Evidence for the role of a peroxidase compound I-type intermediate in. the oxidation of glutathione, NADH, ascorbate, and dichlorofluorescin

Transcription

1 Journal of Biological Chemistry Manuscript No. M3:00054 (Revision of April 2003) Evidence for the role of a peroxidase compound I-type intermediate in the oxidation of glutathione, NADH, ascorbate, and dichlorofluorescin by cytochrome c/h 2 O 2 IMPLICATIONS FOR OXIDATIVE STRESS DURING APOPTOSIS Andrew Lawrence, Clare M. Jones, Peter Wardman and Mark J. Burkitt * Gray Cancer Institute, PO Box 100, Mount Vernon Hospital, Northwood, Middlesex, HA6 2JR, United Kingdom Running title: Oxidizing species from cytochrome c/h 2 O 2 * To whom correspondence should be addressed: Dr. M. J. Burkitt, Gray Cancer Institute, PO Box 100, Mount Vernon Hospital, Northwood, Middlesex, HA6 2JR, United Kingdom. Tel: +44 (0) ; Fax: +44 (0) burkitt@gci.ac.uk 1 The abbreviations used are: cyt c, cytochrome c; DCF, 2,7 -dichorofluorescein; DCFH 2, 2,7 dichorofluorescin; DMPO, 5,5-dimethyl-1-pyrroline N-oxide; DTPA, diethylenetriaminepentaacetic acid; EPR, electron paramagnetic resonance; GSH, glutathione (reduced form); GSSG, glutathione (oxidized form); GtPx, glutathione peroxidase; MNP, 2-methyl-2-nitrosopropane; MnSOD, Mn-containing superoxide dismutase; ROS, reactive oxygen species; XO, xanthine oxidase. 1

2 SUMMARY The release of cytochrome c from mitochondria is a crucial step in apoptosis, resulting in the activation of the caspase proteases. A further consequence of cytochrome c release is the enhanced mitochondrial production of superoxide radicals (O 2 ), which are converted to hydrogen peroxide by Mn-superoxide dismutase. Recently, we showed that cytochrome c is a potent catalyst of 2,7 -dichorofluorescin (DCFH 2 ) oxidation to the fluorescent 2,7 -dichorofluorescein (DCF) by these species, leading to the conclusion that DCF fluorescence is a reflection of cytosolic cytochrome c concentration rather than reactive oxygen species levels (Burkitt and Wardman, 2001, Biochem. Biophys. Res. Commun., 282, ). The oxidant generated from cytochrome c has so far not been identified. Several authors have suggested that the hydroxyl radical ( OH) is generated, but others have discussed the possibility of a peroxidase compound I. By examining the effects of various antioxidants (glutathione, ascorbate, NADH) and hydroxyl-radical scavengers (ethanol, mannitol) on the rate of DCFH 2 oxidation by cytochrome c, together with complementary EPR spin-trapping studies, we demonstrate that the hydroxyl radical is not generated. Instead, our findings suggest the formation of a peroxidase compound-i type intermediate, in which one oxidizing equivalent is present as an oxo-ferryl heme species and the other as the protein-tyrosyl radical previously identified by Barr and colleagues (J. Biol. Chem. 271, , 1996). Competition studies involving spin traps indicated that the oxo-ferryl heme component is the active oxidant. These findings provide an improved understanding of the physico-chemical basis to the redox changes that occur during apoptosis. 2

3 INTRODUCTION The interruption of electron flow to cytochrome c oxidase in the mitochondrial respiratory chain caused by the release of cytochrome c (cyt c) 1 during apoptosis results in an increase in the generation of superoxide radicals (O 2 ) at upstream sites (1). Although superoxide can react with aconitase (2,3), GSH (4,5), and nitric oxide (6,7) within the mitochondrial matrix, the vast majority of the radical is expected to undergo MnSOD-catalyzed disproportionation to hydrogen peroxide and oxygen (8,9). Based on observations made using the probe DCFH 2, which can undergo oxidation to the highly fluorescent dichorofluorescein (DCF), several authors have reported increased reactive oxygen species production in both apoptotic and necrotic cells (10-12). However, the direct reaction between H 2 O 2 (which also arises from O 2 by spontaneous disproportionation) and DCFH 2 is very slow. Oxidation of the probe by H 2 O 2 is much more efficient in the presence of a suitable catalyst, such as a peroxidase, ferrous iron or cyt c (13-17). Peroxynitrite, formed in the reaction between O 2 and nitric oxide, can oxidize DCFH 2 directly (18). Recently, we have demonstrated that cyt c is a particularly potent catalyst of DCFH 2 oxidation in the presence of an enzymatic source of O 2 and H 2 O 2. Whereas a 10-fold increase in the rate of O 2 /H 2 O 2 production resulted in only a very modest increase in the rate of DCFH 2 oxidation, the addition of nanomolar amounts of the cytochrome had a profound stimulatory effect. Thus, we have suggested that the increased rate of DCFH 2 oxidation that is observed in apoptotic cells is a reflection of elevated cytosolic cyt c concentration rather than increased O 2 /H 2 O 2 production (17). The H 2 O 2 resulting from the MnSOD-catalyzed disproportionation of O 2 is expected to undergo GtPx-catalyzed reduction to water, accompanied by the oxidation of GSH to 3

4 GSSG (9,19). Although O 2 can oxidize GSH directly, the rate constant for this reaction ( 200 M -1 s -1 ) is too low for it to compete with the rapid conversion of O 2 to H 2 O 2 by Mn-SOD (5). Due to the critical role of the GSH/GSSG couple in the maintenance of intracellular redox conditions (20), there is currently considerable interest in the implications of the extensive GSH oxidation that occurs during apoptosis as a consequence of the above reactions (21-24). For example, the activity of caspase-3, a key apoptotic enzyme that is activated following cyt c release, is known to be particularly sensitive to intracellular redox conditions, being inactivated by excessive H 2 O 2 (11,25,26). On the other hand, there is evidence to suggest that, at least at low concentrations, reactive oxygen species (ROS) play a necessary role in the induction of apoptosis. For example, through their stimulation of mitochondrial cyt c release, involving the mitochondrial permeability transition, ROS are believed to be responsible for the induction of apoptosis by a diverse range of chemical agents (24,27-29). However, the role of ROS in apoptosis is a contentious issue, with some studies reporting that they are not required (30-32). Indeed, it is now apparent that cyt c release from mitochondria can occur by two distinct mechanisms, of which only one appears to involve the permeability transition (33). Moreover, a recent study found that the oxidation of externalised phosphatidylserine by ROS plays a role in the recognition and clearance of apoptotic cells by macrophages (34). Clearly, in order to elucidate the exact role played by such species in apoptosis and other cellular events, it is of vital importance that the observations from experiments employing indicator probes such as DCFH 2 are interpreted with extreme caution. For example, on the strength of its ability to suppress the oxidation of DCFH 2 in dying cells, it has been suggested that the bcl-2 oncoprotein is 4

