Bicarbonate Enhances the Peroxidase Activity of Cu,Zn-Superoxide Dismutase

We examined the effect of bicarbonate on the peroxidase activity of copper-zinc superoxide dismutase (SOD1), using the nitrite anion as a peroxidase probe. Oxidation of nitrite by the enzyme-bound oxidant results in the formation of the nitrogen dioxide radical, which was measured by monitoring 5-nitro-γ-tocopherol formation. Results indicate that the presence of bicarbonate is not required for the peroxidase activity of SOD1, as monitored by the SOD1/H2O2-mediated nitration of γ-tocopherol in the presence of nitrite. However, bicarbonate enhanced SOD1/H2O2-dependent oxidation of tocopherols in the presence and absence of nitrite and dramatically enhanced SOD1/H2O2-mediated oxidation of unsaturated lipid in the presence of nitrite. These results, coupled with the finding that bicarbonate protects against inactivation of SOD1 by H2O2, suggest that SOD1/H2O2 oxidizes the bicarbonate anion to the carbonate radical anion. Thus, the amplification of peroxidase activity of SOD1/H2O2 by bicarbonate is attributed to the intermediary role of the diffusible oxidant, the carbonate radical anion. We conclude that, contrary to a previous report (Sankarapandi, S., and Zweier, J. L. (1999) J. Biol. Chem. 274, 1226–1232), bicarbonate is not required for peroxidase activity mediated by SOD1 and H2O2. However, bicarbonate enhanced the peroxidase activity of SOD1 via formation of a putative carbonate radical anion. Biological implications of the carbonate radical anion in free radical biology are discussed.

Recently, it was reported that the bicarbonate anion (HCO 3 Ϫ ) 1 is required for peroxidase activity and the peroxidase function of copper-zinc superoxide dismutase (SOD1) (1). The peroxidase activity of SOD1 was determined using the electron spin resonance (ESR) technique to monitor the oxidation of spin trap 5,5Ј-dimethyl-1-pyrroline-N-oxide (DMPO) to the DMPO-hydroxyl radical adduct (DMPO-OH). Additional evidence for enhanced peroxidase activity was obtained by measuring the oxidation of 2,2Ј-azino-bis(3-ethylbenzothiazoline)-6sulfonic acid (ABTS) to ABTS . ϩ radical cation (1). The investigators concluded that the bicarbonate anion, anchored to arginine 141 at the active site of SOD1, facilitated the redox cleavage of H 2 O 2 that led to enhanced peroxidase activity (1). In the present study, this interpretation is challenged, and an alternate mechanism for HCO 3 Ϫ -mediated increase in SOD1 peroxidase activity is provided.
Nearly 25 years ago, Hodgson and Fridovich demonstrated that the "copper-bound hydroxyl radical," SOD-Cu 2ϩ -⅐ OH, which is generated in the reaction between SOD1 and H 2 O 2 , oxidizes several anionic ligands including formate, azide, and nitrite anions (2,3). These small molecular weight anionic ligands are presumed to be oxidized by the oxidant formed at the active site of SOD1. The oxidizing potential of this putative SOD-Cu 2ϩ -⅐ OH species is similar to that of the "free" hydroxyl radical, which is considerably higher than those associated with conventional peroxidases (4). The one-electron oxidation potential for the HCO 3 Ϫ /carbonate radical anion (CO 3 . ) couple is ϩ1.59 V (5, 6), which makes oxidation of HCO 3 Ϫ by SOD-Cu 2ϩ -⅐ OH to CO 3 . thermodynamically feasible (7,8). CO 3 . , although less reactive than the hydroxyl radical, is a more selective oxidant that may diffuse over a longer distance, thereby causing oxidative damage to distant biological targets (9,10). We measured the peroxidase activity of SOD1 under anaerobic conditions by using the nitrite anion (NO 2 Ϫ ) as a peroxidase substrate (4). The oxidation of NO 2 Ϫ to the nitrogen dioxide free radical (NO 2 ⅐ ) was measured by monitoring the formation of ␣-tocopheryl quinone (␣-TQ) and 5-nitro-␥-tocopherol (NGTH) (4,(11)(12)(13)(14). Our results clearly demonstrate that the bicarbonate anion is not required for the peroxidase activity of SOD1; however, bicarbonate enhances oxidation of ␣and ␥-tocopherols in the presence and absence of the nitrite anion. In addition, bicarbonate provided protection against H 2 O 2dependent inactivation of SOD1. Taken together, these data suggest that SOD1 and H 2 O 2 are able to oxidize HCO 3 Ϫ to CO 3 . , a more selective oxidant, that leads to amplification of the peroxidase activity of SOD1. Biological implications of the oxidation of HCO 3 Ϫ , a ubiquitous cellular and plasma component, in free radical biology are discussed. EXPERIMENTAL PROCEDURES SOD1 (bovine) was purchased from Roche Molecular Biochemicals. 1,2-Dilauroyl-sn-glycero-3-phosphatidylcholine (DLPC) was purchased from Avanti Polar Lipids (Alabaster, AL). ␣-Tocopherol (␣-TH), ␥-tocopherol (␥-TH), and H 2 O 2 were obtained from Sigma.
