Catalysis of the Haber-Weiss Reaction by Iron-Diethylenetriaminepentaacetate Timothy J. Egan, Steven R. Barthakur, and Philip Aisen Department of Physiology and Biophysics, Albert Einstein College of Medicine, Bronx, New York

ABSTRACT To help settle controversy as to whether the chelating agent diethylenetriaminepentaacetate (DTPA) supports or prevents hydroxyl radical production by superoxide/hydrogen peroxide systems, we have reinvestigated the question by spectroscopic, kinetic, and thermodynamic analyses. Potassium superoxide in DMSO was found to reduce Fe(III)DTPA. The rate constant for autoxidation of Fe(II)DTPA was found (by electron paramagnetic resonance spectroscopy) to be 3.10 M-i s-‘, which leads to a predicted rate constant for reduction of Fe(III)DTPA by superoxide of 5.9 X lo3 M- ’ s- ’ in aqueous solution. This reduction is a necessary requirement for catalytic production of hydroxyl radicals via the Fenton reaction and is confirmed by spin-trapping experiments using DMPO. In the presence of Fe(III)DTPA, the xanthine/xanthine oxidase system generates hydroxyl radicals. The reaction is inhibited by both superoxide dismutase and catalase (indicating that both superoxide and hydrogen peroxide are required for generation of HO.). The generation of hydroxyl radicals (rather than oxidation side-products of DMPO and DMPO adducts) is attested to by the trapping of cr-hydroxethyl radicals in the presence of 9% ethanol. Generation of HO. upon reaction of H,O, with Fe(II)DTPA (the Fenton reaction) can be inhibited by catalase, but not superoxide dismutase. The data strongly indicate that iron-DTPA can catalyze the Haber-Weiss reaction.

INTRODUCTION Studies on the iron complexes of the chelating agent DTPA have provided conflicting evidence about their pro-oxidant activities. Some [l-7] indicate that this complex can lead to the formation of powerful oxidizing agents (hydroxyl radicals or a highly oxidized form of iron such as the ferry1 ion), while others suggest that it prevents or suppresses formation of such species [S-11]. Since DTPA has often been used in experimental studies for its supposed suppression

Address reprint requests and correspondence to: Dr. Philip Aisen, Department Biophysics, Albert Einstein College of Medicine, Bronx, New York 10461. Journal of Inorganic Biochemistry, 48,241-249 (1992) 0 1992 Elsevier Science Publishing Co., Inc., 655 Avenue

of Physiology

and 241

of the Americas,

NY, NY 10010

0162-0134/92/$5.00

242

T. J. Egan et al,

of hydroxyl radical formation and is widely used in medicine as a chelator for Gd(III) 1121, a contrast enhancer in magnetic resonance imaging, we have undertaken to resolve this disagreement. Two recent studies by Yamazaki and Piette [13, 141 have convincingly demonstrated that Fe(II)DTPA produces hydroxyl radicals (HO.) when reacted with H,O, (and not iron-oxo species such as the ferry1 ion). However, this is a stoichiometric reaction between ferrous ion and hydrogen peroxide, and does not establish that iron-DTPA would necessarily cata@ HO. production by cycling of complexed iron between the ferric and ferrous states. Even Fe(I1) desferrioxamine which cannot catalyze HO. production because of its low reduction potential (formal reduction potential, -0.454 V [1.51) and inability to bind Fe(I1) [16] is able to produce this radical in a stoichiometric reaction with H,O, [17]. The aim of the present study has been to determine whether Fe(III)DTPA can catalyze generation of HO. in O;./HzO, systems.

