ADP-iron as a fenton reactant: Radical reactions detected by spin trapping, hydrogen abstraction, and aromatic hydroxylation

ADP-iron as a fenton reactant: Radical reactions detected by spin trapping, hydrogen abstraction, and aromatic hydroxylation

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 277, No. 2, March, pp. 422-428, 1990 ADP-Iron as a Fenton Reactant: Radical Reactions Detected by Sp...

678KB Sizes 1 Downloads 47 Views

ARCHIVES

OF BIOCHEMISTRY

AND

BIOPHYSICS

Vol. 277, No. 2, March, pp. 422-428, 1990

ADP-Iron as a Fenton Reactant: Radical Reactions Detected by Spin Trapping, Hydrogen Abstraction, and Aromatic Hydroxylation John M. C. Gutteridge,l

Imre Zs.-Nagy,2 Lindsay Maidt, and Robert A. Floyd

Oklahoma Medical Research Foundation,

Molecular

Received July 21, 1989, and in revised form November

Toxicology, 825 N.E. 13th Street, Oklahoma City, Oklahoma 73104

16,1989

A mixture of ADP, ferrous ions, and hydrogen peroxide (HzO,) generates hydroxyl radicals (‘OH) that attack the spin trap DMPO (5,5-dimethyl-pyrollidine-Noxide) to yield the hydroxyl free radical spin-adduct, degrade deoxyribose and benzoate with the release of thiobarbituric acid-reactive material, and hydroxylate benzoate to give fluorescent products. Inhibition studies, with scavengers of the ‘OH radical, suggest that the behavior of iron-ADP in the reaction is complicated by the formation of ternary complexes with certain scavengers and detector molecules. In addition, iron-ADP reacting with H202 appears to release a substantial number of ‘OH radicals free into solution. During the generation of ‘OH radicals the ADP molecule was, as expected, damaged by the iron bound to it. Damage to the iron ligand in this way is not normally monitored in reaction systems that use specific detector molecules for ‘OH radical damage. Under certain reaction conditions the ligand may be the major recipient of ‘OH radical damage thereby leading to the incorrect assumption that the iron ligand is a poor Fenton reactant. o isso Academic

Press,

Inc.

It is now well established that the formation of a powerful oxidizing species from superoxide (0,) and hydrogen peroxide ( H202) requires the presence of a metal catalyst with a variable oxidation number. The most likely, although not the only, biological metal to catalyze formation of the hydroxyl radical, or species with similar oxidizing ability, is iron. Iron is usually safely sequesi Permanent address: National Institute for Biological Standards and Control, Blanche Lane, South Mimms, Potters Bar, Herts, EN6 3QG, UK. ’ Present address: Verzar International Laboratory for Experimental Gerontology (VILEG), Italian Section, Department of Gerontological Research, INRCA, Via Birarelli, 8. I-60121 ANCONA, Italy.

tered in proteins to protect it from reacting with oxygen (or its reduction intermediates), and many of these proteins such as transferrin, lactoferrin, haptoglobin, and

hemopexin normally have considerable potential to bind iron ions in ways that hinder or prevent their action in catalyzing radical reactions (l-4). This antioxidant property of many metal-binding proteins has evolved in parallel with the essential need to conserve body iron stores and keep the plasma concentrations of “free” ionic iron and copper at effectively nil (5-7). During severe oxidant stress, however, production of 0, and H202, in excess of the body’s ability to remove them, can result in the release of iron from safe functional storage sites within hemoglobin (&lo), myoglobin (lo), and ferritins (11,12). The released iron can stimulate free radical reactions and can bind to molecules such as desferrioxamine, apotransferrin, and bleomycin. Indeed, the latter compound has been used as a basis for detecting and measuring available “catalytic” iron in biological fluids (5,6). The exact nature of intracellular or extracellular lowmolecular-mass iron in the body is still uncertain but it may well be complexed to molecules such as citrate (13), acetate, or nucleotides and other phosphates (14). However, information on the role of iron complexed to these molecules in catalyzing Fenton reactions is confusing. For example, we and others have shown that ADP-iron will perform as a Fenton reactant (15-18). Flitter et al. found ADP to be as good as EDTA at low iron concentrations (19), whereas others have described ADP-iron as being only half as effective as EDTA-iron (20), either having little or no activity (21,22) or acting as a biological antioxidant (23). The iron chelator EDTA possesses several important properties which make it a useful tool for studying the mechanisms of Fenton chemistry (24-26). However, EDTA is not a biological molecule and parallels cannot necessarily be made with biological iron ligands.

