Mn(III)-desferrioxamine superoxide dismutase-mimic: Alternative modes of action

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 288, No. 1, July, pp. 215-219, 1991

Mn(lII)-Desferrioxamine Superoxide Alternative Modes of Action Stephen M. Hahn,* *Radiation tMolecular

C. Murali

Krishna,*

Dismutase-Mimic:

Amram Samuni,? James B. Mitchell,*

and Angelo RUSSO**~

Oncology Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland Biology, Hebrew University Medical School, Jerusalem 91010, Israel

20892; and

Received October 24, 1990, and in revised form March 5, 1991

Various low-molecular-weight copper chelates have been synthesized to mimic superoxide dismutase (SOD) by catalyzing 02 dismutation. However, in the presence of cellular proteins, such chelates dissociate and thereby lose their SOD-mimetic activity. In contrast, desferrioxamine-Mn(II1) 1: 1 chelate (DF-Mn), an SOD-mimic that affords protection from oxidative damage, reportedly is stable in the presence of serum albumin. DF-Mn, unlike SOD, is reported to permeate the membrane of at least one cell type and can protect cells by detoxifying intracellular 0;. Recently DF-Mn was shown to protect hypoxic cells from H20,-induced damage. Such results suggest that DF-Mn can protect cells from OG-independent damage by alternative mechanisms. This study examines such possibilities. To avoid 0; participation in the damaging process, killing of monolayered V79 Chinese hamster cells was induced in a hypoxic environment by tbutyl hydroperoxide (t-BHP). Damage induced by t-BHP was inhibitable by DF-Mn. DF-Mn was also found to rapidly oxidize iron(bound DNA. Additionally, once DF-Mn oxidizes Fe(I1) or Cu(I), the DF moiety of DFMn dissociates and rapidly binds to Fe(III) or Cu(I1). Without excluding the possibility that DF-Mn protects cells by facilitating the removal of 0;) the present results show that this SOD-mimic can confer protection from cytotoxic processes independent of Of or of OG-derived active species. (c7 1991 Academic Press, Inc.

Superoxide dismutase (SOD)-mediated2 protection of cells exposed to superoxide-derived reactive species and 0; itself has been extensively documented (l-4). Nev1 To whom correspondence should be addressed. ’ Abbreviations used: EPR, electron paramagnetic resonance; t-BHP, t-butyl hydroperoxide; DTPA, diethylenetriamine pentaacetic acid; DF, desferrioxamine (Desferal); X0, xanthine oxidase; HX, hypoxanthine; cyt-c’n, ferricytochrome c; DF-Mn, desferrioxamine-Mn(II1); SOD, superoxide dismutase; DMPO, 5,5-dimethyl-l-pyrroline-N-oxide; PBN, N-tert-butyl-oc-phenylnitrone. 0003.9861/91 $3.00 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

ertheless, difficulty in elevating intra- as well as extracellular SOD levels has prompted a search for nontoxic SOD-mimics which can facilitate 0; removal from the extra-, and most importantly, intracellular compartments. Since copper, iron, and manganese can all be cofactors in the active site of native SOD, attention has focused on low-molecular-weight chelates of these ions (5-9). To be an effective agent, the SOD-mimic should be nonimmunogenic, nontoxic, and stable to dissociation or cell metabolism. Certain copper chelates were found to react rapidly with 0; but proved to be biologically ineffective because of their high rate of reaction with molecular oxygen or dissociation of the chelate. In contrast, manganese chelates were found not to dissociate and to effectively remove 0: radicals (8-10). In particular, manganesebased chelates, such as desferrioxamine-Mn(II1) (DFMn), were found to react rapidly with 0; and to protect cells from various types of oxidative insults (11-13). Recent studies showed that DF-Mn can protect mammalian cells under experimental conditions in which 0, radical participation would be very unlikely (14). This anamolous finding suggested that DF-Mn might function by mechanisms other than as an SOD mimic. In a search for alternative modes of action of DF-Mn, V79 Chinese hamster cells were used as a test system and the injury induced by t-butyl hydroperoxide (t-BHP) under hypoxia was determined. The present results clearly show that DF-Mn can protect cells from Og-independent damage. The possible mechanisms responsible for cell protection are discussed. MATERIALS

