Mechanism of glutathione-mediated DNA damage by the

A Critical Appraisal of the Evolution of N-Nitrosoureas as Anticancer Drugs. C. Thomas Gnewuch and George Sosnovsky. Chemical Reviews 1997 97 (3), 829...
0 downloads 0 Views 495KB Size
Chem. Res. Toxicol. 1992,5, 106-109

106

Mechanism of Glutathione-Mediated DNA Damage by the Antineoplastic Agent 1,3-Bis(2-chloroethyl)-IV-nitrosourea Wilhelm Stahl,+P*Susanne Lenhardt,s Michael Przybylski,*p§and Gerhard Eisenbrandt Fachbereich Chemie, Lebensmittelchemie und Umwelttoxikologie, Universitaet Kaiserslautern, Erwin-Schroedinger-Strasse,0-6750 Kaiserslautern, Germany, and Fakultaet fuer Chemie, Universitaet Konstanz, Universitaetsstrasse 10,D-7750 Konstanz, Germany Received July 9,1991

S-[(2-Chloroethyl)carbamoyl]glutathione(SCCG), a compound formed during the decomposition of BCNU in the presence of GSH, induces DNA damage in a human lymphoblastoid cell line. This GSH conjugate was shown by direct fast atom bombardment mass spectrometric analysis to transfer an aminoethyl group to the N-7 position of guanosine. The resulting N7-(aminoethy1)guanosine adduct readily undergoes depurination. From these model studies, DNA aminoethylation appears to represent a plausible explanation as the major cause for the DNA-damaging effects exerted by SCCG. Introduction The tripeptide glutathione (GSH)' is involved in a number of important cellular processes and, among other biological effects, plays an important role as a detoxification system. A large number of exogenous and endogenous toxic compounds are conjugated to GSH and, after further biotransformation, excreted as nontoxic mercapturic acids ( I , 2). In a few cases, however, involvement of GSH in toxification mechanisms has been demonstrated. Several groups have revealed covalent glutathione conjugation of certain xenobiotics as a first step in their toxification; among others, halogenated alkene derivatives have been reported to be activated in this way (3-6). Guengerich and co-workers investigated the genotoxicity of 2-haloethane thioethers of GSH and showed chemical reactions of the GSH conjugates with DNA bases to be responsible for their mutagenicity (7-9). The adduct N7-ethyl-2-(S-glutathionyl)guanosine has been identified as a reaction product of guanosine and 1,Zdibromoethane in the presence of GSH and GSH transferase. We recently reported mutagenic activity of a glutathionyl adduct formed upon incubation of GSH with the anticancer agent 1,3-bis(2-chloroethyl)-N-nitrosourea (BCNU; see Figure 1). The reaction product S-[(2chloroethyl)carbamoy1]glutathione (SCCG) proved to be mutagenic in the Ames test and appeared to contribute to the mutagenic and carcinogenic effects of BCNU (10). In the present report results of our studies on the elucidation of the mechanism of the genotoxic action of SCCG are summarized. Experimental Procedures Materials. Guanine, guanoaine, and reduced glutathione were obtained from Fluka (Neu-Ulm, Germany), and 2-chloroethyl isocyanate was from Aldrich (Steinheim, Germany). All other chemicals were of analytical-gradepurity from Merck (Darmstadt, Germany). N'-(2-aminoethyl)guanine was synthesized as described previously (11). S-[(2-chloroethyl)carbamoyl]glutathionewas prepared from equimolar amounts of GSH and 2-chloroethyl isocyanate in +

Universitaet Kaiserslautern.

* Present address: Institut fuer Physiologische Chemie I, Univ-

ersitaet Duesseldorf, Moorenstrasse 5,D-4000 Duesseldorf, Germany. 8 Universitaet Konstanz.

