- Email: [email protected]
Antioxidant effect of manganese
ARCHIVES
OF BIOCHEMISTRY
Vol. 299, No. 2, December,
Antioxidant Mariagrazia
AND
BIOPHYSICS
pp. 330-333,
1992
Effect of Manganese
Coassin,* Fulvio Ursini,?’
and Albert0 Bindolit
*Department of Biological Chemistry, University of Padua; TDepartment of Chemical Science and Technology, University of Udine; and jCNR-Center for Mitochondrial Physiology, Padua, Italy
Received
May
14, 1992, and in revised
form,
August
10, 1992
The antioxidant effects of manganese and other transition metals were studied as the inhibition of microsomal lipid peroxidation and crocin bleaching by peroxyl radicals. The peroxyl radical scavenging capacity was measured by competition kinetics analysis. While Zn(II), Ni(II), and Fe(I1) were almost completely ineffective, Mn(I1) and Co(I1) showed a free radical scavenging capacity, exhibiting relative rate constant ratios respectively of 0.513 and 0.287. This indicates that Mn(I1) is by far the most active. Therefore, the chain-breaking antioxidant capacity of Mn(I1) seems to be related to the rapid quenching of peroxyl radicals according to the reaction R-00. + Mn(I1) + H’ + ROOH + Mn(II1). The antioxidant mechanism is discussed considering the different reduction potentials of the examined cations. $121992
Academic
Press,
Inc.
Although some transition metal ions are active catalysts of lipid peroxidation, an antioxidant effect has been also reported (1). Both these effects have been attributed to the redox capability of metal ions, even though possible physical-chemical modifications of membranes have been also suggested (1). For example, while an observed antioxidant effect of Zn(II), Mn(II), and Co(I1) has been attributed to iron displacement from its binding sites in membranes (l), some of the d-block elements are also theoretically able to react with free radicals, playing a scavenger role. Manganese has been reported to inhibit lipid peroxidation both in vitro (2-9) and in vivo (10). However, the mechanism of this antioxidant effect has not been completely clarified. According to the definition, a free radical scavenger antioxidant must react with a peroxidation-driving radical and the rate constant of this reaction must be far higher than the rate constant of the reaction of the same radical with the biological molecule to be protected (i.e., a poly1 To whom
correspondence
should
be addressed.
Fax: 39 49 807-3310.
unsaturated fatty acid, PUFA’) (11, 12). Among free radicals involved in the lipoperoxidative process (principally, carbon-centered lipid radical, lipid peroxyl radical, and lipid alkoxyl radical), only scavenging of peroxyl radical seems relevant for an effective antioxidant effect. In fact, carbon-centered lipid radical reacts with oxygen at a diffusion-controlled rate and lipid alkoxyl radical reacts with lipids almost as fast as with free radical scavengers, in both cases frustrating a possible antioxidant reaction (11, 12). On the other hand, scavenging of the lipid peroxyl radical, driving the peroxidation chain reaction by reaction with PUFA, leads to an antioxidant effect since its reaction with biological target molecules is rather slow is possible (approx 10’ Mm’ s-l) and the competition (11, 12). In this paper we report the capacity of Mn(I1) to interact with hydroperoxyl radical in comparison with other metal ions and Trolox C. The results provide an explanation for the observed antioxidant effect of this metal ion at physiological concentration. MATERIALS
AND
METHODS
The reactivity of the metal ions with peroxyl radicals was measured by competition kinetics of crocin bleaching in the presence of peroxyl radicals generated by thermal decomposition of a diazo compound (13). This procedure was based on that previously described by Bors et al. (14). The test was carried out at 4O’C in a medium containing 137 mM NaCl, 2.7 mM KCl, 1.5 mM KH,PO,, 8.1 mM Na,HP04 (pH 7.4), 12 PM crocin, and increasing concentrations of divalent transition metal ions or antioxidants. The peroxyl radical generating reaction was started by adding ABAP (azobis-2.amidinopropane hydrochloride), from a fresh, ice-cold 0.6 M solution, to obtain a final concentration of 10 mM. The rate of crocin bleaching was recorded at 440 nm. The bleaching rate was linear 1.5 min after the addition of the diazo compound and the rate from min 2 to min 4 was used for calculations. Bleaching rates were plotted as V,/V, versus [A]/[C], according to the equation
VdVa = 1 + K./K. lAl/lCl,
’ Abbreviations used: PUFA, polyunsaturated fatty acid; ABAP, azohis-2-amidinopropane hydrochloride; MDA, malondialdehyde; DTPA, diethylene triamine pentaacetic acid.
330
0003.9861/92 Copyright All
rights
(c> 1992 of reproduction
$5.00
by Academic Press, Inc. in any form reserved.
