Coordination chemistry and the carcinogenicity and mutagenicity of chromium(VI) Paul O'Brien and Guofang Wang Department of Chemistry, Queen Mary College, University of London, Mile End Road, London E1 4NS, England Abstract Chromate is a known carcinogen, it is only in recent years that the molecular mechanisms by which this toxicity may be expressed have been investigated. The toxicity of chromate may be mediated by the reaction of chromium(VI) with glutathione (GSH) to generate relatively stable chromium(V) complexes and other more reactive intermediates. The conditions favouring the formation of such complexes have been studied. Reactive intermediates generated during the reduction of chromate by GSH include thionyl radicals and at least two relatively stable chromium(V) species (g -1.996 and g -1.986). Mixtures of chromium(VI) and glutathione and a chromium(V) complex of glutathione, which we have isolated from the reaction (g = 1.996), are capable of causing strand breaks in bacteriophage PM2 DNA. In contrast a chromium(Ill) complex of GSSG, one of the final products of the reaction between GSH and chromium(VI), does not damage DNA in closed circle assays. These observations support the suggestion that reactive intermediates generated during the reduction of chromium(VI) provide one route by which the genotoxicity of chromate may be expressed.

Introduction

There is considerable current interest in the carcinogenicity and mutagenicity of chromium(VI) (Connett and Wetterhahn, 1983; Bianchi and Levis, 1987). Chromate is a well established carcinogen in animals (Langard, 1983; Glaser, 1986) and epidemiological evidence points to it being responsible for lung and nasal cancers in workers exposed to chromate, notably eleclIo platers (Gray, 1987) and those involved with chromium containing pigments (Langard and Vigander, 1983). At a molecular level it is of interest that the chromate ion, CrO4 2-, the dominant form of chromium(VI) in neutral aqueous solutions, can readily cross cellular membranes via nonspecific anion carriers (Connett and Wetterhahn, 1983; Kortenkamp et al. 1987; O'Brien and Ozolins 1989a; Wiegand et a/.,1985). Detailed studies of model systems support the suggestion of a facile uptake mechanism for chromate (Wiegand et al., 1985; Debbetto 1988) and the widespread use of 51Cr labelled chromate to tag erythrocytes (Gray and Sterling, 1950) is based on the fact that once within such cells chromium, in a reduced form, is immobilized. In contrast, it is in general difficult for chromium(III) complexes to enter cells; although certain ligands may greatly facilitate uptake (Kortenkamp et al., 1987). The two step process leading to toxicity has been termed the 'uptake- reduction' model (Connett and Wetterhahn, 1983). Chromium(VI) has been shown to damage DNA, both in vivo and in vitro, in a number of ways including: binding to DNA and chromatin (Cupo and Wetterhahn, 1985a), causing intrasWand crosslinks (Tsapakos et al., 1981), crosslinking to proteins (Fornace et al., 1981) and by

causing strand breaks (Cupo and Wetterhahn, 1985b; Kawanishi et aL, 1986; Aiyar eta/., 1989; Kortenkamp et al. 1989). Some years ago Cupo and Wetterhahn (1985b) observed that more strand breaks were observed in chick embryo hepatocytes in which glutathione (7-glutamylcysteinylglycine, GSH) had been induced. A number of groups of workers have hence become interested in substantiating the hypothesis that the interaction of chromate with GSH within cells generates species capable of causing strand breaks in DNA (Aiyaret a/., 1989; Barrett et al., 1985; Goodgame and Joy, 1986; Kawanishi et aL, 1986; Kitagawa et al., 1988; O'Brien et al., 1987, 1989b,). In this article we will particularly emphasise the possibility that intermediates generated in the reduction of chrominm(VI) lead to the generation of strand breaks. It is important to point out that in taking this approach we are emphasising one end point for chromate toxicity and it may be that in vivo toxicity is actually expressed via another end points such as crosslinking. In this context here are a number of papers which discuss the binding of chromium0II) to DNA and nucleotides (Tsapakos and Wetterhahn, 1983; Beyersmann and Koster, 1987). We have been studying the complexes formed by glutathione and chromium for a number of years (Abdullah et al., 1985; Barrett et al., 1985), and have become interested to discover if the ultimate genotoxic form of chromium is an intermediate chromium complex involving glutathione, in one of the oxidation states ((V), (IV) or (III)), potentially formed after the initial absorption and reduction of chromate. We are particularly interested in a hypothesis in which genotoxicity is expressed by via a relatively stable chromium(V) species (Barrett et al., 1985; O'Brien et al., 1987, 1989b). However, another possibility