Table I. Single-Strand Break Factors (SSF) in Namalva Cells at Different Concentrations of SCCG" SCCG concn, pM control 0.05 0.1

SSF

SCCG concn, UM 0.5

SSF

1.1

1.0

17.5

11.8

4.6

SSF values were determined according to Meyn et al. (14).

aqueous acetone, pH 3 (in 70% yield), and was recrystallized from methanol/water. Structure and homogeneity of SCCG were established by 'H NMR, fast atom bombardment (FAB) mass spectra, and elemental analysis. Before use, solutions of SCCG in water or culture medium were neutralized with sodium hydroxide. Determination of DNA Single-Strand Breaks. Namalva cells (3 X l@/mL) were grown in RMPI medium 1640 (GIBCO, BRL, Eggenstein, Germany) (12)and were incubated for 1 h with different concentrations of SCCG. Induction of single-strand breaks (SSB) was determined by means of the alkaline filter elution technique as described by Stem1 et al. (13). Single-strand break factors were calculated according to Meyn et al. (14). Reaction of SCCG with Guanosine. Guanosine (10pmol) and SCCG (50 rmol) dissolved in 10 mL of 50 mM sodium cacodylate buffer were incubated at 37 "C, pH 7.2,and aliquota of the reaction mixture were taken at different time intervals for HPLC analysis (see Figure 2). HPLC analyses were carried out on a 250 x 4.6 cm Shandon hypersil ODS 5-pm reversed-phase column eluted with 0.05 M ammonium dihydrogen phosphate as a mobile phase at 1.0 mL/min (void volume approximately 1.5 mL), using UV detection at 260 nm or fluorescence detection at excitation/emission wavelengths of 300/430 nm. Separations and isolation of reaction products were also performed on a 250 X 4.6 cm, 5-rum Kontron Spherisorb ODS column; "+minoethyl)guanine and guanosine were used as reference compounds. Fast Atom Bombardment Mass Spectrometry. The reaction of SCCG with guanosine was directly analyzed by time-dependent, dynamic fast atom bombardment mass spectrometry (FAB-MS). Experiments were carried out with a modified Finnigan MAT 312/AMD-5000 double-focusing spectrometer equipped with a high-field magnet and a 25-kV cesium thermionic primary ion source. Equimolar solutions (approximately 10 nmol) of SCCG and guanosine in 40 mM Tris-HC1buffer (pH 7.8)were supplemented with 10% (v/v) glycerol as a matrix and deposited (ca. 5 pL) on a stainless steel FAB target probe, thermostated Abbreviations: BCNU, 1,3-bis(2-chloroethyl)-N-nitroeouree; HECNU, 3-(2-hydroxyethyl)-l-(2-chloroethyl)-N-nitrosourea; SCCG, S-[(2chloroethyl)carbamoyI]glutathione;GSH, glutathione; FAB, fast atom bombardment; SSB, single-strand breaks.

0893-228~/92/2705-0106$03.00/0 0 1992 American Chemical Society

-

Chem. Res. Toxicol., Vol. 5, No. 1, 1992 107

Glutathione-Mediated DNA Damage by BCNU

CI

9.

uNH 0

m i 2 1351413 m l z 32'1413

i

1-

5I

SCCG 413

CI 3

6

9

tIm1n

28)

GSH

I

CI

GsKNH 0

SCCG Figure 1. Decomposition pathway of BCNU to 2-chloroethyl isocyanate, and formation of the SCCG adduct with GSH. 2

t.138 h

t =17 h r

0

I

7 t (mid

1L

Figure 2. HPLC analysis (UV detection, 260 nm) of the fluorescent reaction products of SCCG with guanosine. (Left) Incubation mixture at start of reaction; (middle) reaction products after 17 h; (right)readion products after 138h. Peaks: 1,guanine;

2, guanosine; 3, W-(aminoethy1)guanine;4, unidentified fluorescent decomposition product. For HPLC conditions see Experimental Procedures.

at approximately 40 "C (15). After insertion of the probe, acquisition of mass spectra was carried out continuously for approximately 20 min at a scan cycle time of 25 s, using a Finnigan 98-200 data system.