MATERIALS

AND METHODS

DTPA was obtained from Sigma and used to prepare Fe(III)DTPA in two ways. In the first, the acid form of DTPA was dissolved in water and 0.5 equivalents of FeCl, .6H,O (Fisher Scientific) were added to the solution. The pH was then slowly raised to 7.4 with dilute NaOH to give a 10 mM Fe(III)DTPA stock solution. In the second, DTPA (20 mM stock, pH 7.4) and FeCI, (10 mM stock in 0.1 M HCl) were added to the reaction mixture under study (so that the complex was formed in situ). Both methods gave identical results. For spin-trapping experiments Fe(II)DTPA.was prepared in situ in the same way except that Fe(NH,),(SO,), .6H,O (Aldrich) in 0.1 M HCl was used. In the studies on Fe(II)DTPA autoxidation the complex was prepared by adding Fe(NH,),(SO,),*6H,O in 0.1 M HCl to the acid form of DTPA together with the acid form of MOPS (3-[N-morpholinolpropane-sulphonic acid) (Sigma) under an atmosphere of nitrogen. The pH was slowly raised to 7.4 under the N2 atmosphere and a sample removed to record the EPR spectrum (t = 0). The solution was then poured into an open beaker and stirred vigorously. Aliquots were removed at fixed intervals in order to obtain EPR spectra. Oxidation of Fe(II)DTPA by H,O, (Baker) was observed in essentially the same way except that the H,O, was added under N, and the solution was not exposed to air. Reduction of Fe(III)DTPA by 0;. was also studied in this manner, but the stock Fe(III)DTPA solution (0.025 ml) was added to 10 mM KO, (Aldrich) dissolved in 0.600 ml pure dimethyl sulfoxide (DMSO) (Baker). DMPO (5,5-dimethyl-1-pyrroline N-oxide) for spin-trapping was obtained from Sigma and was purified as described elsewhere [lg] and stored in small aliquots in the dark at 203 K. Solutions were prepared for each experiment by mixing stock solutions of the appropriate reagents (H,O,, xanthine (Calbiochem), xanthine oxidase (Calbiochem), superoxide dismutase (Sigma), catalase (Sigma), MOPS buffer adjusted to pH 714, DTPA stock solution, Fe(II1) stock solution (or Fe(III)DTPA stock solution), or Fe(B) stock solution). In all experiments MOPS buffer was used only after passage down a Chelex100 column to remove extraneous metal ions. Electron paramagnetic resonance (EPR) spectra of Fe(II1) complexes were obtained by mixing all appropriate solutions with an equal volume of 5.78 M

IRON-DTPA

CATALYSIS OF HABER-WEISS

REACTION

243

NaClO, to forestall aggregation of the iron upon freezing (which would weaken the signal due to spin-spin interactions). These solutions were then transferred to precision quartz EPR sample tubes (Wilmad), frozen in liquid nitrogen, and spectra recorded at 77 K. Spin-trapping reactions were performed at room temperature (296 K) with solutions containing 0.13 M DMPO using precision quartz capillaries (1.2 mm i.d., Wilmad) as sample tubes. All EPR spectra were obtained with a Bruker ER 200 D-SRC spectrometer. Instrumental conditions are given in the relevant figure captions. RESULTS AND DISCUSSION In order for an iron complex (Fe-L) to be able to produce hydroxyl radicals it must be able to undergo the Fenton reaction (Eq. (1)): Fe(II)-L

+ H,O, -+ Fe(III)-L

+ OH-+

HO*.

(1)

As mentioned, this reaction has been convincingly demonstrated in the case of Fe(II)DTPA [13,14]. Using EPR spectroscopy we confirmed that the Fe(II)DTPA (0.132 mM) is completely oxidized to Fe(III)DTPA by 6.5 mM H,O, within the time taken to freeze the sample in liquid nitrogen (2.5 min). However, in order for the iron complex to be able to catalyze the production of HO. radicals the resulting Fe(III)DTPA must be able to be reduced back to the Fe(I1) form. In the metal catalyzed Haber-Weiss reaction this is accomplished by superoxide (Eq. (2)): Fe(III)-L

+ 0;.

-+ Fe(I1) - L + 0,.

(2)

It has been reported [193 that hydrogen peroxide itself can produce HO. radicals by reaction with Fe(II1) complexes and it was originally believed that this was due to the ability of H,O, to reduce Fe(III)-L (Eq. (3)): Fe(III)-L

+ H,O, --f Fe( 11)-L + H+ + HO, * .

(3)

However, in two recent studies [20, 211 Gutteridge et al. have shown that HO. production is blocked by superoxide dismutase and that is 0;. which acts as the reducing agent. The superoxide is apparently generated as a product of the reaction between hydrogen peroxide and Fe(III)-L without reduction of the iron by the H,O,. These studies confirm that reduction of iron by superoxide is the key step in the catalytic production of HO.. Reduction of Fe(II1) complexes by superoxide in aqueous solution is complicated by the very rapid dismutation of 0, . , but this problem can be avoided by performing the reaction in DMSO. Figure 1 shows that 10 mM KO, in pure DMSO can reduce 0.132 mM Fe(III)DTPA (0.025 ml aqueous solution added to 0.600 ml DMSO solution). However, this does not prove that the reduction can occur in aqueous solution. Furthermore, the addition of the aqueous solution (of Fe(III)DTPA) causes dismutation of the superoxide, resulting in OH- production, so the effective pH in the reaction mixture is probably very high. As a result of this high pH the ferric species being reduced is probably different from that at neutral pH. After mixing with aqueous NaClO, prior to freezing, the