5 an antioxidant (10), yet this observation can be accounted for solely by the fact that the protein prevents the release of cyt c from mitochondria (17). In the apoptotic cell, there is also the possibility that cyt c may catalyze the oxidation of GSH (and other molecules) by hydrogen peroxide, which may have implications for the redox-sensitive events that occur downstream of cyt c release (35). The purpose of the investigations reported here was to examine in mechanistic detail, using biomimetic models, the interactions that can occur between cyt c, O 2 /H 2 O 2, DCFH 2, GSH and other redox cofactors in the apoptotic cell. On the basis of our findings, we have proposed a mechanism for the activation of cyt c to an oxidizing species in the presence of O 2 /H 2 O 2 and elucidated the mechanism of its interaction with DCFH 2, GSH, NADH and ascorbate. We propose that these reactions explain many of the redox changes observed in apoptotic cells. Although we accept that there is good evidence that ROS act as signaling species in the induction of apoptosis, our findings suggest that the redox changes reported in the literature occur largely as a consequence of apoptosis, and probably serve to inactivate the process, favoring a shift toward a necrotic mode of cell death. EXPERIMENTAL PROCEDURES Materials Horse heart cytochrome c, ascorbic acid, 2,7 -dichorofluorescein, 2,7 - dichorofluorescin diacetate, ferrous sulfate, NADH, chelating resin ( Chelex 100 ), MOPS buffer, GSH, GSSG, DTPA, DMPO, hypoxanthine, MNP and xanthine oxidase were purchased from Sigma-Aldrich (Poole, England). DCFH 2 was prepared from 2,7 - dichorofluorescin diacetate by alkaline hydrolysis (13). The spin trap DMPO was purified by vacuum distillation (Kugelrohr) and stored under N 2 at 80 C. Sodium dihydrogen 5

6 phosphate and disodium hydrogen phosphate were from BDH (Poole, England). Stock solutions of 0.2 M phosphate buffer and 0.2 MOPS buffer (both ph 7.4), as well as the hypoxanthine stock solution and water used in all dilutions, were treated with chelating resin to remove contaminating metal ions using the batch method (36). Hydrogen peroxide was purchased from Aldrich (Gillingham, England) as a 27.5 % wt solution and standardized by titration against a 0.1 N potassium permanganate standard solution, the concentration of which was routinely checked by titration against sodium oxalate (both from Aldrich)(37). For EPR experiments requiring both ferricytochrome c and ferrocytochrome c (Fig. 8), a 25 mm stock solution of the cytochrome in 0.1 M phosphate buffer (ph 7.4) was divided into two aliquots. To one aliquot was added ascorbic acid (in phosphate buffer, adjusted to ph 7.4) to give final concentrations of cyt c and ascorbic acid of mm and 50 mm, respectively. The other aliquot was diluted to mm with phosphate buffer. Both solutions were then incubated on ice for 2 h, before purification through Micro Bio-Spin columns (Bio-Gel P6), obtained from BioRad (Hemel Hempstead, England). An aliquot of the eluant from the ascorbate-reduced cyt c was then diluted using 200 mm ascorbic acid (in N 2 -purged phosphate buffer) and its concentration determined from the resultant absorption at 550 nm (ε = 27,700 M -1 cm -1 ). The original eluant was then used for reaction with H 2 O 2, as indicated in the legend to Fig. 8B (Fe 2+ - cyt c). For the corresponding reaction using Fe 3+ -cyt c (Fig. 8A), the eluant from the column loaded with cyt c that had not been reduced with ascorbate was used. To ensure that an identical concentration of Fe 3+ -cyt c was used, a small aliquot of the eluant was 6

7 reduced by the addition of solid sodium dithionite and the absorption at 550 nm measured. Fluorescence assay of DCFH 2 oxidation The generation of DCF from DCFH 2 was monitored by fluorescence spectroscopy, as reported previously (17). All incubations contained 10 µm DCFH 2, which was incubated with either 4 mu ml -1 XO and 0.5 mm hypoxanthine (to generate superoxide plus hydrogen peroxide) or simply 250 µm H 2 O 2. Reactions using the XO-hypoxanthine system were performed in 0.1 M MOPS buffer (ph 7.4) containing 1 mm DTPA. Reactions involving bolus H 2 O 2 addition were performed in 0.1 M phosphate buffer, ph 7.4. Other reagents were included as specified in the appropriate figure legends. Reactions were initiated by the final addition of cyt c at the concentrations indicated. Fluorescence values were converted to DCF concentrations using a standard curve prepared using known amounts of DCF. Data shown are representative of at least three independent experiments. Electron paramagnetic resonance spectroscopy EPR Spectra were recorded on a Bruker EMX spectrometer (Bruker UK Ltd, Coventry, England) equipped with a highsensitivity cylindrical cavity (X-band), operating at a modulation frequency of 100 khz and the settings specified in the appropriate figure legends. All reaction mixtures were performed in Eppendorf tubes and transferred to a quartz flat-cell following the final addition of cyt c. The spectrometer field calibration was routinely checked using the signal from a dilute solution of Fremy's salt [a(n) = G (38)] and was accurate to within ± 0.05 G. Hyperfine coupling constants (in gauss; 1 G = 10 4 tesla) were determined from spectral simulations performed using software available through the Internet ( and described elsewhere (39). 7

8 RESULTS To mimic the production of O 2 radicals by cyt c-depleted mitochondria, we used xanthine oxidase (XO) with hypoxanthine as substrate. The incubation of DCFH 2 with XO led to the generation of only a very modest amount of DCF over the 25 min recording period (Fig. 1, trace a). However, the inclusion of 25 nm cyt c resulted in a massive enhancement in the rate of DCFH 2 oxidation (Fig. 1, trace e), as reported previously (17). The ability of cyt c to catalyze DCFH 2 oxidation was severely compromised when GSH was included at a concentration of 7.5 mm (Fig. 1, trace c), which is typical of reported intracellular concentrations (40-42). Moreover, the addition of GSH without the cytochrome led to a moderate increase in rate of DCFH 2 oxidation by XO (Fig. 1, trace b), which accounted for essentially all of the DCF generated when the cyt c was added together with the thiol (Fig. 1, trace c). Much higher concentrations of cyt c were needed to stimulate DCFH 2 oxidation in the presence of GSH. For example, at a concentration of 250 nm, which is representative of the concentrations needed to trigger the final stages of apoptosis (43-45), the cytochrome caused almost a 3-fold increase in the yield of DCF after 25 min (Fig. 1, traces b and d). Since GSH is known to undergo oxidation to GSSG during apoptosis (1), it was considered appropriate to examine the ability of cyt c to oxidize DCFH 2 in the presence of both GSH and GSSG, at concentrations reflecting those achieved in apoptotic cells following release of the cytochrome into the cytosol. Cellular levels of GSH and GSSG are often reported in terms of a redox potential (1,20,46). This gives a more accurate reflection of the prevailing redox environment in the cell than the simple ratio [GSH]/[GSSG] (even though the couple is not at thermodynamic equilibrium due to the 8