Results Incubation of Namalva cells with different concentrations of SCCG led to extensive DNA strand breakage in this tumor cell line. Formation of single-strand breaks was dose dependent, with effects still observable at 0.1 pM concentrations (Table I). These data demonstrate that SCCG is a potent DNA-damaging agent. Guanosine adduct formation was studied by incubating SCCG with guanosine a t pH 7.2 (37"C) for a total period of 139 h. HPLC analyses at different times (Figure 2) revealed the formation of a single major reaction product (peak3) detectable at 260 nm, which eluted with a definite retention time of approximately 3.5 min corresponding to authentic N-(aminoethy1)guanine. The same product was obtained using a second, octadecylsilyl reversed-phased HPLC column with a retention time of 4.4 min (see Experimental Procedures). The identification of "-(aminoethy1)guanine was ascertained by isolation of the HPLC peak and FAB-MSanalysis, which yielded molecular and fragment ions identical with that of the authentic compound. Only minor or trace additional reaction products were detectable after prolonged incubation times; peak 4

Figure 3. Direct FAB-MS analysis after 8 min of the mixture of reaction products upon incubation of SCCG with guanosine. Ions at m/z 413/415,308, and 284 represent MH+ of SCCG, GSH, and guanosine, respectively. (A) W - (2-Aminoethy1)guanosine; (B) RT?-(2-aminoethyl)guanine;(C) S-[(2-hydroxyethyl)carbamoyll-GSH. Time dependences of the ion intensity ratios, m/z 195/413 (0)and m/z 327/413 (m), are shown in the insert.

represents an unidentified decomposition product which was also found upon spontaneous degradation of SCCG. Molecular information to derive a possible biological reaction pathway inducing DNA fragmentation was obtained by direct, dynamic FAB mass spectrometricanalysis (15, 16) of the reaction products between SCCG and guanosine. The time course of the reaction was monitored directly for approximately 20 min by recording intensity ratios of the corresponding molecular ion signals (insert in Figure 3). The FAB mass spectrum taken at 8 min reaction time (Figure 3) clearly shows protonated molecular ions [MH+] of (aminoethy1)guanosine (m/z 327) and (aminoethy1)guanine (m/z 195). Furthermore, MH+ ions of GSH (m/z 308), of SCCG, and of guanosine are also prominent (m/z 413/415;m / z 284). In addition, a major ion MH+ most probably results from hydrolytic formation of S-[(2-hydroxyethyl)carbamoyl]-GSH(m/z 395). The time dependence of the ion intensity ratio m/z 195/413 indicates a rapid formation of the N-(aminoethyl)guanine by depurination from (aminoethy1)guanine (see insert in Figure 3); however, a direct comparison with the reaction monitored by HPLC is not possible due to the different solvent system and concentration, and higher pH, used for the FAB-MS experiment. Since progressive evaporation of the solvent precluded analyses beyond approximately 25 min with the mass spectrometric system employed, an additional FAB spectrum was obtained with a aliquot of the reaction mixture after 72 h which clearly revealed the predominant molecular ion of the aminoethylated guanine. No ions were detected indicative for formation of an ethyl-bridged cross-link between glutathione and guanosine or guanine.

Discussion Several cases have been reported so far (3,8)in which toxic and mutagenic effects of alkyl halogenated compounds could be attributed to the formation of reactive glutathionyl derivatives or their metabolic successors. We have recently described the formation and isolation of

Stahl e t al.

