared by the published procedures of Carson and Wong (1964) and van Bladeren et al. (1980) or du Vigneaud and Patterson (1936) and the modifications of McKinney et al. (1957). All products were considered pure by at least two thinlayer chromatographic systems, and the structures were consistent with the respective ‘H-NMR spectra obtained. Identification via mass spectrometry was obtained by either plasma desorption ionization mass spectrometry (Baldwin and McLafferty, 1973; Cotter. 1979) or fast atom bombardment (FAB) quadrupole mass spectrometry (Surman and Vickerman, 1981; Barber ef al., 1981). Synthesis of9(2-chloroethyl)-glutathione. To approximately 200 ml of liquid ammonia in a three-neck flask, which was fitted with a mechanical stirrer and sodalime tube, was added 7.68 g (0.025 mol) glutathione (GSH). The flask was kept at -35°C by cooling in a dry ice-ethanol bath. Sodium was introduced into the flask in small portions to the first permanent discharge of blue color. A total of 2.89 ml (0.035 mol) of I-bromo-2chloroethane was added in small portions over a I-hr period with the reaction temperature held at -35°C. After evaporation of the ammonia, the off-white solid was lyophilized. Thin-layer chromatography in n-butanol: acetic acid:HzO (25:4:10) Avicel of the solid residue showed a ninhydrin positive spot corresponding to Rf = 0.44. A sample placed in distilled water at room temperature for 25 hr showed a spot corresponding to synthetic .S-(2-hydroxyethyl)glutathione (R, = 0.21) prepared in a similar manner as above except with 2bromoethanol instead of I-bromo-2-chloroethane. MS(FAB): 396 (M + Na)‘, 2.0% 394 (M + Na)+, 6.5%. 330 (GSH + Na)+, 0.6%, 308 (GSH + H)+ 208, 11.4%.
388
VADI, SCHASTEEN.
Preparation and isolation of plasmid DNA. Plasmid pBR322 (Sutcliffe, 1978) was grown and amplified in Escherichia coli strain RRI as described (Norgard et al., 1979; Vadi and Reed, 1983). Radioactive plasmid was obtained by adding [methyl-SH]tbymidine to the medium (1 to 2 mCi/liter) just prior to the amplification step, which was initiated by addition of chloramphenicol (Clewell, 1972). Following 10 hr of amplification, the cells were sedimented by centrifugation, incubated with lysozyme, and treated with a salt detergent solution. Chromosomal DNA and proteins were removed by centrifugation, and remaining nucleic acids were precipitated with a mixture of sodium trichloroacetate:ethanol. The pellet of nucleic acids was resuspended in TENbuffer (10 mM Tris-HCl, I mM EDTA, 10 mM NaCl, pH 7.5) and treated with RNase A (Vadi and Reed, 1983) before reprecipitation with NaTCA:ethanol. The final product, consisting of pBR322, was routinely purified on a cesium chloride gradient containing EtdBr (ethidium bromide) (Radloff et al., 1967). All steps involving EtdBr were performed under red light, thus ensuring a high yield of supercoiled plasmid DNA. The isolation procedure has been described in detail elsewhere (Vadi and Reed, 1983). Agarose gel electrophoresis. The various forms of pBR322 were separated by agarose gel electrophoresis with 0.7% agarose in a horizontal slab gel system (McDonnell et al., 1977). DNA bands were visualized by staining the gel in 5 Irglml EtdBr solution and illuminating from below with ultraviolet light. The amount of DNA in each band was quantified by liquid scintillation counting, and the proportion of supercoiled DNA was calculated as the percentage of total DNA in that channel (Vadi and Reed, 1983). Incubation conditions. pBR322 was incubated with 2haloethylcysteine analogs or glutathione conjugates at 37°C for the times indicated. The reaction mixtures contained 0.1 M cacodylate, pBR322 (0.2 rdgl), 7 mM Tris, pH 7.5, 0.7 mM EDTA, 7 mM NaCI, and 10 mM concentration of the analog studied, unless otherwise stated. S-(2-Haloethyl)-L-cysteine analogs were added in DMSO. The final concentration of DMSO as 10% in all samples including control incubations. Glutathione conjugates were added in TEN-buffer. S-(2-Haloethyl)-glutathione was checked by thin-layer chromatography for purity [<5% of s-(2-hydroxy-ethyl)-glutathione as the major contaminant] prior to each incubation. Reactions were stopped by spin dialysis through Sephacryl S-200 equilibrated with TEN-buffer (Neal and Florini, 1973; Vadi and Reed, 1983). DNA was precipitated by addition of l/IO vol 3 M sodium acetate and 5 vol of ethanol, and samples were kept at -50°C for at least 1.5 hr. Pelleted DNA was resuspended in TENbuffer and analyzed for supercoiled plasmid by agarose gel electrophoresis as described above. Denaturation ojDNA. In some experiments the DNA was denatured by heating at neutral pH (Vadi and Reed, 1983). Following spin dialysis and precipitation, the