78

Coordination chemistry and the carcinogenicity and mutagenicity of chromium(Vl)

d J' G p9

9 E'"c

H

:

'

9

[11 Erythrocytes

I

4.0

0,,

[2]

=7 I

I

I

t

[3] [4] I

I

3.0

2.0

t41 t31 o,,

0

ppm

p

/C-CH-CH2-CH2-C-NH-CH-C-NH-CH2 - C H0

NH2

CH2 [21

[11

OH

SH F i g u r e 1 1H nmr spectra of erythrocytes. Results of a typical experiment: initial spectrum of erythrocytes, 5-rain

aquisition; and a 5-min aquisition, 5 rain after the addition of chromate 0.1 raM, pH = 7.0. is that DNA is damaged by the in vivo generation of hydroxyl radicals from some reaction involving GSH, chromium in one of its higher oxidation states and molecular oxygen (Kawanishi et al. 1986; Aiyar et al. 1989). In this paper we review, with reference to the more recent literature, some of our studies of the reaction of glutathione with chromate. Two aspects of this chemistry are of interest: firstly any effects of reaction conditions, especially buffer on the pathway for the reduction of

chromate, and secondly the role of chromium(V) complexes in this reaction and their relevence to the damage of DNA

Chromium(VI)

and I n t a c t E r y t h r o c y t e s

The 1H NMR spectrum of GSH within intact human erythrocytes can readily be recorded the method has been described in detail by Rabenstein et 02. (1985). We have used this method to show that chromate readily passes into

Paul O'Brien and Guofang Wang

79

%

0.7

0.6

0.5

0.4

A

0.3

0.2

0.1

0

200

400

600

800

Figure 2 Typical non linear fits of Aobs (650 nm) vs. time; x- axis time seconds; y-axis absorbance. Fitted using (A) [GSH] = 0.2 mole dm "3, klobs = 1.72 x 10 -2 and k2obs = 2.92 x 10"3; 03) [GSH] = 0.5 mole dm "3, klo~ = 7.97x lO-2 and k2obs = 6.23 x 10" 3 (pH = 7.0, 20~ [Cr04 2-] = 1 x 10 -3 mole dm'3 ). The fit is quite poor at the higher GSH concentration, we attribute this to errors associated with the very rapid nature of the first step of the reaction. Only values of kiths at [GSH] < 0.4 mole dm "3 were used in determining the functionality of klobs.

intact erythrocytes and within erythrocytes interacts with GSH (O'Brien and Ozolins, 1987, 1989a). A typical result is shown in Figure 1. These experiments show not only that chromate can enter erythrocytes, but also that once within these cells it reacts with glutathione.

Mechanisms in the Reduction of Chromate by Glutathione In our work we have carefully investigated (O'Brien and Ozolins, 1989b) the reaction of chromate with glutathione in the absence of buffer. Our results revealed a more complicated series of reactions for this system than has previously been appreciated. On mixing solutions of glutathione and chromate (pH = 7.00, 20~ [CrO4 2-] = 1 x 10-3 mole dm 3) at concentrations of GSH greater than -0.05 molar a distinct green colour developed. The intermediate generated has a maximum absorbance close to 650 nm. These green solutions of GSH and chromium subsequently decay to give a much less intense purple colour characteristic of chromium(Ill). We have followed the course of the reaction between GSH and chromate by following the absorbance change at 650 nm. The rate of disappearance of the green species eventually became first order and an estimate of the rate of' disappearance of this species can be obtained from plots o f ln(At-A~ ) for the final stages of the reaction. In general, pseudo first order rate constants were obtained by a nonlinear least squares procedure fitting the absorbance