ANTTnXlnANT
EFFECT
OF
331
MANGANESE
3.0
7d tl d
11.1 lid
FIG.
oxygen consumption induced by CHP in the presence of divalent metal 1. Microsomal added at the arrow. MDA (nmol/mg protein) formed, determined 5 min after CHP addition, following metal ions (100 FM) were added: (a) Mn(II), (b) Co(II), (c) Zn(II), (d) Ni(II), (e) (b) 100 manganese eflect is reported in the presence of various chelators: (a) 100 pM Mn(II), without manganese and chelators, (d) 100 pM Mn(II) plus 5 mM DTPA.
where V,, is the basal bleaching rate of crocin in the absence of antioxidants, V, is the bleaching rate of crocin in the presence of antioxidants, [A] is the antioxidant concentration, [C] is the crocin concentration, & is the rate constant for the reaction of the peroxyl radical with the antioxidant, and K, is the rate constant for the reaction of the peroxyl radical with crocin. This plot gives a straight line intersecting the ordinate at unity with a slope of K,/K,. Crocin was prepared from saffron (Sigma Chemical Co.) as described by Friend and Mayer (15). Rat liver microsomes were prepared as described by Ernster and Nordenbrand (16) from a 1:4 homogenate in 0.125 M KCI, 15 mM Hepes/ 6.5 mM Tris-HCl (pH 7.4). Microsomes were incubated at a concentration of 1 mg/ml in the same medium used for crocin bleaching and a final volume of 1.36 ml. Oxygen uptake was followed with a platinum electrode assembly of the Clark type (17). Malondialdehyde (MDA) formation was determined with the thiobarbituric acid method according to Buege and Aust (18). Protein was measured by the biuret method (19). RESULTS
The effects ofMn(II), Co(II), Fe(II), Zn(II), and Ni(I1) were tested in rat liver microsomes undergoing peroxidation in the presence of cumene hydroperoxide (CHP) (Fig. 1A). This peroxidation system was adopted to avoid possible interferences related to competition between metal ions, since iron ions are not required for initiation, as in the case of NADPH-ADP system. In the presence of CHP, indeed, initiating free radicals are generated by the interaction of the hydroperoxide with the heme moiety of cytochrome P450 and metal chelators are ineffective (20). While Mn(I1) strongly inhibited lipid peroxidation, measured as oxygen uptake and MDA formation, Co(I1) was much less active and Zn(I1) and Ni(I1) were almost completely ineffective. As expected, Fe(I1) stimulated, although not dramatically, peroxidative oxygen uptake and MDA production. The antioxidant effect of manganese was abolished by diethylenetriaminepentaacetic acid (DTPA), and reduced by about 50% by pyrophosphate (Fig. 1B).
ions and chelated manganese. CHP (100 FM) was is indicated at the end of each curve. In (A) the control without metal ions, (f) Fe(I1). In (B) the FM Mn(I1) plus 5 m&l pyrophosphate, (c) control
The peroxyl radical scavenging capacities of transition metals were analyzed by the competition kinetics of crocin bleaching (Fig. 2). In this test peroxyl radicals were produced by thermal decomposition of the diazo compound ABAP. While Zn(II), Ni(II), and Fe(I1) were almost completely ineffective, Mn(I1) and Co(I1) showed a free radical scavenging capacity, Mn(I1) being by far the most active (Fig. 2). Rate constant ratios of 0.513 and 0.287 were obtained for manganese and cobalt, respectively. Furthermore, these values were in the same range of those obtained, under the same experimental condition, in the presence of the vitamin E analog Trolox C (0.740) (Fig. 2). These results indicate that the rate constant of the interaction of Mn(I1) and Co(I1) with peroxyl radical is high enough to support an antioxidant capacity and that
KciKa
01 0
2
4
a
TrOlOX
0.740
Mn (Ii)
0.513
co (II)
0.267
Fe (II)
0 016
NI (II)
0.005
Zn (II)
0 001
10
lAlilC16 FIG. 2.
Competition kinetic plot of Trolox C and divalent transition metals toward crocin in the ABAP-induced radical reaction. Trolox C and metal ion concentration ranges from 10 to 100 FM. The slope of the straight lines indicates the relative capacity of different metals to interact with peroxyl radical according to Eq. [l] under Materials and Methods. The relative rate constant ratios (KJK,) are also reported.