9 participation of GSH and GSSG in other reactions). The reduction potential for the GSSG/2GSH couple at ph 7.0 and 40 C is 0.24 ± 0.01V (47), from which a value of 0.26 V is obtained at ph 7.4 (48). It is reasonable to assume that the glutathione solution used in the experiments reported in Fig. 1 initially contained per cent GSSG (47). Using the Nernst equation (49), E h = E + RT [GSSG] ln 2 nf [GSH] in which E is the reduction potential for the GSSG/GSH couple at ph 7.4 and 40 C under standard conditions of concentration (approximating to 1 M GSH and 1 M GSSG) and E h is the corresponding potential (half-cell potential) at particular values of [GSH] and [GSSH], it can be shown that the E h of a 7.5 mm glutathione solution that has undergone 1 % oxidation to GSSG is approximately 0.27 V. We examined the ability of cyt c to catalyze DCFH 2 oxidation in the presence of glutathione that had undergone 25, 50 and 75 % oxidation. This was achieved by mixing stock solutions of GSH and GSSG, keeping the total concentration of GSH equivalents ([GSH] + 0.5[GSSG]) constant at 7.5 mm. The resultant E h values estimated for these GSH/GSSG mixtures were calculated to be 0.22 V (25 % GSH oxidation), 0.20 V (50 % GSH oxidation) and 0.17 V (75 % oxidation), which, together with the 1 % oxidised solution, broadly encompass the range of values reported for the GSSG/2GSH couple in cells undergoing proliferation ( 0.24 V), differentiation ( 0.20 V) and apoptosis ( 0.17 V) (20,46). As shown in Fig. 2, the ability of cyt c to stimulate the oxidation DCFH 2 was directly related to the extent of glutathione oxidation. Indeed, under conditions that mimic the redox 9

10 environment of the apoptotic cell, cyt c caused a several-fold increase in the rate of DCFH 2 oxidation (Fig. 2D). The ability of GSH, at a fixed concentration of 7.5 mm, to inhibit the generation of DCF could be overcome by increasing the concentration of DCFH 2 (Fig. 3). This suggests that GSH and DCFH 2 compete for oxidation by a common, cyt c-derived oxidant. The inhibition of DCFH 2 oxidation seen in Fig. 2A is, therefore, a reflection of the failure of the probe, at 10 µm concentration, to compete with GSH (7.5 mm) for reaction with this oxidant. Separate experiments showed that high concentrations of GSSG also inhibit DCFH 2 oxidation, which is believed to be due to its inhibition of XO, presumably via mixed disulfide formation (data not shown). It was noted above that GSH, in the absence of cyt c, actually stimulated the oxidation of DCFH 2, albeit at a modest rate (Fig. 1, trace b), accounting for most of the DCF generated when the cytochrome was added together with 7.5 mm GSH (Fig. 1, trace c). Additional experiments showed that, at this concentration of DCFH 2 (10 µm), the maximum rate of DCF oxidation occurred at a GSH concentration of 1mM, beyond which the addition of further thiol had no effect (data not shown). The stimulation of DCFH 2 oxidation by GSH alone was suppressed by superoxide dismutase (not shown), suggesting the direct involvement of XO-derived superoxide radicals (5). It was next sought to identity of the oxidant responsible for the oxidation of DCFH 2 in the presence of cyt c. The most likely candidates are the hydroxyl radical ( OH), an oxoferryl heme species (cyt c-fe IV =O), or an oxo-ferryl haem species with a second oxidizing equivalent existing as a porphyrin or protein radical-cation (cyt + c-fe IV =O), as encountered in the compound I form of peroxidases (50-53). The fact that DCFH 2 10

11 oxidation occurs at all in the presence of a fold excess of organic buffer and a 50-fold excess of hypoxanthine argues against the role of the free OH radical: if the rate constant for the reaction of OH with these substrates is assumed to be 10 9 M -1 s -1 (54), then it is difficult to see how DCFH 2 could ever compete for oxidation by the radical. In order to investigate DCFH 2 oxidation in the absence of competing organic substrates (including DTPA), additional reactions were carried out in phosphate buffer, using a bolus of H 2 O 2 in place of XO. The incubation of DCFH 2 with H 2 O 2 and cyt c in phosphate buffer resulted in efficient DCF generation (Fig. 4, trace e). As seen with the XO system in MOPS buffer, the addition of GSH resulted in the marked inhibition of DCFH 2 oxidation (Fig. 4, trace b), as did ascorbate and NADH (Fig. 4, traces a and c, respectively). In contrast, the classical hydroxyl-radical scavengers mannitol and ethanol had essentially no effect on the rate of DCF generation (Fig. 4, traces d and f, respectively). In order to provide direct evidence for the interception of an oxidizing species by the above scavengers, complementary electron paramagnetic resonance (EPR) studies were performed. With the exception of ascorbate, which is oxidized to a relatively stable radical that is detectable by direct EPR spectroscopy (55), the spin trap 5,5-dimethyl-1- pyrroline N-oxide (DMPO) was employed. Upon reaction with a short-lived radical (R ), DMPO is converted to a nitroxide radical adduct (Reaction 1), which is relatively longlived and yields a spectrum that can be used to identify the trapped radical (56). N + R N H R (1) O O DMPO DMPO/ R 11