AND REED
DNA was resuspended in 5 mM EDTA and kept at 4°C for 4 to 6 hr. Sucrose was added to a final concentration of IO%, and the samples were heated in a boiling waterbath for 70 sec. The denatured samples were cooled in ice and were immediately analyzed by gel electrophoresis. This procedure has been demonstrated to cause complete denaturation of double stranded linear and nicked circular forms of non-alkylated pBR322 (Vadi and Reed, 1983). Quantification of extent of alkylation. [3H]pBR322 was incubated with (2-chloroethyl)-t.-[3SS]cysteine or (2bromoethyl)-L-[35S]cysteine as described above. Reactions were stopped by precipitation of the DNA. The pellets were washed with a mixture of 15% TEN-buffer, 1.5% 3 M sodium acetate, and 83.5% ethanol, resuspended in TEN-buffer, and reprecipitated. This cycle of precipitating, washing, and resuspending the DNA was repeated at least three times or until the ratio 35S:3Hbecame constant. By knowing the specific activities of the substrates and the DNA, the amount of alkylation per unit DNA was calculated. Electron microscopy. Certain DNA bands were isolated from agarose gels as described (Vadi and Reed. 1983) and prepared for electron microscopy by an aqueous staining technique (Kleinschmidt, 1968). The grids were examined in a Zeiss 10 electron microscope.
RESULTS [3H]pBr322 was incubated with 2-haloethyl-L-cysteine analogs, purified by spin dialysis, and analyzed by agarose gel electrophoresis. Treatment with S-(2-chloroethyl)L-cysteine (CEC) or 9(2-bromoethyl)-L-cysteine (BrEC) resulted in a time-dependent relaxation of the supercoiled form of DNA as determined by a decrease in the amount of DNA migrating as band d in agarose gels (Fig. 1). Band b, which in non-treated samples consists of relaxed circular DNA, disappeared by 120 min (Fig. l), and there was an increase in the proportion of DNA migrating as band c. We have previously shown that in nontreated samples of DNA, the double-stranded linear form of DNA migrates as band c, whereas when DNA has been alkylated with nitrosoureas or dimethyl sulfate prior to gel electrophoresis, band c consists of relaxed circular DNA (Vadi and Reed, 1983). Similar results were obtained when plasmid DNA was treated with 4(2-chloroethyl)-L-cysteine or S-(2-bromoethyl)-L-cysteine (results not
s-(2-HALOETHYL)
Time (min)
0
30
60
ANALOGS
120
FIG. 1. Agarose gel electrophoresis of pBR322 treated with S-(2-bromoethyl)-L-cysteine. Incubations were performed as described under Methods for the times indicated. Bands are designated: (a) upper band; (b) relaxed circle: (c) doubte stranded linear; (d) supercoiied DNA.
shown). Electron microscopy of band c purified from agarose gels of S-(Zchloroethyl)L-cysteine- or S-(2-bromoethyl)-L-cysteinetreated DNA showed that this band consisted of relaxed circles (Fig. 2). S(2-Fluoroethyl)L-cysteine did not interact with plasmid DNA since no change in electrophoretic pattern of the DNA could be detected after treatment with this analog (results not shown). Figures 3 and 4 show the effects of some of the haloethyl+cysteine analogs on the superhelicity of DNA. S-(2-Chloroethyl)-Lcysteine induced a rapid relaxation of plasmid DNA (Fig. 3). Blocking the carboxyl group of CEC by esterification reduced the rate of relaxation to some extent, whereas acetylation of the free amine of CEC dramatically decreased interaction with DNA. Thus, N-acetyl-S-(2-chloroethyl)-L-cysteine (CENAcC) was able to lower only slightly the proportion of superhelical DNA present. The methyl ester of this analog (CENAc M.E.) did not interact with DNA. Similar results were obtained with the S (2-bromoethyl)-L-cysteine analogs (Fig. 4).