332
COASSIN,
URSINI,
Mn(I1) and Co(R) actually fulfill the requirements of active chain-breaking antioxidants (11, 12). While the chelating agent DTPA completely removed the antioxidant effect of Mn(II), pyrophosphate further increased the inhibition of the crocin bleaching, giving rise to a rate constant ratio of 1.496, about double of that obtained with Trolox C (compare Figs. 2 and 3). This stimulation was less marked in the presence of ADP, the most likely physiological chelator. Upon interaction of the peroxyl radical with Mn(II)-pyrophosphate complex, the formation of the Mn(III)-pyrophosphate complex was spectroscopically confirmed (Fig. 4). It is worth noting that this complex remains stable for days and is easily reduced back by ascorbate or glutathione (not shown).
AND
BINDOLI
400 nm
WAVELENGTH
700 nm
,
/
0.012 0.010
-
e 0.008 i s 0006 4
-
0.004 0.002 -
DISCUSSION
Manganese is an essential transition element for animals and humans and the normal blood level is of 9 yg/ ml (1). It is required for mitochondrial superoxide dismutase synthesis and activates enzymes such as hydrolases and carboxylases (1). The free and total manganese contents in rat liver cells are estimated to be about 0.71 and 34 nmol X ml-l of cell water (1). According to (al), the concentration of Mn(I1) in mammalian liver is about 2.5 X lop5 M and in liver mitochondria, which are able to actively take up and retain manganese (22), it is in the range of 0.4 nmol/mg protein (22). Manganese has been reported to be an inhibitor of lipid peroxidation (a-lo), particularly in microsomes challenged by NADPH/Fe(II)/ ADP or cumene hydroperoxide (23). Furthermore, an analogy between the antioxidant capacity of Mn(I1) and that of phenolic antioxidants has been described (23). The present report provides direct and quantitative evidence of the chain-breaking antioxidant capacity of manganese, which is related to an effective quenching of peroxyl radicals, according to the reaction
KciKa
FIG. 3. Competition kinetic plot of manganese toward crocin in the ABAP-induced radical reaction in the presence of different chelating agents. Manganese(H) concentration ranges from 10 to 100 pM and chelators are 5 mM. The slope of the straight lines indicates the relative capacity of chelated manganese to interact with peroxyl radical according to Eq. [l] under Materials and Methods. The relative rate constant ratios (K,/K,) are also reported.
0.000 FIG. 4. Spectral changes associated with the Mn(III)-pyrophosphate complex formation. The Mn(III)-pyrophosphate complex formation was detected spectroscopically at 40°C in the same medium used for crocin bleaching and in the presence of 100 pM Mn(I1) and 5 mM pyrophosphate. The reaction was started with 10 mM ABAP and spectra were recorded every 20 min.
ROO * + Mn(I1)
+ H+ + ROOH + Mn(II1).
A mechanistic and kinetic rationale for the observed antioxidant effect is suggested by the reducing power of the examined cations. For the transition metal redox couples [M(III)/M(II)] the reduction potential varies in a wide range, being very high for Ni and in the order +1.82, f1.51, and +O.77 for Co, Mn, and Fe, respectively. For Zn the higher oxidation state is not known. Therefore one expects that Ni and Zn are not reactive in the monoelectronic reduction of peroxyl radicals, whereas others are reactive in the sequence Fe > Mn > Co. This accounts for the inhibition of crocin bleaching by Mn(I1) and Co(I1). On the other hand, the antioxidant capacity of Fe(II), although very likely able to reduce hydroperoxyl radicals, is apparently obscured by its interaction also with hydroperoxides, which gives rise to alkoxyl radicals, in turn responsible for the observed crocin bleaching. The rapid oxidation of iron (not shown), while crocin bleaching is not significantly inhibited, is in agreement with this interpretation. The difference in reduction potential appears, therefore, to be the key factor in the divergence between the antioxidant effect of Mn(I1) and the prooxidant effect of Fe(I1). Although both metal ions react with hydroperoxyl radicals, only the more reducing Fe(I1) is also able to promote hydroperoxide heterolytic O-O bond cleavage giving rise to alkoxyl radicals. Hydroperoxides in fact appear resistant to Mn(II), which does not give rise to a Fenton chemistry (24, 25).
