Stimulation by Nitroxides of Catalase-like Activity of Hemeproteins KINETICS AND MECHANISM*

The ability of stable nitroxide radicals to detoxify hypervalent heme proteins such as ferrylmyoglobin (MbFe IV ) produced in the reaction of metmyoglobin (MbFe III ) and H 2 O 2 was evaluated by monitoring O 2 evo- lution, H 2 O 2 depletion, and redox changes of the heme prosthetic group. The rate of H 2 O 2 depletion and O 2 evolution catalyzed by MbFe III was enhanced by stable nitroxides such as 4-OH-2,2,6,6-tetramethyl-piperidi- noxyl (TPL) in a catalytic fashion. The reduction of MbFe IV to MbFe III was the rate-limiting step. Excess TPL over MbFe III enhanced catalase-like activity more than 4-fold. During dismutation of H 2 O 2 , [TPL] and [MbFe IV ] remained constant. NADH caused: ( a ) inhibition of H 2 O 2 decay; ( b ) progressive reduction of TPL to its respective hydroxylamine TPL-H; and ( c ) arrest/inhi-bition of oxygen evolution or elicit consumption of O 2 . Following depletion of NADH the evolution of O 2 resumed, and the initial concentration of TPL was re- stored. Kinetic analysis showed that two distinct forms of MbFe IV might be involved in the process. In summary, by shuttling between two oxidation states, namely nitrox- ide and oxoammonium cation, stable nitroxides enhance the catalase mimic activity of MbFe III , thus facilitating H 2 O 2 dismutation accompanied by O

The heme iron of myoglobin (Mb), 1 the major heme protein in muscle tissue involved in the storage and delivery of oxygen by shuttling between oxymyoglobin (oxyMbFe II ) and Mb, is predominantly in ϩ2 oxidation state. However, oxyMbFe II can undergo oxidation and yield MbFe III and O 2 . , which disproportionates to O 2 and H 2 O 2 (1). Such reactions were also noted for hemoglobin (2). Although no catalytic reactions are normally associated with myoglobin, modest peroxidase activity capable of oxidizing a variety of biological substrates has been identified (3)(4)(5). The reaction between H 2 O 2 and metmyoglobin has been shown to produce a two-electron oxidation product of myoglobin by a heterolytic cleavage of the O-O bond of the peroxide coordinated to the heme.
One oxidation equivalent has been shown to reside on the globin as a transient radical located on an aromatic amino acid residue (3,4). The second and longer lived oxidizing equivalent was shown to be the oxoferryl complex (6,7), Fe IV ϭO, which is denoted as Fe IV throughout the text. Although the lifetime of the globin radical has been estimated to be in the range of 50 -280 ms, the lifetime of the oxoferryl species ranged between minutes and hours (8,9). Both the globin radical and the oxoferryl species are chemically reactive and have access to components in the bulk solution and initiate free radical-mediated reactions that could result in biologic damage (10) by chemical oxidation reactions (11). Considerable efforts have been made to characterize the globin radical. Although earlier studies suggested the globin radical to be a tyrosine phenoxyl radical (12), recent high resolution EPR-spin trapping studies provide strong evidence to identify the amino acid radical on the globin to be centered on tryptophan (13). The globin radical has been proposed to mediate oxidation, peroxidation, and epoxidation of substrates and induce biologic damage (10,14). The oxoferryl moiety on the other hand has also been implicated in oxidation of cellular components and xenobiotics (9).
The decay of the globin radical proceeds through mechanisms that are not conclusively identified. In the absence of other reductants, the ferryl moiety and possibly also the globin radical can oxidize another H 2 O 2 molecule to O 2 . and consequently yield molecular O 2 (Reactions 2 and 3). Reactions of hematin/hydroperoxide to generate alkoxy radicals have been also reported (15). There is, however, no direct evidence for H 2 O 2 -induced decay of the globin radical (Equation 2), and an alternative pathway might involve an intramolecular reduction of globin radical to yield the perferrylMb (MbFe V ). Whatever is the mechanism of globin radical decay, MbFe IV can be reduced to MbFe III by reacting with H 2 O 2 (Equation 3): and thereby confer modest peroxidase activity to myoglobin. The biologic significance of MbFe IV , stems from its ability to oxidize critical targets such as unsaturated fatty acids or mem-brane lipoproteins (4, 16 -18), contribute to post ischemic reperfusion myocardial injury (19), and be converted to an oxidase that can utilize ubiquitous cofactors to generate additional H 2 O 2 (20). Classical antioxidants such as thiols and phenolic derivatives can reduce MbFe IV to MbFe III in a stoichiometric manner (9,21,22). Stable as well as persistent nitroxide free radicals have been previously found to react with ferryl ion (23,24) and decompose H 2 O 2 in the presence of heme proteins without being consumed in the process (25). However, no detailed studies on the mechanism underlying how the nitroxides increase catalase-like activity have been made.