versus time curve directly to a model for consecutive first order reactions, Figure 2 : A ---~ B ---) C with pseudo first order rate constants klo~ and k2o~. The rate of formation of the intermediate appears to follow a rate law klobs = a [ G S H ] x. Attempts to fit more complicated expressions to these results (e.g. kl~os = a[GSI-I]2/(1+ b[GSH]), as used by other workers) produced meaningless fits in which one parameter was essentially undetermined. Moreover a non-linear fit to y = a[GSH]n gave n values close to 2 ( 1.8 + 0.2); hence we have chosen the physically more meaningful model of a reaction second order in glutathione. The rate of disappearance of the intermediate follows a simple first order rate taw with an additional term independent of the concentration of glutathione (k2obs = a + b[GSH]), i A scheme for the reduction is outlined below: Cr(VI) + GSH ---) Cr(VI)-SG + I-I+..... K(kt/k-l) ' (1) Cr(VI)-SG + GSH ---) Cr(IV) + GSSG + H +...... k2 " (2) 2 Cr(IV) ~ Cr(V) + Cr(III) ......... k3 (3) This is similar to the kind of mechanism proposed by Wetterhahn in her studies of a wide range of thiols (Conett and Wetterhahn 1983, 1985, 1986) and that used by Wong and Pennington (1984) in discussing the reduction of chromate by cysteine. Taking a pre-equilibrium in the thiolate ester and a stationary state in chromium(IV) we can derive the following expression for the rate of appearance of chromium(V): rate = Kk2[GSH]2[Cr(V!)]/(2+2K[GSH])

80

Coordination chemistry and the carcinogenicity and mutagenicity of chromium(VI)

Table 1 Rate constants derived from fits of klobs and

k2obs. klobs = 0.35 (+ 0.1) x [GSH] 2 s -1 k2obs = 1.5 (+ 0.2) x 10-3 + 9.1 (+ 0.6) x [GSH] x 10-3 S"1 kiobs fitted by least squares to y = ax z k2obs fitted by least squares to y = ax + b

We are probably in the limit that 2K > 2 (K --20, Connett and Wetterhahn, 1986) this rate law becomes first order in glutathione. One possible explanation for the experimentally observed second order rate law is that the rate determining electron transfer (2) is general base catalysed, with glutathione being the only general base present in significant concentration equation (2) can be rewritten as: Cr(VI)-SG + 2GSH ---) Cr(IV) + GSSG + H + + GSH (2') Another possible explanation is that the reaction we are following proceeds via the interaction of free GSH with the bis thiolate ester of chromate, if this were present in a small quantity, in preequilibrium, then the additional steps would be : Cr(VI)-SG + GSH --) Cr(VI)-(SG)2 + I-I+ ....... K2 (1') Cr(VI)-(SG)'2 + GSH ---) Cr(IV) + GSSG + GSH....k2" (2")

If the quantity of the bis thiolate ester was small then we would be in the limit : rate = K2k2" [Cr-SG][GSH] 2 again a simple second order dependence on the glutathione concentration. Our most recent work vide infra (O'Brien and Ozolins, 1987; O'Brien et al., 1989b) suggests that the green intermediate isolated from this may be a bis glutathione species with thiolate coordinated GS-, an observation which supports the above suggestion. We are at present trying to study further the kinetics of this reaction by varying the pH and ionic strength. One problem with such work is the relatively drastic effect of any buffers on the reaction. The disappearance of the chromium(V) spectrum follows a rate law first order in glutathione with a distinct intercept. A scheme which could account for this observation is given below: Cr(V) + nL --* Cr(V)Ln ............. fast (4) Cr(V)Ln + GSH --~ Cr(V)Ln GSH ........ K' (5) Cr(V)LnGSH ---) Cr(III) .............. k6 (6) Cr(V)Ln ---) Cr(III) ................. k7 (7) rate = - kT[Cr(V)Ln] - krK'[GSH] [Cr(V)Ln]/(I+K'[GSH]) A l i g a n d c a p t u r e s c h r o m i u m ( V ) to form an intermediate (4) (ligand exchange reactions at the d 1 chromium(V) centre are likely to be extremely rapid). In this context the ligand (L) could represent carboxylate f u n c t i o n s on G S H or G S S G or in other w o r k e r s experiments a molecule of buffer. The intermediate complex then formed may decompose either by reaction with a further molecule of glutathione, postulated above as involving a rapid preequilibrium (1 > K'), to give a term

60

CCr 30

~Cr -1

-21.-

300

500 (rim) Figure 3 Electronic spectrum and circular dichroism of the chromium(Ill) complex.