AND

METHODS

Chemicals. Desferrioxamine (DF) was a gift from Ciba Geigy; hypoxanthine (HX) was purchased from Calbiochem; MnOz was purchased from Aldrich Chemical Co.; xanthine oxidase (EC 1.1.3.22, xanthine: oxygen oxidoreductase) (X0) grade III from buttermilk, superoxide dismutase, t-butyl hydroperoxide, and ferricytochrome c (cyt-c”‘) were obtained from Sigma; H,Oz was bought from Fisher. Xanthine oxidase was further purified on a G25 Sephadex column. All chemicals were prepared and used without further purification. The 1:l green complex 215

216

HAHN ET AL.

of DF-Mn(II1) (final concentration IO-35 mM) was prepared with minor modifications as previously reported (9, 10). Briefly, 10% excess Mn02 and DF were mixed in water and stirred overnight. The green DF-Mn solution was centrifuged and stored at 4OC. The complex was assayed spectrophotometrically using ?Onrn = 109 MS’ cm-’ and e320”m= 1570 Mm’ cm-‘. EPR spectroscopy was used to assay the Mn(I1) content in the DF-Mn preparation, whereas the total manganese content was determined after complete reduction by excess ascorbate. Distilled-deionized water was used throughout all experiments. Hydrogen peroxide assay. H202 was assayed using a YSI Model 21 industrial analyzer (Yellow Springs Instruments) equipped with a selective electrode for HzOz. Electron paramagnetic resonance. EPR spectra were recorded with a Varian E9 X-band spectrometer, with field set at 3357 G, modulation frequency of 100 kHz, nonbroadening modulation amplitude, and nonsaturating microwave power. For aerated experiments, samples (0.050.1 ml) of solutions were drawn by a syringe into a gas-permeable teflon capillary of 0.8 mm inner diameter and 0.05 mm wall thickness (Zeus Industrial Products, Inc., Raritan, NJ). Each capillary was folded twice, inserted into a narrow quartz tube which was open at both ends (25 mm i.d.), and then placed horizontally into the EPR cavity. For hypoxic experiments, the reagents were made hypoxic using argon gas in separate arms of a two-arm H-shaped sealed test-tube. The test tube was connected to a flat EPR cell. To detect short-lived transients, either 0.1 M DMPO or 75 mM PBN spin traps were included in the reaction mixture. The EPR spectrophotometer was interfaced to an IBM PC through an analog-to-digital converter and a data translation hardware (DT2801), and the spectra were digitalized by using commercial acquisition software, enabling subtraction of background signals. Cell survival analysis. Survival of Chinese hamster V79 cells in tissue culture was assessed by clonogenic assay (14). Cells were incubated in complete F-12 medium 16-24 h prior to experimental procedures and then exposed to t-BHP. Following treatment, cells were trypsinized, rinsed, counted, and plated in triplicate for macroscopic colony formation. Following appropriate plating periods, colonies were fixed, stained, and lastly counted with the aid of a dissecting microscope. For hypoxic experiments the cells were plated into specially designed, glass flasks sealed with soft rubber stoppers. Nineteen-gauge needles were pushed through to act as entrance and exit ports for humidified gas mixture of 95% N2/5% CO, (Matheson Gas Products). Each flask was also equipped with a ground glass side arm vessel which when rotated and inverted could deliver 0.2 ml of medium containing t-BHP. Stoppered flasks were connected in series and mounted on a reciprocating platform and gassed at 37°C for 45 min prior to and throughout the experiment. After 45 min the effluent gas phase was
V/u = 1 + hat X [DF-MnI/k,yt,+6,peroxide X [W-~“‘1. ill rate constant deterKnowing kc, c+superoxide, the catalytic mined for DF-Mn at pH 7 is 1.5 X lo7 M-l s-l. Hypoxic cytoprotective effect of DF-Mn. Cytotoxic effects of t-BHP on various cell systems have been previously reported (16-18) and were found to be inhibitable by the SOD-mimic DF-Mn (14). To investigate Og-independent damage, the effect of DF-Mn on cells treated with t-BHP was examined under hypoxic conditions. The hypoxic experiments eliminate the possibility of constituitive oxygen formation through H202 dismutation by cellular catalase. V79 cells were exposed to 3 mM t-BHP in the presence or absence of 0.5 mM DF-Mn. DF-Mn fully protected the cells from t-BHP cytotoxicity (Fig. 1A). To determine if DF alone provided protection, V79 cells were exposed to 3 mM t-BHP for 45 or 120 min (Fig. 1B). DF was added simultaneously with the addition of t-BHP. DF provided modest protection at 120 min but no protection at 45 min. Unlike DF-Mn, DF requires an incubation time of at least 2 h for protection to be observed. DF-Mn reaction with DNA-Fe(H). The protection afforded by DF both under aerobic and hypoxic conditions indicates that redox-active metals are critically important in cellular oxidative damage. To study if DF-Mn protection results from interfering with transition metal-mediated damage, the SOD-mimic was hypoxically mixed with iron in 50 mM Hepes buffer, pH 7.2, at 22”C, and the reaction mixture was followed spectrophotometrically. The reaction was studied both in the absence and in the presence of 0.1 mg/ml salmon DNA. The respective absorption spectra of 0.5 mM Fe(I1) (with DNA) and 0.5