ANTIOXIDANT
EFFECT
Manganese forms stable complexes with several chelators. It is well known that Mn(I1) can be easily transformed to Mn(II1) by the superoxide anion (26) and Mn complexes are important catalysts of superoxide dismutation in some Lactobacillaceae (26). The inhibiting effect of the strongly chelating ligand DTPA on the reactivity of Mn(I1) could be ascribed to the lack of a facile way for the electron transfer from the metal to the hydroperoxyl radical, for example via coordination. On the contrary, pyrophosphate ions could stabilize, as usual, the higher coordination state of the metal ion, thus decreasing the reduction potential of the Mn(III)/Mn(II) couple. Although favoring the interaction with hydroperoxyl radicals in solution, pyrophosphate decreases the inhibitory action of Mn(I1) on lipid peroxidation (Fig. 1B). This discrepancy is likely related either to a displacement, caused by pyrophosphate, of the Mn(I1) bound to the sites of the membranes, where lipid peroxidation occurs, or to a lower accessibility of Mn-pyrophosphate complex itself to membranes. Mn(III), the product of the antioxidant effect of Mn(II), in the presence of water may disproportionate into Mn(IV) and Mn(I1) and subsequently a two-electron oxidation by Mn(IV) might take place. Higher oxidation states of manganese could be formed in biological systems and the presence of this oxidized species could account for the toxic effect of high levels of manganese. Nevertheless, in biological systems several electron transfer processes are operating and are able to recycle manganese back to its reduced form. The reported concentrations of manganese in mammalian cells are compatible with the reported antioxidant capacity and particularly in mitochondria this metal ion fulfils the requirements of a physiologically relevant antioxidant. ACKNOWLEDGMENTS The useful discussion and suggestions of M. Bressan are gratefully acknowledged. This work was supported by the National Research Council of Italy, Special Project Raisa, subproject 4 no 631.
REFERENCES 1. Halliwell, B., and Gutteridge, J. M. C. (1989) Free Radicals in Biology and Medicine, 2nd ed., Clarendon Press, Oxford.
333
OF MANGANESE
2. Thiele, E. H., and Huff, J. W. (1960) Arch. Biochem. Biophys. 88, 203-207. 3. Tam, B. K., and McCay, P. B. (1970) J. Biol. Chem. 245, 22952300. 4. Pfeifer, P. M., and McCay, P. B. (1971) J. Biol. Chem. 246,64016408. 5. Hoffsten, P. E., Hunter, F. E., Jr., Gebicki, J. M., and Weinstein, J. (1962) Rio&em.
Riophys. Res. Commun. 7, 276-280.
6. May, H. E., and Mccay, P. B. (1968) J. Biol. Chen. 243, 22962305. I. Wills, E. D. (1969) Biochem. J. 113,315-332. 8. Takeshige, K., and Minakami, S. (1975) J. Biochem. 77,1067-1073. 9. Bindoli, A., and Cavallini, L. (1980) Bull. Mol. Riol. Med. 5, 1-9. 10. Shukla, G. S., and Chandra, S. V. (1981) Acta Pharmacol. Z’oricol. 48,95-100. 11. Ursini, F., Maiorino, M., and Sevanian, A. (1991) in Oxidative Stress: Oxidants and Antioxidants (Sies, H., Ed.), pp. 319-336, Academic Press, London. 12. Burton, G. W., and Ingold, K. U. (1986) Act. Chem. Res. 19, 194-
201. 13. van Amsterdam, F. T. M., Roveri, A., Maiorino, M., Ratti, E., and Ursini, F. (1992) Free Radical Biol. Med. 12, 183-187. 14. Bars, W., Michael, C., and Saran, M. (1984) Riochim. Biophys. Acta 796, 312-319.
15. Friend, J., and Mayer, A. M. (1960) Biochim. Biophys. Acta 41, 422-429. 16. Emster, L., and Nordenbrand, K. (1967) in Methods in Enzymology
17.
18. 19. 20. 21. 22. 23. 24.
(Estabrook, W. E., and Pullmann, M. E. Eds.), Vol. 10, pp. 574580, Academic Press, New York. Estabrook, R. W. (1967) in Methods in Enzymology (Estabrook, W. E., and Pullmann, M. E. Eds.), Vol. 10, pp. 41-47, Academic Press, New York. Buege, J. A., and Aust, S. D. (1978) in Methods in Enzymology (Fleischer, S., and Packer, L. Eds.), Vol. 52, pp. 3022310, Academic Press, New York. Gornall, A. G., Bardawill, C. J., and David, M. M. (1949) J. Biol. Chem. 177,751-766. O’Brien, P. J., and Rahimtula, A. (1975) J. Agr. Food Chem. 23, 154-158. Wills, E. D. (1965) Biochim. Biophys. Acta 98, 238-251. Lehninger, A. L. (1965) The Mitochondrion, Benjamin, New York. Cavallini, L., Valente, M., and Bindoli, A. (1984) Inorg. Chim. Acta 91, 117-120. Cheton, P. L. B., and Archibald, F. S. (1988) Free Radical Riol. Med.
5,325-333. 25. Gutteridge, J. M. C., and Bannister, J. V. (1986). Biochem. J. 234, 225-228. 26. Archibald, F. S., and Fridovich, I. (1982) Arch. Biochem. Biophys. 214,452%463.
