- Email: [email protected]
Inhibition of peroxidase-catalyzed reactions by deferoxamine
ARCHIVESOF BIOCHEMISTRYAND BIOPHYSICS Vol. 264, No. 2, August 1, pp. 690~696,1988
Inhibition of Peroxidase-Catalyzed SEYMOUR J. KLEBANOFF’
Reactions by Deferoxamine’
AND
ANN M. WALTERSDORPH
Department of Medicine, RM-16, University of Washington, Seattle WA 981% Received January 20,1988, and in revised form March 24,1988
Phagocytes generate superoxide (05) and hydrogen peroxide (HzOz) and their interaction in an iron-catalyzed reaction to form hydroxyl radicals (OH) (Haber-Weiss reaction) has been proposed. Deferoxamine chelates iron in a catalytically inactive form, and thus inhibition by deferoxamine has been employed as evidence for the involvement of OH generated by the Haber-Weiss reaction. We report here that deferoxamine also inhibits reactions catalyzed by the peroxidases of phagocytes, i.e., myeloperoxidase (MPO) and eosinophil peroxidase (EPO). The reactions inhibited include iodination in the presence and absence of chloride and the oxidation of guaiacol. Iodination by MPO and Hz02 is stimulated by chloride due to the intermediate formation of hypochlorous acid (HOCl). Iodination by reagent HOC1 also is inhibited by deferoxamine with the associated consumption of HOCl. Iron saturation of deferoxamine significantly decreased but did not abolish its inhibitory effect on iodination by MPO + H202 or HOCI. Deferoxamine did not affect the absorption spectrum of MPO, suggesting that it does not react with or remove the heme iron. The conversion of MPO to Compound II by HzOz was not seen when H202 was added to MPO in the presence of deferoxamine, suggesting either that deferoxamine inhibited the formation of Compound II by acting as an electron donor for MPO Compound I or that deferoxamine immediately reduced the Compound II formed. Iodination by stimulated neutrophils also was inhibited by deferoxamine, suggesting an effect on peroxidase-catalyzed reactions in intact cells. Thus deferoxamine has multiple effects on the formation and activity of phagocyte-derived oxidants and therefore its inhibitory effect on oxidantdependent damage needs to be interpreted with caution. o 1988Academicpress.I,,~.
Phagocytes have been implicated in the production of tissue damage through the release of powerful oxidants (1). Oxygen is reduced by a membrane-associated NADPH oxidase system to the superoxide anion (0~)~ which either dismutates spontaneously or catalyzed by superoxide dis-
mutase (SOD) to form hydrogen peroxide W20d as 20; + 2H+ + O2 + HzOz
H202 is an oxidant with toxic properties. However, its toxicity can be increased markedly by at least two mechanisms. First, H202 reacts with peroxidase (myi Support for these studies was provided by U.S. eloperoxidase [MPO] of neutrophils and Public Health Service Grants AI 07763 and AI 17758. monocytes; eosinophil peroxidase [EPO]) aTo whom correspondence should he addressed. and a halide to form halide oxidation 3Abbreviations used: DTPA, diethylenetriamineproducts such as hypohalous acids or pentaacetic acid; EDTA, ethylenediaminetetraacetic halogens which are powerful oxidants. acid; EPO, eosinophil peroxidase; HrOs, hydrogen Second, H202 can react with Fe2+ (Fenperoxide; HOCl, hypochlorous acid; MPO, myeloperton’s reagent) to form hydroxyl radicals oxidase; O;, superoxide anion; OH, hydroxyl radical; (OH’) as PMA, phorbol myristate acetate; PMN, polymorphonuclear leukocyte; SOD, superoxide dismutase; TCA, trichloroacetic acid.
0963-9861188$3.96 Copyright 6319S3by AcademicPress,Inc. All rights of reproductionin BDYform reserved.
600
H202 + Fe2+ + Fe’+ + OH- + OH
PEROXIDASE
INHIBITION
When the iron concentration is limiting, reduction of the Fe’+ formed is required for the continued formation of OK and this can be accomplished by 0; as 0; + Fe3++ Fe’+ + O2 with the overall reaction being the ironcatalyzed interaction between Hz02 and 05 to form OH’:
This reaction has been termed the ironcatalyzed Haber-Weiss reaction or the OF-driven Fenton mechanism. The formation of OH and HzOz by phagocytes is well established and their interaction to form a strong oxidant such as OH’ has been widely assumed. Questions have been raised in regard to the specificity of some of the assays employed for the measurement of OH’ (e.g., ethylene formation from methional and 2-keto-4-thiomethylbutyric acid (2) and electron paramagnetic resonance spectroscopy (3, 4)) and the formation of other functionally similar radicals, e.g., crypto-OH (5), ferry1 (6-8) or perferryl radicals, has been proposed. Iron binds readily to a variety of biological compounds (proteins, nucleic acids, iron binding proteins, etc.), and thus free iron is present in very low concentrations in biological fluids. Some chelators (e.g., ethylenediaminetetraacetic acid [EDTA]) bind iron in a catalytically active form and stimulate the formation of OH’ by the Haber-Weiss reaction (9-14) whereas others form iron chelates which are catalytically inactive and thus inhibit the iron-catalyzed formation of OH’. Deferoxamine is a powerful iron chelator obtained from Streptomvces pilosus, which is available as Desferol (deferoxamine mesylate, Ciba Pharmaceutical Co., Summit, NJ) for the treatment of acute and chronic iron intoxication. It chelates iron in a catalytically inactive form and thus inhibits OH’ formation by the Haber-Weiss reaction (13-15). Inhibition by deferoxamine has thus been employed as evidence for the biological importance of iron-dependent radical formation (16). Deferoxamine also may limit the availability of OH’ for oxidant damage by di-
BY DEFEROXAMINE
601