700

Paul O'Brien and Guofang Wang

81

g = 1.996

0.9 L I

0.6 < 0.3

!

15 Gauss

I,

I

I

I

l

600

I,

900

~

I

1200

X(nrn) Figure 4 Electronic and EPR spectrum of the solid isolated from the reduction of chromate with glutathione.

first order in glutathione, and also by a path independent of any added glutathione (kT). The fact that there is a glutathione independent path supports our suggestion that the intermediate is a complex of chromium(V) with a reducing agent (i.e. glutathione). Clearly if chromium(IV) were generated from the reduction of chromium(V) this could also disproportionate to generate chromium(V): although we note this possibility we shall not consider it further. Values derived using this model for interpreting klobs and k2obs values are reported in Table 1. The results in Table 1, using interpretation (2') suggest a value of 0.71 mole-2 dm -3 s-1, the second order rate constant for the rate determining electron transfer process (pH 7.0, 20~ no ionic strength control). This value may be compared with the value of 0.2 (+ 0.04) reported by Connett and Wetterhahn (1986) (25~ Tris-HCL 1 mole din-3); given the very different methods used to determine these rate constants this probably constitutes quite good agreement. The similarity in our results may suggest that although Tris may both prevent the observation of chromium(V) complexes, by forming an unstable intermediate (either a chromium(V) complex or a ternary complex involving GSH), and also act as a general base catalysing the reduction, the rate determining step, the initial reduction of the chromium(VI) thiolate ester, is the same in both cases. In related work, Goodgame and Joy (1987) have presented evidence for ternary Cr(V)/ ascorbate/Tris species in the reduction of chromate by ascorbate.

A Chromium(IH) Complex from the Reaction of GSH with Chromate

Chromate (0.1 mole dm -3) and glutathione (1.0 mole dm "3) were allowed to react together for one to two hours, the resulting purple solution reduced in volume and repeatedly chromatographed on Sephadex SP-25 anion exchange resin, eluted with various concentrations of sodium chloride; the columns were protected from light. The chromium(III) containing fraction was readily identified as a single purple/red band which we were unable to resolve into more than one component (O'Brien et al., 1989a). The final eluate was concentrated to a small volume and chromatographed on Sephadex G10 twice in an attempt to desalt the mixture. We were never able to separate the final fraction from all salt. A solid purple material was obtained by evaporating the solution to dryness under reduced pressure. The analytical results can be interpreted in terms of an empirical formulae involving two moles of glutathione and one chromium atom Na2[Cr(GSH)2].2H20 0.7NaCI (CrNa2.7C20H37N6S2CIo.7 (A) RMM = 781) The sodium chloride is almost certainly a coprecipitated impurity. Despite extensive efforts we have been unable to totally desalt this complex. The ligand field strength, as judged from the lowest energy d-d transition (4T2g parentage in Oh) is similar to that of Tris amino acidates of the chromium(III) (Abdullah et al., 1985; Gillard et al., 1974) which suggests that the coordination sphere is N303-N204. The ligand field strength is also similar to that in other GSH