Liposome Preparation-Liposomes were synthesized from DLPC (16). Methanolic solutions of ␣-TH, ␥-TH, ␣-TQ, or NGTH were added to the phospholipid, which was then dried down under a stream of nitrogen and placed in a vacuum dessicator overnight. Multilamellar liposomes were prepared by hydration of the dried lipid with phosphate buffer (200 mM, pH 7.4) containing diethylenetriaminepentaacetic acid (DTPA) (100 M) and thorough mixing. Unilamellar liposomes were prepared from the multilamellar liposomes by freeze-thawing five times using liquid nitrogen and extrusion through a 0.2-m polycarbonate filter (Nucleopore, Pleasanton, CA) five times using an extrusion apparatus (Lipex Biomembranes, Inc., Vancouver, BC).
Tocopherol Measurement-Detection and separation of ␣-TH, ␥-TH, and their oxidation products were performed on a HPLC system equipped with a UV detector. Samples (usually 200 l) were mixed with water (200 l) and ethanol (400 l) and vortexed for 60 s. Heptane (400 l) was added to the samples and vortex-mixed for an additional 60 s. Samples were centrifuged for 10 min at 10,000 rpm to separate the organic and aqueous phases. The organic layer (top) was removed, dried down under a stream of nitrogen, and stored at Ϫ20°C until analysis (Ͻ18 h). The samples were dissolved in methanol and analyzed by HPLC. The mobile phase was methanol/water (95:5) for 10 min, graded to 100% methanol over a 5-min period, and followed by 5 min at 100% methanol. The stationary phase was an analytical C 18 reversed phase column (Partisil ODS-3, 5-m particle size, Whatman Inc., Clifton, NJ). UV detection at ϭ 290 nm was used to quantify ␣-TH, ␥-TH, and NGTH levels, and ϭ 266 nm was used to quantify ␣-TQ.
Measurement of SOD1 Activity-SOD1 activity was measured using the ferricytochrome c (cyt c) reduction assay (4,17). Briefly, xanthine (0.5 mM) and xanthine oxidase (0.05 units/ml) were incubated with cyt c (20 M) in phosphate buffer (200 mM, pH 7.4). The rate of cyt c reduction by the superoxide anion was measured in the presence and absence of SOD1. To investigate the effect of HCO 3 Ϫ and NO 2 Ϫ on SOD1 inactivation by H 2 O 2 , SOD1 (1 mg/ml) was incubated with H 2 O 2 (1 mM) at 37°C in the presence and absence of HCO 3 Ϫ and NO 2 Ϫ . SOD1 activity was measured after diluting an aliquot (5 l) of the reaction mixture with 1 ml of the assay mixture (phosphate buffer containing 1 mM DTPA and 1000 units/ml catalase) (18).