OF BIOCHEMISTRY
Vol. 299, No. 2, December,
Antioxidant Mariagrazia
AND
BIOPHYSICS
pp. 330-333,
1992
Effect of Manganese
Coassin,* Fulvio Ursini,?’
and Albert0 Bindolit
*Department of Biological Chemistry, University of Padua; TDepartment of Chemical Science and Technology, University of Udine; and jCNR-Center for Mitochondrial Physiology, Padua, Italy
Received
May
14, 1992, and in revised
form,
August
10, 1992
The antioxidant effects of manganese and other transition metals were studied as the inhibition of microsomal lipid peroxidation and crocin bleaching by peroxyl radicals. The peroxyl radical scavenging capacity was measured by competition kinetics analysis. While Zn(II), Ni(II), and Fe(I1) were almost completely ineffective, Mn(I1) and Co(I1) showed a free radical scavenging capacity, exhibiting relative rate constant ratios respectively of 0.513 and 0.287. This indicates that Mn(I1) is by far the most active. Therefore, the chain-breaking antioxidant capacity of Mn(I1) seems to be related to the rapid quenching of peroxyl radicals according to the reaction R-00. + Mn(I1) + H’ + ROOH + Mn(II1). The antioxidant mechanism is discussed considering the different reduction potentials of the examined cations. $121992
Academic
Press,
Inc.
Although some transition metal ions are active catalysts of lipid peroxidation, an antioxidant effect has been also reported (1). Both these effects have been attributed to the redox capability of metal ions, even though possible physical-chemical modifications of membranes have been also suggested (1). For example, while an observed antioxidant effect of Zn(II), Mn(II), and Co(I1) has been attributed to iron displacement from its binding sites in membranes (l), some of the d-block elements are also theoretically able to react with free radicals, playing a scavenger role. Manganese has been reported to inhibit lipid peroxidation both in vitro (2-9) and in vivo (10). However, the mechanism of this antioxidant effect has not been completely clarified. According to the definition, a free radical scavenger antioxidant must react with a peroxidation-driving radical and the rate constant of this reaction must be far higher than the rate constant of the reaction of the same radical with the biological molecule to be protected (i.e., a poly1 To whom
correspondence
should
be addressed.
Fax: 39 49 807-3310.
unsaturated fatty acid, PUFA’) (11, 12). Among free radicals involved in the lipoperoxidative process (principally, carbon-centered lipid radical, lipid peroxyl radical, and lipid alkoxyl radical), only scavenging of peroxyl radical seems relevant for an effective antioxidant effect. In fact, carbon-centered lipid radical reacts with oxygen at a diffusion-controlled rate and lipid alkoxyl radical reacts with lipids almost as fast as with free radical scavengers, in both cases frustrating a possible antioxidant reaction (11, 12). On the other hand, scavenging of the lipid peroxyl radical, driving the peroxidation chain reaction by reaction with PUFA, leads to an antioxidant effect since its reaction with biological target molecules is rather slow is possible (approx 10’ Mm’ s-l) and the competition (11, 12). In this paper we report the capacity of Mn(I1) to interact with hydroperoxyl radical in comparison with other metal ions and Trolox C. The results provide an explanation for the observed antioxidant effect of this metal ion at physiological concentration. MATERIALS
AND
METHODS
The reactivity of the metal ions with peroxyl radicals was measured by competition kinetics of crocin bleaching in the presence of peroxyl radicals generated by thermal decomposition of a diazo compound (13). This procedure was based on that previously described by Bors et al. (14). The test was carried out at 4O’C in a medium containing 137 mM NaCl, 2.7 mM KCl, 1.5 mM KH,PO,, 8.1 mM Na,HP04 (pH 7.4), 12 PM crocin, and increasing concentrations of divalent transition metal ions or antioxidants. The peroxyl radical generating reaction was started by adding ABAP (azobis-2.amidinopropane hydrochloride), from a fresh, ice-cold 0.6 M solution, to obtain a final concentration of 10 mM. The rate of crocin bleaching was recorded at 440 nm. The bleaching rate was linear 1.5 min after the addition of the diazo compound and the rate from min 2 to min 4 was used for calculations. Bleaching rates were plotted as V,/V, versus [A]/[C], according to the equation
VdVa = 1 + K./K. lAl/lCl,
’ Abbreviations used: PUFA, polyunsaturated fatty acid; ABAP, azohis-2-amidinopropane hydrochloride; MDA, malondialdehyde; DTPA, diethylene triamine pentaacetic acid.
330
0003.9861/92 Copyright All
rights
(c> 1992 of reproduction
$5.00
by Academic Press, Inc. in any form reserved.