rectly scavenging OH’ (17) or 0; (18, 19). In this study we report that deferoxamine inhibits the peroxidase-H202-halide system thus providing another mechanism for the attenuation of phagocyte-derived oxidant injury by deferoxamine. MATERIALS
AND METHODS
SpeciaZ reagents. Deferoxamine (Desferol) was obtained from Ciba Pharmaceutical Co. and was dissolved in water prior to use. Feroxamine was formed by preincubation of deferoxamine with one or two molar equivalents of ferric chloride for 30 min at room temperature prior to the addition of the other components of the reaction mixture. Guaiacol, human albumin (essentially fatty acid free prepared from fraction V), phorbol 12-myristate 13-acetate (PMA), ethylenediaminetetraacetic acid, disodium salt, dihydrate (EDTA), and diethylenetriaminepentaacetic acid (DTPA) were obtained from Sigma Chemical Co. (St Louis, MO) and H202 (3%) was obtained from Park Davis (Morris Plains, NJ). Sodium hypochlorite (5.25% GP Bleach, Georgia Pacific, Los Angeles, CA) was assayed prior to each experiment by reaction with KI and spectrophotometric measurement of the 1; formed using Eaa = 2.64 X 10’ M-' cm-’ (20). MPO was purified from canine pyometral pus by the method of Agner to the end of step 6 (21) and eosinophil peroxidase (EPO) was purified from horse peripheral blood eosinophils as described by Jorg et al (22). The peroxidases were assayed prior to each experiment by guaiacol oxidation (23). All spectrophotometric analyses were performed with a Beckman DU-70 recording spectrophotometer (Beckman Inst Co, Fullerton, CA). For isolation of polymorphonuclear leukocytes (PMNs), venous blood from normal volunteers was collected just prior to the experiment using 0.2% potassium EDTA as the anticoagulant. The PMNs were isolated by sedimentation in dextran, density gradient centrifugation in Hypaque-Ficoll, and hypotonic lysis of erythrocytes as previously described (24). The preparation, which always contained greater than 97% PMNs with an average purity of 98 to 99%, was suspended in 0.9% sodium chloride at 5 X 10’ PMN/ml. Iodination was measured by the conversion of radioiodide to a trichloroacetic acid (TCA)-precipitable form as previously described (25). The components of the reaction mixture described in the figure or table legends in a final volume of 0.5 ml were tumbled end-over-end for 60 min unless otherwise indicated at 37’C in 12 X 75-mm polystyrene test tubes (Falcon 2054, Be&on-Dickinson, Oxnard, CA) on a Fisher Roto-Rack (Fisher Scientific Co., Pittsburgh, PA). The reaction was stopped by the addition of 1.0 ml of cold 10% TCA and the precipitate was collected by centrifugation (25OOg,5 min, 4°C) and washed four
602
KLEBANOFF
AND WALTERSDORPH
times with 1.0 ml of 10% TCA. The test tube containing the washed precipitate was counted in a gamma scintillation counter. A blank containing the standard salt solution, radioiodide, and albumin was run with each experiment and the counts (less than 0.5% of the total added radioactivity) were subtracted from the experimental values. A standard containing the total ‘%I added to each experimental tube was counted and the percentage iodination was determined as follows cpm experimental - cpm blank x 100. cpm standard Each experimental value was determined in duplicate and the average was employed as a single n for statistical analysis. Statisticalanalysis. The data are expressed as the means + standard error. Statistical differences are determined using Student’s two tailed t test for independent means unless otherwise indicated (not significant [NS], P > 0.05). RESULTS
Iodination by PMA-stimulated human neutrophils was inhibited by deferoxamine. Under the conditions employed in Fig. lA, a significant inhibition was observed at a concentration of 5 X 10V4M, and increased with an increase in deferoxamine concentration. Iodination by stimulated neutrophils is believed to be due predominantly to the oxidation of the added radioiodide by MPO released by degranulation and H202 formed by the respiratory burst (25). In Fig. lB, PMA-stimulated neutrophils were replaced by MPO and HzOz, and, again, iodination was inhibited by deferoxamine in a dose-dependent manner. The deferoxamine could not be replaced in this regard by equivalent concentrations of the chelators EDTA or DTPA. Deferoxamine also inhibited iodination catalyzed by EPO (113 mu) and HzOz (10e4M) under the same conditions as employed in Fig. 1B (without deferoxamine, 78.9 + 8.7%; with low2 M deferoxamine, 2.2 + 0.3%, P < 0.001; with lop3 M deferoxamine, 16.8 f 1.7%, P < 0.01; with 10m4M deferoxamine 50.9 f 2.3%, P < 0.05;n = 3). Iodination by MPO and H202 is stimulated by chloride, particularly when the iodide concentration is relatively low (26). Under these conditions chloride is oxidized by MPO and H202, and the oxidant formed, presumably hypochlorous acid
D.l.rox.mms
(M)
FIG. 1. Inhibitory effect of deferoxamine on iodination. The reaction mixture contained 5 X 10m5M sodium phosphate buffer, pH 7.4, 0.128 M NaCl, 1.2 X lo-’ M KCl, lOmaM CaCl*, 2 X lOwaM MgClr, 8 X lo-@ M NaI (4 nmol; 0.05 &i I?), 2 X lOmaM glucose, 0.25 mg albumin, and deferoxamine (a), EDTA (A), or DTPA (A) at the concentrations indicated. A contained 2.5 X lo6 PMN and 100 ng PMA, B contained 113 mU MPO and lo-’ M HzOz; and C contained lo-’ M HOCl. The results are the means f SE of three to five experiments. The P value for the difference from no chelator is shown where significant (P < 0.05).
(HOCl), oxidizes iodide to form the iodinating species. Figure 1C demonstrates the inhibitory effect of deferoxamine on HOCl-induced iodination using reagent HOC1 as the oxidant.