12 The addition of cyt c together with H 2 O 2 to DMPO caused the appearance of a prominent signal from 5,5-dimethyl-1-pyrrolidone-2-oxyl (DMPOX), the three-electron oxidation product of the spin trap (57). Generation of the nitroxide was dependent on the presence of both cyt c and H 2 O 2, with only very weak, background signals observed when either reactant was omitted (Fig. 5, spectra A to D). With cyt c alone, the background signals consisted of the DMPO adducts of the hydroxyl radical (DMPO/ OH) and a carbon-centered species (six-line signal). The DMPO/ OH adduct is believed to result from the nucleophilic addition of water to the spin trap, preceded or followed by oxidation by the ferricytochrome (57). The identity of the carbon-centered adduct is not known, but is presumed to result from degradation of either the spin trap or cytochrome, as no other organic material was present. The slight increase in the weak, background signal from DMPO/ OH observed following the addition of H 2 O 2 alone to DMPO is believed to reflect the catalysis of OH generation by traces of iron or copper ions, despite treatment of the buffer with chelating resin. The cyt c/h 2 O 2 system was next compared with the Fe 2+ /H 2 O 2 system (the Fenton reaction), which is known to generate the OH radical (58). As shown in Fig. 5, spectrum E, the addition of Fe 2+ to H 2 O 2 resulted in the appearance of a prominent signal from the DMPO/ OH adduct, which is clearly different from the corresponding reaction using cyt c (Fig. 5, spectrum A). When GSH was included, the signal from DMPOX was replaced by a prominent signal from the glutathionyl radical adduct, DMPO/ SG (Fig. 6, spectrum A), indicating one-electron oxidation of the thiol to the glutathionyl radical (5). This signal was not observed when H 2 O 2 was omitted from the reaction system (not shown). When NADH was incubated with cyt c and H 2 O 2, a signal from the superoxide adduct (DMPO/ OOH) 12

13 was detected, together with a signal from DMPO/ OH, which is formed during the decomposition of DMPO/ OOH (Fig. 6, spectrum B) (5). This is believed to reflect the one-electron oxidation of NADH, followed by the rapid transfer of an electron to molecular oxygen from the resultant NAD radical (k = M -1 s -1 )(59,60). When H 2 O 2 was omitted from this reaction system, only trace levels of the same signals were observed (not shown), presumably due to the very slow, direct oxidation of NADH by cyt c and subsequent oxygen reduction by NAD and the reduced cytochrome. In contrast to GSH and NADH, ethanol was not oxidized by cyt c/h 2 O 2. Instead, only the signal from DMPOX, was observed, even though its formation was suppressed by the addition of excess ethanol (Fig. 6, spectra C and D). In contrast, when ethanol was included in the Fe 2+ /H 2 O 2 system, a prominent signal from the α-hydroxyethyl radical adduct was detected (Fig. 6, spectrum E), reflecting oxidation of the alcohol by the OH radical (58). These findings provide further support for the notion that the active oxidant generated from cyt c/h 2 O 2 is not the OH radical, which would have oxidized ethanol to the α- hydroxyethyl radical, forming the DMPO/ CH(OH)CH 3 adduct, as confirmed using the Fe 2+ /H 2 O 2 system. Instead, it is proposed that a less powerful, more selective oxidant is formed. Indeed, ascorbate was also oxidized by cyt c/h 2 O 2, as evidenced by the increase in the intensity of the signal from the ascorbate radical (Fig. 7, spectrum A). The background signal, which is invariably seen in ascorbate solutions, is believed to reflect basal autoxidation, which would be catalyzed by contaminating metal ions (Fe or Cu), despite treatment with chelating resin (Fig. 7, spectrum B)(61). Hydrogen peroxide alone caused the doubling of this signal (Fig. 7, spectrum D). 13

14 Having ruled out the generation of OH, our attention turned to the peroxidase-type properties of cyt c. Using the spin trap 2-methyl-2-nitrosopropane (MNP), Mason and colleagues have reported the trapping and EPR detection of a tyrosyl radical on cyt c following its reaction with H 2 O 2 (52,62). To explore in greater depth the mechanism of cyt c activation by H 2 O 2, we examined the ability of both the ferrous and ferric forms of the protein to form the tyrosyl radical. When H 2 O 2 was added to Fe 3+ -cyt c in the presence of MNP, an EPR spectrum characteristic of an immobilized nitroxide was detected, indicating the formation of a spin adduct between MNP and a protein-centered radical (Fig. 8). Non-specific proteolytic digestion of the sample with pronase resulted in the sharpening of the spectral features, revealing hyperfine coupling to a nitrogen nucleus (a N = 15.5 G)(data not shown). This isotropic spectrum is identical to that reported by Mason and colleagues, which they identified unambiguously as a tyrosyl radical adduct on the basis of superhyperfine coupling detected by use of the deuterated spin trap (d 9 - MNP)(52,62). At low concentrations of H 2 O 2, the signal detected using Fe 3+ -cyt c was always stronger than that observed using the same concentration of the ferrocytochrome, Fe 2+ -cyt c. However, the difference in the yield of adduct became less marked as the concentration of the peroxide was increased (Fig. 8). Although MNP can clearly intercept the cyt c-tyrosyl radical, the spin trap failed to prevent DCFH 2 oxidation by the cyt c/h 2 O 2 couple. In fact, MNP enhanced the rate of DCF formation (Fig. 9, trace e). In contrast, the spin trap DMPO afforded DCFH 2 a degree of protection from oxidation by cyt c/h 2 O 2 (Fig. 9, traces a to c). 14

15 DISCUSSION We reported recently that cyt c is a potent catalyst of DCFH 2 oxidation in the presence of an enzymatic source of superoxide radicals and H 2 O 2, suggesting that the increased DCF production often seen in apoptotic cells can be attributed to an increased cytosolic concentration of the cytochrome rather than an actual change in reactive oxygen species status (17). Mitochondria isolated from apoptotic cells have, in fact, been demonstrated to generate superoxide at enhanced rates, reflecting the disruption of electron transfer caused by cyt c release (1). However, as a reporter of ROS, DCF is poorly sensitive to changes in the concentration of superoxide and H 2 O 2, the product of its disproportionation (17). In the present study we have extended these investigations and shown that the rate of DCFH 2 oxidation is highly sensitive to GSH status and, therefore, cellular redox conditions. Specifically, we have shown that GSH and DCFH 2 compete for oxidation by the same oxidizing species. Although the ability of the cyt c/h 2 O 2 couple to initiate biomolecular oxidation has been known for several years, the identity the oxidizing species has not been established (50,51,63-65). Several workers have suggested formation of the hydroxyl radical through redox cycling, involving the initial reduction of the ferricytochrome by either superoxide or hydrogen peroxide (Reactions 2 to 4) (50,51,63). cyt c-fe III + O 2 cyt c-fe II + O 2 (2) cyt c-fe III + H 2 O 2 cyt c-fe II + O H + (3) cyt c-fe II + H 2 O 2 cyt c-fe III + OH + OH (4) 15