ANTTnXlnANT
EFFECT
OF
331
MANGANESE
3.0
7d tl d
11.1 lid
FIG.
oxygen consumption induced by CHP in the presence of divalent metal 1. Microsomal added at the arrow. MDA (nmol/mg protein) formed, determined 5 min after CHP addition, following metal ions (100 FM) were added: (a) Mn(II), (b) Co(II), (c) Zn(II), (d) Ni(II), (e) (b) 100 manganese eflect is reported in the presence of various chelators: (a) 100 pM Mn(II), without manganese and chelators, (d) 100 pM Mn(II) plus 5 mM DTPA.
where V,, is the basal bleaching rate of crocin in the absence of antioxidants, V, is the bleaching rate of crocin in the presence of antioxidants, [A] is the antioxidant concentration, [C] is the crocin concentration, & is the rate constant for the reaction of the peroxyl radical with the antioxidant, and K, is the rate constant for the reaction of the peroxyl radical with crocin. This plot gives a straight line intersecting the ordinate at unity with a slope of K,/K,. Crocin was prepared from saffron (Sigma Chemical Co.) as described by Friend and Mayer (15). Rat liver microsomes were prepared as described by Ernster and Nordenbrand (16) from a 1:4 homogenate in 0.125 M KCI, 15 mM Hepes/ 6.5 mM Tris-HCl (pH 7.4). Microsomes were incubated at a concentration of 1 mg/ml in the same medium used for crocin bleaching and a final volume of 1.36 ml. Oxygen uptake was followed with a platinum electrode assembly of the Clark type (17). Malondialdehyde (MDA) formation was determined with the thiobarbituric acid method according to Buege and Aust (18). Protein was measured by the biuret method (19). RESULTS
The effects ofMn(II), Co(II), Fe(II), Zn(II), and Ni(I1) were tested in rat liver microsomes undergoing peroxidation in the presence of cumene hydroperoxide (CHP) (Fig. 1A). This peroxidation system was adopted to avoid possible interferences related to competition between metal ions, since iron ions are not required for initiation, as in the case of NADPH-ADP system. In the presence of CHP, indeed, initiating free radicals are generated by the interaction of the hydroperoxide with the heme moiety of cytochrome P450 and metal chelators are ineffective (20). While Mn(I1) strongly inhibited lipid peroxidation, measured as oxygen uptake and MDA formation, Co(I1) was much less active and Zn(I1) and Ni(I1) were almost completely ineffective. As expected, Fe(I1) stimulated, although not dramatically, peroxidative oxygen uptake and MDA production. The antioxidant effect of manganese was abolished by diethylenetriaminepentaacetic acid (DTPA), and reduced by about 50% by pyrophosphate (Fig. 1B).
ions and chelated manganese. CHP (100 FM) was is indicated at the end of each curve. In (A) the control without metal ions, (f) Fe(I1). In (B) the FM Mn(I1) plus 5 m&l pyrophosphate, (c) control
The peroxyl radical scavenging capacities of transition metals were analyzed by the competition kinetics of crocin bleaching (Fig. 2). In this test peroxyl radicals were produced by thermal decomposition of the diazo compound ABAP. While Zn(II), Ni(II), and Fe(I1) were almost completely ineffective, Mn(I1) and Co(I1) showed a free radical scavenging capacity, Mn(I1) being by far the most active (Fig. 2). Rate constant ratios of 0.513 and 0.287 were obtained for manganese and cobalt, respectively. Furthermore, these values were in the same range of those obtained, under the same experimental condition, in the presence of the vitamin E analog Trolox C (0.740) (Fig. 2). These results indicate that the rate constant of the interaction of Mn(I1) and Co(I1) with peroxyl radical is high enough to support an antioxidant capacity and that
KciKa
01 0
2
4
a
TrOlOX
0.740
Mn (Ii)
0.513
co (II)
0.267
Fe (II)
0 016
NI (II)
0.005
Zn (II)
0 001
10
lAlilC16 FIG. 2.
Competition kinetic plot of Trolox C and divalent transition metals toward crocin in the ABAP-induced radical reaction. Trolox C and metal ion concentration ranges from 10 to 100 FM. The slope of the straight lines indicates the relative capacity of different metals to interact with peroxyl radical according to Eq. [l] under Materials and Methods. The relative rate constant ratios (KJK,) are also reported.