cyt-c”’ reFerricytochrome c reduction assay. The SOD-inhibitable duction assay was used to determine rate constants of reaction with 0, (15). Superoxide radicals were generated at 25 + 0.2”C in aerated phosphate buffer containing 5 mM HX and lo-80 @M cyt-c”‘. To protect xanthine oxidase and avoid dissociating the SOD-mimic, 50 PM DF rather than DTPA or EDTA was included in the reaction system. The reaction was started by addition of 0.01 U/ml X0 and the rate of cyt-c”’ reduction was spectrophotometrically fallowed at 550 nm.

RESULTS

Superoxide reaction with DF-Mn. The reaction rate of DF-Mn with 0; was determined using the SOD-inhibitable cyt-c’r’ reduction assay. Formation of cyt-c” was monitored in the absence (V) and in the presence (u) of varying DF-Mn concentrations. Data were analyzed by plotting V/u as a function of [DF-Mn], and k,,, was calculated according to

control

45 min

control

120 min

Mn-DF 120 min

FIG. IA. Protection from t-BHP-induced cytotoxicity by DF-Mn. Monolayered Chinese hamster V79 cells in full growth medium at 37°C were exposed under 95% N2/5% CO, to 3 mM t-BHP for various periods of time in the presence of control, no additives, and 0.5 mM DF-Mn.

Mn-DESFERRIOXAMINE-MODES

217

OF ACTION

O.D.

45 min

45

min

120

min

120

min

FIG. 1B. Protection from t-BHP-induced cytotoxicity by desferrioxamine. Monolayered Chinese hamster V79 cells infull growth medium at 37°C were exposed under 95% N2/5% CO2 to 3 mM t-BHP for various periods of time in the presence of control 45 min, 0.5 mM desferrioxamine, control 120 min, and 0.5 mM desferrioxamine.