82

Coordination chemistry and the carcinogenicity and mutagenicity of chromium(Vl)

l

in Water EPR Spectrum of Green Intermediate (in the presence of DMPO)

g = 1.996

g = "1.986

[GSH] = 3.0 x 10 `3 mole dm -3 "~r

100 Gauss

Figure 5 EPR spectrum of the intermediate in aqueous solution. (For conditions see O'Brien and Wang, 1989.) and GSSG complexes of chromium(III) (Connet and Wetterhahn, 1983; Wiegand et al. 1984, 1986) some of which have been isolated from biological systems (Wiegand et al., 1986). The properties of the material we have isolated can be summarized as follows. The complex is a readily soluble ionic material, suggesting a monomeric material (RMM -800). It is difficuk, on the basis of our results to rule out some level of polymerization. The ligand field strength is similar to that of the Tris amino acidates of chromium(III) suggesting an N303 chromophore. There is no band in the electronic spectrum in the position associated with the S-Cr charge transfer transition. The complex contains no free -SH groups, as judged from infra red spectroscopy of the solid and Eliman's assay.

The spectral properties of this complex, Figure 3, may be of use in elucidating chromium speciation in samples isolated from in vitro and in vivo experiments. The complex described above has similar spectroscopic properties to the final product of the reduction of chromate by GSH, either from the intermediate complex we have described or from solutions of chromate and GSH. In particular the strong negative Cotton effect at 500 nm seems to be present in all these species. This is quite different from the optical properties of Tris amino acidate complexes at these wavelenghths. A Chromium(V) Intermediate We first observed the formation of chromium(V) during the reduction of chromate by glutathione some years ago

Paul O'Brien and Guofang Wang

(Barrett et al., 1985). We have now isolated a complex of glutathione and chromium(V) Na4Cr(GSH)4.8H20 from the reaction of glutathione and chromate (O'Brien and Ozolins, 1987; O'Brien et al., 1989b). The compound is identified as a chromium(V) complex by its EPR and electronic spectrum Figure 4. These properties are typical of a chromium(V) complex. On dissolution in water the complex decomposes to chromium(III) containing species, the rate of decomposition is retarded by low concentrations of GSH and then catalysed by higher concentrations. This observation is consistant with the kind of mechanism outlined above, i.e. glutathione both acts as a complexing agent stabilizing chromium(V) and as a reducing agent. Spectroscopic studies of the solid suggest that it may contain coordinated carboxylate and thiolate functions. The complex generates both the g = 1.996 and the g = 1.986 chromium(V) species (Barrett et al., 1985; Kitagawa eta/., 1988; Shi and Dalai, 1988; Aiyar et al., 1989) on dissolution in water, when the complex is dissolved in 0.5 molar GSH only the g = 1.996 species is observed. If the complex is incubated with the spin trap 5,5-dimethyl- 1-pyrroline-N-oxide (DMPO) (Finkelstein et al., 1980) a spectrum typical of the glutathionyl radical can be observed (Aiyar et al., 1989; Shi et al., 1989), Figure 5. In the absence of added glutathione chromate is generated during the decomposition of this solid species which may suggest that the complex contains both GSSG and GSH. A scheme for the decomposition of this compound is shown in Figure 6.

83

1989). Mixtures of chromium(VI) and glutathione and the chromium(V) complex detailed above both cause strand breaks, However, the chromium(M) complex of GSSG described above caused no damage to DNA. There may be a number of reasons for these apparent discrepancies, one particularly interesting hypothesis concerns the role of any radicals generated during the reduction. Thionyl radicals are now well established by products of the reaction, vide supra, and Rowley and Halliwell (1982) have demonstrated that thiols can lead to the generation of OH, which is well known for its ability to produce DNA strand breaks. Traces of iron, potentially present in vivo but removed by Aiyar et al. (1989) may assist in the generation of OH'. Such mechanisms have been discussed in detail by Kortenkamp (1989). In the context of the above discussion it is of interest that the simple chromium(V) complex oxochromium(V) chloride (CROCI4) can generate hydroxyl radicals (O'Brien and Wang, 1989). On dissolution in an aqueous solution of the spin trap 5,5-dimethyl-l-pyrroline-N-oxide (DMPO) (Finkelstein et al., 1980) the tyical spectrum of the hydroxyl radical was generated (aN= arI = 14.7 Gauss, and g = 2.005). These results of this work clearly show that the reduction of chromium(V) in aqueous solution can generate hydroxyl radicals. Although hydroxyl radicals have not been observed in the reduction of chromate by glutathione, this observation could be related to the known ability of GSH to trap hydroxyl radicals (Rowley and HalliweU, 1982). The radicals trapped in any such experiments will be a complex function of the conditions of the experiment and it is difficult to exclude the possibility of hydroxide radical formation in vivo on the basis of spin trapping.