422 All

0003-9861/90 $3.00 Copyright 0 1990 by Academic Press, Inc. rights of reproduction in any form reserved.

RADICAL

REACTIONS

OF ADP-IRON

AS A FENTON

423

REACTANT

-Li*

6

,lOG

,

E

I FIG. 1. ESR spectra recorded under standard conditions described in the methods section. (A) Spectrum control system. (B) The effect of addition of Ferrozine (0.588 mM) before the HzOz. (C) The effect of addition shown under (A). (D) The effect of the addition of 50 mM deoxyribose to the system shown under (A). (E) and mM, respectively, of ADP to the system shown under (A). Horizontal bar represents 10 G (1 mT). Instrumental methods section. The percentage of control (A) of each of the treatments are as such: (b) 2.4%, (C) 70.9%, (D)

In the present study we compare the ability of the Fenton mixture, ferrous salt-ADP and hydrogen peroxide, to damage different detector molecules. Damage was detected as hydrogen atom abstraction leading to the release of thiobarbituric acid-reactive (TBA)3 material from deoxyribose and benzoate, aromatic hydroxylation of benzoate to give fluorescent dihydroxy products, and hydroxyl free radical addition to the spin trap DMPO to give the DMPO-OH adduct. MATERIALS

AND

METHODS

2-Deoxy-D-ribose, CuZn superoxide dismutase (bovine erythrocyte), catalase (bovine liver), and albumin (human, fatty acid free) were purchased from Sigma Chemicals, St. Louis. ADP was purchased from Boehringer-Mannheim. All other chemicals were of the highest purity available from Aldrich Chemicals, Milwaukee. Deoxyribose degradation as well as benzoate degradation were carried out as previously described (27) using the TBA reaction. Benzoate hydroxylation was measured as changes in fluorescence (28). Briefly, reactions were carried out in new clean plastic tubes in a total volume of 1.0 ml. The reaction mixture contained the following reactants at the final reaction concentrations: detector molecule (benzoate or de-

3 Abbreviations used: TBA, thiobarbituric acid-reactive; DMPO, 5,5-dimethyl-pyrollidine-N-oxide; EPR, electron paramagnetic resonance; 4-POBN, ol-(4-pyridyl-1-oxide)-N-tert-butylnitrone, PBN, Ntert-butyl-o-phenylnitrone; SOD, superoxide dismutase.

of the DMPO-OH adduct in the of 50 mM mannitol to the system (F) The effect of adding 4 and 50 parameters as described in the 56.4%, (E) 54.3%, and (F) 16.2%.

oxyribose) 1 mM, ADP 1.0 mM, ferrous salt 0.1 mM, hydrogen peroxide 0.25 mM, and buffer 12.5 mM, pH 7.4. Where appropriate 0.1 ml of inhibitor was added to the reaction when EDTA was present or ferrous salts were used alone; the ferrous salt was the last reagent added to the reaction. Buffer solutions consisted of sodium phosphate salts or sodium bicarbonate (12.5 mM) carefully adjusted to pH 7.4. Samples were incubated at 37°C for 1 h. TBA reactivity was determined by adding 0.2 ml of 1% w/v TBA in 0.05 M NaOH, 0.2 ml of trichloroacetic acid 2.5% w/v, and heating at 100°C for 10 min. The resulting pink chromaphore was read at 532 nm against appropriate blanks and controls. Fluorescent products of benzoate were determined in incubated samples by excitation at 305 nm and fluorescence emission measured at 408 nm. All the results shown are the mean of three or more separate assays that differed by +5%. Spin-trapping experiments were carried out essentially as previously described (15, 16) using a total volume of 0.1 ml containing as final concentrations 2 mM ADP, 0.1 mM Fe& (in 0.12 mM HCl), 0.01 ml of test material, 50 mM DMPO, 3 mM H202, and buffer (50 mM NaCl, 12.5 mM NaHC03, pH 7.1). The EPR spectra were obtained on a Varian E-9 X-band spectrometer operating in the first derivative mode. Typical instrument settings were as follows: scan range 100 G, field set 3230 G, time constant 0.3 s, scanning time 4 min, modulation amplitude 2.0 G, modulation frequency 100 kHz, receiver gain 1.25 X lo”, microwave power 25 mW, microwave frequency 9.1 GHz, temperature 20°C. The instrument was calibrated with potassium peroxylamine disulfonate which has an A n = 1.302 mT in saturated bicarbonate solution.