PEROXIDASE
INHIBITION
The inhibitory effect of deferoxamine on iodination by MPO + H202 and by HOC1 was statistically significant at concentrations down to 5 X 10e4 and 2 X low4 M, respectively, under the conditions employed in Figs. 1B and 1C. However, sensitivity to deferoxamine was dependent on the experimental conditions. Thus when the H202 concentration employed in Fig. 1B (10e4M) was lowered to lo-’ M, deferoxamine significantly inhibited iodination by the MPO-HzOz-chloride system at concentrations down to 2 X 10m5 M (41.1 f 2.9% iodination without deferoxamine vs 29.3 + 1.4% with deferoxamine, n = 3, P < 0.05). Similarly when the HOC1 concentration was lowered from 10e4M (Fig. 1C) to 10e5 M, deferoxamine significantly inhibited iodination at a concentration of 10m5M (33.0 + 1.1% iodination without deferoxamine vs 19.9 f 4.6% with deferoxamine, n = 4, P < 0.05). In contrast, the concentration of deferoxamine required to inhibit iodination by MPO and H202 was unaffected by a lo-fold increase in MPO concentration. The consumption of HOC1 during its incubation with deferoxamine was indicated by the fall in HOC1 levels as assayed by its reaction with KI and the measurement of the I; formed (20). When 50 nmol of HOC1 was incubated with 50 nmol of deferoxamine in 10m2M sodium phosphate buffer, pH 7.0,22.6 + 5.9 nmol of HOC1 was recovered as compared to 50.4 f 0.5 nmol when the HOC1 was incubated in buffer alone (n = 5, P < 0.002). When the amount of deferoxamine added was increased to 5 pmol, essentially all of the HOC1 was lost (1.5 + 0.7 nmol recovered, n = 5, P < 0.001). Iodination in the presence and absence of chloride over the pH range 5.0 to 8.0 is shown in Fig. 2. In the presence of chloride, iodination was high over the pH range 5.0 to 7.0 but fell as the pH was increased to 8.0 under the conditions employed. When chloride was deleted, iodination was low at neutral and alkaline pH and rose as the pH was lowered to 5.0, presumably reflecting the more effective direct oxidation of iodide at this pH. Deferoxamine was inhibitory over the entire pH range (5.0-8.0) and in the presence or absence of chloride. The inhibition by de-
BY DEFEROXAMINE
603
0 53
6.0
7.0 PH
FIG. 2. Effect of pH. The reaction mixture contained lo-’ M sodium phosphate buffer at the pH indicated, 8 X 10v6M NaI (4 nmol; 0.05 pCi ‘%I), 113 mU MPO, 10m4M H,Or, and where indicated 0.1 M NaCi (solid lines) and 2 X lo-’ M deferoxamine (open symbols). The results are the means of three experi-
feroxamine at pH 5.0 indicates that its effects on peroxidase-catalyzed reactions are not limited to those in which HOC1 is an intermediate, as it also can inhibit iodination resulting from the direct oxidation of iodide by MPO and H202. Deferoxamine also inhibited the oxidation of guaiacol by MPO and H202 (Fig. 3). Conversion of deferoxamine to its ironsaturated form (feroxamine) by preincubation with 2 M equivalents of FeS+significantly decreased but did not abolish the inhibition of MPO + H202- or HOCl-induced iodination and completely prevented the inhibition of PMN + PMA-induced iodination under the conditions employed in Table I. The concentration of deferoxamine employed in the PMN + PMA system was limited by the inhibitory effect of FeCb, alone at concentrations higher than that employed in Table I. The effect of deferoxamine on the absorption spectrum of MPO is shown in Fig. 4. The incubation of deferoxamine with MPO for periods up to 15 min had no effect
604
KLEBANOFF
AND WALTERSDORPH
the spectrum remained that of MPO. When deferoxamine was added following the formation of Compound II by the addition of H202 to MPO, the spectrum was converted back to that of MPO (data not shown). DISCUSSION
FIG. 3. Inhibition of MPO-catalyzed guaiacol oxidation by deferoxamine. To a l-cm light path cuvette was added 10-s M sodium phosphate buffer, pH 7.0, 3.3 X lo-’ M HaOr, 180 mU MPO, 0.0043 ~1 guaiacol, deferoxamine at the concentrations indicated, and water to a final volume of 3.0 ml. The cuvette components were rapidly mixed and the absorbancy was determined at 470 nm for l-2 min. The optical density change per minute was calculated from the initial rate and converted to units of peroxidase activity added to the cuvette as previously described (23).
on the absorption spectrum of MPO under the conditions employed in Fig. 4A. When H202 was added to MPO in the absence of deferoxamine, spectral changes occurred which suggested the formation of MPO Compound II; i.e., there was a shift in the Soret band from 430 to 456 nm, a loss in absorption at 570 nm, and the appearance of a peak at approximately 630 nm (Fig. 4B) (27). In contrast, when HzOzwas added to MPO in the presence of deferoxamine,
The studies reported here indicate an inhibitory effect of deferoxamine on MPO-catalyzed reactions. Both iodination and guaiacol oxidation by MPO and Hz02 were inhibited, with the effect on iodination being evident in both the presence and absence of chloride. In the absence of chloride, iodide is oxidized directly by MPO and H202, whereas in its presence, a chloride oxidation product, presumably HOCl, is a required intermediate for optimum iodination. Reagent HOC1 reacts with iodide in the absence of MPO and H202 to induce iodination and this reaction is inhibited by deferoxamine. Iodination by intact neutrophils, which is believed to be due largely to the MPO system, also was inhibited by deferoxamine, suggesting that deferoxamine can affect peroxidase-catalyzed reactions in the intact cell. What is the mechanism of deferoxamine inhibition of MPO-catalyzed reactions? Deferoxamine is a chelating agent with high affinity for iron and MPO contains heme iron essential for its activity. This raises the possibility that deferoxamine inactivates MPO by the chelation of the
TABLE I EFFECTOF IRON SATURATIONOF DEFEROXAMINE Iodination (W) Addition None DF FeC& DF + FeC&
PMN + PMA 62.6 + 44.2 f 58.9 + 59.0 +
2.3” 1.3 <0.001* 2.4 NS 2.7 NS
MPO + Ha02 44.5 + 12.5 + 46.8 + 33.4 +
3.3 1.5
HOC1 32.4 + 8.0 + 32.5 + 21.5 +
1.6 1.5
Note. The reaction mixture was as described in Fig. 1 except that 2.5 X lo6 PMN + 100 ng PMA, 113 mU MPO + lo-’ M H202, and lo-’ M HOC1 were employed and deferoxamine (DF, 5 X 10e4M for PMN + PMA, lo-’ M for MPO + HaOr and HOCl), FeCla (lo-’ M for PMN + PMA, 2 X 1O-4M for MPO + HzOz and HOCl), or both DF and FeC18were added where indicated. DF and FeCls were preincubated for 30 min at room temperature. a Mean k SE of four experiments. *P value for the difference from no addition. ‘P value for the difference from DF alone.
PEROXIDASE
.
.
.
.
220
288
316
364
412
460
506
556
INHIBITION
604
552
BY DEFEROXAMINE
700
Wavelength
220
288
316
605
.
.
.
.
.
.
*
.
364
412
460
sba
555
604
652
700
(nm)
FIG. 4. The effect of deferoxamine on the absorption spectrum of MPO. In A, the l-cm light path cuvette contained lo-* M deferoxamine (-), 45 units MPO ( - * * ), or lo-* M deferoxamine + 45 units MPO (- - -). In B the cuvette contained 45 units MPO (-), 45 units MPO + 10m4M H202 (- - -), or 45 units MPO + lo-’ M Hz02 + lo-’ M deferoxamine ( - - - ). Total volume, 1 ml.
heme iron and the finding that iron-saturated feroxamine was less effective than deferoxamine in the inhibition of MPO tends to support this mechanism. However, a number of lines of evidence mitigate against MPO-iron chelation as the mechanism of inhibition. The addition of inhibitory concentrations of deferoxamine had no effect on the absorption spectrum of MPO, suggesting that deferoxamine neither binds to the heme iron nor removes it from the MPO molecule. Further, the concentration of deferoxamine which was inhibitory varied not with the MPO but with the H202 concentration. Iodination by MPO and HzOz is stimulated by chloride particularly at neutral or alkaline pH and the mechanism proposed is the oxidation of chloride by the peroxidase system to HOC1 and the oxidation of iodide by HOC1 to form the iodinating species. Iodination occurs on the addition of reagent HOC1 to iodide in a protein containing solution and this iodination reaction, which occurs in the absence of MPO, also is inhibited by deferoxamine. The inhibitory concentration of deferoxamine varied with the HOC1 concentration. Finally, the concentration of deferoxamine required for inhibition was considerably higher than that of the heme iron or the trace amounts of iron contamination in the reaction mixture (approximately 5 X lo-* M).