Studies of Strand Breaks in Closed Circle DNA Assays

Studies of the formation of strand breaks in DNA may help in assessing these various hypotheses. In vitro studies of the formation of strand breaks in closed circle DNA (Scovell, 1986) may help in assessing these various hypotheses. Two independent studies have shown that the combination of hydrogen peroxide and chromate can produce strand breaks in double strand DNA (Aiyar et al., 1989; Kawanishi et al., 1986). Work from our laboratories has shown that a combination of chromate and GSH in the absence of 1-1202can cause strand breaks in the supercoiled circular DNA of bacteriophage PM2 (Kortenkamp et al.,

SOLID

f

"Cr(V)" ~

Cr(V)GSI-In

Conclusions

Our work allows a number of conclusions concerning the possible role of GSH in mediating the toxicity of chromate. 1. Compared to other reducing agents, such as L-cysteine and D-penicillamine the chromium(V) complexexs generated by GSH are relatively stable. 2. Other reactive intermediates, such as thionyl radicals are generated during this reduction. 3. Mixtures of glutathione and chromium(VI) can cause strand breaks in closed circle DNA assays, as can an intermediate complex isolated from this reaction.

Cr(IV)

> Cr(III) products

Figure 6 Reaction scheme for the decomposition of the intermediate. Some of the potential pathways from the solid glutathione intermediate to chromium(Ill), note that the chromium(Ill) products produced by the various pathways could be different.

84

Coordination chemistry and the carcinogenicity and mutagenicity of chromium(VI)

4. One chromium(III) containing product of the reduction of chromate by GSH a chromium(III) GSSG complex does not apparently damage DNA in in vitro assays. 5. Chromium(V) complexes, specifically [CROCI4]- can generate hydroxyl radicals. The reaction of GSH with chromate is complex and the exact species generated in these reactions are not yet fully identified, further work is needed to understand the chemistry of this system and the following points are particularly significant. 1. What are the special features of GSH that lead to the relative stabilization of the chromium(V) intermediates? 2. All of our work to date has tended to emphasise the potential role of reactive intermediates in generating strand breaks. It is equally possible that the toxicity of chromate is caused by a crosslinking lesion. Reactive intermediates of chromium, as generated by GSH, should be investigated to prove the preferred site of binding to DNA or initially nucleotides or nucleosides. Our studies to date suggest that a very complicalod coordination chemistry may underlie the toxicity of chromate. Acknowledgements W e w o u l d like to thank the C a n c e r R e s e a r c h C a m p a i g n for their g e n e r o u s support o f this project, D r P.A. H a m i l t o n ( Q . M . C . ) for his assistance with the fitting o f dam by n o n - l i n e a r methods. W e also thank P r o f e s s o r J.O. E d w a r d s ( D e p a r t m e n t o f Chemistry, B r o w n University) for m a n y helpful discussions c o n c e r n i n g this w o r k and in particular for first suggesting general base ea!_alysis as an explanation o f o u r k i n e t i c r e s u l t s on c h r o m a t e r e d u c t i o n . W e are especially grateful to D r A. K o r t e n k a m p and P r o f e s s o r D. B e y e r s m a n n o f the U n i v e r s i t a t B r e m e n , their enthusiastic collaboration has m a d e m u c h o f the a b o v e w o r k possible.