RESULTS

When a ferrous salt and ADP are mixed with hydrogen peroxide at pH 7.1 a strong DMPO-OH adduct is

424

GUTTERIDGE

TBA 0.4 A532nm

ET AL.

U.LUTBA A532nm

-

0.30 -

0.3 0.20-

t

12.5mM HCOj

I

6.25mM 6.25mM

I

I

POZ HCO;

12.5mM PO:

I

1.0

,

3.0

a

,

5.0

I

,

7.0

I

,

9.0

ADP (mM)

FIG. 2. Deoxyribose degradation by a ferrous salt and hydrogen peroxide. Bicarbonate ions (12.5 mM, pH 7.4) were added to phosphate ions (12.5 mM, pH 7.4) in the presence and absence of ADP (1 mM).

FIG. 3. Deoxyribose degradation (A 532nm) by a ferrous salt and hydrogen peroxide in phosphate ions (12.5 mM pH 7.4) and bicarbonate ions (12.5 mM, pH 7.4) as a function of ADP concentration.

rapidly formed in the presence of either phosphate or carbonate ions (l&16) in agreement with previous studies (Fig. 1A). Stabilizing the ferrous ions in the ferrous state with ferrozine, a more powerful iron chelator than ADP, greatly inhibited formation of the DMPO-OH adduct (Fig. 1B). The ADP-iron-hydrogen peroxide mixture degrades deoxyribose and benzoate with the release of thiobarbituric acid-reactive products in the presence of phosphate and bicarbonate ions. However, phosphate and bicarbonate ions substantially affect the yield of TBA-reactive material released from deoxyribose (Fig. 2). In the absence of ADP but in the presence of bicarbonate ions, ferrous salt and hydrogen peroxide were less effective in degrading deoxyribose than in the presence of phosphate ions (Fig. 2). When ADP, at a reaction concentration of 1 mM, was added it greatly stimulated deoxyribose damage, but further addition of ADP decreased this stimulation (Fig. 3) although at a final concentration of 9 mM stimulation was still greater than in the absence of ADP. Addition of ADP to the Fenton mixture in phosphate buffer decreased deoxyribose degradation over all concentrations tested (Fig. 3). Competitive scavenging of OH radicals by ADP may possibly explain the results of spin trapping experiments which show that at substantially higher concentrations of ADP there is a decreased formation of the DMPO-OH adduct (data not shown).

In the process of scavenging OH radicals ADP may form carbon- or nitrogen-centered radicals but attempts to detect these with the spin traps 4-POBN and PBN in a short time were not successful as shown in Fig. 4. The spin-trapped radical spectrum more pronounced in B has coupling constants of AN = 15.3, A: = 1.7 G, which

C FIG. 4. ESR spectra recorded under conditions described in the Materials and Methods section except that receiver gain was 6.3 X lo3 (A and B) or 6.3 X lo4 (C) and the time constant was 1 s, using 4-POBN or PBN as spin trap. (A) Control system with 4-POBN, i.e., in presence of 2 mM ADP. (B) The effect of addition of +14 mM ADP to the system shown under (A). (C) The same system as shown under (B) but 4-POBN was replaced by PBN. Horizontal bar represents 10 G (1 mT).