Our findings suggest that deferoxamine is oxidized by MPO and H202 and thus would be expected to compete with other electron donors for oxidation by the MPO system. MPO reacts with H202 to form a highly unstable, catalytically active complex called Compound I, which reacts with a variety of electron donors to reform MPO with oxidation of the electron donor. In the absence of a suitable electron donor or in the presence of an excess of H202, Compound I is converted to Compound II which is inactive as a catalyst of iodide or chloride oxidation. The formation of Compound II on the addition of Hz02 to MPO was not seen in the presence of deferoxamine, which is compatible with deferoxamine acting as an electron donor in the peroxidase system. However, the spectrum of Compound II is converted to that of MPO by deferoxamine, raising the possibility that the absence of the spectrum of Compound II on the addition of H202 to MPO in the presence of deferoxamine may be due in part or totally to the reduction of Compound II by deferoxamine rather than to the lack of its formation. Deferoxamine also reacts with and degrades HOCl, a product of the MPO-HzOz-chloride system and a probable intermediate in the MPOdependent iodination reaction under our experimental conditions. Although these findings suggest that deferoxamine inhibits iodination in part by the degrada-
606
KLEBANOFF
AND
tion of HOCl, iodination and the oxidation of guaiacol is inhibited by deferoxamine in the absence of chloride, indicating a more general effect on peroxidase-catalyzed reactions. Deferoxamine is oxidized to a nitroxide free radical by horseradish peroxidase and H202, by hypoxanthine and xanthine oxidase, and by OH’ generated by the photolysis of H202 (28). It appears likely therefore that deferoxamine inhibits MPO-catalyzed reactions either by competing with other electron donors for oxidation by MPO and HzOz or by reacting with oxidants formed by the MPO system such as HOCl. Our finding that the conversion of deferoxamine to feroxamine by the addition of iron decreased its inhibitory effect on MPO-catalyzed reactions suggests that the iron chelate is less susceptible to oxidation than is the iron-free molecule. Feroxamine also was less effective than deferoxamine as a scavenger of 05 (18) whereas feroxamine and deferoxamine were equally effective as OH’ scavengers (17). OH’ formation by the iron-catalyzed Haber-Weiss reaction was inhibited by deferoxamine but not by feroxamine at relatively low concentrations; however, when concentrations were high both deferoxamine and feroxamine were inhibitory (18). Among the evidence given in support of a role for OH’ in tissue damage induced by phagocytes is the inhibitory effect of deferoxamine, based on the iron requirement for OH’ formation by the HaberWeiss reaction and on the strong ironchelating properties of deferoxamine. The studies reported here indicate that deferoxamine also can inhibit MPO-catalyzed reactions and thus data obtained with deferoxamine in relation to the toxic mechanisms of phagocytes should be interpreted with caution, particularly when the deferoxamine concentration is high relative to that of available iron. ACKNOWLEDGMENT We gratefully acknowledge the expert secretarial assistance of Ms. Caroline Wilson. REFERENCES 1. KLEBANOFF, S. J. (1988) in Inflammation: Basic Principles and Clinical Correlates (Gallin,
WALTERSDORPH
2. 3.
4.
5.
J. I., Goldstein, I. M., and Snyderman R., Eds.), pp. 391-444, Raven Press, New York. KLEBANOFF, S. J., AND ROSEN, H. (1978). J. Exp. Med. 148.490-506. BRITIGAN, B. E., ROSEN, G. M., CHAI, Y., AND COHEN, M. S. (1986) J. Biol. Chem. 261, 4426-4431. BRITIGAN, B. E., COHEN, M. S., AND ROSEN, G. M. (1987) J. Leukocyte Biol. 41,349-362. YOUNGMAN, R. J. (1984) Trends Biochem Sci 9,
280-283. 6. KOPPENOL, W. H., AND LIEBMAN, J. F. (1984) J. Phys. Chem 88,99-101. 7. RUSH, J. D., AND KOPPENOL, W. H. (1986) J. Biol Chem. 261,6730-6733. 8. RUSH, J. D., AND BIELSKI, B. H. J. (1986) J. Amer. Chem. Sot. 108,523-525. 9. BUETTNER, G. R., OBERLEY, L. W., AND CHAN LEUTHAUSER, S. W. H. (1978) Phdochem Pb
tobiol 28,693-695. 10. MCCORD, J. M., AND DAY, E. D., JR. (1978) FEBS
L&t. 86,139-142. 11. HALLIWELL, B. (1978) FEBS L&t. 92,321-326. 12. HALLIWELL, B. (1978) FEBS L&t. 96,238-242. 13. GUTTERIDGE, J. M. C., RICHMOND, R., AND HALLIWELL, B. (1979) Biochem J. 184,469-472. 14. ROSEN, H., AND KLEBANOFF, S. J. (1981) Arch,
Bbchem Biophys. 208,512-519. 15. CEDERBAUM, A. I., AND DICKER, E. (1983) Bio&cm J. 210,107-113. 16. HALLIWELL, B., AND GUTTERIDGE, J. M. C. (1986)
Arch. Biochem. Biophys. 246,501-513. 17. HOE, S., ROWLEY, D. A., AND HALLIWELL, B. (1982) Chem.-Bid Interact. 41,75-81. 18. SINACEUR, J., RIBIBRE, C., NORDMANN, J., AND NORDMANN, R. (1984) Biochem Phawuxol33, 1693-1694. 19. HALLIWELL, B. (1985) Biochem. Pharmad 34,
229-233. 20. AWTREY, A. D., AND CONNICK, R. E. (1951) .I.
Amer. Chem Sot. 73.1842-1843. 21. AGNER, K. (1958) Acta Chew. Scund 12,89-94. 22. JORG, A., PASQUIER, J.-M., AND KLEBANOFF, S. J. (1982) Biochem Biophys. Acta 701,185-191. 23. KLEBANOFF, S. J., WALTERSDORPH, A. M., AND ROSEN, H. (1984) in Methods in Enzymology (Packer, L., Ed.), Vol. 105, pp. 399-403, Academic Press, San Diego. 24. CLARK, R. A., AND KLEBANOFF, S. J. (1979) J. Immunol 122,2605-2610. 25. KLEBANOFF, S. J., AND CLARK, R. A. (1977) J. Lab. Clin Med 89,675-686. 26. KLEBANOFF, S. J. (1970) in Biochemistry of the Phagocytic Process (Schultz, J., Ed.), pp. 89-110, North-Holland, Amsterdam. 27. HOOGLAND, H., VAN KUILENBURG, A., VAN RIEL C., MUIJSERS, A. O., AND WEVER, R. (1987) B&him. Biophya Acta 916,75-82. 28. MOREHOUSE, K. M., FLITTER, W. D., AND MASON, R. P. (1987) FEBS Lett. 222.246-250.