Measurement of Copper and Zinc Release during Inactivation of SOD1 by H 2 O 2 -To investigate the effect of HCO 3
Ϫ on the release of copper and zinc from the enzyme during H 2 O 2 -mediated SOD1 inactivation, SOD1 (1 mg/ml) was incubated with H 2 O 2 (1 mM) in Chelextreated phosphate buffer (200 mM, pH 7.4) at 37°C in the presence of 4-pyridylazaresorcinol (PAR, 100 M). The change in absorbance due to Cu 2ϩ -and Zn 2ϩ -PAR complex formation was followed for 3 h at 500 nm. The amount of zinc released during this process was calculated by adding 1.6 mM of nitrilotriacetic acid at the end of the reaction. The decrease in absorbance corresponds to Zn 2ϩ -nitrilotriacetic acid complex formation (19). The amount of copper released during this process was calculated by the subsequent addition of 0.8 mM EDTA to the incubation mixture as previously reported (19). The molar extinction coefficient of the Cu 2ϩ -and Zn 2ϩ -PAR complex remained unchanged up to 250 mM HCO 3 Ϫ . ESR Measurements-The formation of ␣-tocopheroxyl (␣-T ⅐ ) and 5-nitro-␥-tocopheroxyl (NGT ⅐ ) radicals was monitored by ESR (4,20). ESR measurements were performed at ambient temperature using a Varian E-109 spectrometer operating at 9.5 GHz (X-band) and 100 kHz field modulation. A typical incubation for ESR experiments consisted of SOD1 (10 mg/ml), H 2 O 2 (10 mM); bicarbonate (25 mM); and ␣-TH, ␥-TH, or NGTH (2 mM) incorporated into DLPC liposomes (100 mM) in 0.25 ml of phosphate buffer (200 mM, pH 7.4) containing DTPA (100 M). The reaction was initiated by the addition of H 2 O 2 . Samples were immediately analyzed in a quartz flat cell.

HPLC Analysis of ␣-TH, ␥-TH, and Their Oxidation and
Nitration Products-␣-TH and ␥-TH were used as probes to monitor the peroxidase activity of SOD1 in the presence of H 2 O 2 (4). In order to limit CO 2 contamination, all solutions were degassed and stored in a nitrogen glove box. After a minimum of 48 h, experiments were performed in the nitrogen glove box. A buffer concentration of 200 mM phosphate was used to prevent changes in pH after the addition of bicarbonate solutions. The pH was determined both before and after the termination of all reactions to ensure that the bicarbonate effect was not due to pH changes. Fig. 1 shows typical HPLC traces obtained when analyzing large unilamellar DLPC liposomes containing ␣-TH, ␥-TH, NGTH, or ␣-TQ (structures shown in insets). Fluorescence detection ( ex ϭ 275 nm, em ϭ 320 nm) was used to detect both ␥-TH (B) and ␣-TH (D) with retention times of 13.7 and 15.9 min, respectively. NGTH (A), the nitration product of ␥-TH, was observed to have a retention time of 18.0 min at 290 nm. ␣-TQ (C), the two-electron oxidation product of ␣-TH, was observed to have a retention time of 11.4 min at 266 nm. All peaks were verified using authentic standards.
The Effect of Approximately 10 M ␣-TH was consumed during the first hour of incubation, after which tocopherol depletion occurred at a slower rate (5 M during the next hour and 6 M over the next 3 h). This result can be explained in terms of the reaction between ␣-TH and NO 2 ⅐ by SOD1/H 2 O 2 (4). The addition of HCO 3 Ϫ to this system markedly enhanced ␣-TH oxidation and ␣-TQ formation. When 3 mM HCO 3 Ϫ was added to the solution, little change in ␣-TH consumption occurred during the first hour. However, by the fourth hour of incubation, a small increase in consumption was observed. With increasing HCO 3 Ϫ concentration, a significant increase in ␣-TH consumption ( Ϫ had little effect on NGTH formation. Fig. 3C shows the effect of the presence or absence of NO 2 Ϫ and HCO 3 Ϫ on ␥-TH depletion and NGTH formation after a 4-h incubation in the presence of SOD1 (1 mg/ml) and H 2 O 2 (1 mM). Approximately 5 M ␥-TH was consumed over a 4-h period, possibly due to NO 2 Ϫ contamination (as evidenced by a slight increase in the formation of NGTH, control t ϭ 4 h). In the presence of added NO 2 Ϫ (1 mM), 20 M ␥-TH was consumed, and 10 M NGTH was formed. In the presence of HCO 3 Ϫ (20 mM), 20 M ␥-TH was also consumed, but NGTH formation was negligible. When both HCO 3 Ϫ and NO 2 Ϫ were present, approximately 90% of ␥-TH was consumed, with a concomitant formation of approximately 8 M NGTH. This paradoxical result (i.e. enhanced ␥-TH depletion and decreased NGTH formation) can be explained if CO 3 . also caused the oxidation of NGTH.
These reactions suggest that HCO 3 Ϫ is not required for SOD1/ H 2 O 2 /NO 2 Ϫ -mediated nitration of ␥-TH. The presence of HCO 3 Ϫ , however, enhanced the oxidation of ␥-TH and possibly NGTH.