RADICAL

REACTIONS

OF ADP-IRON

ADP A DF

ADP

I

AD

AD AMP

AD $MP II

7

1

AS A FENTON

REACTANT

425

The hydroxyl radical scavengers mannitol and thiorea partly inhibited damage to deoxyribose and benzoate by iron-ADP although considerably greater inhibition was seen in bicarbonate buffer. Urea included as a control for nonspecific scavenging effects did not inhibit in any of the detector systems (Tables I and II). Mannitol (50 mM) inhibited formation of the DMPO-OH adduct by about 30% when added with 50 mM DMPO (Fig. 1). Under similar conditions 50 mM deoxyribose inhibited formation of the DMPO-OH adduct by 45% (Fig. 1). Formate was effective at inhibiting benzoate damage detected as release of TBA reactivity as well as benzoate hydroxylation. However, it failed to protect deoxyribose from damage under similar reaction conditions (Tables I and II). Addition of superoxide dismutase (SOD), catalase, or albumin to the Fenton mixture would be expected to show inhibition only for catalase. However, in a phosphate buffer some inhibition was seen for SOD in its native form and this was slightly increased when the protein was heat-denatured (Table I). Inhibition by SOD, and an enhancement following heat-denaturation, was greatly increased when bicarbonate ions were present as was inhibition by albumin, which was included as a control for nonspecific protein scavenging effects (Table II). When ADP and EDTA were tested at equal molarity with the same ratio of ligand to iron ions (lO:l), ADP was subsequently more effective than EDTA at degrading deoxyribose (Table III).

FIG. 5.

HPLC separation of products from ADP after incubation at 37°C with a ferrous salt (1 mM) and hydrogen peroxide (1 mM) in 12.5 mM bicarbonate buffer pH 7.4. The separations were obtained on a 15 cm column of LiChrosorb RP18 (10 pM) with a solvent system consisting of KHzPOl (0.01 M), tertiary butyl phosphate (0.01 M), and CH&N 8% v/v at a flow rate of 0.6 ml/min. Peaks were detected at 254 nm, sensitivity X 0.1. Trace 1, standards of Ad, AMP, and ADP; trace 2,l mM ADP; trace 3,l mM ADP incubated with a ferrous salt and hydrogen peroxide for 1 min at 37°C; trace 4, as above except incubation was for 35 min; trace 5, as above except incubation was for 70 min. Ad, adenosine; ADP, adenosine-5’-diphosphate; AMP, adenosine-5’monophosphate.

corresponds to the 4-POBN hydroxyl free radical adduct (29). The reason why 4-POBN traps more hydroxyl free radical when ADP is higher in concentration in contrast to the results obtained with DMPO is at present unresolved. Damage to the ADP molecule was observed by changes in uv absorbance after incubation of ADP-Fe2’ with HzOz and separation of the products by HPLC. Figure 5 shows that after 1 and 35 min incubation with iron, ADP concentrations fall while AMP and adenosine concentrations increase. In addition, several unidentified products also increase. Incubation for 70 min showed little change in adenosine, AMP, and ADP from that which had already occurred at 35 min (Fig. 5). We assume the slight increase in all three compounds to be due to the decrease in sample volume by evaporation.

DISCUSSION

Formation of a DMPO-OH adduct clearly shows that there is a substantial yield of OH radicals released into free solution when an ADP-ferrous iron complex is mixed with hydrogen peroxide (15-18,32). This Fenton mixture also damages deoxyribose with the release of TBA reactivity, and aromatically hydroxylates benzoate as demonstrated by an increase in fluorescence. These findings contrast sharply with previous reports that claim an ADP-iron complex is insufficiently reactive to be a biological Fenton reactant (21, 22) or that ADP is a biological antioxidant (23) protecting against Fenton chemistry but they do support reports that ADP-iron is a Fenton reactant (15-19,32). The presence of either phosphate or bicarbonate ions substantially altered radical damage to the deoxyribose molecule with enhanced release of TBA-reactive material occurring in the presence of bicarbonate. The reasons for this are not clear but it is possible that when phosphate is present it forms a ternary complex with the iron-ADP, altering its redox or binding properties as described for other iron complexes (33,34) or, possibly introduces complexed ferric ions into the reaction that might contribute to secondary reactions with hydrogen peroxide (30). Addition of several hydroxyl radical scav-

426

GUTTERIDGE

ET AL.

