Chem. Res. Toxicol. 1990, 3, 595-603
595
Reaction of Chromium(V1) with Hydrogen Peroxide in the Presence of Glutathione: Reactive Intermediates and Resulting DNA Damage Jayshree Aiyar,? Holly J. Berkovits,t Robert A. Floyd,f and Karen E. Wetterhahn*it Departments of Chemistry and Biochemistry, Dartmouth College, Hanover, New Hampshire 03755, and Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma 73104 Received July 3, 1990
T h e reaction of chromium(V1) with hydrogen peroxide was studied in t h e presence of glutathione. In vitro, reaction of chromium(V1) with hydrogen peroxide alone led t o production of hydroxyl radical as the significant reactive intermediate, while reaction of chromium(V1) with glutathione led t o formation of two chromium(V)-glutathione complexes and t h e glutathione thiyl radical. Incubation of chromium(V1) with glutathione prior t o addition of hydrogen peroxide led t o formation of peroxochromium(V) species and a dramatic increase in hydroxyl radical production over t h a t detected in the reaction of chromium(V1) with hydrogen peroxide alone. In contrast, addition of chromium(V1) t o a preincubated mixture of glutathione and hydrogen peroxide led to a decrease in hydroxyl radical production over t h a t obtained in t h e reaction of chromium(V1) with hydrogen peroxide. When pBR322 DNA was added t o the above reactions, the extent of chromium(V1)-induced DNA strand breakage correlated with the relative amount of hydroxyl radical formed. Reaction of chromium(V1) with calf thymus DNA in the presence of a preincubated mixture of glutathione and hydrogen peroxide led t o detection of the 8hydroxydeoxyguanosine adduct, whose formation correlated with t h a t of hydroxyl radical production. No significant chromium-DNA adduct formation was detected. The results suggest that, in the cellular metabolism of chromium(VI), preformed chromium(V)-glutathione complexes may react with hydrogen peroxide in a Fenton-type manner t o produce hydroxyl radical as the DNA-damaging agent. However, if glutathione reacts with hydrogen peroxide prior to exposure t o chromium(VI), t h e amount of hydroxyl radical generated may not be sufficient t o cause significant DNA damage. Thus, the metabolic pathway of chromium(V1) reduction is important in determining the nature and extent of chromium(V1)-induced DNA damage.
Introduction The DNA-damaging abilities of carcinogenic chromium(V1) compounds have been studied in a variety of in vivo, cultured cell and in vitro systems (1-9). The type and extent of chromium(V1)-induced DNA damage in the cell are likely to be determined by more than one chromium(V1) reduction pathway, and by the sequence in which these pathways occur. In chicken embryo hepatocytes, Cup0 and Wetterhahn (2) found that induction of cytochrome P-450 under conditions in which the glutathione content of the cells had been depleted led to little DNA strand breakage. However, induction of cytochrome P-450 in cells having elevated levels of glutathione led to a greater amount of DNA strand breakage than that seen in the presence of elevated levels of glutathione alone (2), suggesting that a synergism exists between the chromium(V1) reductive pathways involving cytochrome P-450 and glutathione. Sugiyama et al. (4) showed that treatment of Chinese hamster V-79 cells with vitamin B, and chromium(V1) resulted in an increase in chromium(V1)induced DNA strand breaks over that observed upon treatment of the cells with chromium(V1) alone. In vitro *To whom correspondence should be addressed at the Department of Chemistry, Steele Hall, Dartmouth College, Hanover, NH 03755.
Dartmouth College. *Oklahoma Medical Research Foundation.
studies suggested that the increase in DNA strand breakage was due to enhanced chromium(V)-related hydroxyl radical formation in the presence of vitamin B, (4). In contrast, vitamin E is a hydroxyl radical scavenger, and incubation of Chinese hamster V-79 cells with vitamin E prior to treatment with chromium(V1) led to a decrease in chromium(V1)-induced DNA strand breaks ( 5 ) . Previous studies have shown that the nature of chromium(V1)-induced DNA damage is strongly dependent on the reactive intermediates, i.e., chromium(V), hydroxyl radical, or glutathione thiyl radical produced upon reaction of chromium(V1) with reductants such as glutathione and hydrogen peroxide (7-9). Reaction of chromium(V1) with DNA in the presence of glutathione led to formation of chromium-DNA adducts, which correlated with the production of chromium(V)-glutathione complexes (8,9). The chromium-DNA adducts have been characterized as glutathione-chromium-DNA cross-links (10). On the other hand, reaction of chromium(V1) with DNA in the presence of hydrogen peroxide led to generation of hydroxyl radical as the major DNA-damaging species, giving rise to detectable DNA strand breakage and 8-hydroxydeoxyguanosine (7-9). The purpose of the present study was to examine the nature of chromium(V1)-induced DNA damage in the presence of more than one reductant, namely, hydrogen peroxide and glutathione. Such a study is important because, in cells, chromium(V1) is likely to encounter several enzymatic and nonenzymatic reductants 0 1990 American Chemical Society
596
Chem. Res. Toxicol., Vol. 3, No. 6, 1990
(11). In recent studies, Shi and Dalal (12, 13) have observed that reaction of chromium(V1) with several enzymatic and nonenzymatic reductants gives rise to chromium(V) complexes; the "preformed" chromium(V) complexes can then react with hydrogen peroxide to generate hydroxyl radical through a Fenton-like reaction. Our study shows that preformed chromium(V)-glutathione complexes can undergo reaction with hydrogen peroxide, leading to enhanced production of hydroxyl radical over that obtained upon reaction of chromium(V1) with hydrogen peroxide alone. In the event that hydrogen peroxide undergoes rapid reaction with the reductant, e.g., glutathione, prior to reaction with chromium(VI), the formation of chromium(V)-glutathione complexes does not occur, and hydroxyl radical formation may be suppressed due to the decreased availability of hydrogen peroxide. The level of hydroxyl radical produced correlated with the extent of hydroxyl radical mediated DNA damage, i.e., single strand breakage and 8-hydroxydeoxyguanosine observed in these reactions. Thus, this study highlights the importance of the sequence by which the metabolic reduction of chromium(V1) occurs.