Inhibition of Peroxidase-Catalyzed SEYMOUR J. KLEBANOFF’
Reactions by Deferoxamine’
AND
ANN M. WALTERSDORPH
Department of Medicine, RM-16, University of Washington, Seattle WA 981% Received January 20,1988, and in revised form March 24,1988
Phagocytes generate superoxide (05) and hydrogen peroxide (HzOz) and their interaction in an iron-catalyzed reaction to form hydroxyl radicals (OH) (Haber-Weiss reaction) has been proposed. Deferoxamine chelates iron in a catalytically inactive form, and thus inhibition by deferoxamine has been employed as evidence for the involvement of OH generated by the Haber-Weiss reaction. We report here that deferoxamine also inhibits reactions catalyzed by the peroxidases of phagocytes, i.e., myeloperoxidase (MPO) and eosinophil peroxidase (EPO). The reactions inhibited include iodination in the presence and absence of chloride and the oxidation of guaiacol. Iodination by MPO and Hz02 is stimulated by chloride due to the intermediate formation of hypochlorous acid (HOCl). Iodination by reagent HOC1 also is inhibited by deferoxamine with the associated consumption of HOCl. Iron saturation of deferoxamine significantly decreased but did not abolish its inhibitory effect on iodination by MPO + H202 or HOCI. Deferoxamine did not affect the absorption spectrum of MPO, suggesting that it does not react with or remove the heme iron. The conversion of MPO to Compound II by HzOz was not seen when H202 was added to MPO in the presence of deferoxamine, suggesting either that deferoxamine inhibited the formation of Compound II by acting as an electron donor for MPO Compound I or that deferoxamine immediately reduced the Compound II formed. Iodination by stimulated neutrophils also was inhibited by deferoxamine, suggesting an effect on peroxidase-catalyzed reactions in intact cells. Thus deferoxamine has multiple effects on the formation and activity of phagocyte-derived oxidants and therefore its inhibitory effect on oxidantdependent damage needs to be interpreted with caution. o 1988Academicpress.I,,~.
Phagocytes have been implicated in the production of tissue damage through the release of powerful oxidants (1). Oxygen is reduced by a membrane-associated NADPH oxidase system to the superoxide anion (0~)~ which either dismutates spontaneously or catalyzed by superoxide dis-
mutase (SOD) to form hydrogen peroxide W20d as 20; + 2H+ + O2 + HzOz
H202 is an oxidant with toxic properties. However, its toxicity can be increased markedly by at least two mechanisms. First, H202 reacts with peroxidase (myi Support for these studies was provided by U.S. eloperoxidase [MPO] of neutrophils and Public Health Service Grants AI 07763 and AI 17758. monocytes; eosinophil peroxidase [EPO]) aTo whom correspondence should he addressed. and a halide to form halide oxidation 3Abbreviations used: DTPA, diethylenetriamineproducts such as hypohalous acids or pentaacetic acid; EDTA, ethylenediaminetetraacetic halogens which are powerful oxidants. acid; EPO, eosinophil peroxidase; HrOs, hydrogen Second, H202 can react with Fe2+ (Fenperoxide; HOCl, hypochlorous acid; MPO, myeloperton’s reagent) to form hydroxyl radicals oxidase; O;, superoxide anion; OH, hydroxyl radical; (OH’) as PMA, phorbol myristate acetate; PMN, polymorphonuclear leukocyte; SOD, superoxide dismutase; TCA, trichloroacetic acid.
0963-9861188$3.96 Copyright 6319S3by AcademicPress,Inc. All rights of reproductionin BDYform reserved.
600
H202 + Fe2+ + Fe’+ + OH- + OH
PEROXIDASE
INHIBITION
When the iron concentration is limiting, reduction of the Fe’+ formed is required for the continued formation of OK and this can be accomplished by 0; as 0; + Fe3++ Fe’+ + O2 with the overall reaction being the ironcatalyzed interaction between Hz02 and 05 to form OH’:
This reaction has been termed the ironcatalyzed Haber-Weiss reaction or the OF-driven Fenton mechanism. The formation of OH and HzOz by phagocytes is well established and their interaction to form a strong oxidant such as OH’ has been widely assumed. Questions have been raised in regard to the specificity of some of the assays employed for the measurement of OH’ (e.g., ethylene formation from methional and 2-keto-4-thiomethylbutyric acid (2) and electron paramagnetic resonance spectroscopy (3, 4)) and the formation of other functionally similar radicals, e.g., crypto-OH (5), ferry1 (6-8) or perferryl radicals, has been proposed. Iron binds readily to a variety of biological compounds (proteins, nucleic acids, iron binding proteins, etc.), and thus free iron is present in very low concentrations in biological fluids. Some chelators (e.g., ethylenediaminetetraacetic acid [EDTA]) bind iron in a catalytically active form and stimulate the formation of OH’ by the Haber-Weiss reaction (9-14) whereas others form iron chelates which are catalytically inactive and thus inhibit the iron-catalyzed formation of OH’. Deferoxamine is a powerful iron chelator obtained from Streptomvces pilosus, which is available as Desferol (deferoxamine mesylate, Ciba Pharmaceutical Co., Summit, NJ) for the treatment of acute and chronic iron intoxication. It chelates iron in a catalytically inactive form and thus inhibits OH’ formation by the Haber-Weiss reaction (13-15). Inhibition by deferoxamine has thus been employed as evidence for the biological importance of iron-dependent radical formation (16). Deferoxamine also may limit the availability of OH’ for oxidant damage by di-
BY DEFEROXAMINE
601
rectly scavenging OH’ (17) or 0; (18, 19). In this study we report that deferoxamine inhibits the peroxidase-H202-halide system thus providing another mechanism for the attenuation of phagocyte-derived oxidant injury by deferoxamine. MATERIALS
AND METHODS