332
COASSIN,
URSINI,
Mn(I1) and Co(R) actually fulfill the requirements of active chain-breaking antioxidants (11, 12). While the chelating agent DTPA completely removed the antioxidant effect of Mn(II), pyrophosphate further increased the inhibition of the crocin bleaching, giving rise to a rate constant ratio of 1.496, about double of that obtained with Trolox C (compare Figs. 2 and 3). This stimulation was less marked in the presence of ADP, the most likely physiological chelator. Upon interaction of the peroxyl radical with Mn(II)-pyrophosphate complex, the formation of the Mn(III)-pyrophosphate complex was spectroscopically confirmed (Fig. 4). It is worth noting that this complex remains stable for days and is easily reduced back by ascorbate or glutathione (not shown).
AND
BINDOLI
400 nm
WAVELENGTH
700 nm
,
/
0.012 0.010
-
e 0.008 i s 0006 4
-
0.004 0.002 -
DISCUSSION
Manganese is an essential transition element for animals and humans and the normal blood level is of 9 yg/ ml (1). It is required for mitochondrial superoxide dismutase synthesis and activates enzymes such as hydrolases and carboxylases (1). The free and total manganese contents in rat liver cells are estimated to be about 0.71 and 34 nmol X ml-l of cell water (1). According to (al), the concentration of Mn(I1) in mammalian liver is about 2.5 X lop5 M and in liver mitochondria, which are able to actively take up and retain manganese (22), it is in the range of 0.4 nmol/mg protein (22). Manganese has been reported to be an inhibitor of lipid peroxidation (a-lo), particularly in microsomes challenged by NADPH/Fe(II)/ ADP or cumene hydroperoxide (23). Furthermore, an analogy between the antioxidant capacity of Mn(I1) and that of phenolic antioxidants has been described (23). The present report provides direct and quantitative evidence of the chain-breaking antioxidant capacity of manganese, which is related to an effective quenching of peroxyl radicals, according to the reaction
KciKa
FIG. 3. Competition kinetic plot of manganese toward crocin in the ABAP-induced radical reaction in the presence of different chelating agents. Manganese(H) concentration ranges from 10 to 100 pM and chelators are 5 mM. The slope of the straight lines indicates the relative capacity of chelated manganese to interact with peroxyl radical according to Eq. [l] under Materials and Methods. The relative rate constant ratios (K,/K,) are also reported.
0.000 FIG. 4. Spectral changes associated with the Mn(III)-pyrophosphate complex formation. The Mn(III)-pyrophosphate complex formation was detected spectroscopically at 40°C in the same medium used for crocin bleaching and in the presence of 100 pM Mn(I1) and 5 mM pyrophosphate. The reaction was started with 10 mM ABAP and spectra were recorded every 20 min.
ROO * + Mn(I1)
+ H+ + ROOH + Mn(II1).
A mechanistic and kinetic rationale for the observed antioxidant effect is suggested by the reducing power of the examined cations. For the transition metal redox couples [M(III)/M(II)] the reduction potential varies in a wide range, being very high for Ni and in the order +1.82, f1.51, and +O.77 for Co, Mn, and Fe, respectively. For Zn the higher oxidation state is not known. Therefore one expects that Ni and Zn are not reactive in the monoelectronic reduction of peroxyl radicals, whereas others are reactive in the sequence Fe > Mn > Co. This accounts for the inhibition of crocin bleaching by Mn(I1) and Co(I1). On the other hand, the antioxidant capacity of Fe(II), although very likely able to reduce hydroperoxyl radicals, is apparently obscured by its interaction also with hydroperoxides, which gives rise to alkoxyl radicals, in turn responsible for the observed crocin bleaching. The rapid oxidation of iron (not shown), while crocin bleaching is not significantly inhibited, is in agreement with this interpretation. The difference in reduction potential appears, therefore, to be the key factor in the divergence between the antioxidant effect of Mn(I1) and the prooxidant effect of Fe(I1). Although both metal ions react with hydroperoxyl radicals, only the more reducing Fe(I1) is also able to promote hydroperoxide heterolytic O-O bond cleavage giving rise to alkoxyl radicals. Hydroperoxides in fact appear resistant to Mn(II), which does not give rise to a Fenton chemistry (24, 25).