Experimental Section Chemicals. The plasmid DNA pBR322 (Form I), proteinase K, ultrapure sodium dodecyl sulfate (SDS),' and agarose were purchased from Bethesda Research Laboratories, Gaithersburg, MD. Absolute ethanol was purchased from Aaper Chemical Co., Shelbyville, KY. The molecular weight marker h DNA-BstEII digest was purchased from New England Biolabs, Beverly, MA. Atomic absorption standard potassium dichromate solution (K2Cr207,1000 ppm) and sodium dichromate (Na2Cr207.2H20) were purchased from Fisher Scientific Co., Medford, MA. Reduced glutathione (GSH), 5,5'-dithiobis(2-nitrobenzoicacid) (Ellman's reagent), RNase A, type I calf thymus DNA, tris(hydroxymethy1)aminomethane hydrochloride (Tris-HCl),and 3,5diaminobenzoic acid hydrochloride (DABA) were purchased from Sigma Chemical Co., St. Louis, MO. Atomic absorption reference ferric chloride solution (FeCI,, loo0 ppm) was obtained from VWR Scientific Inc., San Francisco, CA. Hydrogen peroxide was obtained as a 30% solution from J. T. Baker Chemical Co., Phillipsburg, NJ. The mixed-bed ion-exchange resin AG 501-X8 (20-50 mesh), Chelex 100 (100-200 mesh, sodium form) analytical grade chelating resin, and AG 50W-X8 (100-200 mesh, sodium form) cation-exchange resin were obtained from Bio-Rad Laboratories, Richmond, CA. Diethylenetriaminepentaacetic acid (DETAPAC), 5,5-dimethyl-l-pyrroline N-oxide (DMPO), and 2,2-diphenyl-l-picrylhydrazyl radical (DPPH) were obtained from Aldrich Chemical Co., Milwaukee, WI. For the EPR studies at 77 K, quartz EPR tubes (4-mm o.d., 3-mm i.d.) were purchased from Wilmad Glass Co. Inc., Buena, NJ. Disodium ethylenediaminetetraacetate (EDTA) and melting point capillaries for the EPR studies at 297 K (Corning brand Pyrex, 1.1-1.2-mm i.d.1 were obtained from Fisher Scientific Co., Springfield, NJ. Activated carbon (50-200 mesh) was purchased from Fisher Scientific Co., Pittsburgh, PA. Reaction of Chromium(V1)with a Preincubated Mixture of Hydrogen Peroxide and Glutathione: EPR Detection of Chromium(V) Complexes at 77 K. All solutions were treated with AC 501-X8 inn-exchange resin (4 g/L) for several hours and filtered before use, and all reactions described were carried out in a total volume of 5.0 mL. Glutathione (5.4, 10.8, or 18.0 mM) was preincubated with 18.0 mM hydrogen peroxide for 2 min, after which chromium(V1) (1.8 mM, 0.9 mM Na2Cr2O7.2H20)was added and the incubation was continued for a further period of 28 min in 25 mM Tris-HC1, pH 8.0, 37 "C. A t various times, aliquots (500 MI,)of the reaction were frozen in liquid nitrogen and their Abbreviations: DETAPAC, diethylenetriaminepentaacetic acid; DMPO, 5,5-dimethyl-l-pyrroline N-oxide; DPPH, 2,2-diphenyl-l-picrylhydrazyl radical; EDTA, disodium ethylenediaminetetraacetate:GSH, reduced glutathione;GSSG, oxidized glutathione; 8-OH-dG,E-hydroxy2'-deoxyguanosine; SDS, sodium dodecyl sulfate.