16 However, whereas cyt c/h 2 O 2 was found to oxidize GSH, NADH and ascorbate to their respective radicals, the oxidizing intermediate showed no reactivity toward ethanol. The fact that no DMPO/ CH(OH)CH 3 was detected suggests that the initially-formed DMPO/ OH adduct was generated by inverted spin-trapping, rather than by the trapping of free OH. This involves the initial one-electron oxidation of the spin trap to a radical cation, followed by the nucelophilic addition of water (57,66,67). The failure to detect the initially-formed DMPO/ OH adduct is believed to reflect its rapid oxidation to the corresponding nitrone and, ultimately, DMPOX (Scheme 1). In a related study, we reported the oxidation of DMPO to DMPOX by Ce IV without the detection of DMPO/ OH (57). Support for the proposal that this latter reaction involves the initial formation of DMPO/ OH is provided by the work of Clémont and colleagues who, using the DMPO analogue 5-(diethylphosphoryl)-5-methyl-1-pyrroline N-oxide (DEPMPO), observed, together with DEPMPOX, a weak signal from DEPMPO/ OH (68). Given that oxygen-radical adducts of DMPO are less stable than those of DEPMPO, we consider it reasonable to suggest that, as in the Ce IV /DMPO system, DMPO/ OH is an intermediate in the oxidation of DMPO by cyt c/h 2 O 2. The finding that ethanol suppressed the signal from DMPOX is proposed to reflect its competition with water for nucleophilic addition to the DMPO radical cation, resulting in the generation of an EPR-silent nitrone (Scheme 1). An alternative explanation for the ability of ethanol to suppress DMPOX formation, without generation of the α-hydroxyethyl radical, is that the alcohol may undergo a direct, two-electron oxidation to acetaldehyde. However, this seems unlikely because ethanol failed to protect DCFH 2 from oxidation by cyt c/h 2 O 2, which would be the case had the alcohol reacted directly 16

17 with the oxidant generated from the cyt c/h 2 O 2 couple. Moreover, the finding that very high concentrations of ethanol are required to suppress DMPOX formation is supportive of competition between the alcohol and water for nucleophilic addition to the DMPO radical cation. As mentioned above, Mason and colleagues have recently reported the spin trapping of a protein tyrosyl radical on H 2 O 2 -activated cyt c (52,62). In addition to considering the possible role of the OH radical in the formation of this species, these workers discussed the possible involvement of a peroxidase compound I-type intermediate, in which one oxidizing equivalent is stored as an oxo-ferryl haem (Fe IV =O) species and the other as a porphyrin π-radical cation (52,53). Rapid electron transfer from the protein to porphyrin moiety would result in generation of the tyrosyl radical. Although such reactions have been demonstrated to occur during the interaction of H 2 O 2 with metmyoglobin (69-71), there is no spectroscopic evidence for the formation of an oxo-ferryl intermediate on cyt c (52). This may be because the heme iron in cyt c is hexacoordinate, with no coordinated water molecule that can be displaced by H 2 O 2. However, as discussed by Mason and colleagues, compound I formation from cyt c may still be possible because its sixth ligand is a methionine residue that is readily displaced (52). Our observation that the amount of cyt c tyrosyl-radical trapped by MNP when added to H 2 O 2 is approximately halved when the ferrocytochrome is used instead of the ferricytochrome (except when [H 2 O 2 ]>>[cyt c]) indicates that cyt c-fe III is at least one reaction step closer to the tyrosyl radical than cyt c-fe II. This is consistent with the formation of a compound I-type species according to Reaction 5. cyt c-fe III + H 2 O 2 cyt + c-fe IV =O + H 2 O (5) 17

18 The cyt c-based radical cation shown in Reaction 5, which may be initially located on the porphyrin ring, would undergo rapid deprotonation to give the neutral tyrosyl radical. When starting with the ferrocytochrome, one-half equivalent of H 2 O 2 would be consumed in the initial oxidation up to the ferric state. This would explain why the difference in yield of the tyrosyl radical adduct is less prominent when the peroxide is present in large excess. Had generation of the OH radical been a requirement for generation of the tyrosyl radical, the yield of the adduct would have been greater in the cyt c-fe II /H 2 O 2 system. The observation that DCFH 2 oxidation is not suppressed by MNP suggests that the cyt c-based tyrosyl radical is not responsible for DCFH 2 oxidation. The reason why MNP had, in fact, the opposite effect, namely the stimulation of DCFH 2 oxidation, remains to be determined. One possible explanation is that cyt c/h 2 O 2 may oxidize MNP to a nitroso radical cation (72), which would be expected to oxidize DCFH 2, but this will require experimental validation. In contrast, the finding that DCFH 2 oxidation is partially suppressed by DMPO indicates that the probe and the spin trap compete for oxidation by a common cyt c-dervived oxidant, which cannot be the tyrosyl radical. Moreover, the fact that GSH and NADH compete with DMPO for oxidation by cyt c/h 2 O 2 (indicated by their prevention of DMPOX generation), and that GSH, NADH and ascorbate prevent DCF generation from DCFH 2, indicates that they too react with this oxidant. On this basis, we propose that DCFH 2, DMPO, GSH, NADH, and ascorbate are oxidized by the oxo-ferryl heme component of cyt c compound I, and that the tyrosyl radical serves as a less reactive radical sink, as seen with horse metmyoglobin (71) (Scheme 2). We cannot, of course, rule out some oxidation of GSH, ascorbate, NADH, and even DCFH 2, 18

19 by the tyrosyl radical. This is because the one-electron reduction potentials (E at ph 7) for the GS,H + /GSH, ascorbate radical,h + /ascorbate monoanion, and NAD,H + /NADH couples [approximately 0.8, 0.3, and 0.3 V, respectively (73-75)] are lower than the expected value for the tyrosyl radical [ 0.9 V, the value for the tyrosyl radical,h + /tyrosine couple (76)]. Without knowledge of the reduction potential for the DCFH,H + /DCFH 2 couple, we cannot say whether oxidation by the cyt c tyrosyl radical is thermodynamically feasible. However, the finding that DMPO partially protects DCFH 2 from oxidation, but is not expected to be oxidized by the tyrosyl radical [the E for DMPO + /DMPO is about 1.9 V in acetonitrile-tetrabutylammonium hexafluorophosphate (67)], suggests this is unlikely. The fact that an oxo-ferryl heme species on cyt c has never been detected by optical spectroscopy is presumed to reflect the instability of the species. In the absence of a suitable substrate, it is expected the species will be quenched by rapid electron transfer from the protein moiety. In addition to establishing that the oxidant generated during the activation of cyt c by H 2 O 2 is a compound 1-type intermediate and not the hydroxyl radical, this investigation provides important mechanistic insights into the redox reactions that can be expected to occur in apoptotic cells. These redox changes are initiated by the release of cyt c from the mitochondrial electron-transport chain, resulting in increased superoxide generation from upstream sites. Recently, we have established that the second-order rate constant for the oxidation of GSH by the superoxide radical is ~ 200 M -1 s -1, which means that mitochondrial Mn-SOD will always out-compete GSH for reaction with superoxide, resulting in essentially no direct GSH oxidation by the radical (5). Thus, the extensive oxidation of GSH to GSSG that is frequently reported in apoptotic cells can be attributed 19