244

T. J. Egan et al.

FIGURE 1. EPR spectra of Fe(III)DTPA (0.132 mM) 50% in DMSO and 2.89 M in NaCIO,; (b) with and (a) without 5 mM KO,. 9.6 mM NaOH was added to (a) for comparison with (b) since the addition of aqueous 5.78 M NaCIO, to (b) would result in this concentration of OHS- through dismutation of the 0; . . EPR conditions: T, 77 K; modulation frequency, 100 kHz; modulation amplitude, IO G; time constant, 2.00 s; microwave frequency, 9.30 GHz; power, 20.0 mW; gain, 5 X 104.

a

i/f-

concentration of OHis 9.6 mM (assuming complete dismutation of the O;.). It is possible that the reduction only occurs under these extremely basic conditions, but development of a dark purple color immediately upon addition of the Fe(III)DTPA to the KO, solution suggests that reduction occurs prior to mixing with the aqueous NaClO, [22]. The equilibrium constant for Fe(III)DTPA reduction by Oi . can be obtained from a knowledge of the reduction potentials of superoxide and Fe(III)DTPA. Since reduction by superoxide is the reverse of one-electron oxidation by Oz (Eq. (4)) the rate constant for reduction can be estimated from Eq. (5) (as has been pointed out by Wood [23]): calculated

Fe( 11)-L + 0,

2 Fe(III)-L

K = k,,‘k,,

+ 0;

.,

(4) (9

where K is the equilibrium constant and k, and k, are the forward and reverse rate constants, respectively. We have been unable to find a measured value for k, (for Fe(III)DTPA) in a search of the literature, but it has been estimated [24] to be less than 104 M- ’ SC’ (at pH 7.0). Determination of the rate constant for Fe(II)DTPA autoxidation, combined with knowledge of the value of K should allow calculation of k,. The autoxidation of Fe(II)DTPA was studied by following the growth of the EPR signal amplitude at g’ = 4.27 (no change in signal linewidth was observed; Fig. 2). A semilog plot of the g’ = 4.27 signal amplitude (inset, Fig. 2) gives a pseudo first-order rate constant of 0.047 min..’ which leads to a second-order rate constant of 3.10 Mm-’ s ’ (given an aqueous O2 concentration of 0.254 mM under one atmosphere of air [23]). The results are in excellent agreement with a report in which Fe oxidation was followed using acidic thiocyanate [25]. Vandegaer et al. [26] have reported a standard oxidation potential for Fe(lI)DTPA in the pH range 6-8 of -0.034 V. Using Eqs. 6 and 5 (and the standard reduction potential of +O.O34V) this leads to a value for k, of 5.9x 10’ M-’ s-l, which is in excellent agreement with the experimental

IRON-DTPA

CATALYSIS OF HABER-WEISS

REACTION

245

Time / minute

L-_-

1200

I

I

1400 Magnetic

evidence

1600 Field

/

Gauss

above [24]. For comparison the rate reduction at pH 7.0 is 1.9 x lo6 M-’ s-l [24].

mentioned

Fe(III)EDTA

_I

1600

FIGURE 2. Autoxidation of 0.132 mM Fe(II)DTPA monitored by EPR. Aliquots of reaction mixture were removed at time intervals, diluted by 50% with 5.78 M NaClO,, transferred to EPR sample tubes, and frozen in liquid nitrogen. Instrumental conditions are as for Figure 1, except the gain was 5 X 105; times at which aliquots were taken, (a) 0, (b) 7.5, (c) 1.5,(d) 22.5, (e) 37.5, (f) 45 and (g) 90 min. Inset: a semilog plot of the amplitude data together with data from two similar runs.

K= exp((F/nRT)[E”(O,(aq)/O;

constant

0) - E”(Fe(III)DTPA/Fe(II)DTPA)]}

for

(6)

where F is Faraday’s constant, n is the number of electrons involved in the redox reaction, R is the gas constant, T is the absolute temperature, and the E”s are the respective standard reduction potentials; E”(O,(aq)/O; *>= -0.160 V [23]. The reduction potential for the Fe(III)DTPA2-/Fe(II)DTPA3couple is constant between pH 6 and 8 owing to the predominance of the species Fe(II1) DTPA2- and Fe(II)DTPA3- in this pH range so that no protons are transferred during the redox reaction [26]. Using the stability constants for Fe(III)DTPA and Fe(II)DTPA reported by Martell and Smith [27] and the speciation program provided in Martell and Motekaitis [28] this is clearly illustrated in Figures 3(a) and (b). These arguments suggest, but do not prove, that 0; . can reduce Fe(III)DTPA. In order to further investigate this possibility, spin-trapping experiments were performed using the spin-trap DMPO. Figure 4 confirms that Fe(II)DTPA can react with H,O, in a Fenton reaction, producing HO. radicals. The DMPO - OH. signal is fully formed within the time that it takes to mix the reagents and record the spectrum. Addition of superoxide dismutase has no effect on the production of HO * radicals as 0; * is not involved in this stoichiometric Fenton reaction (the superoxide might be expected to recycle some of the Fe(III)DTPA formed in the Fenton reaction back to Fe(I1) which could undergo further Fenton chemistry, but this is apparently negligible under the conditions of this experiment). As expected,