Aiyar et al. EPR spectra recorded at 77 K using a Varian V-4500 spectrometer with 100-kHz field modulation, 8.3-G modulation amplitude, 250-G sweep width, 300-ms time constant, 0.8-mW output microwave power, 2.5-min scan time, 9.212-9.215-GHz microwave frequency, and 5.0 X lo3 gain. A Hewlett Packard Model X532B frequency meter was used for calibration, and 2,2-diphenyl-1picrylhydrazyl radical (DPPH; g = 2.0036 f 0.0003) was used as a reference standard. The intensities of the EPR signals were determined as described earlier (9). Reaction of Chromium(V1) with Hydrogen Peroxide, Glutathione, or a Combination of Both Reductants: EPR Detection of Chromium(V), Glutathione Thiyl, and Hydroxyl Radical Intermediates at 297 K. All solutions were treated with AG 501-X8 ion-exchange resin (4 g/L) for several hours and filtered before use, and all reactions described were carried out in a total volume of 1.0 mL. (A) For the reaction of chromium(V1) with hydrogen peroxide, chromium(V1) (2.7 mM, 1.35 mM Na2Cr207.2H,0) was incubated with 27.0 mM hydrogen peroxide in 25 mM Tris-HC1, pH 8.0,24 "C. (B) For the reaction of chromium(V1) with glutathione, chromium(V1) (2.7 mM, 1.35 mM Na2Cr2O7*2H2O) was incubated with 27.0 mM glutathione in 25 mM Tris-HC1, pH 8.0, 24 "C. (C) For reactions in which chromium(V1) was added to a mixture of hydrogen peroxide and glutathione, glutathione (8.1,16.2, or 27.0 mM) was preincubated with 27.0 mM hydrogen peroxide for 2 min, after which chromium(V1) (2.7 mM, 1.35 mM Na2Cr207-2H20) was added, and the incubation was continued for a further period of 28 min in 25 mM Tris-HCI, pH 8.0, 24 "C. Alternatively, chromium(V1) (2.7 mM, 1.35 mM Na2Cr2O7.2H20)and 27.0 mM glutathione were preincubated for 2 min and hydrogen peroxide (27.0 mM) added for a further incubation period of 28 min in 25 mM Tris-HC1, pH 8.0,24 "C. All reactions (A-C) were carried out in the presence or absence of 0.10 M DMPO. The DMPO was purified by a slight modification of the method of Buettner and Oberley (14), as described earlier (8). The EPR spectra of the reactions were recorded at 297 K (24 "C) using a Bruker ESP-300 spectrometer with 100-kHz field modulation, 1.0-G modulation amplitude, 1 x IO5 receiver gain, 100-G sweep width, 5.12-ms time constant, 2-mW output microwave power, and 9.425-9.445-GHz microwave frequency. A Hewlett Packard Model X532B frequency meter was used for calibration, and 2,2-diphenyl-l-picrylhydrazyl radical (DPPH; g = 2.0036 f 0.0003) was used as a reference standard. The intensities of the EPR signals were determined by doubleintegration analysis of the first derivative spectra, as described earlier (8). Reaction of Calf Thymus DNA with Chromium(V1)in the Presence of Hydrogen Peroxide, Glutathione, or a Combination of Both Reductants. All buffer solutions were treated with AG 501-X8 ion-exchange resin and filtered before use. Solutions of calf thymus DNA (-2 mg/mL) were prepared in 25 mM Tris-HCI, pH 8.0 at 37 "C, and purified of contaminating iron by three successive Chelex 100 ion-exchange column chromatography steps (5-mL bed volume, 1-cm id., 1 mL/min flow rate), followed by dialysis against the solution buffer containing 1 mM DETAPAC for -12 h, and by dialysis against solution buffer for -12 h at 4 "C. DNA concentrations were determined by UV spectral analysis on a Hewlett Packard 8451A diode array spectrophotometer a t 260 nm using t260 = 6600 M-' cm-'. Iron concentrations were determined by atomic absorption spectroscopy using a wavelength of 248.3 nm on a Perkin-Elmer 503 atomic absorption spectrophotometer equipped with an HGA-2100 graphite furnace. The settings for the HGA furnace were as follows: drying time of 40 s at 110 "C, charring time of 60 s a t 1000 "C, and atomizing time of 10 s at 2700 "C. This method was sensitive to iron concentrations of 2.5 ppb (44.8 nM). Atomic absorption analysis of reaction mixtures containing DNA treated for iron removal indicated that 0.05) from that obtained upon reaction of chromium(V1) with hydrogen peroxide in the absence of glutathione (- 1213-fold above control), and the level stayed constant as the concentration of glutathione in the reaction mixture was increased (Figure 5). Thus the presence of glutathione in the reaction did not increase the level of chromium bound to DNA. This is in contrast to the reaction of DNA with chromium(V1) in the presence of glutathione alone which led to high levels (between 30-fold and 800-fold above control, depending on the concentration of glutathione used) of chromium binding to DNA (9). The results are consistent with a low level of chromium(V) complex formation leading to a low level of chromium-DNA adducts. 8-Hydroxydeoxyguanosine Adduct. The level of 8-OH-dG obtained upon reaction of chromium(V1) with DNA in the presence of a preincubated mixture of glutathione and hydrogen peroxide was significantly greater than that obtained in control incubations consisting of DNA, chromium(V1) and glutathione, DNA and glutathione, or DNA, glutathione, and hydrogen peroxide (Figure 6). However, the level of 8-OH-dG was less than that obtained in the reaction of chromium(V1) with DNA in the presence of hydrogen peroxide and no added glu-
Figure 6. Level of the 8hydroxydeoxyguanosine adduct produced upon reaction of calf thymus DNA with chromium(V1) in the presence of a preincubated mixture (2 min) of hydrogen peroxide and glutathione. Calf thymus DNA (1.8 mM) was incubated with (A) 1.8 mM chromium(V1) (0.9 mM Na2CrzO7.2H20)and in-