ESR Detection of Tocopheroxyl Radicals: Enhancement by HCO 3 Ϫ -To probe the involvement of tocopheroxyl radical during HCO 3 Ϫ -enhanced oxidation of tocopherols, we used direct ESR. The addition of H 2 O 2 (10 mM) to an incubation mixture containing ␣-TH (2 mM) and SOD1 (10 mg/ml) produced a seven-line ESR spectrum (Fig. 4A, a) characteristic of the ␣-T ⅐ radical. In the presence of HCO 3 Ϫ , there was a marked increase in signal intensity (Fig. 4A, b). Due to a decreased signal-tonoise ratio, a high modulation amplitude was used that restricted our ability to resolve couplings from the methyl protons at carbon-8 and methylene protons at carbon-4. The computer simulation of the ␣-T ⅐ radical is shown in Fig. 4A, c (a CH 3 5 (3H) ϭ 5.66 G; a CH 3 7 (3H) ϭ 4.49 G) (20,22,23). When either SOD1 or H 2 O 2 was excluded, there was no detectable ESR signal (data not shown).
Incubation of DLPC liposomes containing NGTH with SOD1/ H 2 O 2 produced a four-line spectrum (Fig. 4B, a) with a 1:3:3:1 intensity ratio, typical for an electron interacting with three equivalent protons. The addition of 25 mM HCO 3 Ϫ resulted in the enhancement of this signal (Fig. 4B, b). The computer simulation of this spectrum (dotted lines) is shown in Fig. 4B, c (a CH 3 7 (3H) ϭ 4.35 G). At this high modulation, the nitrogen coupling could not be resolved.
DLPC liposomes containing ␣-TH (60 M) were incubated with SOD1 (1 mg/ml) and H 2 O 2 (1 mM). Fig. 4C shows HPLC traces of samples incubated in the absence (Fig. 4C, a, 58 Ϯ 1 M ␣-TH) or presence (Fig. 4C, b, (Fig. 4D, a) and to 14 Ϯ 1 M in the presence of HCO 3 Ϫ (25 mM, Fig. 4D, b). The ESR and HPLC results clearly demonstrate that the HCO 3 Ϫ anion enhanced SOD1/H 2 O 2 -mediated oxidation of tocopherols via a radical-mediated pathway. It is likely that an oxidant, possibly the CO 3 .  (Fig. 5). The addition of NO 2 Ϫ (1 mM) resulted in an increase in absorbance at 234 nm, corresponding to the formation of conjugated dienes as a function of time. HCO 3 Ϫ (25 mM) significantly enhanced conjugated diene formation in the presence of NO 2 Ϫ . Fig. 6A (inset) shows the rate of superoxide-dependent cyt c reduction, monitored as an increase in absorbance at 550 nm (a). In the presence of SOD1, the increase in absorbance was inhibited (f). After incubation of SOD1 with H 2 O 2 for 4 h, SOD1-mediated inhibition of cyt c reduction was less effective (b). The presence of HCO 3 Ϫ (c), NO 2 Ϫ (d), or HCO 3 Ϫ and NO 2 Ϫ (e) prevented the inactivation of SOD1 by H 2 O 2 . SOD1 activity was calculated from the initial rate of cyt c reduction and is shown as a percentage of the rates of cyt c observed in the absence (a) or in the presence (f) of SOD1 (n ϭ 3 Ϯ S.D.). Inactivation of SOD1 was accompanied by the release of copper and zinc from the active site as determined by the PAR assay. NO 2 Ϫ (1 mM) and/or HCO 3 Ϫ (20 mM) inhibited the release of copper and zinc during H 2 O 2 -mediated inactivation of SOD1 (Fig. 6B).

DISCUSSION
Bicarbonate Anion as a Peroxidase Substrate-The oxidative inactivation of SOD1 caused by H 2 O 2 has been attributed to the "peroxidase activity" (2, 3). This "peroxidase activity" results in the generation of a highly potent, "site-specific" hydroxyl radical-like species that inactivates the enzyme by oxidative attack on a histidine residue (His-61) at the active site (24 -27) (Scheme 1).
The putative oxidant, SOD-Cu 2ϩ -⅐ OH, reacts with several electron donor compounds to form the corresponding radical while protecting the enzyme from oxidative inactivation. HCO 3 Ϫ partially protects against H 2 O 2 -induced inactivation of SOD1 and inhibits copper release. This may be attributed to the oxidation of HCO 3 Ϫ to CO 3 . by the enzyme-bound oxidant (2)(3)(4).