246

T. J. Egan et al.

2

4

10

6 PH (A)

FIGURE 3. (A) Speciation plot for Fe(III)DTPA, (a) FeDTPAH _, (b) FeDTPA’ -, and (c) Fe(OH)DTPA’-. (B) Speciation plot for Fe(II)DTPA, (a) 2

4

6 PH 03)

6

10

Fe”, (b) FeDTPAH’-, Cd) Fe(OH)DTPA”-. Fe(OH), DTPA’ ~.

(cl FeDTPA;’ -, and (e)

when added one minute prior to the Fe(II)DTPA, catalase prevents formation of HO.. This is due to the disproportionation of H,O, to 0, and H,O before reaction with the Fe(II)DTPA. When the catalase and Fe(II)DTPA were added simultaneously the DMPO-OH. signal was weakened but not completely destroyed. When superoxide and hydrogen peroxide were generated by the xanthine (0.221 mM)-xanthine oxidase (0.06 units/ml) system, it was found that a DMPOOH. signal formed in the presence of Fe(III)DTPA (see Fig. 5). This signal strengthens over a period of at least 90 minutes (spectra in Fig. 5 were recorded 24 min after mixing). This continuing development of the spectrum suggests ongoing catalysis of HO. production. A DMPO-OH. signal can be produced without formation of HO. [29] via the formation of spin-trapped superoxide which converts to DMPO-OH.. This does not appear to be the case in our experiments, since catalase completely inhibits formation of the DMPO-OH. signal, thus implicating H202 in the reaction (see Fig. 5). Further evidence that

IRON-DTPA

CATALYSIS OF HABER-WEISS

Jc

d

I

3450

I

3475 Magnetic

3500 Field

/

I

3525

Gauss

REACTION

247

FIGURE 4. Hydroxyl radical trapping in the Fenton reaction using DMPO; (a) blank (H,O, + DMPO, but no Fe); (b) H,O, + FeCIIlDTPA, (cl H,O, + Fe(II)DTPA + superoxide dismutase; (d) H,O, + Fe(II)DTPA + catalase. EPR conditions: T, 296 K; modulation frequency, 100 kHz; modulation amplitude, 1.25 G; time constant, 2.00 s; microwave frequency, 9.770 GHz; power, 2.5 mW; gain, 5 x 105.

HO. radicals are produced is provided by the spectrum obtained in the presence of 9% ethanol (see Fig. 5) showing clear evidence for an cr-hydroxyethyl radical [30] which is produced by reaction between HO. and ethanol. Since HO. formation requires formation of Fe(I1) (as discussed above) and the iron-DTPA was added as the ferric complex, superoxide must be able to reduce the Fe(III)DTPA in this.system. This is supported by the observation that superoxide dismutase greatly decreases the DMPO-OH. signal (see Fig. 5). A recent report 1311 has shown that DMPO can inhibit dismutation of 0; a, thus preventing HO + production. Addition of superoxide dismutase then increases the amount of hydroxyl radical trapped. This effect was not observed in our studies, probably because of the higher concentration of xanthine oxidase which results in more rapid production of superoxide (which was reported to counteract the effect of DMPO on dismutation [31]). It should be noted that absolute rate constants for HO. production cannot be obtained by following the growth of the DMPO-OH* signal because the trapping of HO. is not stoichiometric and the observed rate of DMPO-OH *

3450

3475 Magnetic

3500 Field

/

Gauss

3525

FIGURE 5. Hydroxyl radical trapping with DMPO using xanthine/xanthine oxidase to generate superoxide and hydrogen peroxide: (a) blank (no Fe(III)DTPA); (b) Fe(IIIlDTPA added to the system; (cl Fe(III)DTPA + superoxide dismutase; (d) Fe(IIIlDTPA + catalase; (e) Fe(IIIlDTPA + ethanol. All spectra were recorded 24 min after mixing. Instrumental conditions are as for Figure 4.