108 Chem. Res. Toxicol., Vol. 5, No. 1, 1992 COOH

COOH

H2N-

NH-COOH

,,reo L NH

0

7

-

H2N-

HCI

\

tl -

NH-COOH

/

N

r,

COOH H2N-

NH-COOH HNvCO

f

0

NH

w

HOH2C

0

OH OH

-

/ iH

chloroethylamine that subsequently binds to N', or whether SCCG directly interacts with guanosine to an "-bound intermediate that releases glutathione and C02 to give N7-(aminoethyl)guanosine, is not clear at present. However, since 2 4 (hydroxyethyl)carbamoyl]glutathione was identified as a further reaction product in the system, a direct aqueous hydrolysis of SCCG appears unlikely. SCCG might thus function as a transport form for the potent alkylating agent 2-chloroethylamine. Comparison of BCNU with its water-soluble noncarbamoylating congener 3-(2-hydroxyethy1)-1-(2-chloroethy1)nitrosourea (HECNU), which is not able to form SCCG, clearly shows that the potential to form SCCG is not relevant for anticancer effectiveness of these compounds. It has, however, been shown that BCNU is a much stronger carcinogen and much more toxic than HECNU (18). This might well result from its potential to generate monofunctional alkylating (aminoethylating) intermediates in the presence of GSH in addition to those generated from the 2-(chloroethyl)-N-nitrosoureido part of BCNU. Moreover, by virtue of its strong carbamoylating potential, BCNU is an extremely potent inhibitor of glutathione reductase. This effect might play a key role for induction of pulmonary fibrosis observed in about 30% of patients treated with BCNU (19). Acknowledgment. We thank K. Haegele for expert assistance with the mass spectrometry analyses. This work has been supported in part by the Fonds der chemischen Industrie, Frankfurt, Germany, and the University of Konstanz.

__

NH z

References (1) Sies, H., and Ketterer, B., E&. (1988) Glutathione conjugation,

mechanisms and biological significance, Academic Press, London. (2) Larsson, A., Orrenius, S., Holmg-ren, A., and Mannervik, B., MS. (1983)Functions of glutathione, biochemical, physiological, toxicological and clinical aspects, Raven Press, New York. OH OH +

GSH

C02

Figure 4. Reaction pathway of guanosine with SCCG,leading to formation of the N'-(2-aminoethyl)guanine adduct.

SCCG as a glutathionyl adduct upon incubation of GSH with the anticancer agent BCNU (10, 17). SCCG forms by reaction of BCNU or its decomposition product, 2chloroethyl isocyanate, with the thiol group of GSH. SCCG proved to be a potent mutagen in Salmonella typhimurium strains TA 100 and TA 1535, indicating point mutations (10). The data obtained in the present study with a human lymphoblastoid line cell show that SCCG also is a potent DNA-damaging agent. HPLC and direct, dynamic FAB mass spectrometric analyses (15,16) provide evidence for a chemical reaction of SCCG with DNA bases, as demonstrated for guanosine. We therefore propose a new mechanism for induction of DNA damage by SCCG as shown in Figure 4. An aminoethyl group is transferred from the carbamoyl-GSH derivative SCCG to the N-7 position of guanosine. The glycosidic bond of the guanosine-N-7 adduct is subsequently cleaved, resulting in the liberation of W-(aminoethy1)guanine. In vivo, this would create an apurinic site, susceptible to strand break induction. W-(Aminoethy1)guanine has also been found to undergo rapid ring opening of the imidazole ring (11). The resulting ring-opened formamidopyrimidine lesion might be removed by a DNA glycosylase, leaving an apurinic site. Whether SCCG aminoethylates guanosine by liberating