SCHEME 2
Unlike the copper-bound hydroxyl radical at the active site, CO 3 . is a freely diffusible oxidant that can oxidize target molecules at a distance (9,10). We attribute the increased peroxidase activity of SOD1 (observed in the presence of HCO 3 Ϫ ) to the oxidative reactions catalyzed by CO 3 . . 3 . -In contrast to CO 2 . , a reducing radical formed from oxidation of the formate anion, CO 3 . is an oxidizing radical (28,29 reduces ⅐ NO to NO Ϫ , CO 3 . oxidizes ⅐ NO to NO 2 Ϫ (30). CO 3 . forms a conjugate acid and base form, and this radical is mostly unprotonated and exists as an anion at physiological pH levels (31,32).

Oxidative Reactions of CO
The HCO 3 Ϫ -induced depletion of tocopherols in the presence of SOD1 and H 2 O 2 reported in the present study can be explained based on the following reactions. CO 3 .
Ϫ -dependent oxidation of tocopherols and unsaturated fatty acid in the presence of SOD1 and H 2 O 2 . We attribute this to a fast electron transfer reaction between CO 3 . and NO 2 Ϫ that results in the formation of NO 2 ⅐ (33).
NO 2 ⅐ can abstract the phenolic hydrogen atom from the tocopherols to form the tocopheroxyl radical (34). NO 2 ⅐ reacts with ␣-T ⅐ to form the corresponding quinone ␣-TQ. Alternatively, NO 2 ⅐ reacts with the linoleate-type fatty acid (k ϭ 10 5 to 10 6 M Ϫ1 s Ϫ1 ) to form the lipid-conjugated diene.
ϩ cation radical has been postulated as an intermediate (7,36 Biological Implications-Hodgson and Fridovich (38) reported that the addition of carbonate anion to an aerobic xanthine/xanthine oxidase system produced luminescence. This effect was attributed to a dimerization of CO 3 . radical anion (39) formed from the reaction between ⅐ OH and CO 3 2Ϫ (33) as follows.
In the pioneering studies of the peroxidase activity of SOD1, some experiments were performed in carbonate buffer (2,3). It is likely that H 2 O 2 -induced inactivation of SOD1 was partially protected by the carbonate anion.
Bicarbonate has been shown to alter the reactivity of reactive nitrogen species (40 -48). Peroxynitrite, a potent oxidant formed from a diffusion-controlled reaction between O 2 . and ⅐ NO, has been proposed to react with CO 2 to generate an intermediate that consists of a caged radical pair, CO 3 . /NO 2 ⅐ . The chemiluminescence detected in the reaction between ONOO Ϫ and CO 2 was attributed to the formation of a carbonate anion radical (49). The CO 2 -peroxynitrite reaction facilitated the nitration and oxidation of tyrosine (39). This scenario is similar to the bicarbonate-enhanced reactivity of SOD1/ H 2 O 2 /NO 2 Ϫ reported in the present study. HCO 3 Ϫ is abundant in biological systems. The plasma concentration of HCO 3 Ϫ is approximately 25 mM (50). Recently, an extracellular SOD1 has been identified in atherogenesis (51). The peroxidase activity of this extracellular SOD1 has been suggested to play an important role in the oxidative modification of lipid. This activity may be amplified in the presence of HCO 3 Ϫ . There is emerging evidence that links familial amyotrophic lateral sclerosis to mutations in the gene encoding cytosolic SOD1 (52). The mutations in SOD1 induce a gain in function that is associated with increased peroxidase activity (53). HCO 3 Ϫ may influence the radical reactions catalyzed by these mutations of SOD1. Regulation of oxidative reactions by HCO 3 Ϫ in other physiologically relevant inflammatory processes (e.g. myeloperoxidase) may also be important (54,55).
In conclusion, in contrast to a previous report (1), evidence presented in this study strongly suggests that HCO 3 Ϫ is not an absolute requirement for eliciting the peroxidase activity of SOD1. However, HCO 3 Ϫ enhances the peroxidase activity of SOD1 by acting as a "sacrificial" electron donor and in so doing forms a CO 3 . radical intermediate, a selective yet potent biological oxidant. The involvement of CO 3 . in biological oxidation seems more ubiquitous (56) and should be taken into consideration in future studies.