The unique feature of the antioxidative capacity of nitroxides is their catalytic nature. By undergoing 1-electron transfer reactions, nitroxides "shuttle" among three oxidation states, being readily reduced to cyclic hydroxylamines or oxidized to oxoammonium cations, as shown in Scheme 1 for 2,2,6,6,tetramethylpiperidine-N-oxyl (TPO). All three forms can be present in the tissues. Earlier studies showed that by flip-flopping between the two forms of the nitroxide/oxoammonium redox pair, nitroxides facilitate superoxide dismutation in a catalytic manner, i.e. they possess superoxide dismutase mimetic activity. The midpoint redox potentials of nitroxide/oxoammonium pair for many nitroxides range between 700 and 900 mV versus normal hydrogen electrode (26,27). In the presence of a 2-electron reductant such as NADH, the oxoammonium cation can be directly reduced to hydroxylamine (27). Chemically, the oxoammonium species has also been isolated as an end product in hemoglobin/H 2 O 2 -mediated oxidation (24).
Several independent studies using whole animals, isolated organs, or cell culture identified nitroxides as capable of inhibiting oxidative biologic damage imposed by diverse types of oxidative insults (28 -32). More recent studies demonstrated the inhibitory effects of nitroxides on oxidant pathways in methyl prednisone-induced apoptosis of thymocytes (33). Several mechanisms underlying the protective activity of nitroxides have been elucidated, including oxidation of redox-active reduced transition metals (28,29), catalytic removal of O 2 .
(27), detoxification of xenobiotic-derived semiquinones (34), and termination of radical chain reactions (35). The present study concentrates on the kinetics and mechanism of the involvement of nitroxides in redox processes mediated by heme proteins, such as Mb. The results demonstrate the catalytic fashion by which nitroxides can facilitate the catalase-like activity of MbFe IV and substantiate the potential role of nitroxides in detoxification of H 2 O 2 and reactive species such as ferryl.
Electron Paramagnetic Resonance-EPR measurements were carried out by transferring the reaction mixture to gas-permeable teflon capillary tube (Zeus Industries, Orangeburg, SC) of 0.81-mm inner diameter, 0.38-mm wall thickness, and 15-cm length. Each capillary was folded twice, inserted into a narrow quartz tube that was open at both ends (2.5-mm inner diameter), and then placed in the EPR cavity. EPR parameters were as follows: modulation amplitude, 1 G; time constant, 0.13 s; scan time, 30 min; gain, 2.5 ϫ 10 3 ; modulation frequency, 100 kHz; microwave power, 10 mW.
Cell Culture-Chinese hamster V79 lung fibroblasts were cultured in Ham's F-12 medium supplemented with 10% fetal calf serum and antibiotics. Cell survival was assessed by clonogenic assay. Stock cultures of exponentially growing cells were trypsinized, rinsed, and plated (7 ϫ 10 5 cells/dish) into a number of 100-cm 2 Petri dishes and incubated 16 h at 37°C prior to use in experimental protocols. Immediately prior to treatment, medium was removed from the plates, the cell monolayer was rinsed twice with phosphate-buffered saline, and fresh Ham's F-12 medium without fetal calf serum was added. Cells were exposed to hydrogen peroxide (final concentration, 750 M) for 1 h in the absence or the presence of TPL (50 M) and MbFe III (50 M). Following treatment, the cells were washed twice with phosphate-buffered saline, trypsinized, counted, and plated for macroscopic colony formation. For each dose determination, cells were plated into triplicate dishes, and each experiment was repeated a minimum of two times. Plates were incubated 7 days, after which colonies were fixed with methanol/acetic acid (3:1), stained with crystal violet, and counted. Colonies containing Ͼ50 cells were scored. Error bars represent S.D. of the mean. The three-electron reduced form of TPL, namely the cyclic amine, did not enhance O 2 evolution. The cyclic hydroxylamine TPL-H progressively enhanced oxygen evolution mediated by MbFe III , concomitant with the hydroxylamine's rapid oxidation to TPL (data not shown).