TABLE

I

Radical Damage Stimulated by ADP-Ferrous Ions and Hydrogen Peroxide in the Presence of a Phosphate Buffer, pH 7.4 Deoxyribose degradation TBA-reactivity

Inhibitors 1. 2. 2. 2.

Blank Control +MannitoI +Thiourea

2.

+Urea

L2

10 mM 0.5 mM

0.5 mM

2. +Formate 10 mM 2. +SOD 0.01 mg/ml 2. +SOD heatdenatured 2. +Catalase 0.01 mdml 2. +Catalase heatdenatured 2. +Albumin 0.01 m&l 2. +Albumin heatdenatured

nm

0.005 0.238 0.127 0.133 0.252 0.258 0.173

Benzoate degradation TBA-reactivity

% Inhibition

4,~ nm

47% 44% NS NS 28%

0.005 0.059 0.039 0.039 0.061 0.036 0.045

0.162

32%

0.095

% Inhibition

Benzoate hydroxylation Fluorescence units

% Inhibition

36% 36% NS 41% 26%

2 138 80 80 139 68 74

42% 42% NS 57% 46%

0.043

30%

71

48%

62%

0.021

67%

61

55%

0.268

NS

0.060

NS

138

NS

0.247

NS

0.061

NS

131

NS

0.245

NS

0.060

NS

172

NS

Note. Final concentrations are shown. NS, not significant. Fe’+, 0.1 mM, H,O*, 0.25 mM.

Na,PO,/Na,HPO,,

engers to the Fenton mixture gave an interesting pattern of inhibition. Mannitol and thiourea were partly inhibitory in all the detector systems but this activity greatly increased in the presence of bicarbonate ions. Mannitol and thiourea are known to have properties other than radical scavenging that can affect Fenton reactions. Both can bind metal ions (14) and react or complex with hydrogen peroxide (14,31). The autoxidation of ferrous ions bound to ADP is greatly decreased (15), prolonging the ability of ferrous ions to react with hydrogen peroxide. The poor reactivity of formate ions in Fenton systems has been claimed to provide a means of differentiating hydroxyl radicals from other oxidizing species such as the ferry1 ion (FeO”) in that the inability of formate to protect against damage implies that the species is not ‘OH (20). However, this assumption must be questioned from the data presented here since, under identical reaction conditions, formate protects the detector molecule benzoate but not deoxyribose, implying that the idiosyncrasy lies with the detector molecule. It is likely that the binding properties of detector molecules and scavengers are involved in this anomaly as previously suggested (26, 27). If the ADP-iron complex binds to deoxyribose but not to benzoate then formate, as a poor metal-binding molecule, would be unable to influence site-specific damage to deoxyribose. However, mannitol and thiourea could influence binding and effectively compete with deoxyribose (26,27) by complexing with iron-ADP and directing damage onto themselves.

12.5 mM; Benzoate,

1 mM; Deoxyribose,

1 mM; ADP, 1.0 mM;

Addition of increasing concentrations of ADP to the Fenton mixture decreased formation of the DMPO-OH adduct and decreased damage to the deoxyribose molecule, suggesting that competitive scavenging of OH radicals was occurring. When Fenton chemistry is driven by iron attached to a ligand, then it can be expected that the ligand would sustain substantial damage from generated radicals. Here, we might expect that phosphate and organic radicals are formed from ADP, although they were not trapped in the time observed, but ADP was degraded into fragments, two of which were identified as adenosine and AMP. It is also possible that fragments cleaved from ADP still retain an iron-binding potential and may contribute to Fenton reactions in either a major or minor way. When a substantial number of OH radicals escape scavenging by the iron ligand (chelator) and are released into free solution, where they can react with added scavengers at rates predicted from their known second order rate constants, then the chelator may be said to function in a radiomimetic way as has been observed for EDTA (27). Here, ADP-iron (or fragments cleaved from it) acted as an effective Fenton reactant. However, damage to detector molecules and patterns of inhibition by OH radical scavengers suggests that mechanisms dependent on both metal binding (site specific) and the release of OH radicals into free solution (radiomimetic) contributed to the reactive properties of ADP-iron. The reaction between a ferrous complex and hydrogen peroxide leading to OH radical formation is inhibited by catalase but not by SOD. Here, however, we observed

RADICAL

REACTIONS

OF ADP-IRON TABLE

Radical Damage Stimulated

by ADP-Ferrous

Ions and Hydrogen

1. Blank

2.