creasing concentrations-5.4, 10.8, or 18.0 mM-of glutathione (open triangles), (B) increasing concentrations-5.4,10.8, or 18.0 mM-of glutathione (closed squares),(C) preincubated mixtures ( 2 min) of 18.0 mM hydrogen peroxide + increasing concentrations-5,4, 10.8, or 18.0 mM-of glutathione (closed triangles), or (D) 1.8mM chromium(V1) (0.9 mM Na2Cr207.2H20) and preincubated mixtures (2 min) of 18.0mM hydrogen peroxide + increasing concentrations-5.4,10.8, or 18.0 mM-of glutathione (closed circles). Samples were repeatedly dialyzed and ethanol precipitated as described under Experimental Section. The 8hydroxydeoxyguanosine adduct (8-OH-dG)was determined by HPLC analysis with electrochemical detection, and DNA, by microfluorometric assay. The control 8-OH-dG/DNAratio [2.68 (k1.26) X 10-5nmol of 8-OH-dG/nmolof DNA-P] represents the basal level of adduct present in samples consisting of calf thymus DNA alone. Results are the mean f SD of three to six determinations. tathione, and the level of 8-OH-dG decreased as the concentration of glutathione in the reaction was increased. The trend in 8-OH-dG formation mirrored that of hydroxyl radical production in these reactions (Figure 2), indicating that formation of 8-OH-dG was most likely hydroxyl radical mediated. Chromium(V1)-Induced DNA Strand Breaks. The extent of chromium(V1)-induced DNA strand breakage was determined by agarose gel electrophoresis of reaction mixtures consisting of supercoiled (form I) pBR322 DNA in the presence or absence of chromium(VI), glutathione, and hydrogen peroxide. The pBR322 DNA used in this study consisted of 64.1 f 1.9% supercoiled (form I) DNA and 33.9 f 1.0% nicked circular (form 11) DNA, as determined by densitometry of control reactions consisting of DNA alone, DNA and hydrogen peroxide, or DNA, chromium(VI), and glutathione (Figure 7). Reaction of pBR322 DNA with chromium(V1) in the presence of hydrogen peroxide led to an increase in the relative amount of form I1 DNA to 59% (Figure 7). When glutathione was preincubated (2 min) with hydrogen peroxide and DNA prior to addition of chromium(VI), the relative increase in the amount of form I1 DNA stayed a t approximately 59% (Figure 7), although under these conditions, a decrease in hydroxyl radical production was observed relative to that seen in the reaction of chromium(V1) with hydrogen peroxide alone (Figure 1). This may be due to the lower sensitivity of the nicking assay as compared to the detection of radical species by EPR spectroscopy. Preincubation of pBR322 DNA with chromium(V1) and glutathione for 2 min prior to addition of hydrogen peroxide led to a dramatic increase in the amount of nicked circular form 11 DNA (72%) relative to control DNA samples (34%) (Figure 7 ) , which correlated with the dramatic in-
600 Chem. Res. Toxicol., Vol. 3, No. 6 , 1990
Aiyar et al.
80 T
60
YO 40
20
0 A
B
C
D
Figure 7. 1,evel of DNA strand breakage produced upon reaction of pBR322 plasmid DNA with chromium(V1) in the presence of
hydrogen peroxide and glutathione. Reactions were carried out in 50 mM Tris-HC1,pH 8.0,37 "C. pBR322 plasmid DNA (0.10 mM DNA-P) was preincubated at 37 "C for 2 min with GSH (18 mM) and either H202(18 mM) (C) or K2Cr207[0.9 mM, 1.8 mM Cr(VI)](D). Addition of Cr(V1) (C) or H202(D) was followed by further incubation for 28 min at 37 "C. Reaction A represents the mean i SD of control reactions consisting of DNA alone, DNA + H202,or DNA + Cr(V1) + GSH; reaction B consisted of a 30-min incubation of pBR322 with Cr(V1) and H20z: Samples were analyzed by agarose gel electrophoresis,as described under Experimental Section, and DNA was visualized with ethidium bromide, photographed and quantitated by densitometry. Closed bars: Form I (supercoiled) DNA. Hatched bars: Form I1 (nicked circular) DNA. Results are the mean i SD of two determinations. crease in hydroxyl radical production during the reaction (Figure 1).
Discussion We have shown that when more than one reductant, e.g., glutathione and hydrogen peroxide, is available for reaction with chromium(VI), the level of DNA-damaging reactive intermediates, e.g., hydroxyl radical, produced is dependent on the sequence in which chromium(V1) reacts with each of the reductants. If chromium(V1) reacts with glutathione to form chromium(V)-glutathione complexes prior to reaction with hydrogen peroxide, a dramatic increase in hydroxyl radical production is observed relative to that seen upon reaction of chromium(V1) with hydrogen peroxide alone. Thus, the chromium(V)-glutathione complexes may undergo a Fenton-type reaction with hydrogen peroxide to generate hydroxyl radical. Reaction of the chromium(V)-glutathione complexes with hydrogen peroxide is rapid, since no EPR spectroscopic evidence of chromium(V)-glutathione complex formation or of glutathione thiyl radical formation is seen upon addition of hydrogen peroxide to the reaction of chromium(V1) with glutathione. However, the high level of hydroxyl radical generated by reaction of chromium(V)-glutathione complexes with hydrogen peroxide is accompanied by formation of two chromium(V) complexes whose E P R signals have characteristic g values of 1.979 and 1.974. Although the chromium(V) EPR signal at g = 1.974 was very weak, the g value suggested that the signal represented the tetraperoxochromium(V) species (g = 1.9735) reported by Kawanishi et al. ( 7 ) . The g = 1.979 chromium(V) species may be formed by removal of one or more of the peroxo ligands from the g = 1.974 tetraperoxochromium(V) species as described earlier ( 7 ) . On the other hand, it may represent "mixed complexes" such as a peroxo-chromium(V)-glutathione complex or a peroxo-chromium(V)-Tris complex analogous to the ascorbate-chromium(V)-Tris complex kaV = 1.976) described by Goodgame and Joy (21).