DF-Mn hypoxic solutions monitored before and right after mixing are presented in Fig. 2. Following mixing, the optical absorption bands at 320 and 640 nm, characteristic of the DF-Mn green complex, disappeared, whereas the absorbance at 440 nm corresponding to DFFe(II1) appeared. A similar result was obtained upon mixing the two reagents in the absence of DNA. To elucidate the fate of the manganese, the EPR spectra were recorded for 0.5 mM DF-Mn before and after mixing with 0.5 mM DNA-Fe(I1) under hypoxic conditions. Typical spectra as illustrated in Fig. 3 indicate the reduction of EPRsilent Mn(II1) and the formation of Mn(I1). The disappearance of DF-Mn absorbance and the appearance of DNA-Fe(III)-DF and the evolution of the Mn(II) signal took place instantaneously, being too rapid for kinetic study without using a rapid mixing technique. DF-Mn reaction with DNA-C&II). Copper( unlike Fe(II), is not anticipated to reduce the bound Mn(III), yet it might displace manganese from its DF chelate. To examine such a possibility, 0.5 mM DF-Mn was hypoxitally mixed with 5 mM Cu(I1) in the presence of 0.1 mg/ ml salmon DNA. The change in OD320nm and OD640nm and the appearance of the Mn(I1) EPR signal were followed. Mixing DNA-Cu(I1) with DF-Mn resulted in the disappearance of the typical absorption spectrum of the green complex accompanied by the generation of Mn(I1). However, the level of Mn(I1) produced by Cu(I1) was only 60-70% of that formed by Fe(I1) (compare Fig. 3 traces c and d).

580

480

380

280

Wavelength, FIG. 2. Optical spectra changes Fe(I1). Solutions of 0.5 mM DF-Mn mM F(B) + 0.1 mg/ml salmon DNA (a) DF-Mn; (b) DNA-Fe(I1); (c) 1:l

680

nm

upon mixing DF-Mn with DNAin 10 mM Hepes, pH 7.2, and 0.5 in water were mixed hypoxically. mixture of DF-Mn + DNA-Fe(H).

mM

DISCUSSION Previous observations indicate that cytoprotection provided by DF generally requires l-2 h preincubation

(19). This cytoprotection agrees with a previous conclusion that intracellular “redox-active” metals, and particularly iron, are instrumental in HzOz and t-BHP-mediated cytotoxicities (16-18). Since DF slowly enters the cell, previous failures of DF to provide protection more than likely resulted from too short preincubation periods. Our data confirm these observations and suggest that in order to observe modest protection, a 2-h incubation time is

a

DF-Mn

DF-Mn

+ ascorbate

+ DNA-Fe”

-1 2840

3340

3090 FIELD

35’41

3840

(GAUSS)

FIG. 3. EPR spectral changes upon mixing DF-Mn with DNA-Fe(B). Solutions of 0.5 mM DF-Mn in 10 mM Hepes, pH 7.2, and 0.5 mM Fe(B) + 0.1 mg/ml salmon DNA in water were mixed hypoxically. (a) DFMn; (b) DF-Mn + ascorbate; (c) DF-Mn + DNA-Fe(B); (d) DF-Mn + DNA-Cu(I1).

218

HAHN

necessary. Likewise, native SOD does not protect because it cannot readily enter into the cell. However, the chelate DF-Mn exerted its protective effect immediately after being added to the cell suspension (Fig. IA) as was observed previously for aerated test systems (14). This observation suggests that DF-Mn can rapidly permeate into the cell, allowing free DF to accumulate more quickly in the intracellular space. In addition, neither native SOD nor catalase show hypoxic cytoprotection as expected in the absence of 0; and Hz02 (14). Mn(I1) or Mn(IV) themselves have been reported not to protect, suggesting that cytoprotection is due to the complex itself rather than to its dissociation products (11, 20). Mechanisms of damage. Oxidative biological damage mediated by transition metals and HzOz is commonly attributed to species produced during the redox cycling of the metals. The critical step in this cycle is the reaction of peroxide with the reduced metal, yielding a peroxo complex,

ET AL.