SpeciaZ reagents. Deferoxamine (Desferol) was obtained from Ciba Pharmaceutical Co. and was dissolved in water prior to use. Feroxamine was formed by preincubation of deferoxamine with one or two molar equivalents of ferric chloride for 30 min at room temperature prior to the addition of the other components of the reaction mixture. Guaiacol, human albumin (essentially fatty acid free prepared from fraction V), phorbol 12-myristate 13-acetate (PMA), ethylenediaminetetraacetic acid, disodium salt, dihydrate (EDTA), and diethylenetriaminepentaacetic acid (DTPA) were obtained from Sigma Chemical Co. (St Louis, MO) and H202 (3%) was obtained from Park Davis (Morris Plains, NJ). Sodium hypochlorite (5.25% GP Bleach, Georgia Pacific, Los Angeles, CA) was assayed prior to each experiment by reaction with KI and spectrophotometric measurement of the 1; formed using Eaa = 2.64 X 10’ M-' cm-’ (20). MPO was purified from canine pyometral pus by the method of Agner to the end of step 6 (21) and eosinophil peroxidase (EPO) was purified from horse peripheral blood eosinophils as described by Jorg et al (22). The peroxidases were assayed prior to each experiment by guaiacol oxidation (23). All spectrophotometric analyses were performed with a Beckman DU-70 recording spectrophotometer (Beckman Inst Co, Fullerton, CA). For isolation of polymorphonuclear leukocytes (PMNs), venous blood from normal volunteers was collected just prior to the experiment using 0.2% potassium EDTA as the anticoagulant. The PMNs were isolated by sedimentation in dextran, density gradient centrifugation in Hypaque-Ficoll, and hypotonic lysis of erythrocytes as previously described (24). The preparation, which always contained greater than 97% PMNs with an average purity of 98 to 99%, was suspended in 0.9% sodium chloride at 5 X 10’ PMN/ml. Iodination was measured by the conversion of radioiodide to a trichloroacetic acid (TCA)-precipitable form as previously described (25). The components of the reaction mixture described in the figure or table legends in a final volume of 0.5 ml were tumbled end-over-end for 60 min unless otherwise indicated at 37’C in 12 X 75-mm polystyrene test tubes (Falcon 2054, Be&on-Dickinson, Oxnard, CA) on a Fisher Roto-Rack (Fisher Scientific Co., Pittsburgh, PA). The reaction was stopped by the addition of 1.0 ml of cold 10% TCA and the precipitate was collected by centrifugation (25OOg,5 min, 4°C) and washed four
602
KLEBANOFF
AND WALTERSDORPH
times with 1.0 ml of 10% TCA. The test tube containing the washed precipitate was counted in a gamma scintillation counter. A blank containing the standard salt solution, radioiodide, and albumin was run with each experiment and the counts (less than 0.5% of the total added radioactivity) were subtracted from the experimental values. A standard containing the total ‘%I added to each experimental tube was counted and the percentage iodination was determined as follows cpm experimental - cpm blank x 100. cpm standard Each experimental value was determined in duplicate and the average was employed as a single n for statistical analysis. Statisticalanalysis. The data are expressed as the means + standard error. Statistical differences are determined using Student’s two tailed t test for independent means unless otherwise indicated (not significant [NS], P > 0.05). RESULTS
Iodination by PMA-stimulated human neutrophils was inhibited by deferoxamine. Under the conditions employed in Fig. lA, a significant inhibition was observed at a concentration of 5 X 10V4M, and increased with an increase in deferoxamine concentration. Iodination by stimulated neutrophils is believed to be due predominantly to the oxidation of the added radioiodide by MPO released by degranulation and H202 formed by the respiratory burst (25). In Fig. lB, PMA-stimulated neutrophils were replaced by MPO and HzOz, and, again, iodination was inhibited by deferoxamine in a dose-dependent manner. The deferoxamine could not be replaced in this regard by equivalent concentrations of the chelators EDTA or DTPA. Deferoxamine also inhibited iodination catalyzed by EPO (113 mu) and HzOz (10e4M) under the same conditions as employed in Fig. 1B (without deferoxamine, 78.9 + 8.7%; with low2 M deferoxamine, 2.2 + 0.3%, P < 0.001; with lop3 M deferoxamine, 16.8 f 1.7%, P < 0.01; with 10m4M deferoxamine 50.9 f 2.3%, P < 0.05;n = 3). Iodination by MPO and H202 is stimulated by chloride, particularly when the iodide concentration is relatively low (26). Under these conditions chloride is oxidized by MPO and H202, and the oxidant formed, presumably hypochlorous acid
D.l.rox.mms
(M)
FIG. 1. Inhibitory effect of deferoxamine on iodination. The reaction mixture contained 5 X 10m5M sodium phosphate buffer, pH 7.4, 0.128 M NaCl, 1.2 X lo-’ M KCl, lOmaM CaCl*, 2 X lOwaM MgClr, 8 X lo-@ M NaI (4 nmol; 0.05 &i I?), 2 X lOmaM glucose, 0.25 mg albumin, and deferoxamine (a), EDTA (A), or DTPA (A) at the concentrations indicated. A contained 2.5 X lo6 PMN and 100 ng PMA, B contained 113 mU MPO and lo-’ M HzOz; and C contained lo-’ M HOCl. The results are the means f SE of three to five experiments. The P value for the difference from no chelator is shown where significant (P < 0.05).
(HOCl), oxidizes iodide to form the iodinating species. Figure 1C demonstrates the inhibitory effect of deferoxamine on HOCl-induced iodination using reagent HOC1 as the oxidant.
PEROXIDASE
INHIBITION
The inhibitory effect of deferoxamine on iodination by MPO + H202 and by HOC1 was statistically significant at concentrations down to 5 X 10e4 and 2 X low4 M, respectively, under the conditions employed in Figs. 1B and 1C. However, sensitivity to deferoxamine was dependent on the experimental conditions. Thus when the H202 concentration employed in Fig. 1B (10e4M) was lowered to lo-’ M, deferoxamine significantly inhibited iodination by the MPO-HzOz-chloride system at concentrations down to 2 X 10m5 M (41.1 f 2.9% iodination without deferoxamine vs 29.3 + 1.4% with deferoxamine, n = 3, P < 0.05). Similarly when the HOC1 concentration was lowered from 10e4M (Fig. 1C) to 10e5 M, deferoxamine significantly inhibited iodination at a concentration of 10m5M (33.0 + 1.1% iodination without deferoxamine vs 19.9 f 4.6% with deferoxamine, n = 4, P < 0.05). In contrast, the concentration of deferoxamine required to inhibit iodination by MPO and H202 was unaffected by a lo-fold increase in MPO concentration. The consumption of HOC1 during its incubation with deferoxamine was indicated by the fall in HOC1 levels as assayed by its reaction with KI and the measurement of the I; formed (20). When 50 nmol of HOC1 was incubated with 50 nmol of deferoxamine in 10m2M sodium phosphate buffer, pH 7.0,22.6 + 5.9 nmol of HOC1 was recovered as compared to 50.4 f 0.5 nmol when the HOC1 was incubated in buffer alone (n = 5, P < 0.002). When the amount of deferoxamine added was increased to 5 pmol, essentially all of the HOC1 was lost (1.5 + 0.7 nmol recovered, n = 5, P < 0.001). Iodination in the presence and absence of chloride over the pH range 5.0 to 8.0 is shown in Fig. 2. In the presence of chloride, iodination was high over the pH range 5.0 to 7.0 but fell as the pH was increased to 8.0 under the conditions employed. When chloride was deleted, iodination was low at neutral and alkaline pH and rose as the pH was lowered to 5.0, presumably reflecting the more effective direct oxidation of iodide at this pH. Deferoxamine was inhibitory over the entire pH range (5.0-8.0) and in the presence or absence of chloride. The inhibition by de-