In an earlier study (91, we showed that reaction of chromium(V1) with hydrogen peroxide did not proceed in a catalytic Fenton-type reaction involving the tetraperoxochromium(V) intermediate. The present study suggests that while the reaction of chromium(V1) with hydrogen peroxide may not lead to significant hydroxyl radical production, preformed chromiumtv) complexes such as chromium(V)-glutathione complexes may undergo efficient Fenton-type reactions with hydrogen peroxide to generate hydroxyl radical. The ability of chromium(V) complexes to react with hydrogen peroxide more efficiently than chromium(V1) has been documented in other studies ( 4 , 12, 13). Sugiyama et al. ( 4 ) have shown vitamin B, enhances the formation of both hydroxyl radical and tetraperoxochromate(V) during the reaction of sodium chromate with hydrogen peroxide. Like glutathione, vitamin B2 reacts with chromium(V1) to form a chromium(V) intermediate (gav = 1.9771, and this species, in a manner similar to that of the chromium(V)-glutathione complexes described in our study, has been postulated to be responsible for increased hydroxyl radical and tetraperoxochromium(V) production when the reaction of chromium(V1) with hydrogen peroxide is carried out in the presence of vitamin B2 ( 4 ) . A similar observation has been made by Shi and Dalal (12) upon addition of hydrogen peroxide to the glutathione reductase catalyzed reduction of chromate in the presence of NADPH. In this case, a chromium(V)-NADPH intermediate (gav= 1.9785) was shown to undergo a Fenton-type reaction with hydrogen peroxide, resulting in hydroxyl radical production. One explanation of these results is that formation of the tetraperoxochromium(V) species, which may be the species that is directly responsible for production of hydroxyl radical ( 7 ) , is facilitated by preforming chromium(V) complexes which can then undergo facile ligand displacement reactions with hydrogen peroxide to give the tetraperoxochromium(V) species. This interpretation would be consistent with our detection of peroxochromium(V) species upon addition of hydrogen peroxide to a preincubated mixture of chromium(V1) and glutathione. Thus, one or more of the chromium(V)-glutathione complexes may undergo a ligand substitution reaction with hydrogen peroxide, giving rise to the tetraperoxochromium(V) species with possible oxidation of glutathione:
-
CrV(GSH), + ( n + 4)H,02 CrV(02)43- (2n)H20+ (n/2)GSSG
+
+ 8H+,
n = 1-4 (1)
Kawanishi et al. showed that reaction of the tetraperoxochromium(V) species with hydrogen peroxide to generate hydroxyl radical involved superoxide formation, since addition of superoxide dismutase (SOD) to the reaction resulted in 90% inhibition of production of the DMPOhydroxyl radical adduct (7). Thus, if the preformed chromium(V)-glutathione complexes were quantitatively converted to the tetraperoxochromium(V) species (eq 1) which is then responsible for hydroxyl radical production, one would expect the reaction of chromium(V)-glutathione complexes with hydrogen peroxide to be SOD-inhibitable. Shi and Dalal have reacted chromium(V1) with various reductants including NADPH, several diols, a-hydroxycarboxylic acids, and glutathione to form chromium(V) complexes and have shown that each of these chromium(V) complexes reacts with hydrogen peroxide efficiently to generate hydroxyl radical in a Fenton-type fashion (13). However, they found that reaction of the preformed chromium(V) complexes with hydrogen peroxide to form
Cr( V I ) Reaction with H,O,IGSH and D N A Damage hydroxyl radical did not involve superoxide formation (13). Their results are in contrast to those obtained by Kawanishi et al. ( 7 ) for reaction of the tetraperoxochromium(V) species with hydrogen peroxide. On the basis of Shi and Dalal’s results, it is unlikely that the preformed chromium(V)-glutathione complexes described in our study are quantitatively converted to tetraperoxochromium(V) by the ligand-exchange mechanism of eq 1, since tetraperoxochromium(V) reacts with hydrogen peroxide to form superoxide (7, 13). The more probable mechanism is one involving direct Fenton-type reaction of chromium(V)-glutathione complexes with hydrogen peroxide (13). However, a small amount of chromium(V)-glutathione species may participate in the reaction described by eq 1, since we have detected a low level of peroxochromium(V) species in these reactions. Addition of chromium(V1) to a preincubated mixture of glutathione and hydrogen peroxide resulted in decreased hydroxyl radical production and decreased DNA damage. Thus, the order of addition of a series of reductants, e.g., glutathione and hydrogen peroxide, to chromium(V1) is of prime importance in determining the extent to which reactive intermediates, e.g., hydroxyl radical, are produced. Although no chromium(V)-glutathione complex formation or thiyl radical formation was seen, very low levels of a chromium(V) species was formed upon reaction of chromium(V1) with a preincubated mixture of glutathione and hydrogen peroxide, as visualized by a weak axial chromium(V) EPR signal a t 77 K. The parallel component of the axial signal could not be clearly distinguished a t the low level of chromium(V) detected in these reactions; however, on the basis of earlier studies (7, 9), the chromium(V) EPR spectrum obtained in the present study was identified as that of the tetraperoxochromium(V) complex. The low level of tetraperoxochromium(V) complex formation correlated with the low level of chromium-DNA adducts detected in our study. These results are consistent with our earlier studies in which we showed that reaction of chromium(V1) with hydrogen peroxide led to a low level of tetraperoxochromium(V) and related chromium-DNA adduct formation, whereas reaction of chromium(V1) with glutathione led to a high level of chromium(V)-glutathione complexes and related chromium-DNA adduct formation (8, 9).