20 solely to the removal of H 2 O 2 (produced by Mn-SOD) by glutathione peroxidase. Initially, the cyt c compound I will be largely quenched by GSH, but as this is converted to GSSG the species will begin to attack other targets, including ascorbate and NADH (as well as DCFH 2 in experimental systems). It is well known that the activity of caspase-3, a key enzyme in apoptosis, is very sensitive to redox conditions (11,25,26). The inactivation of caspase-3 by H 2 O 2 may be of little consequence once the protease has cleaved its target proteins (e.g., polyadp-ribose polymerase and ICAD, the inhibitor of the caspase-activated deoxyribonuclease). However, in cells having already disturbed redox conditions for other reasons, cyt c compound I could inactivate the protease before apoptosis is fully underway (e.g., at the level of the zymogen), thereby preventing cell death. For example, hydrogen peroxide and/or superoxide-generating xenobiotics are known to inhibit apoptosis by their inactivation of caspase-3 (11). Moreover, there is evidence to suggest that cancerous and pre-cancerous cells are under persistent oxidative stress (77-79), which could contribute to chemo-resistance through the cyt c-dependent reactions revealed herein. Finally, from a methodological standpoint, it is evident that the level of DCF fluorescence in apoptotic cells is a function of both free [cyt c] and [GSH]/[GSSG], being relatively insensitive to superoxide and hydrogen peroxide levels. Acknowledgments The authors are grateful to Drs. M. Naylor and P. Thomson for assistance with DMPO purification.this work was supported by Cancer Research UK Programme Grant C134/A

21 References 1. Cai, J., and Jones, D. P. (1998) J. Biol. Chem. 273, Flint, D. H., Tuminello, J. F., and Emptage, M. H. (1993) J. Biol. Chem. 268, Hausladen, A., and Fridovich, I. (1994) J. Biol. Chem. 269, Wefers, H., and Sies, H. (1983) Eur. J. Biochem. 137, Jones, C. M., Lawrence, A., Wardman, P., and Burkitt, M. J. (2002) Free Radic. Biol. Med. 32, Huie, R. E., and Padmaja, S. (1993) Free Rad. Res. Commun. 18, Goldstein, S., and Czapski, G. (1995) Free Radic. Biol. Med. 19, Turrens, J. F. (1997) Bioscience Reports 17, Cadenas, E., and Davies, K. J. A. (2000) Free Radic. Biol. Med. 29, Kane, D. J., Sarafian, T. A., Anton, R., Hahn, H., Gralla, E. B., Valentine, J. S., Örd, T., and Bredesen, D. E. (1993) Science 262, Samali, A., Nordgren, H., Zhivotovsky, B., Peterson, E., and Orrenius, S. (1999) Biochem. Biophys. Res. Commun. 255, Higuchi, Y., and Yoshimoto, T. (2002) Arch. Biochem. Biophys. 400, LeBel, C. P., Ischiropoulos, H., and Bondy, S. C. (1992) Chem. Res. Toxicol. 5, Royall, J. A., and Ischiropoulos, H. (1993) Arch. Biochem. Biophys. 302, Zhu, H., Bannenberg, G. L., Moldéus, P., and Shertzer, H. G. (1994) Arch. Toxicol. 68,

22 16. Rota, C., Chignell, C. F., and Mason, R. P. (1999) Free Radic. Biol. Med. 27, Burkitt, M. J., and Wardman, P. (2001) Biochem. Biophys. Res. Commun. 282, Kooy, N. W., Royall, J. A., and Ischiropoulous, H. (1997) Free Rad. Res. 27, Chance, B., Sies, H., and Boveris, A. (1979) Physiol. Rev. 59, Schafer, F. Q., and Buettner, G. R. (2001) Free Radic. Biol. Med. 30, Bustamate, J., Tovar-B, A., Montero, G., and Boveris, A. (1997) Arch. Biochem. Biophys. 337, Hampton, M. B., and Orrenius, S. (1998) BioFactors 8, Voehringer, D. W. (1999) Free Radic. Biol. Med. 27, Chandra, J., Samali, A., and Orrenius, S. (2000) Free Radic. Biol. Med. 29, Hampton, M. B., and Orrenius, S. (1997) FEBS Lett. 414, Hampton, M. B., Stamenkovic, I., and Winterbourn, C. C. (2002) FEBS Lett. 517, Atlante, A., Calissano, P., Bobba, A., Azzariti, A., Marra, E., and Passarella, S. (2000) J. Biol. Chem. 275, Carmody, R. J., and Cotter, T. G. (2001) Redox Rep. 6, Yang, J. C., and Cortopassi, G. A. (1998) Free Radic. Biol. Med. 24, Jacobsen, M. D., and Raff, M. C. (1995) Nature 374,

23 31. Muschel, R. J., Bernhard, E. J., Garza, L., McKenna, W. G., and Kock, C. J. (1995) Cancer Res. 55, Jacobsen, M. D. (1996) Trends Biochem. Sci. 21, Gogvadze, V., Robertson, J. D., Zhivotovsky, B., and Orrenius, S. (2001) J. Biol. Chem. 276, Kagan, V. E., Gleiss, B., Tyurina, Y. Y., Tyurin, V. A., Elenström-Magnusson, C., Liu, S.-X., Serinkan, F. B., Arroyo, A., Chandra, J., Orrenius, S., and Fadeel, B. (2002) J. Immmunol. 169, Pan, Z., Voehringer, D. W., and Meyn, R. E. (1999) Cell Death and Differentiation 6, Buettner, G. R. (1988) J. Biochem. Biophys. Methods 16, Belcher, R., and Nutten, A. J. (1970) Quantitative Inorganic Analysis, 3rd Ed., Butterworths, London 38. Faber, R. J., and Fraenkel, G. K. (1967) J. Chem. Phys. 47, Duling, D. (1994) J. Magn. Reson. B 104, Wahlländer, A., Soboll, S., and Sies, H. (1979) FEBS Lett. 97, Meister, A. (1983) in Functions of Glutathione: Biochemical, Physiological, Toxicological and Clinical Aspects (Larson, A., Orrenius, S., Holmgren, A., and Mannervik, B., eds), pp. 1-22, Raven, New York 42. Bellomo, G., Vairetti, M., Stivala, L., Mirabelli, F., Richelmi, P., and Orrenius, S. (1992) Proc. Natl. Acad. Sci. USA 89, Li, P., Nijhawan, D., Budihardjo, I., Srinivasula, S. M., Ahmad, M., Alnemri, E. S., and Wang, X. (1997) Cell 91,