General-The
The Predominant Heme Iron Species-The distribution of the heme iron among its various oxidation states was determined by recording the uv-visible spectra as a function of time and deriving the respective concentrations of MbFe III and MbFe IV according to Whitburn's algorithms (38). Upon the addition of a large excess of H 2 O 2 , as indicated by the arrow, the absorption spectrum of MbFe III was replaced by that of MbFe IV , which persisted until H 2 O 2 depletion (Fig. 1C). TPL did not affect the level of MbFe IV ; however, TPL enhanced the rate of H 2 O 2 depletion (Fig. 1B) associated with a corresponding restoration of MbFe III (Fig. 1C).
The Persistence of TPL-The complete reaction mixture containing MbFe III , H 2 O 2 and TPL was analyzed for the TPL signal using EPR spectroscopy. In the absence of reducing agent, such as NADH, the nitroxide persisted without any appreciable spin loss throughout the complete MbFe III -induced dismutation of a large excess of H 2 O 2 (data not shown). No change of the shape of the EPR spectrum, indicative of any spin-spin interaction between TPL and the heme iron, was observed throughout the course of the reaction. The persistence of TPL concentration indicated that it acts catalytically as a self-replenishing reagent in the dismutation of H 2 O 2 with a concomitant O 2 evolution.
The Catalytic Species-TPL accelerates the catalase-like dis-mutation of H 2 O 2 in a reaction in which TPL is not consumed in the overall process. Furthermore, there is no stoichiometric relationship between TPL and the quantity of the substrate altered, all of which indicates TPL functions as a true catalyst. The observed stoichiometry of 1:2 between O 2 evolution accompanying H 2 O 2 decay mediated by MbFe III with nitroxide provides further evidence for the genuine catalytic nature of the process. Similar catalytic decomposition facilitated by manganese/bicarbonate has been observed in earlier studies (37). Evidently such catalysis implies a time invariance of the concentration of MbFe IV and of the nitroxide (see Fig. 1).    (Fig. 3A). However, at higher NADH concentrations as when [NADH] Ͼ Ͼ [TPL], O 2 was consumed at a rate comparable with that observed in the absence of nitroxide. This effect was also transient, and with the progressive depletion of NADH, the consumption of oxygen ended, and later evolution of O 2 was gradually resumed to the original rate. Successive transient inhibitions of O 2 evolution could be demonstrated by fractionated addition of NADH aliquots, and the inhibition duration increased with the increase of [NADH] added (data not shown). Control experiments showed that the two electron oxidation product of NADH, namely NAD ϩ , had no effect on the rate of O 2 evolution (data not shown). NADH Effect on TPL Level-The effect of NADH on the level of TPL was studied by following the intensity of the EPR signal of the nitroxide throughout the course of the reaction in a sample containing MbFe III and H 2 O 2 . Upon the addition of NADH, the TPL level progressively decreased for several minutes, and when NADH was consumed the EPR signal of [TPL] increased, achieving its original level (Fig. 3C). The restoration of EPR signal of TPL paralleled the restoration of the original rate of O 2 evolution. The transient decrease of EPR signal resulted from the reduction of TPL to its respective hydroxylamine, TPL-H, as this EPR-silent form was oxidizable to TPL by ferricyanide (27). Upon complete consumption of the NADH, the evolution of O 2 was resumed, whereas the original level of nitroxide was restored.

NADH Effect on H 2 O 2 Decay-Upon
NADH Oxidation-The oxidation of NADH by the MbFe III (25 M) and H 2 O 2 (1 mM) in the presence and the absence of TPL was followed spectrophotometrically at 340 nm, and the results are shown in Fig. 3D. The presence of TPL enhanced the rate of oxidation of NADH.