AND DNA ALKYLATION
389
Once again it was observed that methylation of the carboxyl group slightly reduced the rate of relaxation, whereas acetylation of the primary amine prevented the analogs from interacting with DNA. Bromo-containing analogs were always less reactive with DNA than their chloro-containing counterparts. As expected from the above results with CENAcC, S-(2-chloroethyl)-glutathione did not interact with plasmid DNA as determined by lack of effect on the superhelicity of DNA (Table 1). s-(2-HydroxyethyB-glutathione and 9(2-hydroxyethyl)-L-cysteine also failed to interact with DNA (Table 1). Neutral denaturation of non-alkylated pBR322 resulted in complete loss of doublestranded linear DNA (Fig. 5, channel 1, band c) and formation of single-stranded linear DNA (Fig. 5, channel 2, band e). Electrophoretie patterns identical to control patterns were obtained when pBR322 was treated with N-acetyl-S-(2-chloroethyl)-L-cysteine methyl ester (CENAc M.E.), N-acetyl-S-(2bromoethyl)-L-cysteine (BrENAcC), its methyl ester (BrENAcC M.E.), S(Zchloroethyl)-glutathione, s-(2-hydroxyethyl)-glutathione, or S(2-hydroxyethyl)-L-cysteine followed by neutral denaturation (Fig. 5, channels 6, 9, and 10, or results not shown). These analogs did not induce a relaxation of supercoiled plasmid (Figs. 3 and 4; Table 1) and did not interfere with denaturation. In contrast, the 2-haloethyi+cysteine analogs that were able to induce relaxation of supercoiled plasmid (Figs. 3 and 4) were also able to interfere with denaturation and prevent the double-stranded DNA species in band c from denaturing (Fig. 5, channels 3 to 5, 7 and 8). In these experiments the plasmid was treated for 15 min with 5 mM concentration of the analogs prior to denaturation. When higher concentrations and/ or longer incubations were employed, denaturation resulted in varying degrees of fragmentation of DNA treated with DNA-interacting analogs CEC, CEC M.E., CENAcC, BrEC, or BrEC M.E. (results not shown). pBR322 that was digested with BamHl
390
VADI, SCHASTEEN,
AND REED
FIG. 2. Electron microscopy of S-(2-chloroethyl)-L-cysteine treated pBR322. Plasmid DNA (0.35 rg/ ~1) was incubated with 10 mM S-(2-chloroethyl)+cysteine at 37°C for 2 hr. Gel eiectrophoresis was performed, band c was purified from the gel as described (Vadi and Reed, 1983) and stained (Kleinschmidt, 1968). Magnification X24.480. The presence of only relaxed circular and not linear or supercoiled DNA is based on previous electron microscopic studies with pBR322 (Vadi and Reed, 1983).
S-(2-HALOETHYL)
D
ANALOGS
CENAcC
ME.
CENAcC
391
AND DNA ALKYLATION
(Fig. 7). Approximately 60 to 70% (CEC) or 50% (BrEC) of the radioactivity that precipitated with DNA comigrated with DNA in agarose gels, and when treated DNA was denatured by the neutral sucrose technique prior to gel electrophoresis, 30 to 35% of the radiolabel still corn&rated with DNA. These findings suggest that the radiolabel was covalently bound to DNA rather than associated by either electrostatic interactions or intercalation. The degree of binding was concentration dependent and was higher for CEC than for BrEC at 10 mM substrate concentration (Fig. 7). 35S labeled cysteine or cystine failed to show any detectable binding of
ME BrENAcC
ME
1
loo -BrBrENAcC
CEC
Time,
min
FIG. 3. Effect of S-(2-chloroethyl)-L-cysteine analogs on the superhelicity of pBR322. Incubations were performed as described under Methods for the times indicated in the graph. Values were normalized to 100% superhelicity for non-treated DNA and linear regression was used for data analysis. CEC, S(Z-chloroethyl)-L-cysteine; CEC M.E., S-(2-chloroethyl)-L-cysteine methyl ester; CENAcC, N-acetyl-S-(2-chloroethyl)-Lcysteine; CENAcC M.E., N-acetyl-s-(2-chloroethyl)-Lcysteine methyl ester.