References Abdullah M., Barrett, J. and O'Brien, P. 1985. Synthesis and characterization of some chromium(m) complexes of glutathione. J. Chemical Society, Dalton Transactio~ts, 2085-2089. Aiyar, J., Borges, K.M., Floyd, R.A. and Wetterhahn, K.E. 1989. Role of chromium(V), glutathionyl and hydroxyl radical intermediates in chromium(VI)-inducexl DNA damage. Toxicological and Environmental Chemistry, in press. Barrett, J., O'Brien P. and Swanson F. 1985. Chromium(V) can be generated in the reduction of chromium(V1) by glutathione. Inorg. Chim. Acta, 108, L19-20. Beyersmaon, D. and Koster, A. 1987. On the role of trivalent chromium in chromium toxicity. Toxicological and Environmental Chemistry, 14, 11-22 Bianchi, V. and Levis, A.G. 1987. Recent advances in chromium genotoxicity, Toxicological and Environmental Chemistry, 15, 1-24 Conneu, P.H. and Wetterhahn, K.E. 1983. Metabolism of carcinogenic chromate by cellular constituents. Structure and Bonding (Berlin), 54, 93-124. Connett, P.H. and Wetterhahn K.E. 1985. In vitro reduction of the carcinogen chromate with cellular thiols and carboxylic acids. J. American Chemical Society, 107, 4282-4288. Connett, P.H. and Wetterhahn K.E. 1986. Reaction of chromimn(V1)

with thiols: pH dependence of chromium(v1) thioester formation. J. American Cheatica/Society, 108, 1842-1847. Cupo, D.Y and Weatediahn, ICE. 1985a. Binding of chromitnn to chromatin and DNA from the liver and kidney of rats treated with sodium dichromate and chremium0ll) chloride. Cancer Research. 45, 1146-1151. Cupo D.Y. and Weuerhan ICE., 1985b. Modification of chromimn(Vl)-induced DNA damage in chick e~nbryo hepatocy~, Proc. National Academy of Sciences (USA), 82, 6755-6759. D~euo, P., Arslan, P., Antolini, P. and Luciani 1988. Uptake of chromate by rat thymocyms and role of glutathione in its cytoplasmic rednction. Xenobiotica, 18, 657-664. Finkelstein, E. Rosen, G.M., and Rauckman, E.L 1980. Spin trapping of superoxide and hydroxyl radicals: practical aspects, Arch. Biochemistry and Biophysics, 200, 1-16. Fomace, A3., Seres, D.S., Lechner, J.F. and Harris, C.C. 1981. DNA protein glro~$--l~nkin~by chromium salts. Chemical Biological Interactions, 36, 345-354. Glassex, U., Hoduainer, D., Kloppel, H. and Oldiges, H. 1986. Carcinogenicity of sodium dichromate and chromium oxide aerosols inhaled by male wistar rats. Toxicology, 42, 219-232. Gillard, R.D., Iatmie, S.H., Price, D.C., Phipps, D.A. and Wcick, C.F. 1974. Optically active co-ordination compounds. Part XXXV. Coafigumtions, preparations and reactions of some complexes of amino-acids. J. Chemical Society, Dalton Transactions, 1385-1396. Goodgame, D.IVLL and Joy AM. 1986. Relatively long-lived Cr(V) species are produceA by the action of GSH and carcinogenic Cr(VI). J. Bioinorganlc Chemistry, 26, 219-224. Goodgame, D.M.L and Joy A.M. 1987. EPR study of the Cr(V) and radical species produced in the reduction of Cr(VI) by ascorbate, laorg. Chim. Acta, 135, 115-118, Gray, M. 1987. (~arome platers at risk. Food and Chemical Toxicology, 25, 719-720. Gray S.J. and Sterling K. 1950. The tagging of red cells and plasma proteins with radioactive chromium. J. Clinical Investigation, 29, 1604-1613. Kawanishi, S., Inoue, S. and Sano, S. 1986. Mechanisms of DNA cleavage induced by sodium chromate in the presence of hydrogen peroxide. J. Biological Chemistry, 261, 5952-5958. Kitagawa, S., Seki, H., Kamctani, F. and Sakufi. H. 1988. Epr study on the interaction of hexavalent ~ u m with glutathione or cysteine: production of pentavaleat chromium and its stability. Inorg. Chim. Acta, 152, 251-255. Kortenkamp, A., Beyersmann, D. and O'Brien, P., 1987. Uptake of chromium(HI) complexes by erythrocytes. Toxicological Environmental Chemistry, 14, 23. Kortenkam4a, A., B~emmann, D. O'Briea, P. and Ozolins, Z. 1989. Ganeration of PM2 DNA breaks in the course of reduction of chromium(vI) by gintathione. Mutation Research, 216, 19-26. Koaenkamp, A. 1989. Unterschungenen zur molekularen wirkungsweise voa chromverbindongen als DNA-schadigende agenzie., PhD Thesis, Universitat, Bremen. Langard, S. 1983. The carcinogenicity of chrominmcompoonds in man and animals, Chapter 2. In: D.Burrows (ed.), Chromium: Metabolism and Toxicity, pp.12-30. CR.C. Press Inc., Boca Raton, Florida. Langard, S. and Vigander, T. 1983. Occurence of lung cancer in workers producing chromium pigments. J. Industrial Medicine, 40, 71-74. O'Brien P. and Ozollns Z. 1987. Studies of the reaction of glutathione with chromate: the isolation of an intermediate and experiments on intact e~3~arocytes.Recueil des TravatocChimiques des Pays-Bas, 106, 423. O'Brien, P. and Ozolins. ~ 1989a. 1H spin echo nmr evidence for the interaction of chromium(V]) with glmat~one in intact erythrocytes, lnorg. Chim. Acta, 155, 25-26. O'Brien, P. and Ozolins. Z~ 1989b. Mechanisms in the reduction of chrominm(VI) with glutathione, lnorg. Chim~Acta, 161, 261-266. O'Brien P. and Wang G. 1989. Chromium(V) complexes can generate hydroxyl radicals. Inorg. Chim. Acta, 162, 27. O'Brien, P., Ozolins, ~ and Wang, G. 1989a. A chromium(m)