cytotoxicity since cell killing induced by t-BHP during hypoxia does not involve 0; nor H,O, and is not inhibitable by SOD or catalase. The catalase-like activity of manganese chelates was found to require bicarbonate and Hepes-free solution (28). Using previous experimental conditions no catalase-like activity was anticipated for DF-Mn and none was found (14). The experiments done in the present study would also preempt catalase-like activity. Binding of intracellular metals. Hypoxic protection can result from chelation, by DF originating from DFMn, of critically located labile intracellular metals, such as iron. Other metal-based SOD-mimics have been shown to dissociate in the presence of DTPA, EDTA, or proteins (4, 5) and are prone to dissociate in a biological environment. The DF-Mn preparation contains about 8-20% Mn(II), as judged by EPR (lo), and the preparation does contain varying amounts of free DF. It is, therefore, possible that the uncoordinated fraction of the chelate would bind “labile” intracellular iron rendering it “redox inactive.” Moreover, the results presented in trace d of Fig. 3 L - M(“-I)+ + H202 + L - Mb-l)+ _ H 0 2 27 [21 indicate that intracellular copper ions could readily displace Mn(II1) from its DF chelate. Consequently, chelated that can generate a higher oxidation state of the metal copper will be redox-inactive and less cytotoxic. In the and/or ‘OH radical (21-24), case of copper reaction with DF-Mn, a lower EPR signal was observed since Mn(II1) dismutated to yield Mn(II) and the EPR-silent insoluble Mn(IV) as previously disL - M(“-‘I+ - HzOz + L - M”+ + OH- + ‘OH [3a] cussed (10). L - Mb-l)+ - H202 + L _ Mb+l)+ + 20H-. Reoxidation of reduced metals. It is plausible that DFPbl bound Mn(ITI) reoxidizes intracellular reduced iron and copper. The commonly proposed mechanism accounting Therefore, any reagent which can bind the metal to render for the metal-catalyzed HzO,-induced cell damage involves it redox inactive or destroy H202 may prevent Reaction the redox reactions of unbound or chelated metal ions. In [2] and afford protection. particular, redox reactions of metals coordinated to bioIt has been suggested that various forms of oxidative logical macromolecules can act as catalysts for the Haberstress (including t-BHP) result in alterations of calcium Weiss reaction and lead to site-specific damage (19). Achomeostasis (25-27). This may be an important compocording to that mechanism, metal binding to critical celnent of damage from t-BHP. Protection by DF-Mn lular components is essential for site-specific damage inthrough binding of intracellular calcium cannot be conflicted by deleterious species generated in the vicinity of firmed or excluded by this study. the critical targets. Therefore, reoxidation of reduced Superoxide dismutase and catalase-like activity. DFmetal such as ferrous, Mn has been previously found to afford protection from oxidative damage induced under air (11). Such protection DF-Mn(II1) + Fe(B) + DF-Fe(II1) + Mn(II), [4] can result from the removal of 0: radicals from intracellular compartments where levels of endogenous SOD would be expected to block Reaction [3] and diminish the are not sufficiently high. Considering the high k,,, of 0; metal-mediated damage. It is the “redox-active” traces of dismutation, a few micromolar DF-Mn would be sufficient for protection. In several test systems, however, up to 1 iron or copper, rather than the total cellular metal, which are instrumental in cell injury. Yet, the extent and ability mM DF-Mn was required to achieve significant protection of intracellular “redox-active” (sometimes called “labile”) (12,13). This observation, together with the hypoxic protective effect of DF-Mn, suggests alternative modes of transition metals to participate in redox reaction are still ill-defined. Consequently, it still is impossible to identify protection. Hypoxic protection from H202 that is afforded by DF-Mn may result from the unlikely but feasible case which cellular metal is oxidized by DF-Mn. However, the rapid oxidation of DNA-bound iron(I1) by DF-Mn supof catalase-facilitated intracellular dismutation of H,02 to generate sufficient oxygen to produce 0; (14). Such an ports such a mechanism (Figs. 2 and 3). Accordingly, Mn(II1) would compete with t-BHP or H202 for the inassumption, however, is untenable in the case of t-BHP

Mn-DESFERRIOXAMINE-MODES

219

OF ACTION

7. Lengfelder, E., Fuchs, C., Younes, M., and Weser, W. (1979) Biochim.

DF-Mn SOD

Biophys.

Hz&

(3

Hz02

CAT L.&-l)+

f

DF-Mn

Acta

567,

492-502.