+Urea

0.5 mM

2. +Formate 10 mM 2. +SOD 0.01 mg/ml 2. +SOD heat-

denatured 2. +Catalase 0.01 &ml 2. +Cat. heat-

denatured

II

Peroxide in the Presence of a Biocarbonate

Buffer, pH 7.4

Benzoate hydroxylation

% Inhibition

Fluorescence units

% Inhibition

68% 82% NS NS 70%

0.006 0.160 0.063 0.061 0.160 0.067 0.072

60% 62% NS 58% 55%

3 314 109 118 304 80 128

65% 62% NS 74% 59%

0.116

77%

0.044

73%

84

73%

0.090

82%

0.036

77%

90

71%

0.215

57%

0.056

65%

111

64%

0.378

23%

0.113

29%

191

39%

0.194

61%

0.056

65%

93

70%

% Inhibition

432 nm

0.007 0.475 0.157 0.091 0.496 0.453 0.151

k2 nm

2. Control 2. +Mannitol 10 mM 2. +Thiourea 0.5 mM

427

REACTANT

Benzoate degradation TBA-reactivity

Deoxyribose degradation TBA-reactivity Inhibitors

AS A FENTON

2. +Albumin 0.01

mdml 2. +Albumin heat-

denatured

Note. Final concentrations are shown. NS, not significant. NaHC03, 12.5

mM;

Benzoate, 1 mM; Deoxyribose, 1 mM; ADP, 1.0

mM;

Fe*+, 0.1

mM, H202, 0.25 mM.

some inhibition by SOD, an activity that increased after heat-deactivation of the protein and was further enhanced by the presence of bicarbonate ions. Similar results were obtained for albumin included as a control for nonspecific radical scavenging effects. Heat-deactiva-

TABLE III Deoxyribose Degradation by Ferrous Ions and Hydrogen Peroxide: A Comparison of ADP and EDTA as Iron Ligands for OH Radical Formation

Ratio 1O:l

Phosphate

Bicarbonate

(12.5 mM) pH 7.4

(12.5 mM) pH 7.4

Thiobarbituric acid reactivity b2 nm ADP (1 mM):

Ferrous Ions (0.1

mM)

0.224

0.476

Ferrous Ions (0.1 mM)

0.137

0.098

63%

485%

EDTA

(1 mM):

% Enhancement in deoxyribose degradation by ADP when compared with EDTA under the same reaction conditions

Note. Final reaction concentrations are shown. The ferrous salt was added to the reaction mixture when all other reagents were present.

tion of a protein can increase the exposure of thiol groups and these groups may be responsible for the increased scavenging of OH radicals. Using the iron-ADP reaction it has recently been shown that the sulfur-containing amino acids are substantial scavengers of OH radicals (31). The data presented here shows that the biological nucleotide ADP has the potential to act as a potent iron ligand for Fenton chemistry and, when complexed with iron, depending on the molar ratio of iron to ADP, can be considerably more effective than an equimolar concentration of the iron chelator EDTA. We further show that considerable care must be taken when defining a Fenton reactant. Experiments that attempt to show that an added detector molecule is damaged in a specific way depend on either the release of radicals into free solution or a close association between the iron catalyst and the detector molecule. A third possibility, for biological studies is, however, that neither of the above two situations operate for an added detector molecule and that the ligand to which iron is bound is itself the major recipient of radical damage. Under these circumstances it is easy to dismiss an iron complex as a poor or inactive Fenton reactant because damage to the iron ligand is not being monitored for the assay system. Some of the discrepancies reported for ADP-iron as a Fenton reactant in the literature may arise from its ability to act in one or all of the three functions described above.

428

GUTTERIDGE

ACKNOWLEDGMENTS The research reported was supported in part by NIH Grants CA42854, ES04296, and NS23307. J.M.C.G. was an OMRF Greenberg Scholar from July to December 1988. I.Zs.-N. expresses appreciation to OMRF for the opportunity to conduct research as a visiting scholar.