Paul O'Brien and Guofang Wang complex of oxidised glutathione isolated from the reduction of chromium(VI) with glutathione, lnorg. Chim. Acta, in press. O'Brien, P., Pratt, J., Swanson, F., Thornton, P. and Wang, G. 1989b. The Isolation and Characterization of a Chromium(V) Containing Complex from the Reaction of Glutathione with Chromate. lnorg. Chim. Acta, in press. Rabenstein, D.L, Brown, D.W. and McNeil, C,J. 1985. 1H nmr studies of glutathione in intact erythrocytes. Analytical Chemistry, 26, 219. Rowley, D.A. and Halliwell B. 1982. Superoxide-dependent formation of hydroxyl radicals in the presence of thiol compounds. FEBS Letters, 138, 33-36. Scovell, W.M., 1986. Supercoiled DNA. J. ChemicalEducation, 63, 562-565 Shi, X. and N.S.Dalal, N.S. 1988. On the mechanism of the chromate reduction by glutathione: ESR evidence for the gluthione radical and an isolable chromium(V) intermediate. Biochemical and Biophysical Research Communications, 156, 137-142. Tsapakos, M.J. and Wetterhahn, K.E. 1983. Interaction of chromium with nucleic acids. ChemicalBiological Interactions, 46, 265277.

85

Tsapakos, M.J., Hampton, T.H. and Wetterhahn, ICE. 1981 The carcinogen chromate induces DNA crosslinks in rat liver. Y. Biological Chemistry, 256, 3623-3626. Wiegand, H.J., Ottenwalder, H. and Bok H.M. 1985. Fast uptake kinetics in vitro of 51Cr(V1) by red blood cells of man and rat. Toxicology, 57, 31-34 Wiegand, H.J., Ottenwalder, H. and Bolt, H.M. 1984. The reduction of chrominm(VI) to chromium0II) by glutathione: an intrcellular redox pathway in the metabolism of the carcinogen chromate. Toxicology, 33, 341-348. Wiegand, H3., Ottenwalder, H. and Bolt, H.M. 1986. Disposition of a soluble chromate in the isolated rat liver. Xenobiotica, 16, 839-844. Wong, D.W.L and Pennington, D.E. 1984. Stoichiometry, kinetics and mechanisms of the chromium(VI) oxidation of L-cysteine at neutral pH. lnorg. Chem., 23, 2528. (Manuscript No. 194: received July, and accepted for publication September 26, 1989.)