40BrEC
% B yj
x\
30-
ME
BrEC
5 cn’
z 20
restriction enzyme (Vadi and Reed, 1983) prior to CEC treatment and denaturation gave no indication of CEC-induced DNA interstrand crosslink formation in linear DNA (Fig. 6, channels 5 and 6). Thus, linear DNA did not display the same alteration by CEC treatment as supercoiled plasmid which induced to relax and also showed interference of denaturation (Fig. 5, Channels 3 to 5, 7 and 8). Treatment of pBR322 with (2-chloroethyl)L-[35S]-cysteine or (2-bromoethyl)-L-[35S]cysteine resulted in binding of 35Slabel to DNA
t
100 0
30
60
120 Time,
mm
FIG. 4. Effect of s-(2-bromoethyl)-L-cysteine analogs on the superhelicity of pBR322. See legend to Fig. 3 for details. BrEC, 4(2-bromoethyl)-L-cysteine; BrEC M.E., .S-(2-bromoethyl)-L-cysteine methyl ester; BrENAcC, Nacetyl-S-(2-bromoethyl)-L-cysteine; BrENAcC M.E., Nacetyl-S-(2-bromoethyI)-L-cysteine methyl ester.
392
VADI, SCHASTEEN, TABLE
EFFECT OFS(2-CHLOROETHYL)-GSH,
AND REED 1
&(&HYDROXYETHYL)-GSH, ONTHESUPERHELICITYOF
AND 9('i'-HYDROXYETHYL)-L-CYSTEINE
vBR322
% Supercoiled DNA Time (mm)
s-(2-0HEt)GSH
S-(2-CIEt)GSH
0 30 60 120
78.7 76.5 77.0 75.8
f + + k
3.2 4.4 4.0 3.9
78.7 77.9 77.0 78.2
f + + +
s-(2-0HEt)cys
3.2 4.3 5.9 3.3
78.7 77.5 75.5 74.2
rt k * t
3.2 2.1 3.9 1.3
Note. Incubations contained 10 tIIM substrate and were performed as described under Methods for the times indicated. N values are X + SD of three experiments.
radiolabel to DNA (results not shown). (2Hydroxyethyl)-L-[35S]cysteine coprecipitated with DNA to a certain extent, but the inter-
1
2
3
4
5
action with DNA decreased with incubation time, and the radiolabel did not corn&ate with DNA in agarose gels.
b
7
8
9
10
FIG. 5. Agarose gel electrophoresis of pBR322 treated with S(2-haloethyl)-L-cysteine analogs and denatured by a neutral sucrose denaturation technique. pBR322 was incubated with 5 mM concentration of the analogs for I5 min. The experimental conditions were as described under Methods. Channels I and 2 represent control pBR322 before and after denaturation, respectively. Channels 3 to IO represent pBR322 which was treated with haloethyl cysteine analogs as indicated in the figure prior to neutral sucrose denaturation. See legends to Figs. 3 and 4 for explanation of abbreviations.
.S-(Z-HALOETHYL)
Barn
ANALOGS
Hl
AND DNA ALKYLATION
393
CONTROL
FIG. 6. Denaturation of BumHI-digested pBR322 treated with S-(Z-chloroethyI)-L-cysteine. Supercoiled and BamHl digested pBR322 were treated with 3 mM CEC prior to denaturation and agarose gel electrophoresis. Channels 1 and 2: Control BamH 1-digested pBR322; native and denatured, respectively. Channels 3 and 4: BarnH l-digested pBR322 treated with CEC for 15 and 30 min. respectively. Channels 5 and 6: Same samples as in channels 3 and 4, after denaturation. Channels 7 and 8: Control pBR322; native and denatured, respectively. Channels 9 and IO: pBR322 treated with CEC for 15 and 30 min, respectively. Channels I1 and 12: Same samples as in channels 9 and 10 after denaturation.