Reaction of chromium(V1) with a preincubated mixture of glutathione and hydrogen peroxide led to an initial increase (less than 1 min into the reaction) in tetraperoxochromium(V) production, followed by its rapid decay over a 30-min period. An overall decrease in hydroxyl radical production was observed over 30 min, and the decrease was proportional to the initial concentration of glutathione in the reaction. No EPR evidence of chromium(V)-glutathione complex formation or of DMPOglutathione thiyl radical adduct formation was obtained. The results suggest that reaction of chromium(V1) with a preincubated mixture of hydrogen peroxide and glutathione leads to an initial increase in tetraperoxochromium (VI production which should lead to increased hydroxyl radical production. However, due to the rapid decay of the tetraperoxochromium(V) intermediate, an overall decrease in hydroxyl radical production is observed over 30 min. When the preincubated mixture (2 min) of glutathione and hydrogen peroxide was assayed for glutathione concentration according to Ellman’s assay (20), the results clearly showed that reaction of glutathione with hydrogen peroxide was rapid, leading to less availability of hydrogen peroxide and glutathione for reaction with chromium(V1). The reaction of glutathione with hydrogen
Chem. Res. Toxicol., Vol. 3, No. 6, 1990 601
peroxide is of biological significance, providing a pathway for the detoxification of peroxides by glutathione, and is catalyzed by peroxidases (22), leading to formation of oxidized glutathione and water. This reaction has also been shown to occur to a significant extent (-80%) within 2 min of incubation of 250 mM glutathione with equimolar hydrogen peroxide at pH 7.6 in the absence of peroxidases or other catalysts (23). However, the reaction in vitro is more complex and results in several products including oxidized glutathione (GSSG; 31% ), the mono- and dioxides of oxidized glutathione [GS(SO)G and G(SO)(SO)G45%], and glutathione sulfinic acid (GSOOH; 6%) (23). In our study of the reaction of chromium(V1)with a preincubated mixture of glutathione and hydrogen peroxide, an initial increase in tetraperoxochromium(V) production may be due to a reductant-mediated reaction by the small percent of unreacted glutathione. However, the rapid reaction of glutathione with hydrogen peroxide during the preincubation period would make both reactants less available for reaction with chromium(V1). One or more of the end products resulting from the reaction of glutathione and hydrogen peroxide (23) may then be responsible for the rapid decay of the tetraperoxochromium(V) species. Both these effects would lead to a net decrease in hydroxyl radical production. In addition, the unreacted glutathione as well the products obtained after the 2-min preincubation of glutathione and hydrogen peroxide may act as hydroxyl radical scavengers. Our studies show that when considering the metabolic activation of chromium(V1) to reactive intermediates by more than one cellular reductant, it is important to examine conditions in the cell prior to exposure to chromium(V1). In chicken embryo hepatocytes, Cup0 and Wetterhahn (2)found that induction of cytochrome P-450 under conditions in which the glutathione content of the cells had been depleted led to little DNA strand breakage. However, induction of cytochrome P-450 in cells having elevated levels of glutathione led to a greater amount of DNA strand breakage than that seen in the presence of elevated levels of glutathione alone (2). In some instances, preformed chromium(V) complexes may be responsible for either directly causing DNA strand breakage or producing hydroxyl radical as the DNA-damaging agent (4, 5, 24). Thus, incubation of Chinese hamster V-79 cells with vitamin B2 and sodium chromate led to increased chromium(V1)-induced DNA single-strand breaks over that obtained upon treatment of the cells with sodium chromate alone (4). Sugiyama et al. ( 5 , 2 4 )showed that treatment of Chinese hamster V-79 cells with vitamin E prior to exposure to chromium(V1) resulted in a decrease in chromate-induced cytotoxicity and DNA strand breaks, together with a decrease in a chromium(V)-glutathione complex detected in these cells. However, it is conceivable that, under certain conditions, reducing agents such as glutathione may react with hydrogen peroxide and protect the cells from chromium(V1)-induced DNA strand breakage and other forms of oxidative damage. Studies have shown that the intracellular concentration of hydrogen peroxide may increase severalfold under certain conditions, e.g., upon irradiation with ultraviolet or visible light (25, 261, during the metabolism of some carcinogens (271,and at sites of inflammation where it is produced by phagocytic cells (28). If preformed chromium(V) complexes are available to react with the increased levels of hydrogen peroxide, then a high level of chromium(V1)-induced oxidative DNA damage may be observed. However, earlier studies have also shown that the steady-state intracellular concentration of hydrogen peroxide is maintained a t a low
602 Chem. Res. Toxicol., Val. 3, No. 6, 1990
Institute, DHHS. The EPR spectrometer was purchased with funding from NSF Grant CHE-8701406.
low .OH, DNA damage
GSH
C
AO2
+ H202 r
0
4
2
\
-
L
k
moderate .OH,
Aiyar et al.