248

T. J. Egan et al.

formation is the net result of the rates of HO - formation, HO * decay by routes other than spin trapping and decay of the DMPO-OH. adduct itself (which is dependent on the concentration and nature of the various radical species present [291X As a result only comparative data on radical production can be obtained. Evidence presented in this work indicates that iron-DTPA complexes support the catalytic formation of hydroxyl radicals by the Haber-Weiss reaction. Our results do not permit a quantitative comparison with other chelating agents but other studies do suggest that DTPA is less active than EDTA. This is consistent with the significantly slower reduction of Fe(III)DTPA by superoxide compared with Fe(III)EDTA. Owing to the larger magnitude of the superoxide dismutation rate relative to the Fe(III)DTPA reduction rate a smaller fraction of the iron would be expected to be reduced. Furthermore, the Fenton reaction of Fe(II)DTPA also appears to be slower than for EDTA 1131. This seemingly lower lability of iron-DTPA may well be the result of lower accessibility of the iron to the bulk solvent compared with iron-EDTA which has been demonstrated by NMR solvent relaxatipn experiments [91. This work was suppolled in part by Grant No. DK 37927 from the National Institutes Health, U.S. Public Health Service.

of

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

J.-C. Sibille, K. Doi, and P. Aisen, J. Biol. Chem. 262, 59 (1987). J. Diguiseppi and I. Fridovich, Arch. Biochcm. Biophys. 205, 323 (1980). G. Cohen and P. M. Sinet, FEBS Lett. 138, 258 (1982). C. C. Winterbourn and H. C. Sutton, Arch. Biochem. Biophys. 235, 116 (1984). P. Gee and A. J. Davison, Biochim. Biophys. Acta 838, 183 (1985). S. Puntarulo and A. I. Cederbaum, Arch. Biochem. Biophys. 258,510 (1987). R. A. Floyd, Can. J. Chem. 60, 1577 (1982). S. E. Fridovich and N. A. Porter, J. Biol. Chem. 256, 260 (1981). E. Graf, J. R. Mahoney, R. G. Bryant, and J. W. Eaton, J. Biol. Chem. 259, 3620 (1984). M. S. Baker and J. M. Gebicki, Arch. Biochem. Biophys. 234, 258 (1984). C. C. Winterbourn and H. C. Sutton, Arch. Biochem. Biophys. 244, 27 (1986). G. L. Wolf, Radiology 172, 709 (1989). I. Yamazaki and L. H. Piette, J. Biol. Chem. 265, 13589 (1990). I. Yamazaki and L. H. Piette, J. Am. Chem. Sot. 113, 7588 (1991). S. R. Cooper, J. V. McArdle, and K. N. Raymond, Proc. Natl. Acad. Sci. USA 75,355l (1978). J. L. Bock and G. Lang, B&him. Biophys. Acta 264, 245 (1972). S. J. Klebanoff, A. M. Waltersdorph, B. R. Michel, and H. Rosen, J. Biol. Chem. 264, 19765 (1989). G. R. Beuttner and L. W. Oberley, Biochem. Biophys. Res. Commun. 83, 69 (1978). N. Uri, Chem. Reu. 50, 375 (1952). J. M. C. Gutteridge, Free Rad. Res. Comms. 9, 119 (1990). J. M. C. Gutteridge, L. Maid& and L. Poyer, Biochem. J. 269, 169 (1990). R. L. Willson, in Iron Metabolism, Ciba Foundation Symposium 51, Elsevier, Amsterdam, 1977, p. 331.

IRON-DTPA

CATALYSIS

OF HABER-WEISS

REACTION

249

23. P. M. Wood, TIBS 12, 250 (1987). 24. G. R. Buettner, T. P. Doherty, and L. K. Patterson, FEBS Lett. 158, 143 (1983). 25. G. Cohen and P. M. Sinet, Deu. Biochem. llA, 27 (1980). 26. J. Vandegaer, S. Chaberek, and A. E. Frost, J. Inorg Nucl. Chem. 11, 210 (1959). 27. A. E. Martell and R. M. Smith, Critical Stability Constants, Vol. 1: Amino Acids, Plenum Press, New York, 1974. 28. A. E. Martell and R. J. Motekaitis, Determination and Use of Stability Constants, VCH Publishers, New York, 1988. 29. A. Samuni, C. M. Krishna, P. Riesz, E. Finkelstein, and A. Russo, Free Rad. Biol. Med. 6, 141 (1989). 30. E. Finkelstein, G. M. Rosen, and E. J. Rauckman, Arch. Biochem. Biophys. 200, 1 (1980). 31. B. E. Britigan, T. L. Roeder, and G. R. Buettner, Biochem. Biophys. Acta 1075, 213 (1991). Received April 21, 1992; accepted May 6, 1992