DISCUSSION In this study we investigated the interactions between a series of S-(2-haIoethyl)-Lcysteine analogs and plasmid DNA. When incubated with supercoiled pBR322, S-(2chloroethyl)-L-cysteine and 9(2-bromoethyl)L-cysteine induced relaxation of the DNA as determined by agarose gel electrophoresis and electron microscopy (Figs. 1 and 2). We suggest that the relaxation is due to strand scission at alkylated labile sites in the DNA. When 35S labeled 9(2-chloroethyl)+cysteine or s-(2-bromoethyl)-L-cysteine was used as the substrate, covalent binding of 35Sto DNA was obtained; S(2-chloroethyl)-L-cysteine resulted in a somewhat higher 35Sbinding than s-(2-bromoethyl)-L-cysteine. This difference could be a reflection of the faster hydroIysis
rate previously noted for the bromo analogs compared to the chloro analogs (Schasteen and Reed, 1983). At least 30 to 35% of this binding was resistant to neutral denaturation of the DNA. No binding of 35S to DNA was obtained with (2-hydroxyethyl)-L-[35S]cysteine, [35S]cysteine, or [35S]cystine substrates which did not induce relaxation of the DNA. As pBR322 was treated with increasing concentrations of DNA-interacting analogs, neutral denaturation resulted in increased fragmentation of the DNA. This finding supports our suggestion that certain analogs were able to create potentially labile sites in the DNA, thus generating sites for spontaneous strand scission. Findings by other workers that support this contention include in vivo studies in which 1,2-dibromoethane (DBE) was administered to mice and a dose-depen-
394
VADI, SCHASTEEN,
AND REED
(2-bromoethyl)GSH (Ozawa and Guengerich, 1983). However, since we now show that CEC relaxed pBR322 rapidly and 5’-(2-chloroethyl)GSH (CEG) essentially had no effect, then metabolism of CEG to CEC could be important relative to the total DNA damage resulting from exposure to 1,2-dichloroethane (DCE) and DBE. Our finding that N-acetylation of the analogs prevented them from interacting with DNA is also in contrast to a report by van Bladeren et al. (198 1) who recorded mutagenic activity with N-acetyl-S-(2-haloethyl)TIME,min L-cysteine methyl esters. A possible explanaFIG. 7. Alkylation of [3H]pBR322 with [35S]CEC or tion for this finding may be the presence of [35S]BrE~. Incubations were as described under Methods for the times indicated in the graph. The amounts of bacterial deacetylases in their system whereas alkylation were calculated from the amount of “S asso- they were absent in the experimental system ciated with the DNA after three precipitations. used in the present studies. The finding that DNA-interacting analogs prevent neutral denaturation of doubledent increase in DNA single strand breaks stranded circular DNA, although no interferand alkali labile sites was observed (White et ence was obtained in the denaturation of al., 1981; Storer and Conolly, 1983). analog-treated linear DNA, can be explained Of particular interest was the finding that in two ways: (1) some analogs do, in fact, S-(2-chloroethyl)-glutathione did not affect interact with both strands of circular DNA, the superhelicity of pBR322 DNA (Table 1). but no such interaction takes place in linear This finding is in contrast with the data DNA; and (2) at the initial stages of alkylobtained with S-(2-chloroethyl)-L-cysteine but ation, DNA becomes relaxed. but no strand not with N-acetyl-S-(2-chloroethyl)-L-cysteine. scission takes place. Relaxed circular DNA Blocking of the amine moiety is known to with both strands covalently closed would affect the hydrolysis enhancement noted with not be denatured to single-stranded DNA alkaline conditions for S-(2-chloroethyl)-Lunder the conditions used. cysteine and eliminates the formation of 3(thiomorpholine)-carboxylic acid (Schasteen ACKNOWLEDGMENTS and Reed, 1983). Incubation of calf thymus DNA with DBE This research was supported by NIEHS Grants ESand GSH in the presence of GSH S-transfer00040 and 1 F32 ES-05192. The authors acknowledge ase results in a product which yields N7the use of the Prophet computer system in molecular ethylguanine when reductively desulfurized modeling and data manipulation. The FAB mass spectra (Ozawa and Guengerich, 1983). Since the analyses by Don Griffin are greatly appreciated. parent DNA adduct contains radiolabel from GSH, the authors assigned it as being 9[2REFERENCES (N7-guanyl)ethyl]GSH. They also concluded that their findings support the hypothesis of BALDWIN, M. A., AND MCLAFFERTY, S. W. (1973). the major interaction of DBE with DNA Direct CI of relatively involatile samples. Org. Mass involving covalent modification by a preSpectrom. 7, 1353-l 356. formed complex of DBE and GSH, i.e., S- BARBER, M., BORDOLI, R. S., SEDGWICK, R. D., AND
S-(2-HALOETHYL)
ANALOGS
TYLER, A. N. (1981). Fast atom bombardment of solids (FAB): A new ion source for mass spectrometry. J. C. S. Chem. Commun. 325-327. CARSON, J. F., AND WONG, F. (1964). The synthesis of L- 1,4-thiazane-3-carboxylic acid 1-oxide. J. Org. Chem. 29, 2203-2205. CLEWELL, D. B. (1972). Nature of ColEl plasmid replication in the presence of chloramphenicol. J. Bucteriol. 110, 667-676.