The Kinetics of MbFe IV Reduction by TPL-The effect of TPL on the rate of O 2 evolution seemed to depend on its direct reaction with the hypervalent iron. To study the kinetics of MbFe IV reaction with TPL, its reduction to MbFe III was spectrophotometrically followed. MbFe IV was prepared by treating 10 M MbFe III with an excess of H 2 O 2 in 50 mM phosphate buffer, pH 7, followed by 500 units/ml catalase to remove unreacted H 2 O 2 . In the absence of nitroxide, MbFe IV decayed slowly to MbFe III ; however, TPL facilitated this reduction. To study the reaction kinetics, the change in OD 408 nm was monitored and ⌬OD ϭ OD ϱ Ϫ OD t was determined, where OD ϱ represented the constant absorbance achieved upon completion of the reaction.
The decay of ⌬OD exhibited a single exponential kinetics irrespective of the nitroxide concentration (see Fig. 4A) even at [TPL] Ͻ [MbFe IV ], and the first order reaction rate constant k obs was evaluated. The reaction was conducted at various concentrations of TPL, and the dependence of k obs on [TPL], which shows a biphasic character, is presented in Fig. 4. The value of k obs linearly increased with the increase of [TPL]; however, at [TPL] Ͻ 100 M, the dependence exhibited a downward curvature. A similar dependence of k obs on [nitroxide] was obtained when TPL was replaced by TPA or TPO as seen in Fig. 4.
The Reaction of H 2 O 2 with TPL ϩ -The initial step of superoxide dismutation catalyzed by TPL involves the generation of TPL ϩ and H 2 O 2 (27). In order to study the reverse reaction, TPL ϩ was generated electrochemically and added into an oxygraph chamber containing deaerated, continually stirred, buffered solution of H 2 O 2 . An instantaneous evolution of O 2 was observed upon the addition of TPL ϩ (data not shown), which indicates that with large excess of H 2 O 2 the reverse reaction does indeed take place.
Cellular Cytoprotection-The protective effect of the combination of MbFe III and TPL against H 2 O 2 injury to mammalian cells was tested. Chinese hamster V79 cells were exposed to 750 M H 2 O 2 for 1 h in the absence or the presence of 50 M MbFe III or 50 M TPL alone or their combination, and H 2 O 2 cytotoxicity was assessed (28,29). Incubation with H 2 O 2 alone caused a 98% loss in cell survival, and modest protection was provided by MbFe III , which can be associated with its pseudoperoxidase activity; whereas, TPL alone was without effect. However, the combination of MbFe III and TPL protected the cells from H 2 O 2 (Fig. 5). The large synergistic activity of TPL on the protective effect of MbFe III is attributed to the facilitation of H 2 O 2 dismutation.

DISCUSSION
The pseudo-peroxidase activity of heme protein (3) is known to involve the intermediacy of MbFe IV (6), a strong oxidant capable of inflicting significant damage by oxidizing a number of biologic targets (10,17,19). Therefore, its detoxification is being attempted using a variety of antioxidants (9). Generally, reagents effective in reducing MbFe IV to MbFe III operate in a stoichiometric manner. Conversely, nitroxide radicals, which shuttle among three oxidation states can detoxify hypervalent metals in a catalytic fashion.
Kinetic Considerations-The predominance of MbFe IV over MbFe III shows that k 1 Ͼ Ͼ k 3 , thus indicating that the reduction of MbFe IV by H 2 O 2 (Reaction 3) is the rate-limiting step in the catalytic cycle of H 2 O 2 dismutation. The rates of O 2 evolution and H 2 O 2 depletion increased about 4 -6-fold in the presence of TPL (see Fig. 1, A and B)  Considering the values of 1.2 ϫ 10 8 , 3.2 ϫ 10 5 , and 1.5 ϫ 10 10 M Ϫ1 s Ϫ1 reported for k 5 , k Ϫ5 , and k 6 (39,40), respectively, the rate of oxygen evolution with an excess of H 2 O 2 would be the same whether TPL ϩ disappears via Reaction 5 or 6.
The higher rate of H 2 O 2 dismutation in the presence of TPL shows that k 4 Ͼ Ͼ k 3 . The time invariance of the concentrations of MbFe IV and TPL provides direct evidence that both reagents act as true catalysts. The intensity of the EPR signal of TPL did not change during the course of the reaction because the steady state concentration of the oxoammonium cation TPL ϩ was below detection. Nevertheless, TPL ϩ was the species mediating the depletion of H 2 O 2 because in the presence of NADH the EPR signal of TPL was lost due to two-electron reduction of TPL ϩ to the respective cyclic hydroxylamine TPL-H via Reaction 7. k 7 NADH ϩ TPL ϩ 3 NAD ϩ ϩ TPL-H REACTION 7 In the presence of NADH alone, the evolution of oxygen was replaced by O 2 consumption because MbFe IV was reduced via Reaction 8, nitroxide accelerates the replenishment of MbFe III and consequently the entire cycle of reactions.