References DNA damage
(1) Hamilton, J. W., and Wetterhahn, K. E. (1986) Chromium-
(VI)-induced DNA damage in chick embryo liver and blood cells in iiiv. Carcinogenesis 7, 2085-2088. ( 2 ) Cupo, D. Y., and Wetterhahn, K. E. (1985) Modification of
GSH
Cr(V) complexes
A+ high .OH, DNA damage
Figure 8. Importance of order of addition of reactants to chromium(V1)-mediated hydroxyl radical production and related DNA damage in the presence of hydrogen peroxide and/or glutathione. Reaction of chromium(V1) (Cr0,2-) with hydrogen peroxide leads to formation of hydroxyl radical which can cause DNA damage in the form of strand breaks and adducts, e.g., 8-hydroxydeoxyguanosine. However, reaction of preformed chromium(V)-glutathione complexes with hydrogen peroxide results in higher levels of hydroxyl radical being produced and a greater extent of DNA damage. Addition of chromium(V1) to a preincubated mixture of glutathione and hydrogen peroxide leads to a decrease in levels of hydroxyl radical and related DNA damage.
level ( 10-8-10-9 M ) by various detoxification reactions, the primary one being its reaction with glutathione which is catalyzed by glutathione peroxidase (29,30). Thus, it is possible that the cellular tripeptide glutathione, whose intracellular concentration may be as high as 10 mM (311, reacts with hydrogen peroxide prior to exposure to chromium(VI1, thus decreasing the extent of chromium(V1)induced DNA strand breakage and other oxidative DNA damage. It is possible that the nature and extent of chromium(V1)-induced DNA damage varies from tissue to tissue, due to differences in the relative levels of glutathione and hydrogen peroxide present in the different tissues; e.g., treatment with chromium(V1) resulted in DNA-protein and DNA-interstrand cross-links with little DNA strand breakage in rat liver, kidney, and lung tissues (32) and chick embryo liver (I), whereas in chick embryo red blood cells, the chromium(V1)-inducedDNA damage was primarily in the form of strand breaks ( I ) . The results obtained in this study and those of other workers have been summarized in Figure 8. Reaction of chromium(V1) with reducing agents such as glutathione gives rise to chromium(V) complexes which can react with hydrogen peroxide in a Fenton-type fashion, leading to significant hydroxyl radical production and related DNA damage. The extent of hydroxyl radical production and related DNA damage observed upon reaction of chromium(V1) with hydrogen peroxide is less than that seen upon reaction of preformed chromium(V) complexes with hydrogen peroxide. Addition of chromium(V1) to a preincubated mixture of glutathione and hydrogen peroxide leads to a decrease in hydroxyl radical production over that obtained upon reaction of chromium(V1) with hydrogen peroxide alone, thus reducing the extent of chromium(VI)-mediated DNA damage. These studies have highlighted the importance of the synergistic interaction of more than one cellular reductant in determining the nature and extent of chromium(V1)-induced DNA damage. Acknowledgment. We thank Dr. Dean E. Wilcox for helping with the EPR studies, Melinda West and Kenton Eneff for performing the HPLC analyses, and Dr. Peter A. Friedman for use of the Jarrell Ash atomic absorption spectrophotometer. This investigation was supported b y USPHS Grants CA34869 (K.E.W.), CA45735 (K.E.W.), and CA42854 (R.A.F.) awarded b y the National Cancer
chromium(V1)-induced DNA damage by glutathione and cytochromes P-450 in chicken embryo hepatocytes. Proc. Natl. Acad. Sci. U.S.A. 82, 6755-6759. (3) Sugiyama, M., Ando, A., Nakao, K., Ueta, H., Hidaka, T., and Ogura, R. (1989) Influence of vitamin B2 on formation of chromium(V), alkali-labile sites, and lethality of sodium chromate(V1) in Chinese hamster V-79 cells. Cancer Res. 49, 6180-6184. (4) Sugiyama, M., Ando, A., and Ogura, R. (1989) Vitamin B2-enhancement of sodium chromate(V1)-induced DNA single strand breaks: ESR study of the action of vitamin B,. Biochem. Biophys. Res. Commun. 