BY DEFEROXAMINE
603
0 53
6.0
7.0 PH
FIG. 2. Effect of pH. The reaction mixture contained lo-’ M sodium phosphate buffer at the pH indicated, 8 X 10v6M NaI (4 nmol; 0.05 pCi ‘%I), 113 mU MPO, 10m4M H,Or, and where indicated 0.1 M NaCi (solid lines) and 2 X lo-’ M deferoxamine (open symbols). The results are the means of three experi-
feroxamine at pH 5.0 indicates that its effects on peroxidase-catalyzed reactions are not limited to those in which HOC1 is an intermediate, as it also can inhibit iodination resulting from the direct oxidation of iodide by MPO and H202. Deferoxamine also inhibited the oxidation of guaiacol by MPO and H202 (Fig. 3). Conversion of deferoxamine to its ironsaturated form (feroxamine) by preincubation with 2 M equivalents of FeS+significantly decreased but did not abolish the inhibition of MPO + H202- or HOCl-induced iodination and completely prevented the inhibition of PMN + PMA-induced iodination under the conditions employed in Table I. The concentration of deferoxamine employed in the PMN + PMA system was limited by the inhibitory effect of FeCb, alone at concentrations higher than that employed in Table I. The effect of deferoxamine on the absorption spectrum of MPO is shown in Fig. 4. The incubation of deferoxamine with MPO for periods up to 15 min had no effect
604
KLEBANOFF
AND WALTERSDORPH
the spectrum remained that of MPO. When deferoxamine was added following the formation of Compound II by the addition of H202 to MPO, the spectrum was converted back to that of MPO (data not shown). DISCUSSION
FIG. 3. Inhibition of MPO-catalyzed guaiacol oxidation by deferoxamine. To a l-cm light path cuvette was added 10-s M sodium phosphate buffer, pH 7.0, 3.3 X lo-’ M HaOr, 180 mU MPO, 0.0043 ~1 guaiacol, deferoxamine at the concentrations indicated, and water to a final volume of 3.0 ml. The cuvette components were rapidly mixed and the absorbancy was determined at 470 nm for l-2 min. The optical density change per minute was calculated from the initial rate and converted to units of peroxidase activity added to the cuvette as previously described (23).
on the absorption spectrum of MPO under the conditions employed in Fig. 4A. When H202 was added to MPO in the absence of deferoxamine, spectral changes occurred which suggested the formation of MPO Compound II; i.e., there was a shift in the Soret band from 430 to 456 nm, a loss in absorption at 570 nm, and the appearance of a peak at approximately 630 nm (Fig. 4B) (27). In contrast, when HzOzwas added to MPO in the presence of deferoxamine,
The studies reported here indicate an inhibitory effect of deferoxamine on MPO-catalyzed reactions. Both iodination and guaiacol oxidation by MPO and Hz02 were inhibited, with the effect on iodination being evident in both the presence and absence of chloride. In the absence of chloride, iodide is oxidized directly by MPO and H202, whereas in its presence, a chloride oxidation product, presumably HOCl, is a required intermediate for optimum iodination. Reagent HOC1 reacts with iodide in the absence of MPO and H202 to induce iodination and this reaction is inhibited by deferoxamine. Iodination by intact neutrophils, which is believed to be due largely to the MPO system, also was inhibited by deferoxamine, suggesting that deferoxamine can affect peroxidase-catalyzed reactions in the intact cell. What is the mechanism of deferoxamine inhibition of MPO-catalyzed reactions? Deferoxamine is a chelating agent with high affinity for iron and MPO contains heme iron essential for its activity. This raises the possibility that deferoxamine inactivates MPO by the chelation of the
TABLE I EFFECTOF IRON SATURATIONOF DEFEROXAMINE Iodination (W) Addition None DF FeC& DF + FeC&
PMN + PMA 62.6 + 44.2 f 58.9 + 59.0 +
2.3” 1.3 <0.001* 2.4 NS 2.7 NS
MPO + Ha02 44.5 + 12.5 + 46.8 + 33.4 +
3.3 1.5
HOC1 32.4 + 8.0 + 32.5 + 21.5 +
1.6 1.5
Note. The reaction mixture was as described in Fig. 1 except that 2.5 X lo6 PMN + 100 ng PMA, 113 mU MPO + lo-’ M H202, and lo-’ M HOC1 were employed and deferoxamine (DF, 5 X 10e4M for PMN + PMA, lo-’ M for MPO + HaOr and HOCl), FeCla (lo-’ M for PMN + PMA, 2 X 1O-4M for MPO + HzOz and HOCl), or both DF and FeC18were added where indicated. DF and FeCls were preincubated for 30 min at room temperature. a Mean k SE of four experiments. *P value for the difference from no addition. ‘P value for the difference from DF alone.
PEROXIDASE
.
.
.
.
220
288
316
364
412
460
506
556
INHIBITION
604
552
BY DEFEROXAMINE
700
Wavelength
220
288
316
605
.
.
.
.
.
.
*
.
364
412
460
sba
555
604
652
700
(nm)
FIG. 4. The effect of deferoxamine on the absorption spectrum of MPO. In A, the l-cm light path cuvette contained lo-* M deferoxamine (-), 45 units MPO ( - * * ), or lo-* M deferoxamine + 45 units MPO (- - -). In B the cuvette contained 45 units MPO (-), 45 units MPO + 10m4M H202 (- - -), or 45 units MPO + lo-’ M Hz02 + lo-’ M deferoxamine ( - - - ). Total volume, 1 ml.
heme iron and the finding that iron-saturated feroxamine was less effective than deferoxamine in the inhibition of MPO tends to support this mechanism. However, a number of lines of evidence mitigate against MPO-iron chelation as the mechanism of inhibition. The addition of inhibitory concentrations of deferoxamine had no effect on the absorption spectrum of MPO, suggesting that deferoxamine neither binds to the heme iron nor removes it from the MPO molecule. Further, the concentration of deferoxamine which was inhibitory varied not with the MPO but with the H202 concentration. Iodination by MPO and HzOz is stimulated by chloride particularly at neutral or alkaline pH and the mechanism proposed is the oxidation of chloride by the peroxidase system to HOC1 and the oxidation of iodide by HOC1 to form the iodinating species. Iodination occurs on the addition of reagent HOC1 to iodide in a protein containing solution and this iodination reaction, which occurs in the absence of MPO, also is inhibited by deferoxamine. The inhibitory concentration of deferoxamine varied with the HOC1 concentration. Finally, the concentration of deferoxamine required for inhibition was considerably higher than that of the heme iron or the trace amounts of iron contamination in the reaction mixture (approximately 5 X lo-* M).