ANTIOXIDANT
EFFECT
Manganese forms stable complexes with several chelators. It is well known that Mn(I1) can be easily transformed to Mn(II1) by the superoxide anion (26) and Mn complexes are important catalysts of superoxide dismutation in some Lactobacillaceae (26). The inhibiting effect of the strongly chelating ligand DTPA on the reactivity of Mn(I1) could be ascribed to the lack of a facile way for the electron transfer from the metal to the hydroperoxyl radical, for example via coordination. On the contrary, pyrophosphate ions could stabilize, as usual, the higher coordination state of the metal ion, thus decreasing the reduction potential of the Mn(III)/Mn(II) couple. Although favoring the interaction with hydroperoxyl radicals in solution, pyrophosphate decreases the inhibitory action of Mn(I1) on lipid peroxidation (Fig. 1B). This discrepancy is likely related either to a displacement, caused by pyrophosphate, of the Mn(I1) bound to the sites of the membranes, where lipid peroxidation occurs, or to a lower accessibility of Mn-pyrophosphate complex itself to membranes. Mn(III), the product of the antioxidant effect of Mn(II), in the presence of water may disproportionate into Mn(IV) and Mn(I1) and subsequently a two-electron oxidation by Mn(IV) might take place. Higher oxidation states of manganese could be formed in biological systems and the presence of this oxidized species could account for the toxic effect of high levels of manganese. Nevertheless, in biological systems several electron transfer processes are operating and are able to recycle manganese back to its reduced form. The reported concentrations of manganese in mammalian cells are compatible with the reported antioxidant capacity and particularly in mitochondria this metal ion fulfils the requirements of a physiologically relevant antioxidant. ACKNOWLEDGMENTS The useful discussion and suggestions of M. Bressan are gratefully acknowledged. This work was supported by the National Research Council of Italy, Special Project Raisa, subproject 4 no 631.
REFERENCES 1. Halliwell, B., and Gutteridge, J. M. C. (1989) Free Radicals in Biology and Medicine, 2nd ed., Clarendon Press, Oxford.
333
OF MANGANESE
2. Thiele, E. H., and Huff, J. W. (1960) Arch. Biochem. Biophys. 88, 203-207. 3. Tam, B. K., and McCay, P. B. (1970) J. Biol. Chem. 245, 22952300. 4. Pfeifer, P. M., and McCay, P. B. (1971) J. Biol. Chem. 246,64016408. 5. Hoffsten, P. E., Hunter, F. E., Jr., Gebicki, J. M., and Weinstein, J. (1962) Rio&em.
Riophys. Res. Commun. 7, 276-280.
6. May, H. E., and Mccay, P. B. (1968) J. Biol. Chen. 243, 22962305. I. Wills, E. D. (1969) Biochem. J. 113,315-332. 8. Takeshige, K., and Minakami, S. (1975) J. Biochem. 77,1067-1073. 9. Bindoli, A., and Cavallini, L. (1980) Bull. Mol. Riol. Med. 5, 1-9. 10. Shukla, G. S., and Chandra, S. V. (1981) Acta Pharmacol. Z’oricol. 48,95-100. 11. Ursini, F., Maiorino, M., and Sevanian, A. (1991) in Oxidative Stress: Oxidants and Antioxidants (Sies, H., Ed.), pp. 319-336, Academic Press, London. 12. Burton, G. W., and Ingold, K. U. (1986) Act. Chem. Res. 19, 194-
201. 13. van Amsterdam, F. T. M., Roveri, A., Maiorino, M., Ratti, E., and Ursini, F. (1992) Free Radical Biol. Med. 12, 183-187. 14. Bars, W., Michael, C., and Saran, M. (1984) Riochim. Biophys. Acta 796, 312-319.
15. Friend, J., and Mayer, A. M. (1960) Biochim. Biophys. Acta 41, 422-429. 16. Emster, L., and Nordenbrand, K. (1967) in Methods in Enzymology
17.
18. 19. 20. 21. 22. 23. 24.
(Estabrook, W. E., and Pullmann, M. E. Eds.), Vol. 10, pp. 574580, Academic Press, New York. Estabrook, R. W. (1967) in Methods in Enzymology (Estabrook, W. E., and Pullmann, M. E. Eds.), Vol. 10, pp. 41-47, Academic Press, New York. Buege, J. A., and Aust, S. D. (1978) in Methods in Enzymology (Fleischer, S., and Packer, L. Eds.), Vol. 52, pp. 3022310, Academic Press, New York. Gornall, A. G., Bardawill, C. J., and David, M. M. (1949) J. Biol. Chem. 177,751-766. O’Brien, P. J., and Rahimtula, A. (1975) J. Agr. Food Chem. 23, 154-158. Wills, E. D. (1965) Biochim. Biophys. Acta 98, 238-251. Lehninger, A. L. (1965) The Mitochondrion, Benjamin, New York. Cavallini, L., Valente, M., and Bindoli, A. (1984) Inorg. Chim. Acta 91, 117-120. Cheton, P. L. B., and Archibald, F. S. (1988) Free Radical Riol. Med.
5,325-333. 25. Gutteridge, J. M. C., and Bannister, J. V. (1986). Biochem. J. 234, 225-228. 26. Archibald, F. S., and Fridovich, I. (1982) Arch. Biochem. Biophys. 214,452%463.