159, 1080-1085. (5) Sugiyama, M., Ando, A., Furuno, A., Burr Furlong, N., Hidaka, T., and Ogura, R. (1987) Effects of vitamin E, vitamin B, and selenite on DNA single strand breaks induced by sodium chromate(V1). Cancer Lett. 38, 1-7. (6) Kortenkamp, A., Ozolins, Z., Beyersmann, D., and O’Brien, P. (1989) Generation of PM2 DNA strand breaks in the course of reduction of chromium(V1) by glutathione. Mutat. Res. 216, 19-26. (7) Kawanishi, S., Inoue, S., and Sano, S. (1986) Mechanism of DNA cleavage induced by sodium chromate(V1) in the presence of hydrogen peroxide. J . Biol. Chem. 261, 5952-5958. (8) Aiyar, J., Borges, K. M., Floyd, R. A., and Wetterhahn, K. E. (1989) Role of chromium(V), glutathione thiyl radical and hydroxyl radical intermediates in chromium(V1)-induced DNA damage. Toxicol. Enuiron. Chem. 22, 135-148. (9) Aiyar, J., Berkovits, H. J., Floyd, R. A., and Wetterhahn, K. E. (1990) Reaction of chromium(V1) with glutathione or with hydrogen peroxide: Identification of reactive intermediates and their role in chromium(V1)-induced DNA damage. Enuiron. Health Perspect. (in press). (10) Borges, K. M., and Wetterhahn, K. E. (1989) Chromium cross-links glutathione and cysteine to DNA. Carcinogenesis 10, 2165-2168. (11) Connett, P. H., and Wetterhahn, K. E. (1983) Metabolism of the carcinogen chromate by cellular constituents. Struct. Bond. 54, 93-124. (12) Shi, X., and Dalal, N. S. (1989) Chromium(V) and hydroxyl radical formation during the glutathione reductase-catalyzed reduction of chromium(V1). Biochem. Biophys. Res. Commun. 163, 627-634. 13) Shi, X., and Dalal, N. S. (1990) On the hydroxyl radical formation in the reaction between hydrogen peroxide and biologically generated chromium(V) species. Arch. Biochem. Biophys. 277, 342-350. 14) Buettner, G. R., and Oberley, L. W. (1978) Considerations in the spin trapping of superoxide and hydroxyl radical in aqueous Biochem. Biosystems using 5,5-dimethyl-l-pyrroline-l-oxide. phys. Res. Commun. 83, 69-74. 15) Kissane, J. M., and Robins, E. (1958) The fluorometric measurement of deoxyribonucleic acid in animal tissues with special reference to the central nervous system. J . Biol. Chem. 233, 184-188. (16) Ciccarelli, R. B., and Wetterhahn, K. E. (1982) Nickel distribution and DNA lesions induced in rat tissues by the carcinogen nickel carbonate. Cancer Res. 42, 3544-3549. (17) Floyd, R. A., Watson, J. J., Harris, J., West, M., and Wong, P. K. (1986) Formation of 8-hydroxydeoxyguanosine, hydroxyl free radical adduct of DNA in granulocytes exposed to the tumor oromoter. tetradeconvlohorbolacetate. Biochem. Biophvs. . - Res. kommun: 137, 841-846 (18) , , Flovd. R. A.. Watson. J. J.. Wone. P. K.. Altmiller. D. H., and Rickaid,’ R. C.’ (1986) Hydroxyl f;ee radical adduct of deoxyguanosine: sensitive detection and mechanisms of formation. Free Radical Res. Commun. 1, 163-172. (19) Floyd, R. A., West, M. S., Eneff, K. L., Hogsett, W. E., and Tingey, D. T. (1988) Hydroxyl free radical mediated formation of 8-hydroxyguanine in isolated DNA. Arch. Biochem. Biophys. 262, 266. (20) Ellman, G. L. (1959) Tissue sulfhydryl groups. Arch. Biochem. Biophys. 82, 70-77.
Cr( V I ) Reaction with H202/GSH and D N A Damage (21) Goodgame, D. M. L., and Joy, A. M. (1987) EPR study of the Cr(V) and radical species produced in the reduction of chromium(V1) by ascorbate. Inorg. Chim. Acta 135, 115-118. (22) Harman, L. S., Carver, D. K., Schreiber, J., and Mason, R. P. (1986) One- and two-electron oxidation of reduced glutathione by peroxidases. J . R i d . Chem. 261, 1642-1648. (23) Finley, J . W., Wheeler, d. L., and Witt, S. C. (1981) Oxidation of glutathione by hydrogen peroxide and other oxidizing agents. J . Agric. Food Chem. 29, 404-407. (24) Sugiyama, M., Ando, A., and Ogura, R. (1989) Effect of vitamin E on survival, glutathione reductase and formation of chromium(V) in Chinese hamster V-79 cells treated with sodium chromate(V1). Carcinogenesis 10, 737-741. (25) Hoffmann, M. E., and Meneghini, R. (1979) DNA strand breaks in mammalian cells exposed to light in the presence of riboflavin and tryptophan. Photochem. Photobiol. 29, 299-303. (26) Parshad, R., Sanford, K. K., Jones, G. M., and Tarone, R. E. (1978) Fluorescent light-induced chromosome damage and its prevention in mouse cells in culture. Proc. Natl. Acad. Sci. U.S.A.
Chem. Res. Toxicol., Vol. 3, No. 6,1990 603 75, 1830-1833. (27) Lorentzen, R. J., and Ts’o, P. P. (1977) Benzo[a]pyrenedione/ benzo[alpyrenediol oxidation-reduction couples and the generation of reactive reduced molecular oxygen. Biochemistry 16, 1467-1473. (28) Simon, R. H., Scoggin, C. H,, and Patterson, D. (1981) Hydrogen peroxide causes the fatal injury to human fibroblasts exposed to oxygen radicals. J . Biol. Chem. 256, 7181-7186. (29) Kaplowitz, N., Aw, T. Y., and Ookhtens, M. (1985) The regulation of hepatic glutathione. Annu. Reu. Pharmacol. Toxicol. 25, 715-744. (30) Boveris, A., and Chance, B. (1973) The mitochondrial generation of hydrogen peroxide. Biochem. J. 134, 707-716. (31) Meister, A,, and Anderson, M. E. (1983) Glutathione. Annu. Rev. Biochem. 52, 711-760. (32) Tsapakos, M. J., Hampton, T. H., and Wetterhahn, K. E. (1983) Chromium(V1)-induced DNA lesions and chromium distribution in rat liver, kidney and lung. Cancer Res. 43, 5662-5667.