COTTER, R. J. (1979). Probe for direct solid sample exposure to reagent gas in Cl mass spectrometry. Anal. Chem. 51, 317-318. Du VIGNEAUD, V.. AND PATTERSON, W. I. (1936). The synthesis of djenkolic acid. J. Biol. Chem. 114, 533538. FABRICANT, J. D., AND CHALMERS. J. H., JR. (1980). Evidence of the mutagenicity of ethylene dichloride and structurally related compounds. In Banbury Reporf 5; EthJplene Dichloride: A Poientiai Health Risk? (8. Ames, P. Infante, and R. Reitz. eds.). Cold Spring Harbor. FISHBEIN,L. ( 1980). Potential carcinogenic and mutagenic industrial chemicals. 1. Alkylating agents. J. Toxicol. Environ. Health 6, 1133-I 177. GUENGERICH, F. P., CRAWFORD. W. M., DOMORADZKI, J. Y.. MACDONALD, T. L., AND WATANABE. P. G. (1980). In vlfro activation of 1,2-dichloroethane by microsomal and cytosolic enzymes. Toxicol. Appl. Pharmacol. 55, 303-3 17. HOOPER. K., GOLD, L. S., AND AMES, B. N. (1980). The carcinogenic potency of ethylene dichloride in two animal bioassays: A comparison of inhalation and gavage studies. In Banbury Report 5: Ethylene Dichloride: A Potential Health Risk? (B. Ames, P. Infante, and R. Reitz. eds.), pp. 65-81. Cold Spring Harbor Laboratory. KLEINSCHMIDT, A. K. (1968). Monolayer techniques in electron microscopy of nucleic acid molecules. In Methods in En~ymolugv (L. Grossman and K. Moldave. eds.), Vol. 12. pp. 36 l-377. Academic Press, New York. LIVESEY, J. C.. AND ANDERS, M. W. (1979). In vitro metabolism of 1,2-dihaloethanes to ethylene. Drug Metab. Dis. 7, 199-203. MCDONNELL, M. W.. SIMON, M. N., AND STUDIER, F. W. (1977). Analysis of restriction fragments of T7 DNA and determination of molecular weights by electrophoresis in neutral and alkaline gels. J. Mol. Biol. 110, 119-146. MCKINNEY, L. L.. WEAKLEY, F. B., ELDRIDGE, A. C., CAMPBELL, R. E., COWAN, J. C., PICKEN, J. C., JR., AND BEISTER, H. E. (1957). S-(Dichlorovinyl)-L-cysteine: An agent causing fatal aplastic anemia in calves. J. Amer. Chem. Sot. Commun. 79, 3932-3933. National Cancer Institute. (1978). Technical Background Irzformation Report #55 on Carcinogenesis Bioassay
AND DNA ALKYLATION
395
of 1,2-Dichloroethane (EDC). U.S. DHEW Publication number (NIH) 78- 136 1. Govt. Printing Office, Washington, D.C. NEAL, M. W., AND FLORINI, J. R. (1973). A rapid method for desalting small volumes of solution. Anal. Biochem. 55, 328-330. NORGARD, M. V., EMIGHOLTZ. K., AND MONAHAN, J. J. (1979). Increased amplification of pBR322 plasmid deoxyribonucleic acid in Escherichia coli K- 12 strains RR1 and X 1776 grown in the presence of high concentrations of nucleoside. J. Bacterial 138, 270-272. OZAWA, N., AND GUENGERICH, F. P. (1983). Evidence for formation of an S-[2-(N-guanyl)ethyl]glutathione adduct in glutathione-mediated binding of the carcinogen 1,2-dibromoethane to DNA. Proc. Natl. Acad. Sci. USA 80, 5266-5270. RADLOFF, R.. BAUER, W., AND VINCCRAD, J. (1967). A dye-buoyant density method for the detection and isolation of closed circular duplex DNA: The closed circular DNA in HeLa Cells. Proc. Nat/. Acad. Sci. USA 57, 1514-1521. RANNUG, U., SUNDVALL, A., AND RAMEL, C. (1978). The mutagenic effect of 1,2-dichloroethane on Salmonella typhimurium. I. Activation through conjugation with glutathione in vitro. Chem. Biol. Interact. 