8. Koppenol, W. H., Levine, F., Hatmaker, T. L., Epp, J., and Rush, J. D. (1986) Arch. Biochem. Biophys. 251, 594-599.

\

9. Darr, D., Zarilla, K. A., and Fridovich, I. (1987) Arch. Biochem. Biophys. 256, 351-355. Biophys. 10. Beyer, W. F. Jr., and Fridovich, I. (1989) Arch. Biochem. 271, 1499156. 11. Darr, D. J., Yanni, S., and Pinnell, S. R. (1988) Free Radical Biol. Med.

4,357-363.

12. Rabinowitch, Radical

Biol.

H. D., Rosen, G. M., and Fridovich, Med.

I. (1989) Free

6, 45-48.

13. Rabinowitch, Biological Damage\

/

H. D., Privalle, C. T., and Fridovich, I. (1987) Free 3, 125-131. 14. Samuni, A., Mitchell, J. B., Samuni, U., DeGraff, W., Krishna, C. M., and Russo, A. (1991), Free Radical Res. Commun., in press. Radical

Biol.

Med.

15. Fridovich, I. (1985) in Handbook of Methods for Oxygen Radical Research (Greenwald, R. A., Ed.), pp. 213-215 CRC Press, Boca Raton, FL. 16. Masaki, N., Kyle, M. E., and Farber, J. L. (1989) Arch. Biochem.

9

SCHEME 1

Biophys.

269,

390-399.

17. Ochi, T., and Miyaura,

tracellular reduced metals as illustrated for Hz02 in Scheme 1. In conclusion, the present findings show that although Mn-based SOD mimic can react with and remove 0; in uitro, its cytoprotective effect can also result from protection against OG-independent damage. ACKNOWLEDGMENTS We thank Drs. I. Fridovich and W. H. Koppenol discussion and insightful comments.

for their helpful

REFERENCES 1. 2. 3. 4.

Fridovich, I. (1986) Arch. Biochem. Biophys. 247, l-11. Weiss, S. J. (1986) Acta Physiol. Stand. (Suppl.) 548, 9-37. McCord, J. M. (1986) Free Radicals Biol. Med. 2, 307-310. Clark, I. A., Chaudhri, G., and Cowden, W. B. (1988) in Oxy-radicals in Molecular Biology and Pathology (Cerutti, P. A., Fridovich, I., McCord, J. M., Eds.), pp. 53-60, A. R. Liss, New York. 5. Brigelius, R., Spottl, R., Bors, W., Lengfelder, E., Saran, M., and Weser, U. (1974) FEBS Lett. 47, 72-75. 6. Weinstein, J., and Bielski, B. H. J. (1980) J. Amer. Chem. Sot. 102, 4916-4919.

S. (1989) Toxicology 55, 69-82.

18. Taher, M. M., and Lakshmaiah, N. (1987) Arch. Biochem. Biophys. 257, 100-106. 19. Samuni, A., Chevion, M., and Czapski, G. (1981) J. Biol. Chem. 256, 12,632-12,635. 20. Archibald, F. S., and Fridovich, I. (1982) Arch. Biochem. Biophys.

214,452-463. 21. Goldstein, S., and Czapski, G. (1986) J. Amer. Chem. Sot. 108, 2244-2250. 22. Rush, J. D., and Koppenol, W. H. (1986) J Biol. Chem. 261,67306733. 23. Thederahn, T. B., Kuwabara, M. D., Larsen, T. A., and Sigman, D. S. (1989) J. Amer. Chem. Sot. 111, 4941-4946. 24. Masarwa, M., Cohen, H., Meyerstein, D., Hickman, D. L., Bakac, A., and Espenson, J. H. (1988) J. Amer. Chem. Sot. 110, 42934297. 25. Jones, D. P., Thor, H., Smith, M. T., Jewell, S. A., and Orrenius, S. (1983) J. Biol. Chem. 258,6390-6393. 26. Elliott, S. J., and Schilling, W. P. (1990) J. Biol. Chem. 265, 103107. 21. Sakaida, I., Thomas, A. P., and Farber, J. L. (1991) J. Biol. Chem. 266, 717-722. 28. Stadtman, E. R., Berlett, B. S., and Chock, P. B. (1990) Proc. Natl. Acad.

Sci. USA,

87,

384-388.