24 44. Zhivotosky, B., Orrenius, S., Brustugun, O. T., and Døskeland, S. O. (1998) Nature 391, Samali, A., Cai, J., Zhivotovsky, B., Jones, D. P., and Orrenius, S. (1999) EMBO J. 18, Kirlin, W. G., Cai, J., Thompson, S. A., Diaz, D., Kavanagh, T. J., and Jones, D. P. (1999) Free Radic. Biol. Med. 27, Rost, J., and Rapoport, S. (1964) Nature 201, Jones, D. P., Carlson, J. L., Mody, Jr., V. C., Cai, J., Lynn, M. J., and Sternberg, Jr., P. (2000) Free Radic. Biol. Med. 28, Wardman, P. W. (1989) J. Phys. Chem. Ref. Data 18, Radi, R., Turrens, J. F., and Freeman, B. A. (1991) Arch. Biochem. Biophys. 288, Ferri, A., and Calza, R. (1995) Biochem. Mol. Biol. Int. 35, Barr, D. P., Gunther, M. R., Deterding, L. J., Tomer, K. B., and Mason, R. P. (1996) J. Biol. Chem. 271, Dunford, H. B. (1987) Free Radic. Biol. Med. 3, Buxton, G. V., Greenstock, C. L., Helman, W. P., and Ross, A. B. (1988) J. Phys. Chem. Ref. Data 17, Ohnishi, T., Yamazaki, H., Iyanagi, T., Nakamura, T., and Yamazaki, I. (1969) Biochim. Biophys. Acta 172, Buettner, G. R. (1987) Free Radic. Biol. Med. 3, Jones, C. M., and Burkitt, M. J. (2002) J. Chem. Soc., Perkin Trans. 2, Burkitt, M. J. (1993) Free Radic. Res. Commun. 18,

25 59. Willson, R. L. (1970) J. Chem. Soc., Chem. Commun., Land, E. J., and Swallow, J. (1971) Biochim. Biophys. Acta 234, Buettner, G. R., and Jurkiewicz, B. A. (1996) Radiat. Res. 145, Qian, S. Y., Chen, Y.-R., Deterding, L. J., Fann, Y. C., Chignell, C. F., Tomer, K. B., and Mason, R. P. (2002) Biochem. J. 363, Cadenas, E., Varsavsky, A. I., Boveris, A., and Chance, B. (1980) FEBS Lett. 113, Cadenas, E., Boveris, A., and Chance, B. (1980) Biochem. J. 188, Harel, S., and Kanner, J. (1988) Free Rad. Res. Commun. 5, Eberson, L., and Nilsson, M. (1993) Acta Chem. Scand. 47, Eberson, L. (1994) J. Chem. Soc., Perkin Trans. 2, Clémont, J.-L., Gilbert, B. C., Rockenbauer, A., and Tordo, P. (2001) J. Chem. Soc., Perkin Trans. 2, Gunther, M. R., Tschirret-Guth, R. A., Witkowska, H. E., Fann, Y. C., Barr, D. P., Ortiz de Montellano, P. R., and Mason, R. P. (1998) Biochem. J. 330, Egawa, T., Shimada, H., and Ishimura, Y. (2000) J. Biol. Chem. 45, Gunther, M. R., Sturgeon, B. E., and Mason, R. P. (2000) Free Radic. Biol. Med. 28, Masui, M., Nose, K., Ohmori, H., and Sayo H. (1982) J. Chem. Soc., Chem. Commun., Wardman, P. (1995) in Biothiols in Health and Disease (Packer, L., and Cadenas, E., eds), pp. 1-19, Marcel Dekker, New York 74. Buettner, G. R. (1993) Arch. Biochem. Biophys. 300,

26 75. Farrington, J. A., Land, E. J., and Swallow, J. (1980) Biochim. Biophys. Acta 590, DePelippis, M. R., Murthy, C. P., Faraggi, M., and Klapper, M. H. (1989) Biochemistry 28, Szatrowski, T. P., and Nathan, C. F. (1991) Cancer Res. 51, Rose, M. L., Rivera, C. A., Bradford, B. U., Graves, L. M., Cattley, R. C., Schoonhoven, R., Swenberg, J. A., and Thurman, R. G. (1999) Carcinogenesis 20, Szaleczky, E., Pronai, L., Nakazawa, H., and Tulassay, Z. (2000) J. Clin. Gastroenterol. 30,

27 FIGURE LEGENDS FIG. 1. GSH suppresses the catalysis of DCFH 2 oxidation by cyt c. DCFH 2 (10 µm) was incubated at 37 C with 4 mu ml -1 XO, 0.5 mm hypoxanthine and 1 mm DTPA in 0.1 M MOPS buffer, ph 7.4, with no further additions (trace a) or in the presence of 7.5 mm GSH (trace b), 25 nm cyt c plus 7.5 mm GSH (trace c), 250 nm cyt c plus 7.5 mm GSH (trace d), or 25 nm cyt c alone (trace e). DCF generation was monitored fluorimetrically. FIG. 2. The rate of cyt c-catalyzed DCFH 2 oxidation is modulated by the GSSG/GSH reduction potential. The ability of cyt c to catalyze the oxidation of DCFH 2 was examined in the presence of GSH and GSSG, mixed in varying proportions to give prevailing reduction potentials of 0.27 V (A), 0.22 V (B), 0.20 V (C), and 0.17 V (D). The concentration of total glutathione equivalents ([GSH] + 0.5[GSSG]) was held constant at 7.5 mm. Incubations, conducted at 37 C in ph 7.4 MOPS buffer, also contained 10 µm DCFH 2 (10 µm), 4 mu ml -1 XO, 0.5 mm hypoxanthine, 1 mm DTPA with no further additions (traces a) or in the presence of 25 nm cyt c (traces b), 100 nm cyt c (traces c), or 250 nm cyt c (traces d). Control incubations contained 250 nm cyt c, but no XO (traces e). DCF generation was monitored fluorimetrically. FIG. 3. DCFH 2 and GSH compete for oxidation by cyt c. DCFH 2 and GSH were incubated at 37 C with 25 nm cyt c, 4 mu ml -1 XO, 0.5 mm hypoxanthine and 1 mm DTPA in 0.1 M MOPS buffer, ph 7.4. The concentration of GSH was held constant at 27