(3) Monks, T., and Lau, S. (1987) Commentary: Renal transport processes and glutathione conjugate mediated nephrotoxicity. Drug Metab. Dispos. 4, 437-441. (4) Koob, M., and Dekant, W. (1991) Bioactivation of xenobiotics by formation of toxic glutathione conjugates. Chem.-Biol.Znteract. 77, 107-136. ( 5 ) van Bladeren, P. J. (1988) Formation of toxic metabolites from drugs and other xenobiotics by glutathione conjugation. Trends Pharmacol. Sci. 9, 295-299. (6) Bruggemann, I. M., Timmink, J. H. M., and van Bladeren, P. J. (1986) Glutathione and cysteine mediated cytotoxicity of allyl and benzyl isothiocyanate. Toxicol. Appl. Pharmacol. 83, 349. (7) Inskeep, P. B., Koga, N., Cmarik, J. L., and Guengerich, F. P. (1986) Covalent binding of 1,2-dihaloalkanesto DNA and stability of the major DNA adduct, S-[2-(N7-guanyl)ethyl]glutathione. Cancer Res. 46, 2839-2844. (8) Koga, N., Inskeep, P. B., Harris, T. M., and Guengerich, F. P. (1986) S-[2-(N"-guanyl)ethyl]glutathione, the major DNA adduct formed from 1,2-dibromoethane. Biochemistry 25, 2192-2198. (9) Guengerich, F. P., Crawford, W. M., Domoradzki, J. Y., Macdonald, D. L., and Watanabe, P. G. (1980) In vitro activation of 1,Z-dichloroethaneby microsomal and cytosolic enzymes. Toxicol. Appl. Pharmacol. 55,307-317. (10) Stahl,W., Denkel, E., and Eisenbrand, G. (1988) Influence of glutathione on the mutagenicity of 2-chloroethyl-nitrosoureas: Mutagenic potential of glutathione derivatives formed from 2chloroethyl-nitrosoureas and glutathione. Mutat. Res. 206, 459-465. (11) Mueller, N., and Eisenbrand, G. (1985) The influence of N7 substituents on the stability of "-alkylated guanosines. Chem.-Biol. Interact. 53, 173-181. (12) Janzowski, C., Jacob, C., Henn, E., Zankl, H., Pool-Zobel,B. L., and Eisenbrand, G. (1989) Investigations on organospecific metabolism and genotoxic effects of the urinary bladder carcinogen N-nitrosobutyl-3-carboxypropylamine(BCNP) and its analogues N-nitrosodibutylamine (NDBA) and N-nitrosobutyl-4-hydroxybutylamine (4-OH-NDBA). Toxicology 59, 195-209.

Chem. Res. Toxicol. 1992,5,109-115 (13) Sterzel, W., Bedford, P., and Eisenbrand, G. (1985)Automated determinationof DNA wing fluorochrome Hoechst 33258. Anal. Biochem. 147,462-467. (14) Meyn, R. E., Grdnina, D. J., and Fletcher, S. E. (1980)Repair of radiation damage in vivo. In Radiation Biology in Cancer Research (Meyn, R. E., and Withers, R., Eds.) pp 95-102,Raven Press, New York. (15) Heidmann, M., Fonrobert, P., Przybylski, M., Platt, K. L., Seidel, A., and Oesch, F. (1988)Conjugation reactions of polyaromatic quinones to mono- and bis-glutathionyladducts: direct analysis by fast atom bombardment mass spectrometry. Biomed. Environ. Mass Spectrom. 14,329-334. (16) Wirth, K.P., Junker, E., Roellgen, F. W., Fonrobert, P., and

109

Przybylski, M. (1989)Effects of cosolvents in glycerol assisted dynamic SIMS. Adv. Mass Spectrom. 11A,430-431. (17) Stahl, W., Krauth-Siegel, R. L., Schirmer, R. H., and Eisenbrand, G. (1987)A method to determine the carbamoylatingpotential of l-(2-chloroethyl)-l-nitrosoureas.ZARC Sci. Publ. 84,

191-193. (18) Eisenbrand, G., and Habs, M. (1980)Chronic toxicity and carcinogenicityof cytostatic N-Nitroso-(2-chloroethyl)ureasafter repeated intravenous application to rats. In Mechanisms of Tolcicity and Hazard Evaluation (Holmstedt et al., Eds.) pp 273-278, Elsevier, Amsterdam. (19) Weiss, R. B., Poster, D. S., and Penta, J. S. (1981)The nitrosoureas and pulmonary toxicity. Cancer Treat. Rev. 8,111-124.