REFERENCES 1. Gutteridge, J. M. C. (1982) &o&em. Sot. Trans. 10,72-73. 2. Gutteridge, J. M. C., Paterson, S. K., Segal, A. W., and Halliwell, B. (1981) Biochem. J. 199,259-261. 3. Gutteridge, J. M. C. (1987) Biochem. Biophys. Acta 917,219-223. 4. Gutteridge, J. M. C., and Smith, A. (1988) Bio&em. J. 256,861865. 5. Gutteridge, J. M. C., Rowley, D. A., and Halliwell, B. (1981) Biochem.J. 199,263-165. 6. Gutteridge, J. M. C., Rowley, D. A., and Halliwell, B. (1982) Biothem. J. 206,605-609. 7. Gutteridge, J. M. C. (1984) Biochem. J. 218,983-985. 8. Gutteridge, J. M. C. (1986) FEBS Lett. 201,291-295. 9. Puppo, A., and Halliwell, B. (1988) Biochem. J. 249,185-190. B. (1988) Free Rd. Res. Commun. 4, 10. Puppo, A., and Halliwell, 415-422. 11. Biemond, P., Van Eijk, H. G., Swaak, A. J. G., and Koster, J. F. (1984) J. Clin. Inuest. 73,1576-1579. 12. Thomas, C. E., Morehouse, L. A., and Aust, S. D. (1985) J. Biol. Chem. 260,3275-3280. B., Aruoma, 0. I., Bonford, 13. Grootveld, M., Bell, J. D., Halliwell, A., and Sadler, P. J. (1989) J. Biol. Chem. 264,4417-4422. 14. Halliwell, B., and Gutteridge, J. M. C. (1986) Arch. Biochem. Biophys. 246,501-514. 15. Floyd, R. A. (1983) Arch. Biochem. Biophys. 225,263-270. 16. Floyd, R. A., and Lewis, C. A. (1983) Biochemistry 22,2645-2649.

ET AL. 17. Floyd, R. A., and Zs-Nagy, I. (1984) Biochem. Biophys. Acta 790, 94-97. 18. Borg, D. C., and Schaich, K. M. (1988) in Oxygen Radicals and Tissue Injury (Halliwell, B., Ed.), pp. 20-26, FASEB, Maryland. 19. Flitter, W., Rowley, 158,310-311.

D. A., and Halliwell,

B. (1983) FEBS Lett.

C. C., and Sutton, 20. Vile, G. F., Winterbourn, Biochem. Biophys. 259,616-626.

H. C. (1987) Arch.

21. Winterbourn, C. C., and Sutton, H. C. (1986) Arch. Biochem. Biophys. 244,27-34. 22. Baker, M. S., and Gebieki, J. M. (1986) Arch. Biochem. Biophys. 246,581-588. 23. Tadolini, B. (1989) Free Rd. Res. Commun. 5,237-243. 24. Walling,

C. (1975) Act. Chem. Res. 8,125-131.

25. Cohen, G., and Sinet, P. M. (1980) in Chemical and Biochemical aspects of superoxide and superoxide dismutase (Bannister, J. V., and Hill, H. A., Eds.), Vol. IIA, pp. 27-37, Elsevier/North Holland, New York. 26. Grootveld,

M., and Halliwell,

B. (1986) Free Rd.

Res. Commun.

1,243-250. 27. Gutteridge,

J. M. C. (1984) Biochem. J. 224,761-767.

28. Gutteridge,

J. M. C. (1987) Biochem. J. 243,709-714.

29. Janzen, E. G., Wang, Y. Y., and Shetty, Chem.Soc. 100,2923-2925.

R. V. (1978) J. Amer.

30. Gutteridge, J. M. C. (1985) FEBS Lett. 185,19-23. 31. Cederbaum, A. I., Dicker, E., Rubin, E., and Cohen, G. (1979) Biochemistry 18,1187-1191. 32. Zs-Nagy, I., and Floyd, R. A. (1984) Biochem. Biophys. Acta 790, 238-250. 33. Burger, R. M., Horwitz, S. B., and Peisach, J. (1985) Biochemistry, 24,3623-3629. 34. Elgavish, 150.

G. A., and Granot,

J. (1979) J. Mugn. Reson. 36, 147-