Mechanism of Ferryl Reduction-The combined presence of MbFe IV , H 2 O 2 , and nitroxide did not result in any detectable modification of the EPR spectrum of the nitroxide or the absorption spectrum of the heme iron. Nevertheless, the biphasic behavior of the reduction of MbFe IV in the presence of increasing concentration of nitroxides (Fig. 4B) suggests that ferrylmyoglobin exists in two forms, which decay via Reactions 9 and 10. Such behavior was also observed in the reaction between MbFe IV and Trolox (41).
k 10 MbFe IV * 3 MbFe III REACTION 10 According to this mechanism the decay of MbFe IV to MbFe III is given by rate Equation 1: In the presence of reductant such as TPL, MbFe IV and MbFe IV * can be reduced through Reactions 4 and 4*, respectively. Ϫ d͓MbFe IV ͔ dt ϭ k obs ͓MbFe IV ͔ ϭ ͭ k 9 k 10 ϩ k 9 k 4 * ͓TPL͔ Rate Equation 2 explains the biphasic dependence of the reaction rate constant of the decay of MbFe IV on [nitroxide] as seen in Fig. 4. The spontaneous decay of MbFe IV to MbFe III in the absence of TPL or any other reductant depends on k 9 , k Ϫ9 and k 10 . In the absence of TPL, rate Equation 2 reduces to rate Equation 1, where k obs ϭ k 9 ⅐k 10 /(k Ϫ9 ϩ k 10 ).
In the presence of a sufficiently high [TPL], rate Equation 2 reduces to rate Equation 3: Ϫ d͓MbFe IV ͔ dt ϭ ͕k 9 ϩ k 4 ͓TPL͔͖͓MbFe IV ͔ (Eq. 3) where k obs ϭ k 9 ϩ k 4 ⅐[TPL]. According to this model k 4 , though not k 9 , depends on the type of reductant reacting with MbFe IV . A linear regression analysis of the experimental points at the linear portion of the biphasic curve seen in Fig. 4 enables the evaluation of k 4 for TPL and k 9 as 1.6 ϫ 10 1 M Ϫ1 s Ϫ1 and 1.5⅐10 Ϫ2 s Ϫ1 , respectively. Similarly, analysis of the kinetics of MbFe IV spontaneous decay in the absence of any reductant would allow the evaluation of k Ϫ9 and k 10 .
In the case of TPA, as seen in Fig. 4, linearity was not approached for the dependence of k obs on [TPA], which suggests that for this nitroxide the condition k 4* ⅐[TPA] Ͼ (k Ϫ9 ϩ k 10 ) was not met.
The Catalytic Turnover Number-In the presence of an excess of H 2 O 2 and TPL, the reduction of each MbFe IV to MbFe III is followed by immediate formation of another MbFe IV via Reaction 1 and is accompanied by depletion of a H 2 O 2 molecule and evolution of O 2 . Under such conditions, [MbFe IV ] is continuously replenished and the catalytic turnover number should equal k obs .
Turnover number ϭ ͑rate of H 2 O 2 decay͒/͓MbFe IV ͔ ϭ k obs ϭ k 9 ϩ k 4 ⅐ ͓TPL͔ (Eq. 4) The results displayed in Fig. 2B indicating a turnover number of 0.65 min Ϫ1 for O 2 with 0.4 mM TPL can be compared with k obs ϭ 1.26 min Ϫ1 seen in Fig. 4 for similar experimental conditions.
In conclusion, the present results further elucidate the role played by nitroxides in detoxifying hypervalent metals and enhancing the catalase-like activity of heme proteins. It is possible that in addition to the previously recognized modes of nitroxide action, these mechanisms underlie the protective biological activity of nitroxide antioxidants. Furthermore, nitroxides having a wide variety of analogues with different ring structures and substituents, which confer them with differences in charge and redox behavior might be helpful tools in elucidating the redox reactions of hypervalent heme species. In addition, covalent linking of nitroxides to proteins at different sites is feasible (42) and would provide an interesting chemical system to extend the study of the redox reactions between nitroxides and myoglobin.