Oxygen-Derived Free Radical and Active Oxygen Complex Formation from Cobalt( I I ) Chelates in Vitro Phillip M. Hanna,* Maria B. Kadiiska,? and Ronald P. Mason National Institute of Environmental Health Sciences, National Institutes of Health, P.O. Box 12233, Research Triangle Park, North Carolina 27709 Received July 8,1991

The electron paramagnetic resonance (EPR) spin trapping technique was used to study the generation of oxygen free radicals from the reaction of hydrogen peroxide with various Co(I1) complexes in pH 7.4 phosphate buffer. The 5,5-dimethyl-l-pyrroline N-oxide (DMPO) spin trap was used in these experiments to detect superoxide and hydroxyl free radicals. Superoxide radical was generated from the reaction of H202with Co(II), but was inhibited when Co(I1) was chelated with adenosine 5'-diphosphate or citrate. Visible absorbance spectra revealed no change in the final oxidation state of the cobalt ion in these samples. The EDTA complex also prevented detectable free-radical formation when H202was added, but visible absorbance data indicated oxidation of the Co(I1) to Co(II1) in this case. The amount of DMPO/'OOH adduct detected by EPR was greatly enhanced when H202reacted with the nitrilotriacetate complex relative to Co(I1) alone, and in addition, concurrent formation of the DMPO/'OH adduct due to slow oxidation of Co(I1) was observed. The hydroxyl radical adduct formation was suppressed by ethanol, but not DMSO, indicating that free hydroxyl radical was not formed. The deferoxamine nitroxide radical was exclusively formed when H202 was added to the Co(I1) complex of this ligand, most probably in a site-specific manner. In the presence of ethylenediamine, Co(I1) bound molecular O2and directly oxidized DMPO to its DMPO/'OH adduct without first forming free superoxide, hydroxyl radical, or hydrogen peroxide. An experiment using 170-enrichedwater revealed that the Co(II)-ethylenediamine complex caused the DMPO to react with solvent water to form the DMPO/'OH adduct. The relevance of these results to toxicological studies of cobalt is discussed.

Introduction Cobalt is an essential element in the human diet but, as with most metals, it is toxic in higher doses (1). The deleterious effects of cobalt toxicity are varied, and large doses Of cobalt Salts me believed to be carcinogenic (1-8, cause DNA damage (443, SupPreM hepatic hemoproteins, especially cytochrome p-450 (7-10), and 'Ontribute to chronic blood disorders (11,12)1among 0 t h effects (12). The chemical mechanisms through which cobalt salts exert their toxicity in vivo may involve, at least in part, the formation of active oxygen species such as the highly reactive hydroxyl radical. Under physiological conditions in vitro, Co(II) has been reported to promote the formation

'F+x"ent

address: InstitUte of Physiology, Academy of Sciences, 1113 Sofia, 'Academician Georgy Bonchev" Street, Building 23,Bulgaria.

of hydroxyl radicals or "reactive species" from hydrogen peroxide, which then proceeded to degrade deoxyribose sugar and hydroxylate aromatic compounds (13, 14). promotionof both lipid peroxidation (14, 15) and sitespecific hydroxyl radical damage to DNA (4,5) by Co(II) and hydrogen peroxide in vitro have alsobeen H ~direct EpR1 ~ ~ ~ using~fiespin~trapping , technique have shown clearly that the superoxide anion, not free hydroxyl is by ~ ~ ( 1 when 1) hydrogen peroxide is added in pH 7.4 phosphate buffer (16). The mechanism through which superoxide is generatedfrom co(n)and Hzozhas yet to be determind, but Abbreviations: EPR, electron paramagnetic resonance; SOD, superoxide dismutase; NTA, nitrilotriacetic acid; ADP, adenosine 5'-diphosphate; EDTA, ethylenediaminetetraacetic acid; DFO, deferoxamine mesylate; DMPO, 5,5-dimethyl-l-pyrrolineN-oxide;DMSO,dimethyl sulfoxide.

This article not subject to U.S. Copyright. Published 1992 by the American Chemical Society