28 7.5 mm while that of DCFH 2 was either 10 µm (trace a), 50 µm (trace b), or 100 µm (trace c). DCF generation was monitored fluorimetrically. FIG. 4. Effects of radical-scavenging antioxidants on cyt c-catalyzed DCFH 2 oxidation. DCFH 2 (10 µm) was incubated at 37 C with 250 nm cyt c, 250 µm hydrogen peroxide, and 0.1 M phosphate buffer, ph 7.4, alone (trace e), or with the addition of either 10 mm ascorbic acid (trace a), 10 mm GSH (trace b), 10 mm NADH (trace c), 10 mm mannitol (trace d), or 10 mm ethanol (trace f). DCF generation was monitored fluorimetrically. FIG. 5. EPR spectra observed during oxidation of the spin trap DMPO by cyt c/ H 2 O 2. A, 0.1 M DMPO plus 100 µm cyt c and 1 mm H 2 O 2. B, 0.1 M DMPO plus 100 µm cyt c. C, 0.1 M DMPO plus 1 mm H 2 O 2. D, 0.1 M DMPO alone. E, 0.1 M DMPO plus 100 µm FeSO 4, 1 mm EDTA and 1 mm H 2 O 2. All reactions were performed in 0.1 M phosphate buffer, ph 7.4, with spectral recording commenced after 5 min. The spectrum in A is assigned to the DMPO oxidation product DMPOX [a(n) = 7.2 G, a(β 2Η) = 4.1 G]. The 4-line signal (1:2:2:1) present in spectrum E is assigned to the hydroxyl radical adduct of DMPO, DMPO/ OH [a(n) = 15.0 G, a(β-h) = 14.6 G]. The instrument settings were: center field, 3480 G; sweep width, 80 G; modulation amplitude, 1 G; microwave power, 20 mw; sweep time; 42 s; time constant 10.2 ms; receiver gain, ; number of accumulated scans, 4. For clarity, spectra B, C, and D are also shown expanded 10-fold in the y-axis. 28

29 FIG. 6. EPR spectra of DMPO-derived nitroxides generated by cyt c/h 2 O 2 in the presence GSH, NADH, and ethanol. A, 100 µm cyt c, 1 mm H 2 O 2, and 5 mm GSH. The 4-line spectrum is assigned to the glutathionyl radical adduct of DMPO [a(n) = 15.1 G, a(β-h) = 16.2 G, a(γ-2h) = 0.6 G]. B, 100 µm cyt c, 1 mm H 2 O 2, and 25 mm NADH. The spectrum consists of signals from the superoxide radical adduct of DMPO [a(n) = 14.1 G, a(β-h) = 11.3 G, a(γ-h) = 1.2 G] and the hydroxyl radical adduct [a(n) = 15.0 G, a(β-h) = 14.6 G]. C and D, 100 µm cyt c, 1 mm H 2 O 2, and either 5 mm or 5 M ethanol, as indicated. Both spectra are assigned to the DMPO oxidation product DMPOX [a(n) = 7.2 G, a(β 2Η) = 4.1 G]. E, 100 µm FeSO 4, 1 mm EDTA, 1 mm H 2 O 2, and 5 M ethanol. The prominent six-line signal is attributed to the α-hydroxyethyl radical adduct of DMPO [a(n) = 15.8 G, a(β-h) = 22.8 G]. All reactions were performed in 0.1 M phosphate buffer, ph 7.4, containing 0.1 M DMPO, with spectral recording commenced after 5 min. The instrument settings used were identical to those given in Fig. 5. FIG. 7. Oxidation of ascorbic acid by cyt c/h 2 O 2. A, 25 µm cyt c, 250 µm H 2 O 2, and 1 mm ascorbate. The doublet EPR signal [a(h) = 1.7 G] is from the ascorbate radical. B, 1 mm ascorbate alone. C, as A, but without H 2 O 2. D, as A, but without cyt c. All reactions were performed in 0.1 M phosphate buffer, ph 7.4, with spectral recording commenced after 5 min. The instrument settings were identical to those given in Fig. 5. FIG. 8. Generation of the cyt c-centered tyrosyl radical by H 2 O 2. A, 835 µm Fe 3+ - cyt c was added to H 2 O 2, at the concentrations indicated, in 50 mm phosphate buffer, ph 7.4, containing 8 mm MNP spin trap. B, as A, but using Fe 2+ -cyt c, which was prepared 29

30 by reducing Fe 3+ -cyt c with ascorbic acid followed by size-exclusion chromatography (see Experimental Procedures for details). The instrument settings were: center field, 3490 G; sweep width, 120 G; modulation amplitude, 5 G; microwave power, 20 mw; sweep time; 84 s; time constant 20.5 ms; receiver gain, ; number of accumulated scans, 4. FIG. 9. Effects of DMPO and MNP on cyt c-catalyzed DCFH 2 oxidation. DCFH 2 (10 µm) was incubated at 37 C with 250 nm cyt c, 250 µm hydrogen peroxide, and 0.1 M phosphate buffer, ph 7.4, alone (trace d), or with the addition of either 200, 100 or 10 mm DMPO (traces a to c, respectively) or 10 mm MNP (trace e). DCF generation was monitored fluorimetrically. SCHEME 1. Proposed mechanism for the generation of DMPOX from DMPO by cyt c/h 2 O 2 and its suppression by ethanol. DMPOX is the two-electron oxidation product of the DMPO hydroxyl radical adduct, DMPO/ OH. As the α-hydroxylethyl radical adduct was not detected upon the inclusion of ethanol, formation of the DMPO/ OH adduct cannot involve the trapping of free OH (dotted arrows). Instead, it is proposed that DMPO/ OH is formed by inverted spin-trapping, involving the nucleophilic addition of water to the DMPO radical cation, formed by one-electron oxidation of the spin trap by cyt c/h 2 O 2 (which is also responsible for further oxidation of DMPO/ OH to DMPOX). Ethanol is proposed to inhibit the formation of DMPOX by competing with water for nucleophilic addition to the DMPO radical cation, eventually forming an EPR-silent nitrone. 30

31 SCHEME 2. Proposed mechanism of cyt c activation by H 2 O 2 and its selective oxidation of substrates. The oxidizing species is proposed to be a compound-1 type intermediate, with one oxidizing equivalent held on the iron center (oxo-ferryl heme) and another as an electron-hole on the protein moiety (giving, through deprotonation, the neutral tyrosyl radical). It is proposed that DMPO, ascorbate, GSH, NADH and DCFH 2 are oxidized by the oxo-ferryl heme species (see text for details). Note that the oxidation of DMPO to DMPOX is not the simple, one-step reaction indicated, but involves several reaction steps (see Scheme 1); similarly, the oxidation of DCFH 2 to DCF is a twoelectron process, involving initial oxidation to a semiquinone (16). 31