Our findings suggest that deferoxamine is oxidized by MPO and H202 and thus would be expected to compete with other electron donors for oxidation by the MPO system. MPO reacts with H202 to form a highly unstable, catalytically active complex called Compound I, which reacts with a variety of electron donors to reform MPO with oxidation of the electron donor. In the absence of a suitable electron donor or in the presence of an excess of H202, Compound I is converted to Compound II which is inactive as a catalyst of iodide or chloride oxidation. The formation of Compound II on the addition of Hz02 to MPO was not seen in the presence of deferoxamine, which is compatible with deferoxamine acting as an electron donor in the peroxidase system. However, the spectrum of Compound II is converted to that of MPO by deferoxamine, raising the possibility that the absence of the spectrum of Compound II on the addition of H202 to MPO in the presence of deferoxamine may be due in part or totally to the reduction of Compound II by deferoxamine rather than to the lack of its formation. Deferoxamine also reacts with and degrades HOCl, a product of the MPO-HzOz-chloride system and a probable intermediate in the MPOdependent iodination reaction under our experimental conditions. Although these findings suggest that deferoxamine inhibits iodination in part by the degrada-
606
KLEBANOFF
AND
tion of HOCl, iodination and the oxidation of guaiacol is inhibited by deferoxamine in the absence of chloride, indicating a more general effect on peroxidase-catalyzed reactions. Deferoxamine is oxidized to a nitroxide free radical by horseradish peroxidase and H202, by hypoxanthine and xanthine oxidase, and by OH’ generated by the photolysis of H202 (28). It appears likely therefore that deferoxamine inhibits MPO-catalyzed reactions either by competing with other electron donors for oxidation by MPO and HzOz or by reacting with oxidants formed by the MPO system such as HOCl. Our finding that the conversion of deferoxamine to feroxamine by the addition of iron decreased its inhibitory effect on MPO-catalyzed reactions suggests that the iron chelate is less susceptible to oxidation than is the iron-free molecule. Feroxamine also was less effective than deferoxamine as a scavenger of 05 (18) whereas feroxamine and deferoxamine were equally effective as OH’ scavengers (17). OH’ formation by the iron-catalyzed Haber-Weiss reaction was inhibited by deferoxamine but not by feroxamine at relatively low concentrations; however, when concentrations were high both deferoxamine and feroxamine were inhibitory (18). Among the evidence given in support of a role for OH’ in tissue damage induced by phagocytes is the inhibitory effect of deferoxamine, based on the iron requirement for OH’ formation by the HaberWeiss reaction and on the strong ironchelating properties of deferoxamine. The studies reported here indicate that deferoxamine also can inhibit MPO-catalyzed reactions and thus data obtained with deferoxamine in relation to the toxic mechanisms of phagocytes should be interpreted with caution, particularly when the deferoxamine concentration is high relative to that of available iron. ACKNOWLEDGMENT We gratefully acknowledge the expert secretarial assistance of Ms. Caroline Wilson. REFERENCES 1. KLEBANOFF, S. J. (1988) in Inflammation: Basic Principles and Clinical Correlates (Gallin,
WALTERSDORPH
2. 3.
4.
5.
J. I., Goldstein, I. M., and Snyderman R., Eds.), pp. 391-444, Raven Press, New York. KLEBANOFF, S. J., AND ROSEN, H. (1978). J. Exp. Med. 148.490-506. BRITIGAN, B. E., ROSEN, G. M., CHAI, Y., AND COHEN, M. S. (1986) J. Biol. Chem. 261, 4426-4431. BRITIGAN, B. E., COHEN, M. S., AND ROSEN, G. M. (1987) J. Leukocyte Biol. 41,349-362. YOUNGMAN, R. J. (1984) Trends Biochem Sci 9,
280-283. 6. KOPPENOL, W. H., AND LIEBMAN, J. F. (1984) J. Phys. Chem 88,99-101. 7. RUSH, J. D., AND KOPPENOL, W. H. (1986) J. Biol Chem. 261,6730-6733. 8. RUSH, J. D., AND BIELSKI, B. H. J. (1986) J. Amer. Chem. Sot. 108,523-525. 9. BUETTNER, G. R., OBERLEY, L. W., AND CHAN LEUTHAUSER, S. W. H. (1978) Phdochem Pb
tobiol 28,693-695. 10. MCCORD, J. M., AND DAY, E. D., JR. (1978) FEBS
L&t. 86,139-142. 11. HALLIWELL, B. (1978) FEBS L&t. 92,321-326. 12. HALLIWELL, B. (1978) FEBS L&t. 96,238-242. 13. GUTTERIDGE, J. M. C., RICHMOND, R., AND HALLIWELL, B. (1979) Biochem J. 184,469-472. 14. ROSEN, H., AND KLEBANOFF, S. J. (1981) Arch,
Bbchem Biophys. 208,512-519. 15. CEDERBAUM, A. I., AND DICKER, E. (1983) Bio&cm J. 210,107-113. 16. HALLIWELL, B., AND GUTTERIDGE, J. M. C. (1986)
Arch. Biochem. Biophys. 246,501-513. 17. HOE, S., ROWLEY, D. A., AND HALLIWELL, B. (1982) Chem.-Bid Interact. 41,75-81. 18. SINACEUR, J., RIBIBRE, C., NORDMANN, J., AND NORDMANN, R. (1984) Biochem Phawuxol33, 1693-1694. 19. HALLIWELL, B. (1985) Biochem. Pharmad 34,
229-233. 20. AWTREY, A. D., AND CONNICK, R. E. (1951) .I.
Amer. Chem Sot. 73.1842-1843. 21. AGNER, K. (1958) Acta Chew. Scund 12,89-94. 22. JORG, A., PASQUIER, J.-M., AND KLEBANOFF, S. J. (1982) Biochem Biophys. Acta 701,185-191. 23. KLEBANOFF, S. J., WALTERSDORPH, A. M., AND ROSEN, H. (1984) in Methods in Enzymology (Packer, L., Ed.), Vol. 105, pp. 399-403, Academic Press, San Diego. 24. CLARK, R. A., AND KLEBANOFF, S. J. (1979) J. Immunol 122,2605-2610. 25. KLEBANOFF, S. J., AND CLARK, R. A. (1977) J. Lab. Clin Med 89,675-686. 26. KLEBANOFF, S. J. (1970) in Biochemistry of the Phagocytic Process (Schultz, J., Ed.), pp. 89-110, North-Holland, Amsterdam. 27. HOOGLAND, H., VAN KUILENBURG, A., VAN RIEL C., MUIJSERS, A. O., AND WEVER, R. (1987) B&him. Biophya Acta 916,75-82. 28. MOREHOUSE, K. M., FLITTER, W. D., AND MASON, R. P. (1987) FEBS Lett. 222.246-250.