20, l-16. SCHASTEEN.C. S., AND REED, D. J. (1983). The hydrolysis and alkylation activities of S-(2-haloethyl)-L-cysteine analogs-Evidence for extended half-life. To,xico/. Appl. Pharmacol. 70, 423-432. SHIH, T.-W., AND HILL, D. L. (1981). Metabolic activation of 1.Zdibromoethane by glutathione transferase and by microsomal mixed function oxidase: Further evidence for formation of two reactive metabolites. Res. Commun. Chem. Pathol. Pharmacol. 33, 449461. SIMMON, V. F. (1980). Review of nonbacterial tests of the genotoxic activity of ethylene dichloride. In Banbuty Report 5. ethylene Dichloride: A Potential Health Risk? (B. Ames, P. Infante, and R. Reitz, eds.), pp. 97-103. Cold Spring Harbor Laboratory. SMIT, W. A., ZEFIROV, N. S., BODRIKOV, I. V., AND KIRMER. M. Z. (1978). Episulfonium ions: Myth and reality. Acct. Chem. Res. 12, 282-288. STORER, R. D., AND CONOLLY, R. B. (1983). Comparative in viva genotoxicity and acute hepatotoxicity of three 1,2-dihaloethanes. Curcinogenesis 4( 1 l), 14911494. SUNDHEIMER, D. W.. WHITE, R. D., BRENDEL, K., AND SIPES,I. G. (1982). The bioactivation of I ,2-dibromoethane in rat hepatocytes: covalent binding to nucleic acids. Carcinogenesis 3, 1129-I 133. SURMAN, D. .l., AND VICKERMAN, J. C. (i981). Fast atom bombardment quadrupole mass spectrometry. J. C. S. Chem. Commun. 324-325. SUTCLIF’FE,J. G. (1978). pBR322 restriction map derived
VADI, SCHASTEEN,
396
from the DNA sequence: Accurate DNA size markers up to 436 1 nucleotide pairs long. Nucleic Acid Rex 5, 2121-2728. VADI, H. V., AND REED, D. J. (1983). Effect of 2chloroethylnitrosoureas on plasmid DNA, including formation of strand breaks and interstrand crosslinks. Chem.
Biol.
Interact.
46, 67-84.
VAN BLADEREN, P. J., BREIMER, D. D., ROTTEVEELSMIJS, G. M. T., DE KNIJFF, P., MOHN, G. R., VAN MEETEREN-WALCHLI, B., BUIJS, W., AND VAN DER GEN, A. (1981). The relation between the structure of vicinal dihalogen compounds and their mutagenic activation via conjugation to glutathione. Carcinogenesis 2(6), 499-505.
VAN BLADEREN, P. J., BUIJS, W.. BREIMER, D. C., AND
AND REED
VAN DER GEN, A. (1980). The synthesis of mercapturic acids and their esters. Eur. J. Med. Chem. 15, 49% 497. VAN BLADEREN, P. J., VAN DER GEN, A., BREIMER, D. D., AND MOHN, G. R. (1979). Stereoselective activation of vicinal dihalogen compounds to mutagens by glutathione conjugation. Biochem. Pharmacol. 28, 2521-2524. WEISBURGER, E. K. (1977). Carcinogenicity studies on halogenated hydrocarbons. Environ. Health Perspect. 21, 2804-28 14. WHITE, R. D., SIPES, I. G., GANDOLFI, A. J., AND BOWDEN, G. T. (198 1). Characterization of the hepatic DNA damage caused by 1,2-dibromoethane using the alkaline elution technique. Carcinogenesis 2, 839-844.